{
  "best_global_step": 9800,
  "best_metric": 2.5283308029174805,
  "best_model_checkpoint": "saves/lntuning/mistral-7b-instruct/train_multirc_1745950270/checkpoint-9800",
  "epoch": 6.525328330206379,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008157272208173587,
      "grad_norm": 8.5,
      "learning_rate": 4.999999876629946e-05,
      "loss": 2.3824,
      "num_input_tokens_seen": 8784,
      "step": 5
    },
    {
      "epoch": 0.0016314544416347174,
      "grad_norm": 5.5,
      "learning_rate": 4.999999375439123e-05,
      "loss": 2.2428,
      "num_input_tokens_seen": 18528,
      "step": 10
    },
    {
      "epoch": 0.002447181662452076,
      "grad_norm": 9.0,
      "learning_rate": 4.9999984887169785e-05,
      "loss": 3.4286,
      "num_input_tokens_seen": 29168,
      "step": 15
    },
    {
      "epoch": 0.0032629088832694347,
      "grad_norm": 5.15625,
      "learning_rate": 4.9999972164636506e-05,
      "loss": 2.5567,
      "num_input_tokens_seen": 40624,
      "step": 20
    },
    {
      "epoch": 0.004078636104086793,
      "grad_norm": 7.90625,
      "learning_rate": 4.999995558679334e-05,
      "loss": 2.8244,
      "num_input_tokens_seen": 49968,
      "step": 25
    },
    {
      "epoch": 0.004894363324904152,
      "grad_norm": 7.0625,
      "learning_rate": 4.999993515364287e-05,
      "loss": 1.5947,
      "num_input_tokens_seen": 60672,
      "step": 30
    },
    {
      "epoch": 0.005710090545721511,
      "grad_norm": 8.625,
      "learning_rate": 4.999991086518822e-05,
      "loss": 3.8766,
      "num_input_tokens_seen": 71408,
      "step": 35
    },
    {
      "epoch": 0.0065258177665388694,
      "grad_norm": 3.9375,
      "learning_rate": 4.999988272143315e-05,
      "loss": 2.6206,
      "num_input_tokens_seen": 82752,
      "step": 40
    },
    {
      "epoch": 0.007341544987356228,
      "grad_norm": 11.875,
      "learning_rate": 4.999985072238199e-05,
      "loss": 2.6362,
      "num_input_tokens_seen": 93296,
      "step": 45
    },
    {
      "epoch": 0.008157272208173586,
      "grad_norm": 6.84375,
      "learning_rate": 4.999981486803969e-05,
      "loss": 2.9237,
      "num_input_tokens_seen": 104592,
      "step": 50
    },
    {
      "epoch": 0.008972999428990946,
      "grad_norm": 1.25,
      "learning_rate": 4.999977515841176e-05,
      "loss": 2.3438,
      "num_input_tokens_seen": 114976,
      "step": 55
    },
    {
      "epoch": 0.009788726649808304,
      "grad_norm": 11.8125,
      "learning_rate": 4.9999731593504344e-05,
      "loss": 3.9838,
      "num_input_tokens_seen": 124944,
      "step": 60
    },
    {
      "epoch": 0.010604453870625663,
      "grad_norm": 10.875,
      "learning_rate": 4.999968417332415e-05,
      "loss": 2.651,
      "num_input_tokens_seen": 134560,
      "step": 65
    },
    {
      "epoch": 0.011420181091443021,
      "grad_norm": 1.7734375,
      "learning_rate": 4.999963289787848e-05,
      "loss": 1.7442,
      "num_input_tokens_seen": 144528,
      "step": 70
    },
    {
      "epoch": 0.012235908312260381,
      "grad_norm": 8.375,
      "learning_rate": 4.999957776717526e-05,
      "loss": 2.3111,
      "num_input_tokens_seen": 155056,
      "step": 75
    },
    {
      "epoch": 0.013051635533077739,
      "grad_norm": 2.796875,
      "learning_rate": 4.9999518781222984e-05,
      "loss": 2.7145,
      "num_input_tokens_seen": 164464,
      "step": 80
    },
    {
      "epoch": 0.013867362753895097,
      "grad_norm": 10.25,
      "learning_rate": 4.9999455940030746e-05,
      "loss": 3.3305,
      "num_input_tokens_seen": 174608,
      "step": 85
    },
    {
      "epoch": 0.014683089974712456,
      "grad_norm": 5.3125,
      "learning_rate": 4.999938924360824e-05,
      "loss": 1.7676,
      "num_input_tokens_seen": 185616,
      "step": 90
    },
    {
      "epoch": 0.015498817195529814,
      "grad_norm": 8.625,
      "learning_rate": 4.999931869196575e-05,
      "loss": 2.6622,
      "num_input_tokens_seen": 195360,
      "step": 95
    },
    {
      "epoch": 0.016314544416347172,
      "grad_norm": 11.8125,
      "learning_rate": 4.999924428511416e-05,
      "loss": 1.9336,
      "num_input_tokens_seen": 207152,
      "step": 100
    },
    {
      "epoch": 0.017130271637164532,
      "grad_norm": 10.625,
      "learning_rate": 4.999916602306494e-05,
      "loss": 2.0811,
      "num_input_tokens_seen": 218016,
      "step": 105
    },
    {
      "epoch": 0.01794599885798189,
      "grad_norm": 1.7734375,
      "learning_rate": 4.999908390583016e-05,
      "loss": 2.7338,
      "num_input_tokens_seen": 229680,
      "step": 110
    },
    {
      "epoch": 0.01876172607879925,
      "grad_norm": 4.40625,
      "learning_rate": 4.999899793342247e-05,
      "loss": 2.3528,
      "num_input_tokens_seen": 238624,
      "step": 115
    },
    {
      "epoch": 0.019577453299616607,
      "grad_norm": 7.28125,
      "learning_rate": 4.999890810585516e-05,
      "loss": 3.1091,
      "num_input_tokens_seen": 248592,
      "step": 120
    },
    {
      "epoch": 0.020393180520433967,
      "grad_norm": 6.40625,
      "learning_rate": 4.999881442314206e-05,
      "loss": 1.8262,
      "num_input_tokens_seen": 259392,
      "step": 125
    },
    {
      "epoch": 0.021208907741251327,
      "grad_norm": 8.6875,
      "learning_rate": 4.9998716885297617e-05,
      "loss": 4.0175,
      "num_input_tokens_seen": 269664,
      "step": 130
    },
    {
      "epoch": 0.022024634962068683,
      "grad_norm": 7.5625,
      "learning_rate": 4.999861549233688e-05,
      "loss": 2.6776,
      "num_input_tokens_seen": 280800,
      "step": 135
    },
    {
      "epoch": 0.022840362182886043,
      "grad_norm": 0.263671875,
      "learning_rate": 4.999851024427548e-05,
      "loss": 1.98,
      "num_input_tokens_seen": 290736,
      "step": 140
    },
    {
      "epoch": 0.023656089403703402,
      "grad_norm": 6.59375,
      "learning_rate": 4.999840114112965e-05,
      "loss": 1.8577,
      "num_input_tokens_seen": 301152,
      "step": 145
    },
    {
      "epoch": 0.024471816624520762,
      "grad_norm": 7.0625,
      "learning_rate": 4.999828818291621e-05,
      "loss": 1.5681,
      "num_input_tokens_seen": 311040,
      "step": 150
    },
    {
      "epoch": 0.025287543845338118,
      "grad_norm": 11.5,
      "learning_rate": 4.999817136965259e-05,
      "loss": 2.5782,
      "num_input_tokens_seen": 322512,
      "step": 155
    },
    {
      "epoch": 0.026103271066155478,
      "grad_norm": 10.1875,
      "learning_rate": 4.9998050701356794e-05,
      "loss": 2.9134,
      "num_input_tokens_seen": 333664,
      "step": 160
    },
    {
      "epoch": 0.026918998286972837,
      "grad_norm": 8.375,
      "learning_rate": 4.999792617804744e-05,
      "loss": 2.7584,
      "num_input_tokens_seen": 345104,
      "step": 165
    },
    {
      "epoch": 0.027734725507790194,
      "grad_norm": 4.71875,
      "learning_rate": 4.9997797799743724e-05,
      "loss": 2.09,
      "num_input_tokens_seen": 356288,
      "step": 170
    },
    {
      "epoch": 0.028550452728607553,
      "grad_norm": 9.6875,
      "learning_rate": 4.999766556646545e-05,
      "loss": 2.6638,
      "num_input_tokens_seen": 365744,
      "step": 175
    },
    {
      "epoch": 0.029366179949424913,
      "grad_norm": 9.375,
      "learning_rate": 4.9997529478232996e-05,
      "loss": 3.0835,
      "num_input_tokens_seen": 376080,
      "step": 180
    },
    {
      "epoch": 0.030181907170242273,
      "grad_norm": 7.53125,
      "learning_rate": 4.9997389535067365e-05,
      "loss": 2.6222,
      "num_input_tokens_seen": 386784,
      "step": 185
    },
    {
      "epoch": 0.03099763439105963,
      "grad_norm": 6.96875,
      "learning_rate": 4.999724573699012e-05,
      "loss": 3.8699,
      "num_input_tokens_seen": 396944,
      "step": 190
    },
    {
      "epoch": 0.031813361611876985,
      "grad_norm": 13.0,
      "learning_rate": 4.9997098084023457e-05,
      "loss": 2.9014,
      "num_input_tokens_seen": 407952,
      "step": 195
    },
    {
      "epoch": 0.032629088832694345,
      "grad_norm": 2.421875,
      "learning_rate": 4.999694657619013e-05,
      "loss": 1.8061,
      "num_input_tokens_seen": 418192,
      "step": 200
    },
    {
      "epoch": 0.032629088832694345,
      "eval_loss": 2.6757116317749023,
      "eval_runtime": 134.4547,
      "eval_samples_per_second": 20.267,
      "eval_steps_per_second": 10.137,
      "num_input_tokens_seen": 418192,
      "step": 200
    },
    {
      "epoch": 0.033444816053511704,
      "grad_norm": 9.25,
      "learning_rate": 4.999679121351352e-05,
      "loss": 2.4379,
      "num_input_tokens_seen": 429056,
      "step": 205
    },
    {
      "epoch": 0.034260543274329064,
      "grad_norm": 2.578125,
      "learning_rate": 4.9996631996017565e-05,
      "loss": 2.0123,
      "num_input_tokens_seen": 440768,
      "step": 210
    },
    {
      "epoch": 0.035076270495146424,
      "grad_norm": 8.375,
      "learning_rate": 4.9996468923726835e-05,
      "loss": 3.8622,
      "num_input_tokens_seen": 451120,
      "step": 215
    },
    {
      "epoch": 0.03589199771596378,
      "grad_norm": 5.25,
      "learning_rate": 4.999630199666647e-05,
      "loss": 1.69,
      "num_input_tokens_seen": 463056,
      "step": 220
    },
    {
      "epoch": 0.03670772493678114,
      "grad_norm": 10.5,
      "learning_rate": 4.999613121486222e-05,
      "loss": 2.3376,
      "num_input_tokens_seen": 474048,
      "step": 225
    },
    {
      "epoch": 0.0375234521575985,
      "grad_norm": 9.8125,
      "learning_rate": 4.999595657834041e-05,
      "loss": 2.4052,
      "num_input_tokens_seen": 483968,
      "step": 230
    },
    {
      "epoch": 0.038339179378415855,
      "grad_norm": 3.875,
      "learning_rate": 4.999577808712798e-05,
      "loss": 1.6723,
      "num_input_tokens_seen": 494032,
      "step": 235
    },
    {
      "epoch": 0.039154906599233215,
      "grad_norm": 12.125,
      "learning_rate": 4.999559574125244e-05,
      "loss": 3.0367,
      "num_input_tokens_seen": 505168,
      "step": 240
    },
    {
      "epoch": 0.039970633820050575,
      "grad_norm": 4.28125,
      "learning_rate": 4.9995409540741934e-05,
      "loss": 1.3688,
      "num_input_tokens_seen": 514768,
      "step": 245
    },
    {
      "epoch": 0.040786361040867934,
      "grad_norm": 2.046875,
      "learning_rate": 4.999521948562516e-05,
      "loss": 1.7807,
      "num_input_tokens_seen": 525776,
      "step": 250
    },
    {
      "epoch": 0.041602088261685294,
      "grad_norm": 5.5,
      "learning_rate": 4.999502557593143e-05,
      "loss": 2.1225,
      "num_input_tokens_seen": 535216,
      "step": 255
    },
    {
      "epoch": 0.042417815482502653,
      "grad_norm": 2.5625,
      "learning_rate": 4.999482781169066e-05,
      "loss": 1.8369,
      "num_input_tokens_seen": 544864,
      "step": 260
    },
    {
      "epoch": 0.04323354270332001,
      "grad_norm": 3.71875,
      "learning_rate": 4.9994626192933324e-05,
      "loss": 1.7563,
      "num_input_tokens_seen": 556624,
      "step": 265
    },
    {
      "epoch": 0.044049269924137366,
      "grad_norm": 5.21875,
      "learning_rate": 4.999442071969054e-05,
      "loss": 2.7922,
      "num_input_tokens_seen": 566320,
      "step": 270
    },
    {
      "epoch": 0.044864997144954726,
      "grad_norm": 7.90625,
      "learning_rate": 4.999421139199397e-05,
      "loss": 1.5266,
      "num_input_tokens_seen": 577312,
      "step": 275
    },
    {
      "epoch": 0.045680724365772085,
      "grad_norm": 2.84375,
      "learning_rate": 4.999399820987592e-05,
      "loss": 2.6298,
      "num_input_tokens_seen": 587840,
      "step": 280
    },
    {
      "epoch": 0.046496451586589445,
      "grad_norm": 3.8125,
      "learning_rate": 4.999378117336924e-05,
      "loss": 1.4227,
      "num_input_tokens_seen": 598224,
      "step": 285
    },
    {
      "epoch": 0.047312178807406804,
      "grad_norm": 10.375,
      "learning_rate": 4.9993560282507415e-05,
      "loss": 2.2756,
      "num_input_tokens_seen": 607856,
      "step": 290
    },
    {
      "epoch": 0.048127906028224164,
      "grad_norm": 12.5,
      "learning_rate": 4.9993335537324495e-05,
      "loss": 2.8238,
      "num_input_tokens_seen": 618656,
      "step": 295
    },
    {
      "epoch": 0.048943633249041524,
      "grad_norm": 3.515625,
      "learning_rate": 4.999310693785516e-05,
      "loss": 1.4007,
      "num_input_tokens_seen": 629056,
      "step": 300
    },
    {
      "epoch": 0.049759360469858877,
      "grad_norm": 8.6875,
      "learning_rate": 4.9992874484134653e-05,
      "loss": 1.9266,
      "num_input_tokens_seen": 641280,
      "step": 305
    },
    {
      "epoch": 0.050575087690676236,
      "grad_norm": 5.375,
      "learning_rate": 4.999263817619882e-05,
      "loss": 1.5467,
      "num_input_tokens_seen": 650992,
      "step": 310
    },
    {
      "epoch": 0.051390814911493596,
      "grad_norm": 10.5625,
      "learning_rate": 4.9992398014084105e-05,
      "loss": 3.6772,
      "num_input_tokens_seen": 662112,
      "step": 315
    },
    {
      "epoch": 0.052206542132310955,
      "grad_norm": 7.6875,
      "learning_rate": 4.999215399782754e-05,
      "loss": 2.7125,
      "num_input_tokens_seen": 672912,
      "step": 320
    },
    {
      "epoch": 0.053022269353128315,
      "grad_norm": 7.78125,
      "learning_rate": 4.999190612746675e-05,
      "loss": 3.8661,
      "num_input_tokens_seen": 682864,
      "step": 325
    },
    {
      "epoch": 0.053837996573945675,
      "grad_norm": 6.4375,
      "learning_rate": 4.999165440303998e-05,
      "loss": 3.0503,
      "num_input_tokens_seen": 692480,
      "step": 330
    },
    {
      "epoch": 0.054653723794763034,
      "grad_norm": 5.28125,
      "learning_rate": 4.999139882458603e-05,
      "loss": 1.3731,
      "num_input_tokens_seen": 703312,
      "step": 335
    },
    {
      "epoch": 0.05546945101558039,
      "grad_norm": 3.703125,
      "learning_rate": 4.9991139392144314e-05,
      "loss": 1.5192,
      "num_input_tokens_seen": 713744,
      "step": 340
    },
    {
      "epoch": 0.05628517823639775,
      "grad_norm": 7.5,
      "learning_rate": 4.999087610575485e-05,
      "loss": 2.8708,
      "num_input_tokens_seen": 722880,
      "step": 345
    },
    {
      "epoch": 0.057100905457215106,
      "grad_norm": 8.125,
      "learning_rate": 4.999060896545824e-05,
      "loss": 2.995,
      "num_input_tokens_seen": 732912,
      "step": 350
    },
    {
      "epoch": 0.057916632678032466,
      "grad_norm": 9.0625,
      "learning_rate": 4.999033797129568e-05,
      "loss": 2.0663,
      "num_input_tokens_seen": 744640,
      "step": 355
    },
    {
      "epoch": 0.058732359898849826,
      "grad_norm": 7.125,
      "learning_rate": 4.999006312330894e-05,
      "loss": 1.2672,
      "num_input_tokens_seen": 754432,
      "step": 360
    },
    {
      "epoch": 0.059548087119667185,
      "grad_norm": 3.421875,
      "learning_rate": 4.998978442154043e-05,
      "loss": 3.1734,
      "num_input_tokens_seen": 766304,
      "step": 365
    },
    {
      "epoch": 0.060363814340484545,
      "grad_norm": 5.75,
      "learning_rate": 4.9989501866033125e-05,
      "loss": 1.9721,
      "num_input_tokens_seen": 777072,
      "step": 370
    },
    {
      "epoch": 0.0611795415613019,
      "grad_norm": 9.8125,
      "learning_rate": 4.998921545683059e-05,
      "loss": 3.5648,
      "num_input_tokens_seen": 787808,
      "step": 375
    },
    {
      "epoch": 0.06199526878211926,
      "grad_norm": 7.46875,
      "learning_rate": 4.9988925193976996e-05,
      "loss": 3.6657,
      "num_input_tokens_seen": 798016,
      "step": 380
    },
    {
      "epoch": 0.06281099600293662,
      "grad_norm": 5.1875,
      "learning_rate": 4.998863107751711e-05,
      "loss": 3.4512,
      "num_input_tokens_seen": 807632,
      "step": 385
    },
    {
      "epoch": 0.06362672322375397,
      "grad_norm": 8.125,
      "learning_rate": 4.998833310749629e-05,
      "loss": 2.8466,
      "num_input_tokens_seen": 816608,
      "step": 390
    },
    {
      "epoch": 0.06444245044457134,
      "grad_norm": 1.8359375,
      "learning_rate": 4.998803128396047e-05,
      "loss": 1.973,
      "num_input_tokens_seen": 826208,
      "step": 395
    },
    {
      "epoch": 0.06525817766538869,
      "grad_norm": 4.84375,
      "learning_rate": 4.9987725606956215e-05,
      "loss": 1.6352,
      "num_input_tokens_seen": 836224,
      "step": 400
    },
    {
      "epoch": 0.06525817766538869,
      "eval_loss": 2.597731351852417,
      "eval_runtime": 134.5236,
      "eval_samples_per_second": 20.257,
      "eval_steps_per_second": 10.132,
      "num_input_tokens_seen": 836224,
      "step": 400
    },
    {
      "epoch": 0.06607390488620606,
      "grad_norm": 4.8125,
      "learning_rate": 4.998741607653066e-05,
      "loss": 2.1235,
      "num_input_tokens_seen": 846160,
      "step": 405
    },
    {
      "epoch": 0.06688963210702341,
      "grad_norm": 2.71875,
      "learning_rate": 4.9987102692731523e-05,
      "loss": 1.5486,
      "num_input_tokens_seen": 856608,
      "step": 410
    },
    {
      "epoch": 0.06770535932784078,
      "grad_norm": 22.25,
      "learning_rate": 4.9986785455607157e-05,
      "loss": 2.1981,
      "num_input_tokens_seen": 867568,
      "step": 415
    },
    {
      "epoch": 0.06852108654865813,
      "grad_norm": 10.9375,
      "learning_rate": 4.9986464365206456e-05,
      "loss": 1.3882,
      "num_input_tokens_seen": 878496,
      "step": 420
    },
    {
      "epoch": 0.06933681376947548,
      "grad_norm": 7.3125,
      "learning_rate": 4.9986139421578956e-05,
      "loss": 2.2494,
      "num_input_tokens_seen": 888784,
      "step": 425
    },
    {
      "epoch": 0.07015254099029285,
      "grad_norm": 3.984375,
      "learning_rate": 4.998581062477477e-05,
      "loss": 1.0806,
      "num_input_tokens_seen": 899392,
      "step": 430
    },
    {
      "epoch": 0.0709682682111102,
      "grad_norm": 7.21875,
      "learning_rate": 4.998547797484458e-05,
      "loss": 2.2443,
      "num_input_tokens_seen": 910112,
      "step": 435
    },
    {
      "epoch": 0.07178399543192757,
      "grad_norm": 7.03125,
      "learning_rate": 4.9985141471839706e-05,
      "loss": 2.548,
      "num_input_tokens_seen": 920000,
      "step": 440
    },
    {
      "epoch": 0.07259972265274492,
      "grad_norm": 3.828125,
      "learning_rate": 4.998480111581203e-05,
      "loss": 4.4676,
      "num_input_tokens_seen": 930064,
      "step": 445
    },
    {
      "epoch": 0.07341544987356229,
      "grad_norm": 9.75,
      "learning_rate": 4.998445690681405e-05,
      "loss": 1.6247,
      "num_input_tokens_seen": 940512,
      "step": 450
    },
    {
      "epoch": 0.07423117709437964,
      "grad_norm": 5.0,
      "learning_rate": 4.9984108844898834e-05,
      "loss": 2.3822,
      "num_input_tokens_seen": 951472,
      "step": 455
    },
    {
      "epoch": 0.075046904315197,
      "grad_norm": 7.0,
      "learning_rate": 4.9983756930120076e-05,
      "loss": 2.2208,
      "num_input_tokens_seen": 961456,
      "step": 460
    },
    {
      "epoch": 0.07586263153601436,
      "grad_norm": 4.96875,
      "learning_rate": 4.9983401162532025e-05,
      "loss": 2.1054,
      "num_input_tokens_seen": 972896,
      "step": 465
    },
    {
      "epoch": 0.07667835875683171,
      "grad_norm": 12.625,
      "learning_rate": 4.998304154218955e-05,
      "loss": 1.9543,
      "num_input_tokens_seen": 984560,
      "step": 470
    },
    {
      "epoch": 0.07749408597764908,
      "grad_norm": 10.5625,
      "learning_rate": 4.998267806914812e-05,
      "loss": 2.9675,
      "num_input_tokens_seen": 994912,
      "step": 475
    },
    {
      "epoch": 0.07830981319846643,
      "grad_norm": 5.1875,
      "learning_rate": 4.998231074346378e-05,
      "loss": 2.2254,
      "num_input_tokens_seen": 1004656,
      "step": 480
    },
    {
      "epoch": 0.0791255404192838,
      "grad_norm": 8.0,
      "learning_rate": 4.998193956519317e-05,
      "loss": 1.6657,
      "num_input_tokens_seen": 1016000,
      "step": 485
    },
    {
      "epoch": 0.07994126764010115,
      "grad_norm": 2.0,
      "learning_rate": 4.9981564534393545e-05,
      "loss": 3.3288,
      "num_input_tokens_seen": 1027104,
      "step": 490
    },
    {
      "epoch": 0.08075699486091852,
      "grad_norm": 9.3125,
      "learning_rate": 4.998118565112272e-05,
      "loss": 2.7015,
      "num_input_tokens_seen": 1035696,
      "step": 495
    },
    {
      "epoch": 0.08157272208173587,
      "grad_norm": 2.046875,
      "learning_rate": 4.998080291543914e-05,
      "loss": 2.1205,
      "num_input_tokens_seen": 1046704,
      "step": 500
    },
    {
      "epoch": 0.08238844930255322,
      "grad_norm": 6.71875,
      "learning_rate": 4.9980416327401826e-05,
      "loss": 2.2885,
      "num_input_tokens_seen": 1056976,
      "step": 505
    },
    {
      "epoch": 0.08320417652337059,
      "grad_norm": 6.9375,
      "learning_rate": 4.998002588707038e-05,
      "loss": 3.1043,
      "num_input_tokens_seen": 1066640,
      "step": 510
    },
    {
      "epoch": 0.08401990374418794,
      "grad_norm": 3.046875,
      "learning_rate": 4.997963159450503e-05,
      "loss": 1.6898,
      "num_input_tokens_seen": 1077552,
      "step": 515
    },
    {
      "epoch": 0.08483563096500531,
      "grad_norm": 4.59375,
      "learning_rate": 4.9979233449766575e-05,
      "loss": 3.3036,
      "num_input_tokens_seen": 1087552,
      "step": 520
    },
    {
      "epoch": 0.08565135818582266,
      "grad_norm": 1.859375,
      "learning_rate": 4.997883145291641e-05,
      "loss": 1.3319,
      "num_input_tokens_seen": 1097904,
      "step": 525
    },
    {
      "epoch": 0.08646708540664003,
      "grad_norm": 8.0625,
      "learning_rate": 4.9978425604016536e-05,
      "loss": 1.5346,
      "num_input_tokens_seen": 1108416,
      "step": 530
    },
    {
      "epoch": 0.08728281262745738,
      "grad_norm": 12.5,
      "learning_rate": 4.9978015903129536e-05,
      "loss": 2.9244,
      "num_input_tokens_seen": 1119776,
      "step": 535
    },
    {
      "epoch": 0.08809853984827473,
      "grad_norm": 10.0,
      "learning_rate": 4.997760235031859e-05,
      "loss": 2.305,
      "num_input_tokens_seen": 1129920,
      "step": 540
    },
    {
      "epoch": 0.0889142670690921,
      "grad_norm": 9.9375,
      "learning_rate": 4.9977184945647473e-05,
      "loss": 2.0366,
      "num_input_tokens_seen": 1140944,
      "step": 545
    },
    {
      "epoch": 0.08972999428990945,
      "grad_norm": 9.0,
      "learning_rate": 4.997676368918055e-05,
      "loss": 2.2726,
      "num_input_tokens_seen": 1152800,
      "step": 550
    },
    {
      "epoch": 0.09054572151072682,
      "grad_norm": 10.6875,
      "learning_rate": 4.9976338580982794e-05,
      "loss": 3.5505,
      "num_input_tokens_seen": 1164656,
      "step": 555
    },
    {
      "epoch": 0.09136144873154417,
      "grad_norm": 6.1875,
      "learning_rate": 4.9975909621119755e-05,
      "loss": 1.8602,
      "num_input_tokens_seen": 1175328,
      "step": 560
    },
    {
      "epoch": 0.09217717595236154,
      "grad_norm": 5.21875,
      "learning_rate": 4.997547680965758e-05,
      "loss": 2.288,
      "num_input_tokens_seen": 1185264,
      "step": 565
    },
    {
      "epoch": 0.09299290317317889,
      "grad_norm": 6.6875,
      "learning_rate": 4.997504014666302e-05,
      "loss": 2.9191,
      "num_input_tokens_seen": 1194912,
      "step": 570
    },
    {
      "epoch": 0.09380863039399624,
      "grad_norm": 5.5,
      "learning_rate": 4.997459963220342e-05,
      "loss": 3.4976,
      "num_input_tokens_seen": 1205792,
      "step": 575
    },
    {
      "epoch": 0.09462435761481361,
      "grad_norm": 6.65625,
      "learning_rate": 4.997415526634671e-05,
      "loss": 2.7631,
      "num_input_tokens_seen": 1216352,
      "step": 580
    },
    {
      "epoch": 0.09544008483563096,
      "grad_norm": 5.65625,
      "learning_rate": 4.99737070491614e-05,
      "loss": 2.0976,
      "num_input_tokens_seen": 1225664,
      "step": 585
    },
    {
      "epoch": 0.09625581205644833,
      "grad_norm": 7.40625,
      "learning_rate": 4.997325498071663e-05,
      "loss": 2.7143,
      "num_input_tokens_seen": 1236912,
      "step": 590
    },
    {
      "epoch": 0.09707153927726568,
      "grad_norm": 5.90625,
      "learning_rate": 4.997279906108211e-05,
      "loss": 2.7869,
      "num_input_tokens_seen": 1248064,
      "step": 595
    },
    {
      "epoch": 0.09788726649808305,
      "grad_norm": 12.375,
      "learning_rate": 4.9972339290328155e-05,
      "loss": 3.1415,
      "num_input_tokens_seen": 1258320,
      "step": 600
    },
    {
      "epoch": 0.09788726649808305,
      "eval_loss": 2.5582432746887207,
      "eval_runtime": 134.6444,
      "eval_samples_per_second": 20.238,
      "eval_steps_per_second": 10.123,
      "num_input_tokens_seen": 1258320,
      "step": 600
    },
    {
      "epoch": 0.0987029937189004,
      "grad_norm": 7.0,
      "learning_rate": 4.9971875668525646e-05,
      "loss": 3.0459,
      "num_input_tokens_seen": 1268864,
      "step": 605
    },
    {
      "epoch": 0.09951872093971775,
      "grad_norm": 8.0625,
      "learning_rate": 4.997140819574609e-05,
      "loss": 2.2181,
      "num_input_tokens_seen": 1279616,
      "step": 610
    },
    {
      "epoch": 0.10033444816053512,
      "grad_norm": 6.21875,
      "learning_rate": 4.997093687206159e-05,
      "loss": 2.3303,
      "num_input_tokens_seen": 1290688,
      "step": 615
    },
    {
      "epoch": 0.10115017538135247,
      "grad_norm": 11.9375,
      "learning_rate": 4.997046169754482e-05,
      "loss": 2.1778,
      "num_input_tokens_seen": 1300656,
      "step": 620
    },
    {
      "epoch": 0.10196590260216984,
      "grad_norm": 1.359375,
      "learning_rate": 4.996998267226905e-05,
      "loss": 2.0096,
      "num_input_tokens_seen": 1311056,
      "step": 625
    },
    {
      "epoch": 0.10278162982298719,
      "grad_norm": 7.96875,
      "learning_rate": 4.996949979630817e-05,
      "loss": 3.1864,
      "num_input_tokens_seen": 1321024,
      "step": 630
    },
    {
      "epoch": 0.10359735704380456,
      "grad_norm": 7.625,
      "learning_rate": 4.996901306973663e-05,
      "loss": 3.224,
      "num_input_tokens_seen": 1330912,
      "step": 635
    },
    {
      "epoch": 0.10441308426462191,
      "grad_norm": 11.625,
      "learning_rate": 4.996852249262949e-05,
      "loss": 2.4455,
      "num_input_tokens_seen": 1339840,
      "step": 640
    },
    {
      "epoch": 0.10522881148543926,
      "grad_norm": 6.21875,
      "learning_rate": 4.996802806506241e-05,
      "loss": 1.9519,
      "num_input_tokens_seen": 1350208,
      "step": 645
    },
    {
      "epoch": 0.10604453870625663,
      "grad_norm": 0.490234375,
      "learning_rate": 4.996752978711164e-05,
      "loss": 1.7961,
      "num_input_tokens_seen": 1360720,
      "step": 650
    },
    {
      "epoch": 0.10686026592707398,
      "grad_norm": 4.15625,
      "learning_rate": 4.996702765885401e-05,
      "loss": 2.6737,
      "num_input_tokens_seen": 1371792,
      "step": 655
    },
    {
      "epoch": 0.10767599314789135,
      "grad_norm": 5.84375,
      "learning_rate": 4.9966521680366964e-05,
      "loss": 1.1267,
      "num_input_tokens_seen": 1381456,
      "step": 660
    },
    {
      "epoch": 0.1084917203687087,
      "grad_norm": 8.125,
      "learning_rate": 4.9966011851728524e-05,
      "loss": 2.1329,
      "num_input_tokens_seen": 1390240,
      "step": 665
    },
    {
      "epoch": 0.10930744758952607,
      "grad_norm": 6.0,
      "learning_rate": 4.996549817301731e-05,
      "loss": 2.0607,
      "num_input_tokens_seen": 1401488,
      "step": 670
    },
    {
      "epoch": 0.11012317481034342,
      "grad_norm": 4.625,
      "learning_rate": 4.9964980644312544e-05,
      "loss": 1.52,
      "num_input_tokens_seen": 1412208,
      "step": 675
    },
    {
      "epoch": 0.11093890203116077,
      "grad_norm": 6.90625,
      "learning_rate": 4.996445926569403e-05,
      "loss": 2.7516,
      "num_input_tokens_seen": 1423232,
      "step": 680
    },
    {
      "epoch": 0.11175462925197814,
      "grad_norm": 5.65625,
      "learning_rate": 4.996393403724218e-05,
      "loss": 2.8692,
      "num_input_tokens_seen": 1432896,
      "step": 685
    },
    {
      "epoch": 0.1125703564727955,
      "grad_norm": 5.28125,
      "learning_rate": 4.9963404959037985e-05,
      "loss": 2.9269,
      "num_input_tokens_seen": 1443216,
      "step": 690
    },
    {
      "epoch": 0.11338608369361286,
      "grad_norm": 11.125,
      "learning_rate": 4.996287203116303e-05,
      "loss": 2.5973,
      "num_input_tokens_seen": 1454352,
      "step": 695
    },
    {
      "epoch": 0.11420181091443021,
      "grad_norm": 7.65625,
      "learning_rate": 4.996233525369951e-05,
      "loss": 1.3081,
      "num_input_tokens_seen": 1464448,
      "step": 700
    },
    {
      "epoch": 0.11501753813524758,
      "grad_norm": 7.625,
      "learning_rate": 4.99617946267302e-05,
      "loss": 1.955,
      "num_input_tokens_seen": 1474704,
      "step": 705
    },
    {
      "epoch": 0.11583326535606493,
      "grad_norm": 7.28125,
      "learning_rate": 4.996125015033846e-05,
      "loss": 1.7467,
      "num_input_tokens_seen": 1486400,
      "step": 710
    },
    {
      "epoch": 0.11664899257688228,
      "grad_norm": 5.71875,
      "learning_rate": 4.996070182460827e-05,
      "loss": 1.7782,
      "num_input_tokens_seen": 1497488,
      "step": 715
    },
    {
      "epoch": 0.11746471979769965,
      "grad_norm": 10.875,
      "learning_rate": 4.996014964962418e-05,
      "loss": 2.5251,
      "num_input_tokens_seen": 1506928,
      "step": 720
    },
    {
      "epoch": 0.118280447018517,
      "grad_norm": 12.875,
      "learning_rate": 4.9959593625471344e-05,
      "loss": 3.5792,
      "num_input_tokens_seen": 1517280,
      "step": 725
    },
    {
      "epoch": 0.11909617423933437,
      "grad_norm": 7.28125,
      "learning_rate": 4.995903375223552e-05,
      "loss": 0.9184,
      "num_input_tokens_seen": 1526864,
      "step": 730
    },
    {
      "epoch": 0.11991190146015172,
      "grad_norm": 12.375,
      "learning_rate": 4.995847003000302e-05,
      "loss": 2.668,
      "num_input_tokens_seen": 1537712,
      "step": 735
    },
    {
      "epoch": 0.12072762868096909,
      "grad_norm": 4.4375,
      "learning_rate": 4.9957902458860804e-05,
      "loss": 1.4034,
      "num_input_tokens_seen": 1548752,
      "step": 740
    },
    {
      "epoch": 0.12154335590178644,
      "grad_norm": 5.9375,
      "learning_rate": 4.995733103889639e-05,
      "loss": 2.2332,
      "num_input_tokens_seen": 1560400,
      "step": 745
    },
    {
      "epoch": 0.1223590831226038,
      "grad_norm": 6.15625,
      "learning_rate": 4.99567557701979e-05,
      "loss": 1.6152,
      "num_input_tokens_seen": 1571696,
      "step": 750
    },
    {
      "epoch": 0.12317481034342116,
      "grad_norm": 3.671875,
      "learning_rate": 4.995617665285403e-05,
      "loss": 1.0118,
      "num_input_tokens_seen": 1582720,
      "step": 755
    },
    {
      "epoch": 0.12399053756423851,
      "grad_norm": 11.25,
      "learning_rate": 4.99555936869541e-05,
      "loss": 3.0687,
      "num_input_tokens_seen": 1593936,
      "step": 760
    },
    {
      "epoch": 0.12480626478505588,
      "grad_norm": 14.1875,
      "learning_rate": 4.995500687258803e-05,
      "loss": 2.5301,
      "num_input_tokens_seen": 1604384,
      "step": 765
    },
    {
      "epoch": 0.12562199200587323,
      "grad_norm": 4.4375,
      "learning_rate": 4.995441620984628e-05,
      "loss": 1.1851,
      "num_input_tokens_seen": 1614704,
      "step": 770
    },
    {
      "epoch": 0.1264377192266906,
      "grad_norm": 5.21875,
      "learning_rate": 4.995382169881996e-05,
      "loss": 3.8308,
      "num_input_tokens_seen": 1623680,
      "step": 775
    },
    {
      "epoch": 0.12725344644750794,
      "grad_norm": 11.375,
      "learning_rate": 4.9953223339600755e-05,
      "loss": 2.9726,
      "num_input_tokens_seen": 1633664,
      "step": 780
    },
    {
      "epoch": 0.12806917366832532,
      "grad_norm": 9.125,
      "learning_rate": 4.995262113228091e-05,
      "loss": 2.441,
      "num_input_tokens_seen": 1643552,
      "step": 785
    },
    {
      "epoch": 0.12888490088914267,
      "grad_norm": 5.6875,
      "learning_rate": 4.995201507695332e-05,
      "loss": 1.8785,
      "num_input_tokens_seen": 1654544,
      "step": 790
    },
    {
      "epoch": 0.12970062810996003,
      "grad_norm": 5.125,
      "learning_rate": 4.995140517371144e-05,
      "loss": 2.0845,
      "num_input_tokens_seen": 1663728,
      "step": 795
    },
    {
      "epoch": 0.13051635533077738,
      "grad_norm": 3.875,
      "learning_rate": 4.995079142264932e-05,
      "loss": 2.0399,
      "num_input_tokens_seen": 1673984,
      "step": 800
    },
    {
      "epoch": 0.13051635533077738,
      "eval_loss": 2.554769277572632,
      "eval_runtime": 134.7236,
      "eval_samples_per_second": 20.227,
      "eval_steps_per_second": 10.117,
      "num_input_tokens_seen": 1673984,
      "step": 800
    },
    {
      "epoch": 0.13133208255159476,
      "grad_norm": 3.96875,
      "learning_rate": 4.995017382386162e-05,
      "loss": 2.7465,
      "num_input_tokens_seen": 1684640,
      "step": 805
    },
    {
      "epoch": 0.1321478097724121,
      "grad_norm": 10.8125,
      "learning_rate": 4.994955237744356e-05,
      "loss": 2.7043,
      "num_input_tokens_seen": 1694576,
      "step": 810
    },
    {
      "epoch": 0.13296353699322946,
      "grad_norm": 6.46875,
      "learning_rate": 4.994892708349101e-05,
      "loss": 3.1384,
      "num_input_tokens_seen": 1705856,
      "step": 815
    },
    {
      "epoch": 0.13377926421404682,
      "grad_norm": 10.8125,
      "learning_rate": 4.994829794210035e-05,
      "loss": 1.7652,
      "num_input_tokens_seen": 1716432,
      "step": 820
    },
    {
      "epoch": 0.13459499143486417,
      "grad_norm": 8.4375,
      "learning_rate": 4.994766495336864e-05,
      "loss": 2.9388,
      "num_input_tokens_seen": 1725936,
      "step": 825
    },
    {
      "epoch": 0.13541071865568155,
      "grad_norm": 8.5,
      "learning_rate": 4.994702811739348e-05,
      "loss": 1.2269,
      "num_input_tokens_seen": 1736256,
      "step": 830
    },
    {
      "epoch": 0.1362264458764989,
      "grad_norm": 8.3125,
      "learning_rate": 4.994638743427308e-05,
      "loss": 2.9047,
      "num_input_tokens_seen": 1746512,
      "step": 835
    },
    {
      "epoch": 0.13704217309731626,
      "grad_norm": 11.0,
      "learning_rate": 4.994574290410624e-05,
      "loss": 2.9808,
      "num_input_tokens_seen": 1756448,
      "step": 840
    },
    {
      "epoch": 0.1378579003181336,
      "grad_norm": 9.9375,
      "learning_rate": 4.9945094526992364e-05,
      "loss": 3.9451,
      "num_input_tokens_seen": 1766528,
      "step": 845
    },
    {
      "epoch": 0.13867362753895096,
      "grad_norm": 11.5625,
      "learning_rate": 4.994444230303142e-05,
      "loss": 1.4968,
      "num_input_tokens_seen": 1778784,
      "step": 850
    },
    {
      "epoch": 0.13948935475976834,
      "grad_norm": 3.8125,
      "learning_rate": 4.994378623232402e-05,
      "loss": 2.9597,
      "num_input_tokens_seen": 1790016,
      "step": 855
    },
    {
      "epoch": 0.1403050819805857,
      "grad_norm": 5.40625,
      "learning_rate": 4.99431263149713e-05,
      "loss": 1.8258,
      "num_input_tokens_seen": 1800608,
      "step": 860
    },
    {
      "epoch": 0.14112080920140305,
      "grad_norm": 5.9375,
      "learning_rate": 4.9942462551075056e-05,
      "loss": 2.5323,
      "num_input_tokens_seen": 1810608,
      "step": 865
    },
    {
      "epoch": 0.1419365364222204,
      "grad_norm": 2.34375,
      "learning_rate": 4.994179494073764e-05,
      "loss": 2.4259,
      "num_input_tokens_seen": 1822896,
      "step": 870
    },
    {
      "epoch": 0.14275226364303778,
      "grad_norm": 12.5625,
      "learning_rate": 4.9941123484062e-05,
      "loss": 3.2869,
      "num_input_tokens_seen": 1835712,
      "step": 875
    },
    {
      "epoch": 0.14356799086385513,
      "grad_norm": 5.875,
      "learning_rate": 4.99404481811517e-05,
      "loss": 2.3777,
      "num_input_tokens_seen": 1845856,
      "step": 880
    },
    {
      "epoch": 0.14438371808467249,
      "grad_norm": 5.71875,
      "learning_rate": 4.9939769032110864e-05,
      "loss": 2.6378,
      "num_input_tokens_seen": 1855360,
      "step": 885
    },
    {
      "epoch": 0.14519944530548984,
      "grad_norm": 12.0625,
      "learning_rate": 4.993908603704423e-05,
      "loss": 3.7733,
      "num_input_tokens_seen": 1865264,
      "step": 890
    },
    {
      "epoch": 0.1460151725263072,
      "grad_norm": 4.875,
      "learning_rate": 4.9938399196057126e-05,
      "loss": 2.6998,
      "num_input_tokens_seen": 1874608,
      "step": 895
    },
    {
      "epoch": 0.14683089974712457,
      "grad_norm": 7.78125,
      "learning_rate": 4.993770850925547e-05,
      "loss": 2.7642,
      "num_input_tokens_seen": 1884736,
      "step": 900
    },
    {
      "epoch": 0.14764662696794192,
      "grad_norm": 3.25,
      "learning_rate": 4.993701397674577e-05,
      "loss": 1.2244,
      "num_input_tokens_seen": 1896112,
      "step": 905
    },
    {
      "epoch": 0.14846235418875928,
      "grad_norm": 3.4375,
      "learning_rate": 4.993631559863515e-05,
      "loss": 2.2194,
      "num_input_tokens_seen": 1908128,
      "step": 910
    },
    {
      "epoch": 0.14927808140957663,
      "grad_norm": 7.5625,
      "learning_rate": 4.9935613375031283e-05,
      "loss": 1.7767,
      "num_input_tokens_seen": 1918528,
      "step": 915
    },
    {
      "epoch": 0.150093808630394,
      "grad_norm": 8.6875,
      "learning_rate": 4.993490730604248e-05,
      "loss": 2.8698,
      "num_input_tokens_seen": 1928496,
      "step": 920
    },
    {
      "epoch": 0.15090953585121136,
      "grad_norm": 7.4375,
      "learning_rate": 4.993419739177761e-05,
      "loss": 3.9938,
      "num_input_tokens_seen": 1939504,
      "step": 925
    },
    {
      "epoch": 0.15172526307202872,
      "grad_norm": 6.90625,
      "learning_rate": 4.9933483632346164e-05,
      "loss": 0.9385,
      "num_input_tokens_seen": 1950464,
      "step": 930
    },
    {
      "epoch": 0.15254099029284607,
      "grad_norm": 7.125,
      "learning_rate": 4.993276602785821e-05,
      "loss": 2.1691,
      "num_input_tokens_seen": 1962032,
      "step": 935
    },
    {
      "epoch": 0.15335671751366342,
      "grad_norm": 2.421875,
      "learning_rate": 4.993204457842441e-05,
      "loss": 1.9701,
      "num_input_tokens_seen": 1972272,
      "step": 940
    },
    {
      "epoch": 0.1541724447344808,
      "grad_norm": 5.59375,
      "learning_rate": 4.993131928415602e-05,
      "loss": 1.7892,
      "num_input_tokens_seen": 1982240,
      "step": 945
    },
    {
      "epoch": 0.15498817195529815,
      "grad_norm": 6.125,
      "learning_rate": 4.993059014516489e-05,
      "loss": 1.6486,
      "num_input_tokens_seen": 1993472,
      "step": 950
    },
    {
      "epoch": 0.1558038991761155,
      "grad_norm": 7.34375,
      "learning_rate": 4.9929857161563464e-05,
      "loss": 2.4397,
      "num_input_tokens_seen": 2003360,
      "step": 955
    },
    {
      "epoch": 0.15661962639693286,
      "grad_norm": 7.25,
      "learning_rate": 4.992912033346477e-05,
      "loss": 2.1575,
      "num_input_tokens_seen": 2015504,
      "step": 960
    },
    {
      "epoch": 0.1574353536177502,
      "grad_norm": 5.40625,
      "learning_rate": 4.992837966098245e-05,
      "loss": 1.8883,
      "num_input_tokens_seen": 2025280,
      "step": 965
    },
    {
      "epoch": 0.1582510808385676,
      "grad_norm": 5.90625,
      "learning_rate": 4.992763514423071e-05,
      "loss": 2.3832,
      "num_input_tokens_seen": 2035136,
      "step": 970
    },
    {
      "epoch": 0.15906680805938495,
      "grad_norm": 5.125,
      "learning_rate": 4.992688678332437e-05,
      "loss": 2.0456,
      "num_input_tokens_seen": 2045664,
      "step": 975
    },
    {
      "epoch": 0.1598825352802023,
      "grad_norm": 6.8125,
      "learning_rate": 4.992613457837884e-05,
      "loss": 1.7361,
      "num_input_tokens_seen": 2056400,
      "step": 980
    },
    {
      "epoch": 0.16069826250101965,
      "grad_norm": 4.96875,
      "learning_rate": 4.992537852951011e-05,
      "loss": 1.2248,
      "num_input_tokens_seen": 2066560,
      "step": 985
    },
    {
      "epoch": 0.16151398972183703,
      "grad_norm": 2.8125,
      "learning_rate": 4.9924618636834785e-05,
      "loss": 2.0636,
      "num_input_tokens_seen": 2077728,
      "step": 990
    },
    {
      "epoch": 0.16232971694265438,
      "grad_norm": 9.125,
      "learning_rate": 4.9923854900470046e-05,
      "loss": 1.8059,
      "num_input_tokens_seen": 2087728,
      "step": 995
    },
    {
      "epoch": 0.16314544416347174,
      "grad_norm": 0.234375,
      "learning_rate": 4.992308732053367e-05,
      "loss": 1.762,
      "num_input_tokens_seen": 2097344,
      "step": 1000
    },
    {
      "epoch": 0.16314544416347174,
      "eval_loss": 2.5650367736816406,
      "eval_runtime": 134.7133,
      "eval_samples_per_second": 20.228,
      "eval_steps_per_second": 10.118,
      "num_input_tokens_seen": 2097344,
      "step": 1000
    },
    {
      "epoch": 0.1639611713842891,
      "grad_norm": 6.6875,
      "learning_rate": 4.992231589714402e-05,
      "loss": 1.7563,
      "num_input_tokens_seen": 2108880,
      "step": 1005
    },
    {
      "epoch": 0.16477689860510644,
      "grad_norm": 3.984375,
      "learning_rate": 4.992154063042007e-05,
      "loss": 1.0112,
      "num_input_tokens_seen": 2117952,
      "step": 1010
    },
    {
      "epoch": 0.16559262582592382,
      "grad_norm": 8.5625,
      "learning_rate": 4.992076152048136e-05,
      "loss": 1.9582,
      "num_input_tokens_seen": 2129728,
      "step": 1015
    },
    {
      "epoch": 0.16640835304674118,
      "grad_norm": 5.375,
      "learning_rate": 4.991997856744807e-05,
      "loss": 2.4562,
      "num_input_tokens_seen": 2140256,
      "step": 1020
    },
    {
      "epoch": 0.16722408026755853,
      "grad_norm": 5.34375,
      "learning_rate": 4.9919191771440905e-05,
      "loss": 2.206,
      "num_input_tokens_seen": 2151088,
      "step": 1025
    },
    {
      "epoch": 0.16803980748837588,
      "grad_norm": 0.08056640625,
      "learning_rate": 4.991840113258122e-05,
      "loss": 1.0637,
      "num_input_tokens_seen": 2161184,
      "step": 1030
    },
    {
      "epoch": 0.16885553470919323,
      "grad_norm": 3.03125,
      "learning_rate": 4.9917606650990933e-05,
      "loss": 2.1335,
      "num_input_tokens_seen": 2171520,
      "step": 1035
    },
    {
      "epoch": 0.16967126193001061,
      "grad_norm": 7.09375,
      "learning_rate": 4.9916808326792566e-05,
      "loss": 2.1058,
      "num_input_tokens_seen": 2181968,
      "step": 1040
    },
    {
      "epoch": 0.17048698915082797,
      "grad_norm": 1.5,
      "learning_rate": 4.9916006160109235e-05,
      "loss": 2.2951,
      "num_input_tokens_seen": 2192800,
      "step": 1045
    },
    {
      "epoch": 0.17130271637164532,
      "grad_norm": 5.3125,
      "learning_rate": 4.991520015106464e-05,
      "loss": 2.0303,
      "num_input_tokens_seen": 2203744,
      "step": 1050
    },
    {
      "epoch": 0.17211844359246267,
      "grad_norm": 3.140625,
      "learning_rate": 4.991439029978308e-05,
      "loss": 1.7062,
      "num_input_tokens_seen": 2212240,
      "step": 1055
    },
    {
      "epoch": 0.17293417081328005,
      "grad_norm": 6.125,
      "learning_rate": 4.9913576606389434e-05,
      "loss": 2.2464,
      "num_input_tokens_seen": 2223072,
      "step": 1060
    },
    {
      "epoch": 0.1737498980340974,
      "grad_norm": 3.859375,
      "learning_rate": 4.991275907100919e-05,
      "loss": 3.3074,
      "num_input_tokens_seen": 2233664,
      "step": 1065
    },
    {
      "epoch": 0.17456562525491476,
      "grad_norm": 5.5625,
      "learning_rate": 4.9911937693768434e-05,
      "loss": 3.6805,
      "num_input_tokens_seen": 2243664,
      "step": 1070
    },
    {
      "epoch": 0.1753813524757321,
      "grad_norm": 8.3125,
      "learning_rate": 4.991111247479382e-05,
      "loss": 1.9175,
      "num_input_tokens_seen": 2254288,
      "step": 1075
    },
    {
      "epoch": 0.17619707969654946,
      "grad_norm": 10.625,
      "learning_rate": 4.9910283414212605e-05,
      "loss": 2.58,
      "num_input_tokens_seen": 2264576,
      "step": 1080
    },
    {
      "epoch": 0.17701280691736684,
      "grad_norm": 2.234375,
      "learning_rate": 4.990945051215265e-05,
      "loss": 1.947,
      "num_input_tokens_seen": 2274832,
      "step": 1085
    },
    {
      "epoch": 0.1778285341381842,
      "grad_norm": 2.125,
      "learning_rate": 4.99086137687424e-05,
      "loss": 1.5949,
      "num_input_tokens_seen": 2285200,
      "step": 1090
    },
    {
      "epoch": 0.17864426135900155,
      "grad_norm": 4.1875,
      "learning_rate": 4.9907773184110874e-05,
      "loss": 1.5938,
      "num_input_tokens_seen": 2298528,
      "step": 1095
    },
    {
      "epoch": 0.1794599885798189,
      "grad_norm": 7.03125,
      "learning_rate": 4.9906928758387715e-05,
      "loss": 3.2581,
      "num_input_tokens_seen": 2309008,
      "step": 1100
    },
    {
      "epoch": 0.18027571580063625,
      "grad_norm": 4.46875,
      "learning_rate": 4.9906080491703146e-05,
      "loss": 2.0334,
      "num_input_tokens_seen": 2319232,
      "step": 1105
    },
    {
      "epoch": 0.18109144302145364,
      "grad_norm": 5.125,
      "learning_rate": 4.990522838418797e-05,
      "loss": 2.0845,
      "num_input_tokens_seen": 2329920,
      "step": 1110
    },
    {
      "epoch": 0.181907170242271,
      "grad_norm": 7.5,
      "learning_rate": 4.9904372435973604e-05,
      "loss": 2.0352,
      "num_input_tokens_seen": 2341984,
      "step": 1115
    },
    {
      "epoch": 0.18272289746308834,
      "grad_norm": 9.6875,
      "learning_rate": 4.990351264719203e-05,
      "loss": 2.1311,
      "num_input_tokens_seen": 2352416,
      "step": 1120
    },
    {
      "epoch": 0.1835386246839057,
      "grad_norm": 18.75,
      "learning_rate": 4.990264901797586e-05,
      "loss": 2.2372,
      "num_input_tokens_seen": 2362768,
      "step": 1125
    },
    {
      "epoch": 0.18435435190472307,
      "grad_norm": 5.625,
      "learning_rate": 4.990178154845826e-05,
      "loss": 1.6287,
      "num_input_tokens_seen": 2373312,
      "step": 1130
    },
    {
      "epoch": 0.18517007912554043,
      "grad_norm": 9.75,
      "learning_rate": 4.9900910238773014e-05,
      "loss": 2.6478,
      "num_input_tokens_seen": 2382832,
      "step": 1135
    },
    {
      "epoch": 0.18598580634635778,
      "grad_norm": 8.6875,
      "learning_rate": 4.990003508905448e-05,
      "loss": 1.573,
      "num_input_tokens_seen": 2392368,
      "step": 1140
    },
    {
      "epoch": 0.18680153356717513,
      "grad_norm": 5.15625,
      "learning_rate": 4.989915609943763e-05,
      "loss": 2.6202,
      "num_input_tokens_seen": 2404096,
      "step": 1145
    },
    {
      "epoch": 0.18761726078799248,
      "grad_norm": 13.0625,
      "learning_rate": 4.9898273270058e-05,
      "loss": 2.8423,
      "num_input_tokens_seen": 2412688,
      "step": 1150
    },
    {
      "epoch": 0.18843298800880987,
      "grad_norm": 5.46875,
      "learning_rate": 4.989738660105174e-05,
      "loss": 2.633,
      "num_input_tokens_seen": 2422624,
      "step": 1155
    },
    {
      "epoch": 0.18924871522962722,
      "grad_norm": 14.375,
      "learning_rate": 4.989649609255559e-05,
      "loss": 3.3181,
      "num_input_tokens_seen": 2432320,
      "step": 1160
    },
    {
      "epoch": 0.19006444245044457,
      "grad_norm": 6.75,
      "learning_rate": 4.989560174470687e-05,
      "loss": 2.152,
      "num_input_tokens_seen": 2443248,
      "step": 1165
    },
    {
      "epoch": 0.19088016967126192,
      "grad_norm": 8.625,
      "learning_rate": 4.989470355764351e-05,
      "loss": 2.6961,
      "num_input_tokens_seen": 2454384,
      "step": 1170
    },
    {
      "epoch": 0.19169589689207928,
      "grad_norm": 13.3125,
      "learning_rate": 4.9893801531504e-05,
      "loss": 2.5541,
      "num_input_tokens_seen": 2465456,
      "step": 1175
    },
    {
      "epoch": 0.19251162411289666,
      "grad_norm": 1.7109375,
      "learning_rate": 4.9892895666427475e-05,
      "loss": 2.472,
      "num_input_tokens_seen": 2476256,
      "step": 1180
    },
    {
      "epoch": 0.193327351333714,
      "grad_norm": 8.6875,
      "learning_rate": 4.9891985962553606e-05,
      "loss": 2.7221,
      "num_input_tokens_seen": 2487296,
      "step": 1185
    },
    {
      "epoch": 0.19414307855453136,
      "grad_norm": 4.75,
      "learning_rate": 4.989107242002269e-05,
      "loss": 3.0473,
      "num_input_tokens_seen": 2496448,
      "step": 1190
    },
    {
      "epoch": 0.19495880577534871,
      "grad_norm": 7.5625,
      "learning_rate": 4.989015503897561e-05,
      "loss": 1.4665,
      "num_input_tokens_seen": 2505856,
      "step": 1195
    },
    {
      "epoch": 0.1957745329961661,
      "grad_norm": 5.0,
      "learning_rate": 4.988923381955383e-05,
      "loss": 2.7923,
      "num_input_tokens_seen": 2515056,
      "step": 1200
    },
    {
      "epoch": 0.1957745329961661,
      "eval_loss": 2.548196315765381,
      "eval_runtime": 134.6665,
      "eval_samples_per_second": 20.235,
      "eval_steps_per_second": 10.121,
      "num_input_tokens_seen": 2515056,
      "step": 1200
    },
    {
      "epoch": 0.19659026021698345,
      "grad_norm": 9.5,
      "learning_rate": 4.988830876189942e-05,
      "loss": 3.3226,
      "num_input_tokens_seen": 2525712,
      "step": 1205
    },
    {
      "epoch": 0.1974059874378008,
      "grad_norm": 7.6875,
      "learning_rate": 4.988737986615503e-05,
      "loss": 1.9326,
      "num_input_tokens_seen": 2535872,
      "step": 1210
    },
    {
      "epoch": 0.19822171465861815,
      "grad_norm": 6.78125,
      "learning_rate": 4.988644713246391e-05,
      "loss": 2.5958,
      "num_input_tokens_seen": 2547248,
      "step": 1215
    },
    {
      "epoch": 0.1990374418794355,
      "grad_norm": 8.5625,
      "learning_rate": 4.988551056096991e-05,
      "loss": 2.4174,
      "num_input_tokens_seen": 2558224,
      "step": 1220
    },
    {
      "epoch": 0.1998531691002529,
      "grad_norm": 6.4375,
      "learning_rate": 4.988457015181743e-05,
      "loss": 3.4297,
      "num_input_tokens_seen": 2568192,
      "step": 1225
    },
    {
      "epoch": 0.20066889632107024,
      "grad_norm": 7.0625,
      "learning_rate": 4.988362590515153e-05,
      "loss": 1.9904,
      "num_input_tokens_seen": 2580112,
      "step": 1230
    },
    {
      "epoch": 0.2014846235418876,
      "grad_norm": 8.25,
      "learning_rate": 4.9882677821117805e-05,
      "loss": 3.4085,
      "num_input_tokens_seen": 2591808,
      "step": 1235
    },
    {
      "epoch": 0.20230035076270494,
      "grad_norm": 7.25,
      "learning_rate": 4.988172589986246e-05,
      "loss": 3.7223,
      "num_input_tokens_seen": 2602336,
      "step": 1240
    },
    {
      "epoch": 0.2031160779835223,
      "grad_norm": 4.40625,
      "learning_rate": 4.9880770141532304e-05,
      "loss": 2.3513,
      "num_input_tokens_seen": 2612096,
      "step": 1245
    },
    {
      "epoch": 0.20393180520433968,
      "grad_norm": 11.5,
      "learning_rate": 4.987981054627472e-05,
      "loss": 2.9924,
      "num_input_tokens_seen": 2622736,
      "step": 1250
    },
    {
      "epoch": 0.20474753242515703,
      "grad_norm": 4.78125,
      "learning_rate": 4.987884711423769e-05,
      "loss": 1.6815,
      "num_input_tokens_seen": 2632320,
      "step": 1255
    },
    {
      "epoch": 0.20556325964597438,
      "grad_norm": 1.859375,
      "learning_rate": 4.9877879845569784e-05,
      "loss": 1.3296,
      "num_input_tokens_seen": 2644160,
      "step": 1260
    },
    {
      "epoch": 0.20637898686679174,
      "grad_norm": 4.0,
      "learning_rate": 4.9876908740420175e-05,
      "loss": 1.6824,
      "num_input_tokens_seen": 2655024,
      "step": 1265
    },
    {
      "epoch": 0.20719471408760912,
      "grad_norm": 5.375,
      "learning_rate": 4.987593379893861e-05,
      "loss": 2.9393,
      "num_input_tokens_seen": 2664848,
      "step": 1270
    },
    {
      "epoch": 0.20801044130842647,
      "grad_norm": 8.5625,
      "learning_rate": 4.987495502127545e-05,
      "loss": 3.3941,
      "num_input_tokens_seen": 2675680,
      "step": 1275
    },
    {
      "epoch": 0.20882616852924382,
      "grad_norm": 6.6875,
      "learning_rate": 4.987397240758162e-05,
      "loss": 1.4506,
      "num_input_tokens_seen": 2686384,
      "step": 1280
    },
    {
      "epoch": 0.20964189575006117,
      "grad_norm": 0.384765625,
      "learning_rate": 4.9872985958008664e-05,
      "loss": 1.4825,
      "num_input_tokens_seen": 2696576,
      "step": 1285
    },
    {
      "epoch": 0.21045762297087853,
      "grad_norm": 4.53125,
      "learning_rate": 4.987199567270871e-05,
      "loss": 1.9586,
      "num_input_tokens_seen": 2707376,
      "step": 1290
    },
    {
      "epoch": 0.2112733501916959,
      "grad_norm": 10.125,
      "learning_rate": 4.9871001551834444e-05,
      "loss": 2.1887,
      "num_input_tokens_seen": 2718768,
      "step": 1295
    },
    {
      "epoch": 0.21208907741251326,
      "grad_norm": 13.9375,
      "learning_rate": 4.98700035955392e-05,
      "loss": 2.6108,
      "num_input_tokens_seen": 2727728,
      "step": 1300
    },
    {
      "epoch": 0.2129048046333306,
      "grad_norm": 10.0625,
      "learning_rate": 4.986900180397686e-05,
      "loss": 3.8105,
      "num_input_tokens_seen": 2739264,
      "step": 1305
    },
    {
      "epoch": 0.21372053185414797,
      "grad_norm": 3.0625,
      "learning_rate": 4.9867996177301926e-05,
      "loss": 1.6473,
      "num_input_tokens_seen": 2749920,
      "step": 1310
    },
    {
      "epoch": 0.21453625907496532,
      "grad_norm": 7.5625,
      "learning_rate": 4.9866986715669464e-05,
      "loss": 2.9277,
      "num_input_tokens_seen": 2759440,
      "step": 1315
    },
    {
      "epoch": 0.2153519862957827,
      "grad_norm": 11.25,
      "learning_rate": 4.9865973419235155e-05,
      "loss": 3.018,
      "num_input_tokens_seen": 2770720,
      "step": 1320
    },
    {
      "epoch": 0.21616771351660005,
      "grad_norm": 9.9375,
      "learning_rate": 4.986495628815526e-05,
      "loss": 2.8937,
      "num_input_tokens_seen": 2781760,
      "step": 1325
    },
    {
      "epoch": 0.2169834407374174,
      "grad_norm": 15.75,
      "learning_rate": 4.986393532258663e-05,
      "loss": 2.7528,
      "num_input_tokens_seen": 2791872,
      "step": 1330
    },
    {
      "epoch": 0.21779916795823476,
      "grad_norm": 5.09375,
      "learning_rate": 4.986291052268671e-05,
      "loss": 1.5995,
      "num_input_tokens_seen": 2803104,
      "step": 1335
    },
    {
      "epoch": 0.21861489517905214,
      "grad_norm": 11.3125,
      "learning_rate": 4.986188188861355e-05,
      "loss": 2.1062,
      "num_input_tokens_seen": 2813296,
      "step": 1340
    },
    {
      "epoch": 0.2194306223998695,
      "grad_norm": 5.03125,
      "learning_rate": 4.9860849420525766e-05,
      "loss": 2.8459,
      "num_input_tokens_seen": 2824960,
      "step": 1345
    },
    {
      "epoch": 0.22024634962068684,
      "grad_norm": 4.625,
      "learning_rate": 4.9859813118582575e-05,
      "loss": 2.1492,
      "num_input_tokens_seen": 2835984,
      "step": 1350
    },
    {
      "epoch": 0.2210620768415042,
      "grad_norm": 5.40625,
      "learning_rate": 4.98587729829438e-05,
      "loss": 2.705,
      "num_input_tokens_seen": 2845744,
      "step": 1355
    },
    {
      "epoch": 0.22187780406232155,
      "grad_norm": 10.75,
      "learning_rate": 4.985772901376983e-05,
      "loss": 2.565,
      "num_input_tokens_seen": 2857344,
      "step": 1360
    },
    {
      "epoch": 0.22269353128313893,
      "grad_norm": 6.0625,
      "learning_rate": 4.9856681211221666e-05,
      "loss": 1.6996,
      "num_input_tokens_seen": 2867776,
      "step": 1365
    },
    {
      "epoch": 0.22350925850395628,
      "grad_norm": 9.1875,
      "learning_rate": 4.985562957546089e-05,
      "loss": 2.4041,
      "num_input_tokens_seen": 2878960,
      "step": 1370
    },
    {
      "epoch": 0.22432498572477363,
      "grad_norm": 5.40625,
      "learning_rate": 4.9854574106649686e-05,
      "loss": 1.7318,
      "num_input_tokens_seen": 2891024,
      "step": 1375
    },
    {
      "epoch": 0.225140712945591,
      "grad_norm": 7.5625,
      "learning_rate": 4.985351480495081e-05,
      "loss": 3.1137,
      "num_input_tokens_seen": 2901456,
      "step": 1380
    },
    {
      "epoch": 0.22595644016640834,
      "grad_norm": 8.375,
      "learning_rate": 4.985245167052762e-05,
      "loss": 2.8767,
      "num_input_tokens_seen": 2912016,
      "step": 1385
    },
    {
      "epoch": 0.22677216738722572,
      "grad_norm": 9.0,
      "learning_rate": 4.9851384703544066e-05,
      "loss": 3.1745,
      "num_input_tokens_seen": 2922192,
      "step": 1390
    },
    {
      "epoch": 0.22758789460804307,
      "grad_norm": 6.25,
      "learning_rate": 4.985031390416469e-05,
      "loss": 2.8386,
      "num_input_tokens_seen": 2933488,
      "step": 1395
    },
    {
      "epoch": 0.22840362182886043,
      "grad_norm": 13.8125,
      "learning_rate": 4.984923927255461e-05,
      "loss": 1.7457,
      "num_input_tokens_seen": 2943280,
      "step": 1400
    },
    {
      "epoch": 0.22840362182886043,
      "eval_loss": 2.5521657466888428,
      "eval_runtime": 134.7023,
      "eval_samples_per_second": 20.23,
      "eval_steps_per_second": 10.119,
      "num_input_tokens_seen": 2943280,
      "step": 1400
    },
    {
      "epoch": 0.22921934904967778,
      "grad_norm": 6.34375,
      "learning_rate": 4.984816080887958e-05,
      "loss": 2.0731,
      "num_input_tokens_seen": 2955792,
      "step": 1405
    },
    {
      "epoch": 0.23003507627049516,
      "grad_norm": 9.0625,
      "learning_rate": 4.9847078513305875e-05,
      "loss": 2.6628,
      "num_input_tokens_seen": 2965968,
      "step": 1410
    },
    {
      "epoch": 0.2308508034913125,
      "grad_norm": 9.0,
      "learning_rate": 4.984599238600043e-05,
      "loss": 2.7593,
      "num_input_tokens_seen": 2975968,
      "step": 1415
    },
    {
      "epoch": 0.23166653071212986,
      "grad_norm": 5.15625,
      "learning_rate": 4.9844902427130716e-05,
      "loss": 1.1594,
      "num_input_tokens_seen": 2986016,
      "step": 1420
    },
    {
      "epoch": 0.23248225793294722,
      "grad_norm": 3.59375,
      "learning_rate": 4.984380863686482e-05,
      "loss": 3.4199,
      "num_input_tokens_seen": 2994768,
      "step": 1425
    },
    {
      "epoch": 0.23329798515376457,
      "grad_norm": 11.6875,
      "learning_rate": 4.984271101537143e-05,
      "loss": 1.5492,
      "num_input_tokens_seen": 3003264,
      "step": 1430
    },
    {
      "epoch": 0.23411371237458195,
      "grad_norm": 9.0625,
      "learning_rate": 4.9841609562819816e-05,
      "loss": 4.1386,
      "num_input_tokens_seen": 3013904,
      "step": 1435
    },
    {
      "epoch": 0.2349294395953993,
      "grad_norm": 3.625,
      "learning_rate": 4.984050427937983e-05,
      "loss": 1.155,
      "num_input_tokens_seen": 3024432,
      "step": 1440
    },
    {
      "epoch": 0.23574516681621666,
      "grad_norm": 11.0,
      "learning_rate": 4.983939516522191e-05,
      "loss": 1.7612,
      "num_input_tokens_seen": 3034624,
      "step": 1445
    },
    {
      "epoch": 0.236560894037034,
      "grad_norm": 3.875,
      "learning_rate": 4.983828222051711e-05,
      "loss": 1.9255,
      "num_input_tokens_seen": 3044288,
      "step": 1450
    },
    {
      "epoch": 0.23737662125785136,
      "grad_norm": 13.0,
      "learning_rate": 4.983716544543705e-05,
      "loss": 2.3639,
      "num_input_tokens_seen": 3054384,
      "step": 1455
    },
    {
      "epoch": 0.23819234847866874,
      "grad_norm": 6.59375,
      "learning_rate": 4.983604484015395e-05,
      "loss": 3.5784,
      "num_input_tokens_seen": 3065136,
      "step": 1460
    },
    {
      "epoch": 0.2390080756994861,
      "grad_norm": 8.75,
      "learning_rate": 4.983492040484064e-05,
      "loss": 2.7678,
      "num_input_tokens_seen": 3076384,
      "step": 1465
    },
    {
      "epoch": 0.23982380292030345,
      "grad_norm": 7.90625,
      "learning_rate": 4.98337921396705e-05,
      "loss": 1.7655,
      "num_input_tokens_seen": 3088800,
      "step": 1470
    },
    {
      "epoch": 0.2406395301411208,
      "grad_norm": 18.125,
      "learning_rate": 4.983266004481753e-05,
      "loss": 3.9605,
      "num_input_tokens_seen": 3099328,
      "step": 1475
    },
    {
      "epoch": 0.24145525736193818,
      "grad_norm": 12.5625,
      "learning_rate": 4.9831524120456316e-05,
      "loss": 2.3588,
      "num_input_tokens_seen": 3109920,
      "step": 1480
    },
    {
      "epoch": 0.24227098458275553,
      "grad_norm": 4.46875,
      "learning_rate": 4.9830384366762026e-05,
      "loss": 1.4966,
      "num_input_tokens_seen": 3119360,
      "step": 1485
    },
    {
      "epoch": 0.24308671180357289,
      "grad_norm": 6.6875,
      "learning_rate": 4.9829240783910436e-05,
      "loss": 2.8051,
      "num_input_tokens_seen": 3128896,
      "step": 1490
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 8.875,
      "learning_rate": 4.982809337207789e-05,
      "loss": 2.3901,
      "num_input_tokens_seen": 3140512,
      "step": 1495
    },
    {
      "epoch": 0.2447181662452076,
      "grad_norm": 12.5,
      "learning_rate": 4.9826942131441337e-05,
      "loss": 3.1811,
      "num_input_tokens_seen": 3150480,
      "step": 1500
    },
    {
      "epoch": 0.24553389346602497,
      "grad_norm": 15.4375,
      "learning_rate": 4.9825787062178315e-05,
      "loss": 3.5,
      "num_input_tokens_seen": 3161408,
      "step": 1505
    },
    {
      "epoch": 0.24634962068684232,
      "grad_norm": 3.09375,
      "learning_rate": 4.9824628164466945e-05,
      "loss": 1.9037,
      "num_input_tokens_seen": 3171472,
      "step": 1510
    },
    {
      "epoch": 0.24716534790765968,
      "grad_norm": 4.21875,
      "learning_rate": 4.982346543848595e-05,
      "loss": 2.3638,
      "num_input_tokens_seen": 3182176,
      "step": 1515
    },
    {
      "epoch": 0.24798107512847703,
      "grad_norm": 4.125,
      "learning_rate": 4.9822298884414626e-05,
      "loss": 2.8021,
      "num_input_tokens_seen": 3193104,
      "step": 1520
    },
    {
      "epoch": 0.24879680234929438,
      "grad_norm": 4.625,
      "learning_rate": 4.982112850243288e-05,
      "loss": 2.4388,
      "num_input_tokens_seen": 3203872,
      "step": 1525
    },
    {
      "epoch": 0.24961252957011176,
      "grad_norm": 9.5,
      "learning_rate": 4.98199542927212e-05,
      "loss": 2.3121,
      "num_input_tokens_seen": 3213440,
      "step": 1530
    },
    {
      "epoch": 0.2504282567909291,
      "grad_norm": 10.125,
      "learning_rate": 4.981877625546066e-05,
      "loss": 2.2994,
      "num_input_tokens_seen": 3224720,
      "step": 1535
    },
    {
      "epoch": 0.25124398401174647,
      "grad_norm": 9.875,
      "learning_rate": 4.981759439083293e-05,
      "loss": 2.2191,
      "num_input_tokens_seen": 3236320,
      "step": 1540
    },
    {
      "epoch": 0.2520597112325638,
      "grad_norm": 5.75,
      "learning_rate": 4.981640869902027e-05,
      "loss": 2.3182,
      "num_input_tokens_seen": 3246928,
      "step": 1545
    },
    {
      "epoch": 0.2528754384533812,
      "grad_norm": 8.25,
      "learning_rate": 4.9815219180205517e-05,
      "loss": 2.1877,
      "num_input_tokens_seen": 3256672,
      "step": 1550
    },
    {
      "epoch": 0.2536911656741985,
      "grad_norm": 0.142578125,
      "learning_rate": 4.9814025834572126e-05,
      "loss": 2.7323,
      "num_input_tokens_seen": 3267200,
      "step": 1555
    },
    {
      "epoch": 0.2545068928950159,
      "grad_norm": 6.96875,
      "learning_rate": 4.981282866230411e-05,
      "loss": 1.7918,
      "num_input_tokens_seen": 3277168,
      "step": 1560
    },
    {
      "epoch": 0.2553226201158333,
      "grad_norm": 9.75,
      "learning_rate": 4.981162766358611e-05,
      "loss": 2.5694,
      "num_input_tokens_seen": 3287488,
      "step": 1565
    },
    {
      "epoch": 0.25613834733665064,
      "grad_norm": 10.0625,
      "learning_rate": 4.9810422838603316e-05,
      "loss": 4.192,
      "num_input_tokens_seen": 3297952,
      "step": 1570
    },
    {
      "epoch": 0.256954074557468,
      "grad_norm": 5.25,
      "learning_rate": 4.9809214187541533e-05,
      "loss": 1.7769,
      "num_input_tokens_seen": 3306912,
      "step": 1575
    },
    {
      "epoch": 0.25776980177828535,
      "grad_norm": 4.90625,
      "learning_rate": 4.980800171058715e-05,
      "loss": 2.2933,
      "num_input_tokens_seen": 3315712,
      "step": 1580
    },
    {
      "epoch": 0.2585855289991027,
      "grad_norm": 6.0625,
      "learning_rate": 4.980678540792715e-05,
      "loss": 1.933,
      "num_input_tokens_seen": 3327312,
      "step": 1585
    },
    {
      "epoch": 0.25940125621992005,
      "grad_norm": 7.625,
      "learning_rate": 4.980556527974909e-05,
      "loss": 1.7199,
      "num_input_tokens_seen": 3338992,
      "step": 1590
    },
    {
      "epoch": 0.2602169834407374,
      "grad_norm": 3.90625,
      "learning_rate": 4.980434132624114e-05,
      "loss": 2.5684,
      "num_input_tokens_seen": 3350720,
      "step": 1595
    },
    {
      "epoch": 0.26103271066155476,
      "grad_norm": 5.78125,
      "learning_rate": 4.980311354759205e-05,
      "loss": 1.7453,
      "num_input_tokens_seen": 3360448,
      "step": 1600
    },
    {
      "epoch": 0.26103271066155476,
      "eval_loss": 2.5495588779449463,
      "eval_runtime": 134.6251,
      "eval_samples_per_second": 20.241,
      "eval_steps_per_second": 10.124,
      "num_input_tokens_seen": 3360448,
      "step": 1600
    },
    {
      "epoch": 0.2618484378823721,
      "grad_norm": 4.9375,
      "learning_rate": 4.980188194399116e-05,
      "loss": 1.7761,
      "num_input_tokens_seen": 3371184,
      "step": 1605
    },
    {
      "epoch": 0.2626641651031895,
      "grad_norm": 5.65625,
      "learning_rate": 4.9800646515628384e-05,
      "loss": 1.6489,
      "num_input_tokens_seen": 3383232,
      "step": 1610
    },
    {
      "epoch": 0.26347989232400687,
      "grad_norm": 1.2578125,
      "learning_rate": 4.979940726269426e-05,
      "loss": 1.8968,
      "num_input_tokens_seen": 3394384,
      "step": 1615
    },
    {
      "epoch": 0.2642956195448242,
      "grad_norm": 6.78125,
      "learning_rate": 4.979816418537988e-05,
      "loss": 1.8214,
      "num_input_tokens_seen": 3404288,
      "step": 1620
    },
    {
      "epoch": 0.2651113467656416,
      "grad_norm": 6.5625,
      "learning_rate": 4.979691728387696e-05,
      "loss": 2.9042,
      "num_input_tokens_seen": 3415440,
      "step": 1625
    },
    {
      "epoch": 0.26592707398645893,
      "grad_norm": 7.6875,
      "learning_rate": 4.979566655837776e-05,
      "loss": 2.726,
      "num_input_tokens_seen": 3426672,
      "step": 1630
    },
    {
      "epoch": 0.2667428012072763,
      "grad_norm": 5.0625,
      "learning_rate": 4.9794412009075184e-05,
      "loss": 2.2014,
      "num_input_tokens_seen": 3436304,
      "step": 1635
    },
    {
      "epoch": 0.26755852842809363,
      "grad_norm": 6.125,
      "learning_rate": 4.979315363616269e-05,
      "loss": 1.8452,
      "num_input_tokens_seen": 3447408,
      "step": 1640
    },
    {
      "epoch": 0.268374255648911,
      "grad_norm": 4.875,
      "learning_rate": 4.979189143983434e-05,
      "loss": 2.2648,
      "num_input_tokens_seen": 3458144,
      "step": 1645
    },
    {
      "epoch": 0.26918998286972834,
      "grad_norm": 4.875,
      "learning_rate": 4.979062542028478e-05,
      "loss": 2.12,
      "num_input_tokens_seen": 3469776,
      "step": 1650
    },
    {
      "epoch": 0.27000571009054575,
      "grad_norm": 6.15625,
      "learning_rate": 4.978935557770923e-05,
      "loss": 2.3385,
      "num_input_tokens_seen": 3480560,
      "step": 1655
    },
    {
      "epoch": 0.2708214373113631,
      "grad_norm": 3.734375,
      "learning_rate": 4.978808191230353e-05,
      "loss": 1.8997,
      "num_input_tokens_seen": 3491264,
      "step": 1660
    },
    {
      "epoch": 0.27163716453218045,
      "grad_norm": 7.46875,
      "learning_rate": 4.9786804424264085e-05,
      "loss": 2.633,
      "num_input_tokens_seen": 3501920,
      "step": 1665
    },
    {
      "epoch": 0.2724528917529978,
      "grad_norm": 0.10693359375,
      "learning_rate": 4.978552311378792e-05,
      "loss": 1.5724,
      "num_input_tokens_seen": 3511776,
      "step": 1670
    },
    {
      "epoch": 0.27326861897381516,
      "grad_norm": 9.6875,
      "learning_rate": 4.978423798107261e-05,
      "loss": 2.3548,
      "num_input_tokens_seen": 3522896,
      "step": 1675
    },
    {
      "epoch": 0.2740843461946325,
      "grad_norm": 6.9375,
      "learning_rate": 4.978294902631635e-05,
      "loss": 2.4052,
      "num_input_tokens_seen": 3533808,
      "step": 1680
    },
    {
      "epoch": 0.27490007341544986,
      "grad_norm": 7.03125,
      "learning_rate": 4.9781656249717914e-05,
      "loss": 1.9794,
      "num_input_tokens_seen": 3544960,
      "step": 1685
    },
    {
      "epoch": 0.2757158006362672,
      "grad_norm": 6.6875,
      "learning_rate": 4.9780359651476645e-05,
      "loss": 2.2846,
      "num_input_tokens_seen": 3554672,
      "step": 1690
    },
    {
      "epoch": 0.27653152785708457,
      "grad_norm": 16.5,
      "learning_rate": 4.977905923179251e-05,
      "loss": 3.55,
      "num_input_tokens_seen": 3565664,
      "step": 1695
    },
    {
      "epoch": 0.2773472550779019,
      "grad_norm": 4.8125,
      "learning_rate": 4.977775499086606e-05,
      "loss": 2.0515,
      "num_input_tokens_seen": 3575744,
      "step": 1700
    },
    {
      "epoch": 0.27816298229871933,
      "grad_norm": 8.5625,
      "learning_rate": 4.97764469288984e-05,
      "loss": 2.1667,
      "num_input_tokens_seen": 3585840,
      "step": 1705
    },
    {
      "epoch": 0.2789787095195367,
      "grad_norm": 6.15625,
      "learning_rate": 4.977513504609127e-05,
      "loss": 1.5532,
      "num_input_tokens_seen": 3596816,
      "step": 1710
    },
    {
      "epoch": 0.27979443674035404,
      "grad_norm": 7.375,
      "learning_rate": 4.9773819342646965e-05,
      "loss": 2.0152,
      "num_input_tokens_seen": 3607584,
      "step": 1715
    },
    {
      "epoch": 0.2806101639611714,
      "grad_norm": 7.96875,
      "learning_rate": 4.97724998187684e-05,
      "loss": 2.5906,
      "num_input_tokens_seen": 3617504,
      "step": 1720
    },
    {
      "epoch": 0.28142589118198874,
      "grad_norm": 5.3125,
      "learning_rate": 4.9771176474659045e-05,
      "loss": 2.6162,
      "num_input_tokens_seen": 3628496,
      "step": 1725
    },
    {
      "epoch": 0.2822416184028061,
      "grad_norm": 3.25,
      "learning_rate": 4.976984931052299e-05,
      "loss": 1.6676,
      "num_input_tokens_seen": 3638848,
      "step": 1730
    },
    {
      "epoch": 0.28305734562362345,
      "grad_norm": 8.875,
      "learning_rate": 4.976851832656489e-05,
      "loss": 2.5647,
      "num_input_tokens_seen": 3648784,
      "step": 1735
    },
    {
      "epoch": 0.2838730728444408,
      "grad_norm": 5.15625,
      "learning_rate": 4.9767183522990004e-05,
      "loss": 4.2479,
      "num_input_tokens_seen": 3657760,
      "step": 1740
    },
    {
      "epoch": 0.28468880006525815,
      "grad_norm": 1.609375,
      "learning_rate": 4.9765844900004176e-05,
      "loss": 2.0166,
      "num_input_tokens_seen": 3667456,
      "step": 1745
    },
    {
      "epoch": 0.28550452728607556,
      "grad_norm": 10.4375,
      "learning_rate": 4.9764502457813834e-05,
      "loss": 2.5445,
      "num_input_tokens_seen": 3679296,
      "step": 1750
    },
    {
      "epoch": 0.2863202545068929,
      "grad_norm": 8.625,
      "learning_rate": 4.9763156196626005e-05,
      "loss": 2.893,
      "num_input_tokens_seen": 3690176,
      "step": 1755
    },
    {
      "epoch": 0.28713598172771027,
      "grad_norm": 8.8125,
      "learning_rate": 4.97618061166483e-05,
      "loss": 2.7165,
      "num_input_tokens_seen": 3699744,
      "step": 1760
    },
    {
      "epoch": 0.2879517089485276,
      "grad_norm": 5.34375,
      "learning_rate": 4.9760452218088915e-05,
      "loss": 2.7786,
      "num_input_tokens_seen": 3710016,
      "step": 1765
    },
    {
      "epoch": 0.28876743616934497,
      "grad_norm": 9.8125,
      "learning_rate": 4.975909450115663e-05,
      "loss": 2.5256,
      "num_input_tokens_seen": 3720224,
      "step": 1770
    },
    {
      "epoch": 0.2895831633901623,
      "grad_norm": 6.53125,
      "learning_rate": 4.975773296606084e-05,
      "loss": 1.9861,
      "num_input_tokens_seen": 3731280,
      "step": 1775
    },
    {
      "epoch": 0.2903988906109797,
      "grad_norm": 10.6875,
      "learning_rate": 4.97563676130115e-05,
      "loss": 1.6704,
      "num_input_tokens_seen": 3742352,
      "step": 1780
    },
    {
      "epoch": 0.29121461783179703,
      "grad_norm": 6.9375,
      "learning_rate": 4.9754998442219166e-05,
      "loss": 2.7467,
      "num_input_tokens_seen": 3752144,
      "step": 1785
    },
    {
      "epoch": 0.2920303450526144,
      "grad_norm": 3.65625,
      "learning_rate": 4.9753625453894984e-05,
      "loss": 1.7765,
      "num_input_tokens_seen": 3761760,
      "step": 1790
    },
    {
      "epoch": 0.2928460722734318,
      "grad_norm": 5.78125,
      "learning_rate": 4.975224864825068e-05,
      "loss": 1.3404,
      "num_input_tokens_seen": 3771952,
      "step": 1795
    },
    {
      "epoch": 0.29366179949424914,
      "grad_norm": 8.0,
      "learning_rate": 4.9750868025498576e-05,
      "loss": 1.8503,
      "num_input_tokens_seen": 3782768,
      "step": 1800
    },
    {
      "epoch": 0.29366179949424914,
      "eval_loss": 2.5513477325439453,
      "eval_runtime": 134.7861,
      "eval_samples_per_second": 20.217,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 3782768,
      "step": 1800
    },
    {
      "epoch": 0.2944775267150665,
      "grad_norm": 7.4375,
      "learning_rate": 4.974948358585158e-05,
      "loss": 2.1661,
      "num_input_tokens_seen": 3794032,
      "step": 1805
    },
    {
      "epoch": 0.29529325393588385,
      "grad_norm": 10.375,
      "learning_rate": 4.9748095329523205e-05,
      "loss": 2.6868,
      "num_input_tokens_seen": 3803696,
      "step": 1810
    },
    {
      "epoch": 0.2961089811567012,
      "grad_norm": 3.25,
      "learning_rate": 4.974670325672752e-05,
      "loss": 1.5509,
      "num_input_tokens_seen": 3815184,
      "step": 1815
    },
    {
      "epoch": 0.29692470837751855,
      "grad_norm": 10.75,
      "learning_rate": 4.974530736767921e-05,
      "loss": 2.5746,
      "num_input_tokens_seen": 3823280,
      "step": 1820
    },
    {
      "epoch": 0.2977404355983359,
      "grad_norm": 7.9375,
      "learning_rate": 4.9743907662593524e-05,
      "loss": 2.7132,
      "num_input_tokens_seen": 3834880,
      "step": 1825
    },
    {
      "epoch": 0.29855616281915326,
      "grad_norm": 5.09375,
      "learning_rate": 4.974250414168633e-05,
      "loss": 2.337,
      "num_input_tokens_seen": 3846688,
      "step": 1830
    },
    {
      "epoch": 0.2993718900399706,
      "grad_norm": 9.125,
      "learning_rate": 4.974109680517407e-05,
      "loss": 2.7045,
      "num_input_tokens_seen": 3856672,
      "step": 1835
    },
    {
      "epoch": 0.300187617260788,
      "grad_norm": 12.1875,
      "learning_rate": 4.973968565327376e-05,
      "loss": 2.9434,
      "num_input_tokens_seen": 3867008,
      "step": 1840
    },
    {
      "epoch": 0.3010033444816054,
      "grad_norm": 5.53125,
      "learning_rate": 4.973827068620303e-05,
      "loss": 2.5616,
      "num_input_tokens_seen": 3876096,
      "step": 1845
    },
    {
      "epoch": 0.3018190717024227,
      "grad_norm": 6.28125,
      "learning_rate": 4.973685190418008e-05,
      "loss": 1.01,
      "num_input_tokens_seen": 3886288,
      "step": 1850
    },
    {
      "epoch": 0.3026347989232401,
      "grad_norm": 8.75,
      "learning_rate": 4.97354293074237e-05,
      "loss": 1.9692,
      "num_input_tokens_seen": 3896096,
      "step": 1855
    },
    {
      "epoch": 0.30345052614405743,
      "grad_norm": 9.625,
      "learning_rate": 4.9734002896153276e-05,
      "loss": 1.9227,
      "num_input_tokens_seen": 3905632,
      "step": 1860
    },
    {
      "epoch": 0.3042662533648748,
      "grad_norm": 5.4375,
      "learning_rate": 4.973257267058877e-05,
      "loss": 2.4571,
      "num_input_tokens_seen": 3915856,
      "step": 1865
    },
    {
      "epoch": 0.30508198058569214,
      "grad_norm": 11.0,
      "learning_rate": 4.973113863095076e-05,
      "loss": 2.8401,
      "num_input_tokens_seen": 3926592,
      "step": 1870
    },
    {
      "epoch": 0.3058977078065095,
      "grad_norm": 6.25,
      "learning_rate": 4.9729700777460384e-05,
      "loss": 2.1013,
      "num_input_tokens_seen": 3937344,
      "step": 1875
    },
    {
      "epoch": 0.30671343502732684,
      "grad_norm": 4.65625,
      "learning_rate": 4.972825911033937e-05,
      "loss": 1.4578,
      "num_input_tokens_seen": 3948224,
      "step": 1880
    },
    {
      "epoch": 0.3075291622481442,
      "grad_norm": 5.25,
      "learning_rate": 4.9726813629810056e-05,
      "loss": 2.2478,
      "num_input_tokens_seen": 3958368,
      "step": 1885
    },
    {
      "epoch": 0.3083448894689616,
      "grad_norm": 8.875,
      "learning_rate": 4.9725364336095326e-05,
      "loss": 2.0339,
      "num_input_tokens_seen": 3968960,
      "step": 1890
    },
    {
      "epoch": 0.30916061668977896,
      "grad_norm": 2.46875,
      "learning_rate": 4.972391122941871e-05,
      "loss": 1.5583,
      "num_input_tokens_seen": 3978656,
      "step": 1895
    },
    {
      "epoch": 0.3099763439105963,
      "grad_norm": 9.0,
      "learning_rate": 4.972245431000428e-05,
      "loss": 1.1397,
      "num_input_tokens_seen": 3989376,
      "step": 1900
    },
    {
      "epoch": 0.31079207113141366,
      "grad_norm": 3.046875,
      "learning_rate": 4.972099357807671e-05,
      "loss": 1.6381,
      "num_input_tokens_seen": 4001264,
      "step": 1905
    },
    {
      "epoch": 0.311607798352231,
      "grad_norm": 6.3125,
      "learning_rate": 4.971952903386127e-05,
      "loss": 3.108,
      "num_input_tokens_seen": 4012576,
      "step": 1910
    },
    {
      "epoch": 0.31242352557304837,
      "grad_norm": 6.25,
      "learning_rate": 4.971806067758381e-05,
      "loss": 2.5796,
      "num_input_tokens_seen": 4022976,
      "step": 1915
    },
    {
      "epoch": 0.3132392527938657,
      "grad_norm": 5.4375,
      "learning_rate": 4.971658850947076e-05,
      "loss": 3.5806,
      "num_input_tokens_seen": 4033664,
      "step": 1920
    },
    {
      "epoch": 0.31405498001468307,
      "grad_norm": 6.25,
      "learning_rate": 4.9715112529749165e-05,
      "loss": 2.894,
      "num_input_tokens_seen": 4045264,
      "step": 1925
    },
    {
      "epoch": 0.3148707072355004,
      "grad_norm": 10.4375,
      "learning_rate": 4.9713632738646624e-05,
      "loss": 2.1021,
      "num_input_tokens_seen": 4056288,
      "step": 1930
    },
    {
      "epoch": 0.31568643445631783,
      "grad_norm": 10.9375,
      "learning_rate": 4.971214913639134e-05,
      "loss": 2.4593,
      "num_input_tokens_seen": 4067472,
      "step": 1935
    },
    {
      "epoch": 0.3165021616771352,
      "grad_norm": 0.23828125,
      "learning_rate": 4.9710661723212104e-05,
      "loss": 1.0891,
      "num_input_tokens_seen": 4078976,
      "step": 1940
    },
    {
      "epoch": 0.31731788889795254,
      "grad_norm": 3.53125,
      "learning_rate": 4.9709170499338295e-05,
      "loss": 2.8066,
      "num_input_tokens_seen": 4088256,
      "step": 1945
    },
    {
      "epoch": 0.3181336161187699,
      "grad_norm": 8.3125,
      "learning_rate": 4.9707675464999895e-05,
      "loss": 2.8469,
      "num_input_tokens_seen": 4098800,
      "step": 1950
    },
    {
      "epoch": 0.31894934333958724,
      "grad_norm": 0.1318359375,
      "learning_rate": 4.970617662042743e-05,
      "loss": 2.5577,
      "num_input_tokens_seen": 4110048,
      "step": 1955
    },
    {
      "epoch": 0.3197650705604046,
      "grad_norm": 9.3125,
      "learning_rate": 4.970467396585206e-05,
      "loss": 2.7401,
      "num_input_tokens_seen": 4120000,
      "step": 1960
    },
    {
      "epoch": 0.32058079778122195,
      "grad_norm": 7.90625,
      "learning_rate": 4.97031675015055e-05,
      "loss": 2.345,
      "num_input_tokens_seen": 4129840,
      "step": 1965
    },
    {
      "epoch": 0.3213965250020393,
      "grad_norm": 2.390625,
      "learning_rate": 4.9701657227620075e-05,
      "loss": 2.5168,
      "num_input_tokens_seen": 4140224,
      "step": 1970
    },
    {
      "epoch": 0.32221225222285665,
      "grad_norm": 7.59375,
      "learning_rate": 4.9700143144428685e-05,
      "loss": 2.7057,
      "num_input_tokens_seen": 4150832,
      "step": 1975
    },
    {
      "epoch": 0.32302797944367406,
      "grad_norm": 3.828125,
      "learning_rate": 4.969862525216482e-05,
      "loss": 2.1708,
      "num_input_tokens_seen": 4161136,
      "step": 1980
    },
    {
      "epoch": 0.3238437066644914,
      "grad_norm": 12.6875,
      "learning_rate": 4.9697103551062556e-05,
      "loss": 2.974,
      "num_input_tokens_seen": 4171392,
      "step": 1985
    },
    {
      "epoch": 0.32465943388530877,
      "grad_norm": 12.1875,
      "learning_rate": 4.9695578041356565e-05,
      "loss": 3.6053,
      "num_input_tokens_seen": 4181984,
      "step": 1990
    },
    {
      "epoch": 0.3254751611061261,
      "grad_norm": 6.21875,
      "learning_rate": 4.969404872328209e-05,
      "loss": 2.3394,
      "num_input_tokens_seen": 4193472,
      "step": 1995
    },
    {
      "epoch": 0.3262908883269435,
      "grad_norm": 3.890625,
      "learning_rate": 4.969251559707498e-05,
      "loss": 2.2157,
      "num_input_tokens_seen": 4205680,
      "step": 2000
    },
    {
      "epoch": 0.3262908883269435,
      "eval_loss": 2.535407781600952,
      "eval_runtime": 134.8752,
      "eval_samples_per_second": 20.204,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 4205680,
      "step": 2000
    },
    {
      "epoch": 0.3271066155477608,
      "grad_norm": 0.1298828125,
      "learning_rate": 4.9690978662971674e-05,
      "loss": 2.6241,
      "num_input_tokens_seen": 4216496,
      "step": 2005
    },
    {
      "epoch": 0.3279223427685782,
      "grad_norm": 4.375,
      "learning_rate": 4.968943792120916e-05,
      "loss": 2.4422,
      "num_input_tokens_seen": 4227232,
      "step": 2010
    },
    {
      "epoch": 0.32873806998939553,
      "grad_norm": 8.9375,
      "learning_rate": 4.9687893372025046e-05,
      "loss": 1.9462,
      "num_input_tokens_seen": 4238512,
      "step": 2015
    },
    {
      "epoch": 0.3295537972102129,
      "grad_norm": 4.75,
      "learning_rate": 4.9686345015657535e-05,
      "loss": 3.0434,
      "num_input_tokens_seen": 4250048,
      "step": 2020
    },
    {
      "epoch": 0.33036952443103024,
      "grad_norm": 11.5,
      "learning_rate": 4.968479285234538e-05,
      "loss": 3.0429,
      "num_input_tokens_seen": 4260640,
      "step": 2025
    },
    {
      "epoch": 0.33118525165184765,
      "grad_norm": 9.5625,
      "learning_rate": 4.9683236882327974e-05,
      "loss": 2.6472,
      "num_input_tokens_seen": 4271248,
      "step": 2030
    },
    {
      "epoch": 0.332000978872665,
      "grad_norm": 2.90625,
      "learning_rate": 4.968167710584526e-05,
      "loss": 1.9837,
      "num_input_tokens_seen": 4281040,
      "step": 2035
    },
    {
      "epoch": 0.33281670609348235,
      "grad_norm": 6.9375,
      "learning_rate": 4.968011352313775e-05,
      "loss": 1.9224,
      "num_input_tokens_seen": 4291520,
      "step": 2040
    },
    {
      "epoch": 0.3336324333142997,
      "grad_norm": 5.78125,
      "learning_rate": 4.967854613444659e-05,
      "loss": 1.0642,
      "num_input_tokens_seen": 4302240,
      "step": 2045
    },
    {
      "epoch": 0.33444816053511706,
      "grad_norm": 10.125,
      "learning_rate": 4.967697494001349e-05,
      "loss": 2.585,
      "num_input_tokens_seen": 4312128,
      "step": 2050
    },
    {
      "epoch": 0.3352638877559344,
      "grad_norm": 16.125,
      "learning_rate": 4.9675399940080736e-05,
      "loss": 3.2917,
      "num_input_tokens_seen": 4322720,
      "step": 2055
    },
    {
      "epoch": 0.33607961497675176,
      "grad_norm": 8.3125,
      "learning_rate": 4.9673821134891226e-05,
      "loss": 2.9071,
      "num_input_tokens_seen": 4332400,
      "step": 2060
    },
    {
      "epoch": 0.3368953421975691,
      "grad_norm": 3.640625,
      "learning_rate": 4.967223852468842e-05,
      "loss": 1.6234,
      "num_input_tokens_seen": 4342928,
      "step": 2065
    },
    {
      "epoch": 0.33771106941838647,
      "grad_norm": 5.5,
      "learning_rate": 4.967065210971639e-05,
      "loss": 3.0376,
      "num_input_tokens_seen": 4354400,
      "step": 2070
    },
    {
      "epoch": 0.3385267966392039,
      "grad_norm": 11.625,
      "learning_rate": 4.966906189021977e-05,
      "loss": 3.7964,
      "num_input_tokens_seen": 4365328,
      "step": 2075
    },
    {
      "epoch": 0.33934252386002123,
      "grad_norm": 9.75,
      "learning_rate": 4.966746786644379e-05,
      "loss": 2.2056,
      "num_input_tokens_seen": 4374512,
      "step": 2080
    },
    {
      "epoch": 0.3401582510808386,
      "grad_norm": 2.515625,
      "learning_rate": 4.966587003863429e-05,
      "loss": 1.5924,
      "num_input_tokens_seen": 4384480,
      "step": 2085
    },
    {
      "epoch": 0.34097397830165593,
      "grad_norm": 7.0625,
      "learning_rate": 4.966426840703765e-05,
      "loss": 4.1598,
      "num_input_tokens_seen": 4396096,
      "step": 2090
    },
    {
      "epoch": 0.3417897055224733,
      "grad_norm": 4.84375,
      "learning_rate": 4.9662662971900875e-05,
      "loss": 2.1955,
      "num_input_tokens_seen": 4406496,
      "step": 2095
    },
    {
      "epoch": 0.34260543274329064,
      "grad_norm": 7.125,
      "learning_rate": 4.9661053733471534e-05,
      "loss": 2.8056,
      "num_input_tokens_seen": 4416368,
      "step": 2100
    },
    {
      "epoch": 0.343421159964108,
      "grad_norm": 9.25,
      "learning_rate": 4.965944069199781e-05,
      "loss": 2.8787,
      "num_input_tokens_seen": 4427360,
      "step": 2105
    },
    {
      "epoch": 0.34423688718492534,
      "grad_norm": 5.375,
      "learning_rate": 4.965782384772842e-05,
      "loss": 1.0717,
      "num_input_tokens_seen": 4437280,
      "step": 2110
    },
    {
      "epoch": 0.3450526144057427,
      "grad_norm": 8.6875,
      "learning_rate": 4.9656203200912734e-05,
      "loss": 2.7308,
      "num_input_tokens_seen": 4447872,
      "step": 2115
    },
    {
      "epoch": 0.3458683416265601,
      "grad_norm": 6.5625,
      "learning_rate": 4.965457875180067e-05,
      "loss": 2.4515,
      "num_input_tokens_seen": 4458784,
      "step": 2120
    },
    {
      "epoch": 0.34668406884737746,
      "grad_norm": 6.5,
      "learning_rate": 4.9652950500642724e-05,
      "loss": 2.4324,
      "num_input_tokens_seen": 4469728,
      "step": 2125
    },
    {
      "epoch": 0.3474997960681948,
      "grad_norm": 3.21875,
      "learning_rate": 4.965131844769001e-05,
      "loss": 1.203,
      "num_input_tokens_seen": 4479872,
      "step": 2130
    },
    {
      "epoch": 0.34831552328901216,
      "grad_norm": 8.875,
      "learning_rate": 4.96496825931942e-05,
      "loss": 2.8917,
      "num_input_tokens_seen": 4489920,
      "step": 2135
    },
    {
      "epoch": 0.3491312505098295,
      "grad_norm": 4.53125,
      "learning_rate": 4.9648042937407566e-05,
      "loss": 2.9721,
      "num_input_tokens_seen": 4499232,
      "step": 2140
    },
    {
      "epoch": 0.34994697773064687,
      "grad_norm": 7.3125,
      "learning_rate": 4.964639948058297e-05,
      "loss": 2.5412,
      "num_input_tokens_seen": 4508528,
      "step": 2145
    },
    {
      "epoch": 0.3507627049514642,
      "grad_norm": 5.34375,
      "learning_rate": 4.9644752222973846e-05,
      "loss": 2.9385,
      "num_input_tokens_seen": 4518880,
      "step": 2150
    },
    {
      "epoch": 0.3515784321722816,
      "grad_norm": 5.46875,
      "learning_rate": 4.964310116483422e-05,
      "loss": 2.8934,
      "num_input_tokens_seen": 4530176,
      "step": 2155
    },
    {
      "epoch": 0.3523941593930989,
      "grad_norm": 5.59375,
      "learning_rate": 4.964144630641872e-05,
      "loss": 1.3564,
      "num_input_tokens_seen": 4540048,
      "step": 2160
    },
    {
      "epoch": 0.3532098866139163,
      "grad_norm": 5.90625,
      "learning_rate": 4.9639787647982525e-05,
      "loss": 2.1106,
      "num_input_tokens_seen": 4549744,
      "step": 2165
    },
    {
      "epoch": 0.3540256138347337,
      "grad_norm": 6.625,
      "learning_rate": 4.963812518978143e-05,
      "loss": 2.9727,
      "num_input_tokens_seen": 4561280,
      "step": 2170
    },
    {
      "epoch": 0.35484134105555104,
      "grad_norm": 12.125,
      "learning_rate": 4.963645893207182e-05,
      "loss": 2.3613,
      "num_input_tokens_seen": 4571232,
      "step": 2175
    },
    {
      "epoch": 0.3556570682763684,
      "grad_norm": 4.5625,
      "learning_rate": 4.963478887511063e-05,
      "loss": 1.265,
      "num_input_tokens_seen": 4580960,
      "step": 2180
    },
    {
      "epoch": 0.35647279549718575,
      "grad_norm": 7.25,
      "learning_rate": 4.963311501915542e-05,
      "loss": 1.8243,
      "num_input_tokens_seen": 4590560,
      "step": 2185
    },
    {
      "epoch": 0.3572885227180031,
      "grad_norm": 4.25,
      "learning_rate": 4.963143736446432e-05,
      "loss": 2.4404,
      "num_input_tokens_seen": 4600048,
      "step": 2190
    },
    {
      "epoch": 0.35810424993882045,
      "grad_norm": 9.5,
      "learning_rate": 4.962975591129603e-05,
      "loss": 2.3032,
      "num_input_tokens_seen": 4610688,
      "step": 2195
    },
    {
      "epoch": 0.3589199771596378,
      "grad_norm": 3.75,
      "learning_rate": 4.962807065990986e-05,
      "loss": 2.5347,
      "num_input_tokens_seen": 4620944,
      "step": 2200
    },
    {
      "epoch": 0.3589199771596378,
      "eval_loss": 2.5436015129089355,
      "eval_runtime": 134.7892,
      "eval_samples_per_second": 20.217,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 4620944,
      "step": 2200
    },
    {
      "epoch": 0.35973570438045516,
      "grad_norm": 9.0625,
      "learning_rate": 4.9626381610565714e-05,
      "loss": 3.3176,
      "num_input_tokens_seen": 4630656,
      "step": 2205
    },
    {
      "epoch": 0.3605514316012725,
      "grad_norm": 6.375,
      "learning_rate": 4.9624688763524043e-05,
      "loss": 3.0733,
      "num_input_tokens_seen": 4642112,
      "step": 2210
    },
    {
      "epoch": 0.3613671588220899,
      "grad_norm": 12.5,
      "learning_rate": 4.962299211904591e-05,
      "loss": 3.3108,
      "num_input_tokens_seen": 4652688,
      "step": 2215
    },
    {
      "epoch": 0.36218288604290727,
      "grad_norm": 3.125,
      "learning_rate": 4.962129167739296e-05,
      "loss": 2.1965,
      "num_input_tokens_seen": 4663840,
      "step": 2220
    },
    {
      "epoch": 0.3629986132637246,
      "grad_norm": 10.25,
      "learning_rate": 4.961958743882742e-05,
      "loss": 2.0496,
      "num_input_tokens_seen": 4674384,
      "step": 2225
    },
    {
      "epoch": 0.363814340484542,
      "grad_norm": 8.1875,
      "learning_rate": 4.961787940361211e-05,
      "loss": 2.2178,
      "num_input_tokens_seen": 4685248,
      "step": 2230
    },
    {
      "epoch": 0.36463006770535933,
      "grad_norm": 6.0625,
      "learning_rate": 4.961616757201043e-05,
      "loss": 2.8622,
      "num_input_tokens_seen": 4696576,
      "step": 2235
    },
    {
      "epoch": 0.3654457949261767,
      "grad_norm": 3.5625,
      "learning_rate": 4.961445194428637e-05,
      "loss": 2.79,
      "num_input_tokens_seen": 4706848,
      "step": 2240
    },
    {
      "epoch": 0.36626152214699403,
      "grad_norm": 10.9375,
      "learning_rate": 4.9612732520704486e-05,
      "loss": 2.7198,
      "num_input_tokens_seen": 4718224,
      "step": 2245
    },
    {
      "epoch": 0.3670772493678114,
      "grad_norm": 6.28125,
      "learning_rate": 4.961100930152994e-05,
      "loss": 3.1937,
      "num_input_tokens_seen": 4728336,
      "step": 2250
    },
    {
      "epoch": 0.36789297658862874,
      "grad_norm": 5.3125,
      "learning_rate": 4.960928228702849e-05,
      "loss": 1.3801,
      "num_input_tokens_seen": 4739264,
      "step": 2255
    },
    {
      "epoch": 0.36870870380944615,
      "grad_norm": 4.90625,
      "learning_rate": 4.960755147746645e-05,
      "loss": 2.3433,
      "num_input_tokens_seen": 4746736,
      "step": 2260
    },
    {
      "epoch": 0.3695244310302635,
      "grad_norm": 4.53125,
      "learning_rate": 4.9605816873110736e-05,
      "loss": 2.3433,
      "num_input_tokens_seen": 4756976,
      "step": 2265
    },
    {
      "epoch": 0.37034015825108085,
      "grad_norm": 0.984375,
      "learning_rate": 4.960407847422883e-05,
      "loss": 1.6484,
      "num_input_tokens_seen": 4766448,
      "step": 2270
    },
    {
      "epoch": 0.3711558854718982,
      "grad_norm": 8.625,
      "learning_rate": 4.960233628108885e-05,
      "loss": 2.4146,
      "num_input_tokens_seen": 4777584,
      "step": 2275
    },
    {
      "epoch": 0.37197161269271556,
      "grad_norm": 2.890625,
      "learning_rate": 4.960059029395942e-05,
      "loss": 2.0387,
      "num_input_tokens_seen": 4788064,
      "step": 2280
    },
    {
      "epoch": 0.3727873399135329,
      "grad_norm": 4.1875,
      "learning_rate": 4.959884051310983e-05,
      "loss": 1.0596,
      "num_input_tokens_seen": 4799568,
      "step": 2285
    },
    {
      "epoch": 0.37360306713435026,
      "grad_norm": 7.4375,
      "learning_rate": 4.959708693880991e-05,
      "loss": 3.1311,
      "num_input_tokens_seen": 4809200,
      "step": 2290
    },
    {
      "epoch": 0.3744187943551676,
      "grad_norm": 3.765625,
      "learning_rate": 4.9595329571330074e-05,
      "loss": 3.1501,
      "num_input_tokens_seen": 4818640,
      "step": 2295
    },
    {
      "epoch": 0.37523452157598497,
      "grad_norm": 8.6875,
      "learning_rate": 4.9593568410941326e-05,
      "loss": 2.0934,
      "num_input_tokens_seen": 4828816,
      "step": 2300
    },
    {
      "epoch": 0.3760502487968023,
      "grad_norm": 4.5,
      "learning_rate": 4.959180345791528e-05,
      "loss": 1.6894,
      "num_input_tokens_seen": 4838864,
      "step": 2305
    },
    {
      "epoch": 0.37686597601761973,
      "grad_norm": 4.96875,
      "learning_rate": 4.9590034712524086e-05,
      "loss": 2.6994,
      "num_input_tokens_seen": 4849728,
      "step": 2310
    },
    {
      "epoch": 0.3776817032384371,
      "grad_norm": 4.1875,
      "learning_rate": 4.958826217504053e-05,
      "loss": 1.34,
      "num_input_tokens_seen": 4860432,
      "step": 2315
    },
    {
      "epoch": 0.37849743045925444,
      "grad_norm": 6.65625,
      "learning_rate": 4.958648584573795e-05,
      "loss": 2.4233,
      "num_input_tokens_seen": 4870784,
      "step": 2320
    },
    {
      "epoch": 0.3793131576800718,
      "grad_norm": 3.015625,
      "learning_rate": 4.958470572489028e-05,
      "loss": 2.6665,
      "num_input_tokens_seen": 4880768,
      "step": 2325
    },
    {
      "epoch": 0.38012888490088914,
      "grad_norm": 6.03125,
      "learning_rate": 4.958292181277203e-05,
      "loss": 1.5832,
      "num_input_tokens_seen": 4891536,
      "step": 2330
    },
    {
      "epoch": 0.3809446121217065,
      "grad_norm": 4.21875,
      "learning_rate": 4.958113410965832e-05,
      "loss": 2.6957,
      "num_input_tokens_seen": 4901680,
      "step": 2335
    },
    {
      "epoch": 0.38176033934252385,
      "grad_norm": 7.75,
      "learning_rate": 4.957934261582481e-05,
      "loss": 2.0524,
      "num_input_tokens_seen": 4913824,
      "step": 2340
    },
    {
      "epoch": 0.3825760665633412,
      "grad_norm": 7.15625,
      "learning_rate": 4.95775473315478e-05,
      "loss": 3.7545,
      "num_input_tokens_seen": 4924528,
      "step": 2345
    },
    {
      "epoch": 0.38339179378415855,
      "grad_norm": 13.0625,
      "learning_rate": 4.9575748257104124e-05,
      "loss": 2.2637,
      "num_input_tokens_seen": 4934784,
      "step": 2350
    },
    {
      "epoch": 0.38420752100497596,
      "grad_norm": 7.3125,
      "learning_rate": 4.9573945392771224e-05,
      "loss": 3.619,
      "num_input_tokens_seen": 4944768,
      "step": 2355
    },
    {
      "epoch": 0.3850232482257933,
      "grad_norm": 13.0,
      "learning_rate": 4.9572138738827134e-05,
      "loss": 2.9085,
      "num_input_tokens_seen": 4954368,
      "step": 2360
    },
    {
      "epoch": 0.38583897544661067,
      "grad_norm": 3.296875,
      "learning_rate": 4.957032829555046e-05,
      "loss": 2.8159,
      "num_input_tokens_seen": 4963616,
      "step": 2365
    },
    {
      "epoch": 0.386654702667428,
      "grad_norm": 7.03125,
      "learning_rate": 4.956851406322039e-05,
      "loss": 2.7959,
      "num_input_tokens_seen": 4973808,
      "step": 2370
    },
    {
      "epoch": 0.38747042988824537,
      "grad_norm": 7.8125,
      "learning_rate": 4.9566696042116704e-05,
      "loss": 1.8125,
      "num_input_tokens_seen": 4984320,
      "step": 2375
    },
    {
      "epoch": 0.3882861571090627,
      "grad_norm": 3.796875,
      "learning_rate": 4.9564874232519766e-05,
      "loss": 2.2002,
      "num_input_tokens_seen": 4995488,
      "step": 2380
    },
    {
      "epoch": 0.3891018843298801,
      "grad_norm": 4.78125,
      "learning_rate": 4.9563048634710516e-05,
      "loss": 2.5364,
      "num_input_tokens_seen": 5006080,
      "step": 2385
    },
    {
      "epoch": 0.38991761155069743,
      "grad_norm": 9.6875,
      "learning_rate": 4.956121924897049e-05,
      "loss": 3.3898,
      "num_input_tokens_seen": 5016464,
      "step": 2390
    },
    {
      "epoch": 0.3907333387715148,
      "grad_norm": 5.71875,
      "learning_rate": 4.955938607558181e-05,
      "loss": 1.7356,
      "num_input_tokens_seen": 5026944,
      "step": 2395
    },
    {
      "epoch": 0.3915490659923322,
      "grad_norm": 7.28125,
      "learning_rate": 4.955754911482715e-05,
      "loss": 2.6464,
      "num_input_tokens_seen": 5037232,
      "step": 2400
    },
    {
      "epoch": 0.3915490659923322,
      "eval_loss": 2.545353889465332,
      "eval_runtime": 134.9038,
      "eval_samples_per_second": 20.2,
      "eval_steps_per_second": 10.103,
      "num_input_tokens_seen": 5037232,
      "step": 2400
    },
    {
      "epoch": 0.39236479321314954,
      "grad_norm": 8.375,
      "learning_rate": 4.9555708366989804e-05,
      "loss": 1.5869,
      "num_input_tokens_seen": 5047536,
      "step": 2405
    },
    {
      "epoch": 0.3931805204339669,
      "grad_norm": 10.9375,
      "learning_rate": 4.9553863832353655e-05,
      "loss": 1.8219,
      "num_input_tokens_seen": 5057408,
      "step": 2410
    },
    {
      "epoch": 0.39399624765478425,
      "grad_norm": 8.5,
      "learning_rate": 4.955201551120313e-05,
      "loss": 1.8402,
      "num_input_tokens_seen": 5068416,
      "step": 2415
    },
    {
      "epoch": 0.3948119748756016,
      "grad_norm": 1.2890625,
      "learning_rate": 4.955016340382328e-05,
      "loss": 2.1584,
      "num_input_tokens_seen": 5079936,
      "step": 2420
    },
    {
      "epoch": 0.39562770209641895,
      "grad_norm": 8.75,
      "learning_rate": 4.954830751049972e-05,
      "loss": 1.3047,
      "num_input_tokens_seen": 5090080,
      "step": 2425
    },
    {
      "epoch": 0.3964434293172363,
      "grad_norm": 8.8125,
      "learning_rate": 4.954644783151864e-05,
      "loss": 1.1211,
      "num_input_tokens_seen": 5101120,
      "step": 2430
    },
    {
      "epoch": 0.39725915653805366,
      "grad_norm": 8.125,
      "learning_rate": 4.954458436716684e-05,
      "loss": 1.7216,
      "num_input_tokens_seen": 5110976,
      "step": 2435
    },
    {
      "epoch": 0.398074883758871,
      "grad_norm": 4.0625,
      "learning_rate": 4.954271711773168e-05,
      "loss": 3.0302,
      "num_input_tokens_seen": 5121840,
      "step": 2440
    },
    {
      "epoch": 0.39889061097968836,
      "grad_norm": 9.0,
      "learning_rate": 4.9540846083501115e-05,
      "loss": 2.4965,
      "num_input_tokens_seen": 5133120,
      "step": 2445
    },
    {
      "epoch": 0.3997063382005058,
      "grad_norm": 5.53125,
      "learning_rate": 4.953897126476369e-05,
      "loss": 2.4533,
      "num_input_tokens_seen": 5144096,
      "step": 2450
    },
    {
      "epoch": 0.4005220654213231,
      "grad_norm": 18.0,
      "learning_rate": 4.9537092661808514e-05,
      "loss": 2.6103,
      "num_input_tokens_seen": 5154848,
      "step": 2455
    },
    {
      "epoch": 0.4013377926421405,
      "grad_norm": 6.4375,
      "learning_rate": 4.9535210274925306e-05,
      "loss": 2.0982,
      "num_input_tokens_seen": 5166112,
      "step": 2460
    },
    {
      "epoch": 0.40215351986295783,
      "grad_norm": 5.125,
      "learning_rate": 4.953332410440435e-05,
      "loss": 2.1108,
      "num_input_tokens_seen": 5177760,
      "step": 2465
    },
    {
      "epoch": 0.4029692470837752,
      "grad_norm": 0.1328125,
      "learning_rate": 4.9531434150536496e-05,
      "loss": 2.8628,
      "num_input_tokens_seen": 5188016,
      "step": 2470
    },
    {
      "epoch": 0.40378497430459254,
      "grad_norm": 7.625,
      "learning_rate": 4.952954041361322e-05,
      "loss": 2.0925,
      "num_input_tokens_seen": 5199744,
      "step": 2475
    },
    {
      "epoch": 0.4046007015254099,
      "grad_norm": 7.875,
      "learning_rate": 4.952764289392655e-05,
      "loss": 3.7824,
      "num_input_tokens_seen": 5210624,
      "step": 2480
    },
    {
      "epoch": 0.40541642874622724,
      "grad_norm": 8.9375,
      "learning_rate": 4.952574159176912e-05,
      "loss": 1.9947,
      "num_input_tokens_seen": 5219344,
      "step": 2485
    },
    {
      "epoch": 0.4062321559670446,
      "grad_norm": 5.21875,
      "learning_rate": 4.952383650743413e-05,
      "loss": 1.0435,
      "num_input_tokens_seen": 5228400,
      "step": 2490
    },
    {
      "epoch": 0.407047883187862,
      "grad_norm": 6.3125,
      "learning_rate": 4.952192764121536e-05,
      "loss": 3.6553,
      "num_input_tokens_seen": 5239056,
      "step": 2495
    },
    {
      "epoch": 0.40786361040867936,
      "grad_norm": 6.8125,
      "learning_rate": 4.9520014993407185e-05,
      "loss": 2.4143,
      "num_input_tokens_seen": 5248640,
      "step": 2500
    },
    {
      "epoch": 0.4086793376294967,
      "grad_norm": 5.21875,
      "learning_rate": 4.951809856430456e-05,
      "loss": 2.0306,
      "num_input_tokens_seen": 5259312,
      "step": 2505
    },
    {
      "epoch": 0.40949506485031406,
      "grad_norm": 8.8125,
      "learning_rate": 4.951617835420303e-05,
      "loss": 2.5674,
      "num_input_tokens_seen": 5269360,
      "step": 2510
    },
    {
      "epoch": 0.4103107920711314,
      "grad_norm": 8.9375,
      "learning_rate": 4.951425436339869e-05,
      "loss": 2.208,
      "num_input_tokens_seen": 5278528,
      "step": 2515
    },
    {
      "epoch": 0.41112651929194877,
      "grad_norm": 7.40625,
      "learning_rate": 4.9512326592188274e-05,
      "loss": 2.838,
      "num_input_tokens_seen": 5288864,
      "step": 2520
    },
    {
      "epoch": 0.4119422465127661,
      "grad_norm": 7.6875,
      "learning_rate": 4.9510395040869054e-05,
      "loss": 2.4106,
      "num_input_tokens_seen": 5298864,
      "step": 2525
    },
    {
      "epoch": 0.41275797373358347,
      "grad_norm": 8.875,
      "learning_rate": 4.9508459709738905e-05,
      "loss": 1.7885,
      "num_input_tokens_seen": 5309824,
      "step": 2530
    },
    {
      "epoch": 0.4135737009544008,
      "grad_norm": 9.6875,
      "learning_rate": 4.950652059909627e-05,
      "loss": 2.6078,
      "num_input_tokens_seen": 5317872,
      "step": 2535
    },
    {
      "epoch": 0.41438942817521823,
      "grad_norm": 9.6875,
      "learning_rate": 4.95045777092402e-05,
      "loss": 2.7714,
      "num_input_tokens_seen": 5327584,
      "step": 2540
    },
    {
      "epoch": 0.4152051553960356,
      "grad_norm": 16.25,
      "learning_rate": 4.950263104047031e-05,
      "loss": 3.3783,
      "num_input_tokens_seen": 5337824,
      "step": 2545
    },
    {
      "epoch": 0.41602088261685294,
      "grad_norm": 11.8125,
      "learning_rate": 4.9500680593086775e-05,
      "loss": 2.3399,
      "num_input_tokens_seen": 5348480,
      "step": 2550
    },
    {
      "epoch": 0.4168366098376703,
      "grad_norm": 6.46875,
      "learning_rate": 4.94987263673904e-05,
      "loss": 2.0048,
      "num_input_tokens_seen": 5358768,
      "step": 2555
    },
    {
      "epoch": 0.41765233705848764,
      "grad_norm": 6.6875,
      "learning_rate": 4.949676836368256e-05,
      "loss": 1.6839,
      "num_input_tokens_seen": 5369472,
      "step": 2560
    },
    {
      "epoch": 0.418468064279305,
      "grad_norm": 7.6875,
      "learning_rate": 4.949480658226518e-05,
      "loss": 1.8178,
      "num_input_tokens_seen": 5381344,
      "step": 2565
    },
    {
      "epoch": 0.41928379150012235,
      "grad_norm": 4.09375,
      "learning_rate": 4.949284102344082e-05,
      "loss": 3.9615,
      "num_input_tokens_seen": 5390688,
      "step": 2570
    },
    {
      "epoch": 0.4200995187209397,
      "grad_norm": 8.125,
      "learning_rate": 4.9490871687512565e-05,
      "loss": 2.7066,
      "num_input_tokens_seen": 5401984,
      "step": 2575
    },
    {
      "epoch": 0.42091524594175705,
      "grad_norm": 4.375,
      "learning_rate": 4.948889857478413e-05,
      "loss": 0.8198,
      "num_input_tokens_seen": 5411792,
      "step": 2580
    },
    {
      "epoch": 0.4217309731625744,
      "grad_norm": 8.875,
      "learning_rate": 4.948692168555978e-05,
      "loss": 1.9961,
      "num_input_tokens_seen": 5421616,
      "step": 2585
    },
    {
      "epoch": 0.4225467003833918,
      "grad_norm": 9.0625,
      "learning_rate": 4.94849410201444e-05,
      "loss": 2.1862,
      "num_input_tokens_seen": 5430976,
      "step": 2590
    },
    {
      "epoch": 0.42336242760420917,
      "grad_norm": 17.875,
      "learning_rate": 4.948295657884341e-05,
      "loss": 3.3456,
      "num_input_tokens_seen": 5442128,
      "step": 2595
    },
    {
      "epoch": 0.4241781548250265,
      "grad_norm": 13.3125,
      "learning_rate": 4.9480968361962835e-05,
      "loss": 3.414,
      "num_input_tokens_seen": 5452736,
      "step": 2600
    },
    {
      "epoch": 0.4241781548250265,
      "eval_loss": 2.5460686683654785,
      "eval_runtime": 134.9209,
      "eval_samples_per_second": 20.197,
      "eval_steps_per_second": 10.102,
      "num_input_tokens_seen": 5452736,
      "step": 2600
    },
    {
      "epoch": 0.4249938820458439,
      "grad_norm": 8.6875,
      "learning_rate": 4.9478976369809305e-05,
      "loss": 1.8162,
      "num_input_tokens_seen": 5463808,
      "step": 2605
    },
    {
      "epoch": 0.4258096092666612,
      "grad_norm": 2.421875,
      "learning_rate": 4.947698060268999e-05,
      "loss": 0.95,
      "num_input_tokens_seen": 5473520,
      "step": 2610
    },
    {
      "epoch": 0.4266253364874786,
      "grad_norm": 2.140625,
      "learning_rate": 4.9474981060912665e-05,
      "loss": 1.9411,
      "num_input_tokens_seen": 5483216,
      "step": 2615
    },
    {
      "epoch": 0.42744106370829593,
      "grad_norm": 8.0,
      "learning_rate": 4.94729777447857e-05,
      "loss": 1.9486,
      "num_input_tokens_seen": 5493136,
      "step": 2620
    },
    {
      "epoch": 0.4282567909291133,
      "grad_norm": 6.90625,
      "learning_rate": 4.947097065461801e-05,
      "loss": 3.2971,
      "num_input_tokens_seen": 5504016,
      "step": 2625
    },
    {
      "epoch": 0.42907251814993064,
      "grad_norm": 8.875,
      "learning_rate": 4.9468959790719125e-05,
      "loss": 2.7721,
      "num_input_tokens_seen": 5514800,
      "step": 2630
    },
    {
      "epoch": 0.42988824537074805,
      "grad_norm": 5.90625,
      "learning_rate": 4.9466945153399146e-05,
      "loss": 3.3625,
      "num_input_tokens_seen": 5524208,
      "step": 2635
    },
    {
      "epoch": 0.4307039725915654,
      "grad_norm": 4.4375,
      "learning_rate": 4.9464926742968755e-05,
      "loss": 1.8359,
      "num_input_tokens_seen": 5535232,
      "step": 2640
    },
    {
      "epoch": 0.43151969981238275,
      "grad_norm": 7.84375,
      "learning_rate": 4.946290455973921e-05,
      "loss": 4.118,
      "num_input_tokens_seen": 5546480,
      "step": 2645
    },
    {
      "epoch": 0.4323354270332001,
      "grad_norm": 10.5625,
      "learning_rate": 4.9460878604022365e-05,
      "loss": 4.2019,
      "num_input_tokens_seen": 5556960,
      "step": 2650
    },
    {
      "epoch": 0.43315115425401746,
      "grad_norm": 3.046875,
      "learning_rate": 4.945884887613065e-05,
      "loss": 1.1823,
      "num_input_tokens_seen": 5566976,
      "step": 2655
    },
    {
      "epoch": 0.4339668814748348,
      "grad_norm": 6.46875,
      "learning_rate": 4.9456815376377055e-05,
      "loss": 3.3858,
      "num_input_tokens_seen": 5577440,
      "step": 2660
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 2.921875,
      "learning_rate": 4.9454778105075195e-05,
      "loss": 1.5755,
      "num_input_tokens_seen": 5587824,
      "step": 2665
    },
    {
      "epoch": 0.4355983359164695,
      "grad_norm": 4.90625,
      "learning_rate": 4.945273706253924e-05,
      "loss": 2.3803,
      "num_input_tokens_seen": 5598656,
      "step": 2670
    },
    {
      "epoch": 0.43641406313728687,
      "grad_norm": 3.3125,
      "learning_rate": 4.9450692249083925e-05,
      "loss": 2.1766,
      "num_input_tokens_seen": 5608656,
      "step": 2675
    },
    {
      "epoch": 0.4372297903581043,
      "grad_norm": 7.5625,
      "learning_rate": 4.9448643665024605e-05,
      "loss": 2.0596,
      "num_input_tokens_seen": 5618912,
      "step": 2680
    },
    {
      "epoch": 0.43804551757892163,
      "grad_norm": 5.6875,
      "learning_rate": 4.944659131067719e-05,
      "loss": 1.4934,
      "num_input_tokens_seen": 5630256,
      "step": 2685
    },
    {
      "epoch": 0.438861244799739,
      "grad_norm": 5.625,
      "learning_rate": 4.944453518635818e-05,
      "loss": 1.5493,
      "num_input_tokens_seen": 5640192,
      "step": 2690
    },
    {
      "epoch": 0.43967697202055633,
      "grad_norm": 5.65625,
      "learning_rate": 4.944247529238465e-05,
      "loss": 2.7586,
      "num_input_tokens_seen": 5649744,
      "step": 2695
    },
    {
      "epoch": 0.4404926992413737,
      "grad_norm": 6.9375,
      "learning_rate": 4.944041162907427e-05,
      "loss": 3.4399,
      "num_input_tokens_seen": 5658464,
      "step": 2700
    },
    {
      "epoch": 0.44130842646219104,
      "grad_norm": 3.921875,
      "learning_rate": 4.943834419674529e-05,
      "loss": 2.9379,
      "num_input_tokens_seen": 5669632,
      "step": 2705
    },
    {
      "epoch": 0.4421241536830084,
      "grad_norm": 12.625,
      "learning_rate": 4.9436272995716506e-05,
      "loss": 2.9962,
      "num_input_tokens_seen": 5679056,
      "step": 2710
    },
    {
      "epoch": 0.44293988090382574,
      "grad_norm": 7.03125,
      "learning_rate": 4.943419802630735e-05,
      "loss": 2.9155,
      "num_input_tokens_seen": 5689376,
      "step": 2715
    },
    {
      "epoch": 0.4437556081246431,
      "grad_norm": 1.6875,
      "learning_rate": 4.94321192888378e-05,
      "loss": 2.4922,
      "num_input_tokens_seen": 5700960,
      "step": 2720
    },
    {
      "epoch": 0.4445713353454605,
      "grad_norm": 8.25,
      "learning_rate": 4.943003678362842e-05,
      "loss": 2.7537,
      "num_input_tokens_seen": 5711680,
      "step": 2725
    },
    {
      "epoch": 0.44538706256627786,
      "grad_norm": 5.90625,
      "learning_rate": 4.942795051100036e-05,
      "loss": 1.5171,
      "num_input_tokens_seen": 5721328,
      "step": 2730
    },
    {
      "epoch": 0.4462027897870952,
      "grad_norm": 6.90625,
      "learning_rate": 4.942586047127536e-05,
      "loss": 2.1331,
      "num_input_tokens_seen": 5732480,
      "step": 2735
    },
    {
      "epoch": 0.44701851700791256,
      "grad_norm": 5.46875,
      "learning_rate": 4.942376666477571e-05,
      "loss": 2.4113,
      "num_input_tokens_seen": 5742896,
      "step": 2740
    },
    {
      "epoch": 0.4478342442287299,
      "grad_norm": 8.25,
      "learning_rate": 4.9421669091824304e-05,
      "loss": 3.3349,
      "num_input_tokens_seen": 5752304,
      "step": 2745
    },
    {
      "epoch": 0.44864997144954727,
      "grad_norm": 12.0625,
      "learning_rate": 4.9419567752744634e-05,
      "loss": 3.7092,
      "num_input_tokens_seen": 5764032,
      "step": 2750
    },
    {
      "epoch": 0.4494656986703646,
      "grad_norm": 11.375,
      "learning_rate": 4.941746264786074e-05,
      "loss": 3.2194,
      "num_input_tokens_seen": 5774960,
      "step": 2755
    },
    {
      "epoch": 0.450281425891182,
      "grad_norm": 3.953125,
      "learning_rate": 4.9415353777497254e-05,
      "loss": 1.6109,
      "num_input_tokens_seen": 5786256,
      "step": 2760
    },
    {
      "epoch": 0.4510971531119993,
      "grad_norm": 9.4375,
      "learning_rate": 4.9413241141979394e-05,
      "loss": 3.1308,
      "num_input_tokens_seen": 5797536,
      "step": 2765
    },
    {
      "epoch": 0.4519128803328167,
      "grad_norm": 5.34375,
      "learning_rate": 4.9411124741632956e-05,
      "loss": 1.163,
      "num_input_tokens_seen": 5807824,
      "step": 2770
    },
    {
      "epoch": 0.4527286075536341,
      "grad_norm": 6.0,
      "learning_rate": 4.940900457678431e-05,
      "loss": 2.4222,
      "num_input_tokens_seen": 5819072,
      "step": 2775
    },
    {
      "epoch": 0.45354433477445144,
      "grad_norm": 3.359375,
      "learning_rate": 4.9406880647760425e-05,
      "loss": 2.2993,
      "num_input_tokens_seen": 5830272,
      "step": 2780
    },
    {
      "epoch": 0.4543600619952688,
      "grad_norm": 5.34375,
      "learning_rate": 4.9404752954888824e-05,
      "loss": 2.4058,
      "num_input_tokens_seen": 5841280,
      "step": 2785
    },
    {
      "epoch": 0.45517578921608615,
      "grad_norm": 1.0078125,
      "learning_rate": 4.940262149849762e-05,
      "loss": 2.1057,
      "num_input_tokens_seen": 5850640,
      "step": 2790
    },
    {
      "epoch": 0.4559915164369035,
      "grad_norm": 9.6875,
      "learning_rate": 4.9400486278915526e-05,
      "loss": 2.2462,
      "num_input_tokens_seen": 5862704,
      "step": 2795
    },
    {
      "epoch": 0.45680724365772085,
      "grad_norm": 7.25,
      "learning_rate": 4.939834729647181e-05,
      "loss": 4.0924,
      "num_input_tokens_seen": 5872752,
      "step": 2800
    },
    {
      "epoch": 0.45680724365772085,
      "eval_loss": 2.5363454818725586,
      "eval_runtime": 134.7289,
      "eval_samples_per_second": 20.226,
      "eval_steps_per_second": 10.117,
      "num_input_tokens_seen": 5872752,
      "step": 2800
    },
    {
      "epoch": 0.4576229708785382,
      "grad_norm": 11.1875,
      "learning_rate": 4.9396204551496326e-05,
      "loss": 3.4319,
      "num_input_tokens_seen": 5882464,
      "step": 2805
    },
    {
      "epoch": 0.45843869809935556,
      "grad_norm": 5.53125,
      "learning_rate": 4.939405804431952e-05,
      "loss": 1.9734,
      "num_input_tokens_seen": 5892048,
      "step": 2810
    },
    {
      "epoch": 0.4592544253201729,
      "grad_norm": 9.125,
      "learning_rate": 4.9391907775272414e-05,
      "loss": 1.755,
      "num_input_tokens_seen": 5901504,
      "step": 2815
    },
    {
      "epoch": 0.4600701525409903,
      "grad_norm": 6.25,
      "learning_rate": 4.9389753744686604e-05,
      "loss": 2.7649,
      "num_input_tokens_seen": 5912832,
      "step": 2820
    },
    {
      "epoch": 0.46088587976180767,
      "grad_norm": 10.8125,
      "learning_rate": 4.938759595289426e-05,
      "loss": 1.9929,
      "num_input_tokens_seen": 5923392,
      "step": 2825
    },
    {
      "epoch": 0.461701606982625,
      "grad_norm": 11.0,
      "learning_rate": 4.938543440022815e-05,
      "loss": 4.088,
      "num_input_tokens_seen": 5933968,
      "step": 2830
    },
    {
      "epoch": 0.4625173342034424,
      "grad_norm": 8.125,
      "learning_rate": 4.938326908702161e-05,
      "loss": 1.915,
      "num_input_tokens_seen": 5943872,
      "step": 2835
    },
    {
      "epoch": 0.46333306142425973,
      "grad_norm": 7.96875,
      "learning_rate": 4.9381100013608554e-05,
      "loss": 2.8769,
      "num_input_tokens_seen": 5954752,
      "step": 2840
    },
    {
      "epoch": 0.4641487886450771,
      "grad_norm": 11.5625,
      "learning_rate": 4.9378927180323485e-05,
      "loss": 3.1543,
      "num_input_tokens_seen": 5965280,
      "step": 2845
    },
    {
      "epoch": 0.46496451586589443,
      "grad_norm": 10.1875,
      "learning_rate": 4.937675058750148e-05,
      "loss": 2.1452,
      "num_input_tokens_seen": 5975792,
      "step": 2850
    },
    {
      "epoch": 0.4657802430867118,
      "grad_norm": 3.59375,
      "learning_rate": 4.937457023547819e-05,
      "loss": 1.2275,
      "num_input_tokens_seen": 5986480,
      "step": 2855
    },
    {
      "epoch": 0.46659597030752914,
      "grad_norm": 9.5625,
      "learning_rate": 4.9372386124589876e-05,
      "loss": 3.0902,
      "num_input_tokens_seen": 5997264,
      "step": 2860
    },
    {
      "epoch": 0.46741169752834655,
      "grad_norm": 4.5625,
      "learning_rate": 4.937019825517333e-05,
      "loss": 1.1095,
      "num_input_tokens_seen": 6006976,
      "step": 2865
    },
    {
      "epoch": 0.4682274247491639,
      "grad_norm": 5.5625,
      "learning_rate": 4.9368006627565954e-05,
      "loss": 3.2964,
      "num_input_tokens_seen": 6016912,
      "step": 2870
    },
    {
      "epoch": 0.46904315196998125,
      "grad_norm": 6.84375,
      "learning_rate": 4.936581124210573e-05,
      "loss": 3.4153,
      "num_input_tokens_seen": 6027248,
      "step": 2875
    },
    {
      "epoch": 0.4698588791907986,
      "grad_norm": 11.1875,
      "learning_rate": 4.9363612099131216e-05,
      "loss": 3.9821,
      "num_input_tokens_seen": 6037616,
      "step": 2880
    },
    {
      "epoch": 0.47067460641161596,
      "grad_norm": 4.4375,
      "learning_rate": 4.936140919898155e-05,
      "loss": 1.3584,
      "num_input_tokens_seen": 6047600,
      "step": 2885
    },
    {
      "epoch": 0.4714903336324333,
      "grad_norm": 6.03125,
      "learning_rate": 4.9359202541996426e-05,
      "loss": 1.8138,
      "num_input_tokens_seen": 6056960,
      "step": 2890
    },
    {
      "epoch": 0.47230606085325066,
      "grad_norm": 2.203125,
      "learning_rate": 4.935699212851616e-05,
      "loss": 2.701,
      "num_input_tokens_seen": 6067408,
      "step": 2895
    },
    {
      "epoch": 0.473121788074068,
      "grad_norm": 4.25,
      "learning_rate": 4.935477795888162e-05,
      "loss": 2.3525,
      "num_input_tokens_seen": 6078240,
      "step": 2900
    },
    {
      "epoch": 0.47393751529488537,
      "grad_norm": 8.0,
      "learning_rate": 4.935256003343426e-05,
      "loss": 2.2479,
      "num_input_tokens_seen": 6087760,
      "step": 2905
    },
    {
      "epoch": 0.4747532425157027,
      "grad_norm": 1.9453125,
      "learning_rate": 4.93503383525161e-05,
      "loss": 2.5845,
      "num_input_tokens_seen": 6098512,
      "step": 2910
    },
    {
      "epoch": 0.47556896973652013,
      "grad_norm": 13.125,
      "learning_rate": 4.934811291646977e-05,
      "loss": 4.7693,
      "num_input_tokens_seen": 6108592,
      "step": 2915
    },
    {
      "epoch": 0.4763846969573375,
      "grad_norm": 6.25,
      "learning_rate": 4.934588372563845e-05,
      "loss": 1.5896,
      "num_input_tokens_seen": 6118816,
      "step": 2920
    },
    {
      "epoch": 0.47720042417815484,
      "grad_norm": 8.875,
      "learning_rate": 4.93436507803659e-05,
      "loss": 1.9661,
      "num_input_tokens_seen": 6129136,
      "step": 2925
    },
    {
      "epoch": 0.4780161513989722,
      "grad_norm": 10.8125,
      "learning_rate": 4.934141408099649e-05,
      "loss": 2.4596,
      "num_input_tokens_seen": 6138768,
      "step": 2930
    },
    {
      "epoch": 0.47883187861978954,
      "grad_norm": 9.4375,
      "learning_rate": 4.9339173627875135e-05,
      "loss": 3.1813,
      "num_input_tokens_seen": 6149824,
      "step": 2935
    },
    {
      "epoch": 0.4796476058406069,
      "grad_norm": 8.625,
      "learning_rate": 4.9336929421347335e-05,
      "loss": 3.1056,
      "num_input_tokens_seen": 6159184,
      "step": 2940
    },
    {
      "epoch": 0.48046333306142425,
      "grad_norm": 4.8125,
      "learning_rate": 4.933468146175918e-05,
      "loss": 1.3328,
      "num_input_tokens_seen": 6168000,
      "step": 2945
    },
    {
      "epoch": 0.4812790602822416,
      "grad_norm": 7.375,
      "learning_rate": 4.933242974945734e-05,
      "loss": 1.4682,
      "num_input_tokens_seen": 6178512,
      "step": 2950
    },
    {
      "epoch": 0.48209478750305895,
      "grad_norm": 4.25,
      "learning_rate": 4.933017428478906e-05,
      "loss": 1.3646,
      "num_input_tokens_seen": 6189360,
      "step": 2955
    },
    {
      "epoch": 0.48291051472387636,
      "grad_norm": 6.5625,
      "learning_rate": 4.932791506810214e-05,
      "loss": 2.1511,
      "num_input_tokens_seen": 6199632,
      "step": 2960
    },
    {
      "epoch": 0.4837262419446937,
      "grad_norm": 8.125,
      "learning_rate": 4.932565209974499e-05,
      "loss": 2.5074,
      "num_input_tokens_seen": 6211280,
      "step": 2965
    },
    {
      "epoch": 0.48454196916551107,
      "grad_norm": 8.4375,
      "learning_rate": 4.93233853800666e-05,
      "loss": 3.1926,
      "num_input_tokens_seen": 6221328,
      "step": 2970
    },
    {
      "epoch": 0.4853576963863284,
      "grad_norm": 10.1875,
      "learning_rate": 4.932111490941651e-05,
      "loss": 2.8718,
      "num_input_tokens_seen": 6231648,
      "step": 2975
    },
    {
      "epoch": 0.48617342360714577,
      "grad_norm": 4.875,
      "learning_rate": 4.9318840688144876e-05,
      "loss": 2.9625,
      "num_input_tokens_seen": 6243296,
      "step": 2980
    },
    {
      "epoch": 0.4869891508279631,
      "grad_norm": 5.5,
      "learning_rate": 4.9316562716602387e-05,
      "loss": 2.5188,
      "num_input_tokens_seen": 6254080,
      "step": 2985
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 8.0625,
      "learning_rate": 4.9314280995140346e-05,
      "loss": 2.2166,
      "num_input_tokens_seen": 6263680,
      "step": 2990
    },
    {
      "epoch": 0.48862060526959783,
      "grad_norm": 6.15625,
      "learning_rate": 4.931199552411063e-05,
      "loss": 2.1349,
      "num_input_tokens_seen": 6274560,
      "step": 2995
    },
    {
      "epoch": 0.4894363324904152,
      "grad_norm": 9.0,
      "learning_rate": 4.930970630386568e-05,
      "loss": 1.594,
      "num_input_tokens_seen": 6285232,
      "step": 3000
    },
    {
      "epoch": 0.4894363324904152,
      "eval_loss": 2.556004285812378,
      "eval_runtime": 134.74,
      "eval_samples_per_second": 20.224,
      "eval_steps_per_second": 10.116,
      "num_input_tokens_seen": 6285232,
      "step": 3000
    },
    {
      "epoch": 0.4902520597112326,
      "grad_norm": 4.5625,
      "learning_rate": 4.9307413334758524e-05,
      "loss": 3.6524,
      "num_input_tokens_seen": 6296272,
      "step": 3005
    },
    {
      "epoch": 0.49106778693204994,
      "grad_norm": 7.0625,
      "learning_rate": 4.930511661714276e-05,
      "loss": 2.1055,
      "num_input_tokens_seen": 6306720,
      "step": 3010
    },
    {
      "epoch": 0.4918835141528673,
      "grad_norm": 4.5,
      "learning_rate": 4.9302816151372576e-05,
      "loss": 2.7039,
      "num_input_tokens_seen": 6316480,
      "step": 3015
    },
    {
      "epoch": 0.49269924137368465,
      "grad_norm": 5.28125,
      "learning_rate": 4.930051193780274e-05,
      "loss": 2.569,
      "num_input_tokens_seen": 6326048,
      "step": 3020
    },
    {
      "epoch": 0.493514968594502,
      "grad_norm": 9.5,
      "learning_rate": 4.929820397678858e-05,
      "loss": 3.0794,
      "num_input_tokens_seen": 6335760,
      "step": 3025
    },
    {
      "epoch": 0.49433069581531935,
      "grad_norm": 10.0,
      "learning_rate": 4.9295892268686015e-05,
      "loss": 2.4776,
      "num_input_tokens_seen": 6345984,
      "step": 3030
    },
    {
      "epoch": 0.4951464230361367,
      "grad_norm": 0.546875,
      "learning_rate": 4.9293576813851536e-05,
      "loss": 1.4657,
      "num_input_tokens_seen": 6356448,
      "step": 3035
    },
    {
      "epoch": 0.49596215025695406,
      "grad_norm": 7.8125,
      "learning_rate": 4.929125761264223e-05,
      "loss": 3.0922,
      "num_input_tokens_seen": 6367440,
      "step": 3040
    },
    {
      "epoch": 0.4967778774777714,
      "grad_norm": 11.3125,
      "learning_rate": 4.928893466541573e-05,
      "loss": 3.0937,
      "num_input_tokens_seen": 6378768,
      "step": 3045
    },
    {
      "epoch": 0.49759360469858877,
      "grad_norm": 15.5625,
      "learning_rate": 4.928660797253027e-05,
      "loss": 2.7171,
      "num_input_tokens_seen": 6390032,
      "step": 3050
    },
    {
      "epoch": 0.4984093319194062,
      "grad_norm": 17.375,
      "learning_rate": 4.928427753434467e-05,
      "loss": 2.3117,
      "num_input_tokens_seen": 6400784,
      "step": 3055
    },
    {
      "epoch": 0.4992250591402235,
      "grad_norm": 10.75,
      "learning_rate": 4.9281943351218286e-05,
      "loss": 2.8343,
      "num_input_tokens_seen": 6410320,
      "step": 3060
    },
    {
      "epoch": 0.5000407863610409,
      "grad_norm": 7.59375,
      "learning_rate": 4.9279605423511095e-05,
      "loss": 3.4688,
      "num_input_tokens_seen": 6420368,
      "step": 3065
    },
    {
      "epoch": 0.5008565135818582,
      "grad_norm": 8.8125,
      "learning_rate": 4.927726375158363e-05,
      "loss": 3.1149,
      "num_input_tokens_seen": 6430912,
      "step": 3070
    },
    {
      "epoch": 0.5016722408026756,
      "grad_norm": 10.25,
      "learning_rate": 4.9274918335797004e-05,
      "loss": 2.3083,
      "num_input_tokens_seen": 6440720,
      "step": 3075
    },
    {
      "epoch": 0.5024879680234929,
      "grad_norm": 11.75,
      "learning_rate": 4.927256917651292e-05,
      "loss": 1.1752,
      "num_input_tokens_seen": 6450800,
      "step": 3080
    },
    {
      "epoch": 0.5033036952443103,
      "grad_norm": 8.4375,
      "learning_rate": 4.927021627409364e-05,
      "loss": 2.1302,
      "num_input_tokens_seen": 6460752,
      "step": 3085
    },
    {
      "epoch": 0.5041194224651276,
      "grad_norm": 3.671875,
      "learning_rate": 4.9267859628902005e-05,
      "loss": 2.0868,
      "num_input_tokens_seen": 6470208,
      "step": 3090
    },
    {
      "epoch": 0.504935149685945,
      "grad_norm": 8.875,
      "learning_rate": 4.9265499241301454e-05,
      "loss": 3.0766,
      "num_input_tokens_seen": 6480768,
      "step": 3095
    },
    {
      "epoch": 0.5057508769067623,
      "grad_norm": 6.40625,
      "learning_rate": 4.926313511165598e-05,
      "loss": 1.9637,
      "num_input_tokens_seen": 6490928,
      "step": 3100
    },
    {
      "epoch": 0.5065666041275797,
      "grad_norm": 9.5,
      "learning_rate": 4.926076724033016e-05,
      "loss": 2.5437,
      "num_input_tokens_seen": 6501856,
      "step": 3105
    },
    {
      "epoch": 0.507382331348397,
      "grad_norm": 3.5625,
      "learning_rate": 4.9258395627689146e-05,
      "loss": 2.0218,
      "num_input_tokens_seen": 6513232,
      "step": 3110
    },
    {
      "epoch": 0.5081980585692144,
      "grad_norm": 5.53125,
      "learning_rate": 4.925602027409868e-05,
      "loss": 2.3945,
      "num_input_tokens_seen": 6523632,
      "step": 3115
    },
    {
      "epoch": 0.5090137857900318,
      "grad_norm": 3.921875,
      "learning_rate": 4.925364117992507e-05,
      "loss": 2.8361,
      "num_input_tokens_seen": 6534176,
      "step": 3120
    },
    {
      "epoch": 0.5098295130108492,
      "grad_norm": 9.5,
      "learning_rate": 4.92512583455352e-05,
      "loss": 2.9379,
      "num_input_tokens_seen": 6544912,
      "step": 3125
    },
    {
      "epoch": 0.5106452402316666,
      "grad_norm": 10.375,
      "learning_rate": 4.9248871771296536e-05,
      "loss": 3.0284,
      "num_input_tokens_seen": 6554656,
      "step": 3130
    },
    {
      "epoch": 0.5114609674524839,
      "grad_norm": 9.5,
      "learning_rate": 4.924648145757711e-05,
      "loss": 3.8549,
      "num_input_tokens_seen": 6564032,
      "step": 3135
    },
    {
      "epoch": 0.5122766946733013,
      "grad_norm": 12.1875,
      "learning_rate": 4.924408740474554e-05,
      "loss": 2.1427,
      "num_input_tokens_seen": 6574544,
      "step": 3140
    },
    {
      "epoch": 0.5130924218941186,
      "grad_norm": 6.15625,
      "learning_rate": 4.924168961317103e-05,
      "loss": 2.6508,
      "num_input_tokens_seen": 6584448,
      "step": 3145
    },
    {
      "epoch": 0.513908149114936,
      "grad_norm": 6.65625,
      "learning_rate": 4.9239288083223334e-05,
      "loss": 2.5556,
      "num_input_tokens_seen": 6593552,
      "step": 3150
    },
    {
      "epoch": 0.5147238763357533,
      "grad_norm": 5.09375,
      "learning_rate": 4.9236882815272803e-05,
      "loss": 1.9787,
      "num_input_tokens_seen": 6602464,
      "step": 3155
    },
    {
      "epoch": 0.5155396035565707,
      "grad_norm": 10.375,
      "learning_rate": 4.9234473809690365e-05,
      "loss": 2.7816,
      "num_input_tokens_seen": 6612000,
      "step": 3160
    },
    {
      "epoch": 0.516355330777388,
      "grad_norm": 3.515625,
      "learning_rate": 4.923206106684752e-05,
      "loss": 2.1682,
      "num_input_tokens_seen": 6623184,
      "step": 3165
    },
    {
      "epoch": 0.5171710579982054,
      "grad_norm": 9.75,
      "learning_rate": 4.922964458711634e-05,
      "loss": 2.6035,
      "num_input_tokens_seen": 6634384,
      "step": 3170
    },
    {
      "epoch": 0.5179867852190227,
      "grad_norm": 3.109375,
      "learning_rate": 4.9227224370869474e-05,
      "loss": 2.0977,
      "num_input_tokens_seen": 6644624,
      "step": 3175
    },
    {
      "epoch": 0.5188025124398401,
      "grad_norm": 3.421875,
      "learning_rate": 4.9224800418480155e-05,
      "loss": 1.5099,
      "num_input_tokens_seen": 6654848,
      "step": 3180
    },
    {
      "epoch": 0.5196182396606575,
      "grad_norm": 7.6875,
      "learning_rate": 4.9222372730322176e-05,
      "loss": 3.129,
      "num_input_tokens_seen": 6665904,
      "step": 3185
    },
    {
      "epoch": 0.5204339668814748,
      "grad_norm": 6.0,
      "learning_rate": 4.921994130676993e-05,
      "loss": 2.2162,
      "num_input_tokens_seen": 6677872,
      "step": 3190
    },
    {
      "epoch": 0.5212496941022922,
      "grad_norm": 7.90625,
      "learning_rate": 4.9217506148198366e-05,
      "loss": 1.5603,
      "num_input_tokens_seen": 6688640,
      "step": 3195
    },
    {
      "epoch": 0.5220654213231095,
      "grad_norm": 9.25,
      "learning_rate": 4.921506725498302e-05,
      "loss": 2.1049,
      "num_input_tokens_seen": 6699264,
      "step": 3200
    },
    {
      "epoch": 0.5220654213231095,
      "eval_loss": 2.5613045692443848,
      "eval_runtime": 134.9035,
      "eval_samples_per_second": 20.2,
      "eval_steps_per_second": 10.104,
      "num_input_tokens_seen": 6699264,
      "step": 3200
    },
    {
      "epoch": 0.5228811485439269,
      "grad_norm": 6.625,
      "learning_rate": 4.9212624627499994e-05,
      "loss": 2.3308,
      "num_input_tokens_seen": 6710016,
      "step": 3205
    },
    {
      "epoch": 0.5236968757647442,
      "grad_norm": 6.75,
      "learning_rate": 4.921017826612597e-05,
      "loss": 1.6122,
      "num_input_tokens_seen": 6720304,
      "step": 3210
    },
    {
      "epoch": 0.5245126029855617,
      "grad_norm": 7.21875,
      "learning_rate": 4.9207728171238223e-05,
      "loss": 2.8545,
      "num_input_tokens_seen": 6730784,
      "step": 3215
    },
    {
      "epoch": 0.525328330206379,
      "grad_norm": 5.46875,
      "learning_rate": 4.920527434321458e-05,
      "loss": 1.2659,
      "num_input_tokens_seen": 6740208,
      "step": 3220
    },
    {
      "epoch": 0.5261440574271964,
      "grad_norm": 8.5,
      "learning_rate": 4.920281678243345e-05,
      "loss": 2.263,
      "num_input_tokens_seen": 6750864,
      "step": 3225
    },
    {
      "epoch": 0.5269597846480137,
      "grad_norm": 10.125,
      "learning_rate": 4.920035548927381e-05,
      "loss": 2.1463,
      "num_input_tokens_seen": 6761072,
      "step": 3230
    },
    {
      "epoch": 0.5277755118688311,
      "grad_norm": 9.1875,
      "learning_rate": 4.919789046411525e-05,
      "loss": 2.7964,
      "num_input_tokens_seen": 6770928,
      "step": 3235
    },
    {
      "epoch": 0.5285912390896484,
      "grad_norm": 3.71875,
      "learning_rate": 4.919542170733787e-05,
      "loss": 2.9093,
      "num_input_tokens_seen": 6781936,
      "step": 3240
    },
    {
      "epoch": 0.5294069663104658,
      "grad_norm": 8.6875,
      "learning_rate": 4.919294921932242e-05,
      "loss": 2.3789,
      "num_input_tokens_seen": 6791632,
      "step": 3245
    },
    {
      "epoch": 0.5302226935312832,
      "grad_norm": 10.9375,
      "learning_rate": 4.919047300045016e-05,
      "loss": 4.3352,
      "num_input_tokens_seen": 6802000,
      "step": 3250
    },
    {
      "epoch": 0.5310384207521005,
      "grad_norm": 10.5,
      "learning_rate": 4.918799305110299e-05,
      "loss": 1.2275,
      "num_input_tokens_seen": 6812976,
      "step": 3255
    },
    {
      "epoch": 0.5318541479729179,
      "grad_norm": 4.28125,
      "learning_rate": 4.918550937166331e-05,
      "loss": 2.0981,
      "num_input_tokens_seen": 6822992,
      "step": 3260
    },
    {
      "epoch": 0.5326698751937352,
      "grad_norm": 4.625,
      "learning_rate": 4.918302196251415e-05,
      "loss": 1.9569,
      "num_input_tokens_seen": 6833680,
      "step": 3265
    },
    {
      "epoch": 0.5334856024145526,
      "grad_norm": 4.53125,
      "learning_rate": 4.91805308240391e-05,
      "loss": 2.5389,
      "num_input_tokens_seen": 6843968,
      "step": 3270
    },
    {
      "epoch": 0.5343013296353699,
      "grad_norm": 8.375,
      "learning_rate": 4.9178035956622326e-05,
      "loss": 2.4846,
      "num_input_tokens_seen": 6855040,
      "step": 3275
    },
    {
      "epoch": 0.5351170568561873,
      "grad_norm": 8.625,
      "learning_rate": 4.917553736064857e-05,
      "loss": 3.2353,
      "num_input_tokens_seen": 6866688,
      "step": 3280
    },
    {
      "epoch": 0.5359327840770046,
      "grad_norm": 6.4375,
      "learning_rate": 4.917303503650314e-05,
      "loss": 2.669,
      "num_input_tokens_seen": 6877808,
      "step": 3285
    },
    {
      "epoch": 0.536748511297822,
      "grad_norm": 7.78125,
      "learning_rate": 4.917052898457194e-05,
      "loss": 1.5405,
      "num_input_tokens_seen": 6887824,
      "step": 3290
    },
    {
      "epoch": 0.5375642385186393,
      "grad_norm": 6.96875,
      "learning_rate": 4.916801920524141e-05,
      "loss": 2.6606,
      "num_input_tokens_seen": 6897936,
      "step": 3295
    },
    {
      "epoch": 0.5383799657394567,
      "grad_norm": 5.09375,
      "learning_rate": 4.916550569889862e-05,
      "loss": 1.7135,
      "num_input_tokens_seen": 6908448,
      "step": 3300
    },
    {
      "epoch": 0.539195692960274,
      "grad_norm": 7.09375,
      "learning_rate": 4.916298846593116e-05,
      "loss": 1.8162,
      "num_input_tokens_seen": 6919680,
      "step": 3305
    },
    {
      "epoch": 0.5400114201810915,
      "grad_norm": 6.9375,
      "learning_rate": 4.916046750672722e-05,
      "loss": 3.4111,
      "num_input_tokens_seen": 6930528,
      "step": 3310
    },
    {
      "epoch": 0.5408271474019088,
      "grad_norm": 10.625,
      "learning_rate": 4.915794282167559e-05,
      "loss": 4.0599,
      "num_input_tokens_seen": 6941360,
      "step": 3315
    },
    {
      "epoch": 0.5416428746227262,
      "grad_norm": 5.8125,
      "learning_rate": 4.915541441116558e-05,
      "loss": 3.2968,
      "num_input_tokens_seen": 6951440,
      "step": 3320
    },
    {
      "epoch": 0.5424586018435436,
      "grad_norm": 3.890625,
      "learning_rate": 4.915288227558711e-05,
      "loss": 3.3474,
      "num_input_tokens_seen": 6961856,
      "step": 3325
    },
    {
      "epoch": 0.5432743290643609,
      "grad_norm": 8.75,
      "learning_rate": 4.915034641533066e-05,
      "loss": 3.1139,
      "num_input_tokens_seen": 6971808,
      "step": 3330
    },
    {
      "epoch": 0.5440900562851783,
      "grad_norm": 6.625,
      "learning_rate": 4.914780683078731e-05,
      "loss": 1.8606,
      "num_input_tokens_seen": 6981968,
      "step": 3335
    },
    {
      "epoch": 0.5449057835059956,
      "grad_norm": 5.71875,
      "learning_rate": 4.9145263522348695e-05,
      "loss": 2.3299,
      "num_input_tokens_seen": 6992224,
      "step": 3340
    },
    {
      "epoch": 0.545721510726813,
      "grad_norm": 9.25,
      "learning_rate": 4.9142716490407e-05,
      "loss": 1.5681,
      "num_input_tokens_seen": 7002768,
      "step": 3345
    },
    {
      "epoch": 0.5465372379476303,
      "grad_norm": 9.8125,
      "learning_rate": 4.914016573535504e-05,
      "loss": 2.1197,
      "num_input_tokens_seen": 7012656,
      "step": 3350
    },
    {
      "epoch": 0.5473529651684477,
      "grad_norm": 6.59375,
      "learning_rate": 4.9137611257586154e-05,
      "loss": 3.2067,
      "num_input_tokens_seen": 7022368,
      "step": 3355
    },
    {
      "epoch": 0.548168692389265,
      "grad_norm": 8.75,
      "learning_rate": 4.9135053057494274e-05,
      "loss": 1.8967,
      "num_input_tokens_seen": 7033024,
      "step": 3360
    },
    {
      "epoch": 0.5489844196100824,
      "grad_norm": 7.28125,
      "learning_rate": 4.913249113547392e-05,
      "loss": 2.9176,
      "num_input_tokens_seen": 7043536,
      "step": 3365
    },
    {
      "epoch": 0.5498001468308997,
      "grad_norm": 6.875,
      "learning_rate": 4.912992549192016e-05,
      "loss": 2.0323,
      "num_input_tokens_seen": 7053824,
      "step": 3370
    },
    {
      "epoch": 0.5506158740517171,
      "grad_norm": 9.125,
      "learning_rate": 4.9127356127228665e-05,
      "loss": 2.8353,
      "num_input_tokens_seen": 7064464,
      "step": 3375
    },
    {
      "epoch": 0.5514316012725344,
      "grad_norm": 1.578125,
      "learning_rate": 4.912478304179564e-05,
      "loss": 1.5215,
      "num_input_tokens_seen": 7075408,
      "step": 3380
    },
    {
      "epoch": 0.5522473284933518,
      "grad_norm": 4.75,
      "learning_rate": 4.9122206236017896e-05,
      "loss": 3.1836,
      "num_input_tokens_seen": 7087296,
      "step": 3385
    },
    {
      "epoch": 0.5530630557141691,
      "grad_norm": 8.125,
      "learning_rate": 4.911962571029282e-05,
      "loss": 1.6045,
      "num_input_tokens_seen": 7098240,
      "step": 3390
    },
    {
      "epoch": 0.5538787829349865,
      "grad_norm": 10.125,
      "learning_rate": 4.9117041465018353e-05,
      "loss": 3.3567,
      "num_input_tokens_seen": 7108736,
      "step": 3395
    },
    {
      "epoch": 0.5546945101558038,
      "grad_norm": 4.5625,
      "learning_rate": 4.911445350059302e-05,
      "loss": 2.4132,
      "num_input_tokens_seen": 7118336,
      "step": 3400
    },
    {
      "epoch": 0.5546945101558038,
      "eval_loss": 2.552403450012207,
      "eval_runtime": 135.0245,
      "eval_samples_per_second": 20.182,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 7118336,
      "step": 3400
    },
    {
      "epoch": 0.5555102373766213,
      "grad_norm": 8.375,
      "learning_rate": 4.9111861817415905e-05,
      "loss": 1.8922,
      "num_input_tokens_seen": 7128304,
      "step": 3405
    },
    {
      "epoch": 0.5563259645974387,
      "grad_norm": 4.9375,
      "learning_rate": 4.91092664158867e-05,
      "loss": 2.5185,
      "num_input_tokens_seen": 7137456,
      "step": 3410
    },
    {
      "epoch": 0.557141691818256,
      "grad_norm": 4.9375,
      "learning_rate": 4.910666729640563e-05,
      "loss": 2.2434,
      "num_input_tokens_seen": 7147712,
      "step": 3415
    },
    {
      "epoch": 0.5579574190390734,
      "grad_norm": 1.828125,
      "learning_rate": 4.910406445937353e-05,
      "loss": 2.4215,
      "num_input_tokens_seen": 7158608,
      "step": 3420
    },
    {
      "epoch": 0.5587731462598907,
      "grad_norm": 3.0625,
      "learning_rate": 4.9101457905191774e-05,
      "loss": 1.5247,
      "num_input_tokens_seen": 7168880,
      "step": 3425
    },
    {
      "epoch": 0.5595888734807081,
      "grad_norm": 7.09375,
      "learning_rate": 4.909884763426233e-05,
      "loss": 2.3167,
      "num_input_tokens_seen": 7178576,
      "step": 3430
    },
    {
      "epoch": 0.5604046007015254,
      "grad_norm": 5.40625,
      "learning_rate": 4.9096233646987736e-05,
      "loss": 2.1785,
      "num_input_tokens_seen": 7189488,
      "step": 3435
    },
    {
      "epoch": 0.5612203279223428,
      "grad_norm": 2.25,
      "learning_rate": 4.9093615943771104e-05,
      "loss": 1.6636,
      "num_input_tokens_seen": 7199584,
      "step": 3440
    },
    {
      "epoch": 0.5620360551431601,
      "grad_norm": 5.5625,
      "learning_rate": 4.909099452501611e-05,
      "loss": 2.2311,
      "num_input_tokens_seen": 7208400,
      "step": 3445
    },
    {
      "epoch": 0.5628517823639775,
      "grad_norm": 9.75,
      "learning_rate": 4.908836939112702e-05,
      "loss": 2.5834,
      "num_input_tokens_seen": 7218704,
      "step": 3450
    },
    {
      "epoch": 0.5636675095847948,
      "grad_norm": 0.15625,
      "learning_rate": 4.908574054250865e-05,
      "loss": 2.2339,
      "num_input_tokens_seen": 7230304,
      "step": 3455
    },
    {
      "epoch": 0.5644832368056122,
      "grad_norm": 7.84375,
      "learning_rate": 4.9083107979566414e-05,
      "loss": 1.4289,
      "num_input_tokens_seen": 7240912,
      "step": 3460
    },
    {
      "epoch": 0.5652989640264295,
      "grad_norm": 0.205078125,
      "learning_rate": 4.908047170270628e-05,
      "loss": 0.7788,
      "num_input_tokens_seen": 7251872,
      "step": 3465
    },
    {
      "epoch": 0.5661146912472469,
      "grad_norm": 7.84375,
      "learning_rate": 4.9077831712334784e-05,
      "loss": 3.3725,
      "num_input_tokens_seen": 7262336,
      "step": 3470
    },
    {
      "epoch": 0.5669304184680642,
      "grad_norm": 7.0,
      "learning_rate": 4.907518800885907e-05,
      "loss": 2.1348,
      "num_input_tokens_seen": 7273344,
      "step": 3475
    },
    {
      "epoch": 0.5677461456888816,
      "grad_norm": 6.25,
      "learning_rate": 4.907254059268681e-05,
      "loss": 2.9166,
      "num_input_tokens_seen": 7282848,
      "step": 3480
    },
    {
      "epoch": 0.568561872909699,
      "grad_norm": 9.375,
      "learning_rate": 4.906988946422628e-05,
      "loss": 2.8431,
      "num_input_tokens_seen": 7293424,
      "step": 3485
    },
    {
      "epoch": 0.5693776001305163,
      "grad_norm": 7.71875,
      "learning_rate": 4.9067234623886315e-05,
      "loss": 1.2115,
      "num_input_tokens_seen": 7304192,
      "step": 3490
    },
    {
      "epoch": 0.5701933273513338,
      "grad_norm": 6.6875,
      "learning_rate": 4.9064576072076316e-05,
      "loss": 3.5245,
      "num_input_tokens_seen": 7315488,
      "step": 3495
    },
    {
      "epoch": 0.5710090545721511,
      "grad_norm": 3.46875,
      "learning_rate": 4.906191380920628e-05,
      "loss": 1.942,
      "num_input_tokens_seen": 7325840,
      "step": 3500
    },
    {
      "epoch": 0.5718247817929685,
      "grad_norm": 11.4375,
      "learning_rate": 4.905924783568675e-05,
      "loss": 2.1367,
      "num_input_tokens_seen": 7334864,
      "step": 3505
    },
    {
      "epoch": 0.5726405090137858,
      "grad_norm": 7.78125,
      "learning_rate": 4.905657815192886e-05,
      "loss": 4.048,
      "num_input_tokens_seen": 7345472,
      "step": 3510
    },
    {
      "epoch": 0.5734562362346032,
      "grad_norm": 0.1796875,
      "learning_rate": 4.90539047583443e-05,
      "loss": 1.5848,
      "num_input_tokens_seen": 7356752,
      "step": 3515
    },
    {
      "epoch": 0.5742719634554205,
      "grad_norm": 11.3125,
      "learning_rate": 4.905122765534534e-05,
      "loss": 2.2789,
      "num_input_tokens_seen": 7366736,
      "step": 3520
    },
    {
      "epoch": 0.5750876906762379,
      "grad_norm": 9.75,
      "learning_rate": 4.9048546843344846e-05,
      "loss": 2.2609,
      "num_input_tokens_seen": 7376416,
      "step": 3525
    },
    {
      "epoch": 0.5759034178970552,
      "grad_norm": 7.4375,
      "learning_rate": 4.9045862322756206e-05,
      "loss": 2.2595,
      "num_input_tokens_seen": 7387456,
      "step": 3530
    },
    {
      "epoch": 0.5767191451178726,
      "grad_norm": 6.4375,
      "learning_rate": 4.904317409399342e-05,
      "loss": 2.6043,
      "num_input_tokens_seen": 7398704,
      "step": 3535
    },
    {
      "epoch": 0.5775348723386899,
      "grad_norm": 9.4375,
      "learning_rate": 4.904048215747104e-05,
      "loss": 2.5046,
      "num_input_tokens_seen": 7408832,
      "step": 3540
    },
    {
      "epoch": 0.5783505995595073,
      "grad_norm": 9.125,
      "learning_rate": 4.90377865136042e-05,
      "loss": 3.227,
      "num_input_tokens_seen": 7419120,
      "step": 3545
    },
    {
      "epoch": 0.5791663267803246,
      "grad_norm": 12.0,
      "learning_rate": 4.90350871628086e-05,
      "loss": 2.0458,
      "num_input_tokens_seen": 7429712,
      "step": 3550
    },
    {
      "epoch": 0.579982054001142,
      "grad_norm": 7.4375,
      "learning_rate": 4.903238410550052e-05,
      "loss": 2.4662,
      "num_input_tokens_seen": 7440480,
      "step": 3555
    },
    {
      "epoch": 0.5807977812219594,
      "grad_norm": 14.1875,
      "learning_rate": 4.90296773420968e-05,
      "loss": 3.1046,
      "num_input_tokens_seen": 7451840,
      "step": 3560
    },
    {
      "epoch": 0.5816135084427767,
      "grad_norm": 9.5,
      "learning_rate": 4.902696687301486e-05,
      "loss": 1.7273,
      "num_input_tokens_seen": 7461696,
      "step": 3565
    },
    {
      "epoch": 0.5824292356635941,
      "grad_norm": 10.4375,
      "learning_rate": 4.902425269867268e-05,
      "loss": 2.9585,
      "num_input_tokens_seen": 7471568,
      "step": 3570
    },
    {
      "epoch": 0.5832449628844114,
      "grad_norm": 6.0625,
      "learning_rate": 4.902153481948883e-05,
      "loss": 2.2035,
      "num_input_tokens_seen": 7482512,
      "step": 3575
    },
    {
      "epoch": 0.5840606901052288,
      "grad_norm": 3.296875,
      "learning_rate": 4.901881323588244e-05,
      "loss": 1.3552,
      "num_input_tokens_seen": 7492016,
      "step": 3580
    },
    {
      "epoch": 0.5848764173260461,
      "grad_norm": 1.078125,
      "learning_rate": 4.90160879482732e-05,
      "loss": 0.9002,
      "num_input_tokens_seen": 7501328,
      "step": 3585
    },
    {
      "epoch": 0.5856921445468636,
      "grad_norm": 4.90625,
      "learning_rate": 4.9013358957081405e-05,
      "loss": 1.5126,
      "num_input_tokens_seen": 7511952,
      "step": 3590
    },
    {
      "epoch": 0.5865078717676809,
      "grad_norm": 15.3125,
      "learning_rate": 4.901062626272789e-05,
      "loss": 2.6654,
      "num_input_tokens_seen": 7523520,
      "step": 3595
    },
    {
      "epoch": 0.5873235989884983,
      "grad_norm": 7.84375,
      "learning_rate": 4.900788986563406e-05,
      "loss": 2.0952,
      "num_input_tokens_seen": 7533408,
      "step": 3600
    },
    {
      "epoch": 0.5873235989884983,
      "eval_loss": 2.552873373031616,
      "eval_runtime": 134.8449,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 7533408,
      "step": 3600
    },
    {
      "epoch": 0.5881393262093156,
      "grad_norm": 10.75,
      "learning_rate": 4.9005149766221915e-05,
      "loss": 1.7189,
      "num_input_tokens_seen": 7543760,
      "step": 3605
    },
    {
      "epoch": 0.588955053430133,
      "grad_norm": 6.4375,
      "learning_rate": 4.9002405964914e-05,
      "loss": 2.3758,
      "num_input_tokens_seen": 7554320,
      "step": 3610
    },
    {
      "epoch": 0.5897707806509503,
      "grad_norm": 9.0625,
      "learning_rate": 4.899965846213346e-05,
      "loss": 3.6809,
      "num_input_tokens_seen": 7564672,
      "step": 3615
    },
    {
      "epoch": 0.5905865078717677,
      "grad_norm": 9.4375,
      "learning_rate": 4.899690725830399e-05,
      "loss": 2.1018,
      "num_input_tokens_seen": 7574608,
      "step": 3620
    },
    {
      "epoch": 0.591402235092585,
      "grad_norm": 5.5,
      "learning_rate": 4.899415235384985e-05,
      "loss": 1.428,
      "num_input_tokens_seen": 7586704,
      "step": 3625
    },
    {
      "epoch": 0.5922179623134024,
      "grad_norm": 9.375,
      "learning_rate": 4.899139374919589e-05,
      "loss": 2.3859,
      "num_input_tokens_seen": 7596992,
      "step": 3630
    },
    {
      "epoch": 0.5930336895342198,
      "grad_norm": 10.6875,
      "learning_rate": 4.898863144476752e-05,
      "loss": 4.1084,
      "num_input_tokens_seen": 7607376,
      "step": 3635
    },
    {
      "epoch": 0.5938494167550371,
      "grad_norm": 10.0625,
      "learning_rate": 4.898586544099072e-05,
      "loss": 2.4244,
      "num_input_tokens_seen": 7616608,
      "step": 3640
    },
    {
      "epoch": 0.5946651439758545,
      "grad_norm": 3.609375,
      "learning_rate": 4.898309573829204e-05,
      "loss": 1.9677,
      "num_input_tokens_seen": 7626416,
      "step": 3645
    },
    {
      "epoch": 0.5954808711966718,
      "grad_norm": 5.40625,
      "learning_rate": 4.898032233709862e-05,
      "loss": 3.2209,
      "num_input_tokens_seen": 7637104,
      "step": 3650
    },
    {
      "epoch": 0.5962965984174892,
      "grad_norm": 4.625,
      "learning_rate": 4.8977545237838123e-05,
      "loss": 2.4833,
      "num_input_tokens_seen": 7648816,
      "step": 3655
    },
    {
      "epoch": 0.5971123256383065,
      "grad_norm": 7.6875,
      "learning_rate": 4.8974764440938836e-05,
      "loss": 2.2217,
      "num_input_tokens_seen": 7660336,
      "step": 3660
    },
    {
      "epoch": 0.5979280528591239,
      "grad_norm": 10.0625,
      "learning_rate": 4.897197994682959e-05,
      "loss": 1.9646,
      "num_input_tokens_seen": 7670336,
      "step": 3665
    },
    {
      "epoch": 0.5987437800799412,
      "grad_norm": 12.0,
      "learning_rate": 4.8969191755939786e-05,
      "loss": 2.4625,
      "num_input_tokens_seen": 7680000,
      "step": 3670
    },
    {
      "epoch": 0.5995595073007586,
      "grad_norm": 10.125,
      "learning_rate": 4.8966399868699396e-05,
      "loss": 2.3734,
      "num_input_tokens_seen": 7690912,
      "step": 3675
    },
    {
      "epoch": 0.600375234521576,
      "grad_norm": 4.71875,
      "learning_rate": 4.8963604285538965e-05,
      "loss": 1.7672,
      "num_input_tokens_seen": 7703616,
      "step": 3680
    },
    {
      "epoch": 0.6011909617423934,
      "grad_norm": 1.9140625,
      "learning_rate": 4.8960805006889604e-05,
      "loss": 2.2219,
      "num_input_tokens_seen": 7711952,
      "step": 3685
    },
    {
      "epoch": 0.6020066889632107,
      "grad_norm": 9.4375,
      "learning_rate": 4.8958002033183004e-05,
      "loss": 2.6358,
      "num_input_tokens_seen": 7723216,
      "step": 3690
    },
    {
      "epoch": 0.6028224161840281,
      "grad_norm": 7.9375,
      "learning_rate": 4.8955195364851414e-05,
      "loss": 3.7151,
      "num_input_tokens_seen": 7732192,
      "step": 3695
    },
    {
      "epoch": 0.6036381434048455,
      "grad_norm": 1.7421875,
      "learning_rate": 4.895238500232766e-05,
      "loss": 2.1522,
      "num_input_tokens_seen": 7742144,
      "step": 3700
    },
    {
      "epoch": 0.6044538706256628,
      "grad_norm": 5.375,
      "learning_rate": 4.8949570946045143e-05,
      "loss": 2.3937,
      "num_input_tokens_seen": 7753600,
      "step": 3705
    },
    {
      "epoch": 0.6052695978464802,
      "grad_norm": 3.921875,
      "learning_rate": 4.89467531964378e-05,
      "loss": 1.367,
      "num_input_tokens_seen": 7763248,
      "step": 3710
    },
    {
      "epoch": 0.6060853250672975,
      "grad_norm": 6.4375,
      "learning_rate": 4.894393175394019e-05,
      "loss": 3.0416,
      "num_input_tokens_seen": 7773424,
      "step": 3715
    },
    {
      "epoch": 0.6069010522881149,
      "grad_norm": 3.578125,
      "learning_rate": 4.8941106618987406e-05,
      "loss": 1.6572,
      "num_input_tokens_seen": 7783248,
      "step": 3720
    },
    {
      "epoch": 0.6077167795089322,
      "grad_norm": 7.09375,
      "learning_rate": 4.893827779201512e-05,
      "loss": 2.6385,
      "num_input_tokens_seen": 7794048,
      "step": 3725
    },
    {
      "epoch": 0.6085325067297496,
      "grad_norm": 6.875,
      "learning_rate": 4.893544527345957e-05,
      "loss": 3.5935,
      "num_input_tokens_seen": 7803920,
      "step": 3730
    },
    {
      "epoch": 0.6093482339505669,
      "grad_norm": 7.40625,
      "learning_rate": 4.8932609063757563e-05,
      "loss": 2.344,
      "num_input_tokens_seen": 7816048,
      "step": 3735
    },
    {
      "epoch": 0.6101639611713843,
      "grad_norm": 7.84375,
      "learning_rate": 4.8929769163346484e-05,
      "loss": 2.4508,
      "num_input_tokens_seen": 7827056,
      "step": 3740
    },
    {
      "epoch": 0.6109796883922016,
      "grad_norm": 6.9375,
      "learning_rate": 4.892692557266429e-05,
      "loss": 1.0042,
      "num_input_tokens_seen": 7837392,
      "step": 3745
    },
    {
      "epoch": 0.611795415613019,
      "grad_norm": 7.53125,
      "learning_rate": 4.8924078292149464e-05,
      "loss": 2.4875,
      "num_input_tokens_seen": 7847888,
      "step": 3750
    },
    {
      "epoch": 0.6126111428338363,
      "grad_norm": 17.75,
      "learning_rate": 4.892122732224114e-05,
      "loss": 3.0358,
      "num_input_tokens_seen": 7858064,
      "step": 3755
    },
    {
      "epoch": 0.6134268700546537,
      "grad_norm": 5.28125,
      "learning_rate": 4.8918372663378944e-05,
      "loss": 2.3247,
      "num_input_tokens_seen": 7867648,
      "step": 3760
    },
    {
      "epoch": 0.614242597275471,
      "grad_norm": 4.875,
      "learning_rate": 4.89155143160031e-05,
      "loss": 2.3061,
      "num_input_tokens_seen": 7878496,
      "step": 3765
    },
    {
      "epoch": 0.6150583244962884,
      "grad_norm": 8.875,
      "learning_rate": 4.891265228055441e-05,
      "loss": 3.2971,
      "num_input_tokens_seen": 7890016,
      "step": 3770
    },
    {
      "epoch": 0.6158740517171059,
      "grad_norm": 2.578125,
      "learning_rate": 4.890978655747424e-05,
      "loss": 1.7368,
      "num_input_tokens_seen": 7901056,
      "step": 3775
    },
    {
      "epoch": 0.6166897789379232,
      "grad_norm": 8.625,
      "learning_rate": 4.89069171472045e-05,
      "loss": 2.0773,
      "num_input_tokens_seen": 7911296,
      "step": 3780
    },
    {
      "epoch": 0.6175055061587406,
      "grad_norm": 5.625,
      "learning_rate": 4.890404405018772e-05,
      "loss": 2.245,
      "num_input_tokens_seen": 7920896,
      "step": 3785
    },
    {
      "epoch": 0.6183212333795579,
      "grad_norm": 9.4375,
      "learning_rate": 4.8901167266866934e-05,
      "loss": 2.376,
      "num_input_tokens_seen": 7931168,
      "step": 3790
    },
    {
      "epoch": 0.6191369606003753,
      "grad_norm": 11.8125,
      "learning_rate": 4.88982867976858e-05,
      "loss": 2.0185,
      "num_input_tokens_seen": 7941328,
      "step": 3795
    },
    {
      "epoch": 0.6199526878211926,
      "grad_norm": 6.65625,
      "learning_rate": 4.889540264308852e-05,
      "loss": 2.8358,
      "num_input_tokens_seen": 7950560,
      "step": 3800
    },
    {
      "epoch": 0.6199526878211926,
      "eval_loss": 2.542593240737915,
      "eval_runtime": 134.8495,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 7950560,
      "step": 3800
    },
    {
      "epoch": 0.62076841504201,
      "grad_norm": 6.1875,
      "learning_rate": 4.889251480351986e-05,
      "loss": 2.726,
      "num_input_tokens_seen": 7960672,
      "step": 3805
    },
    {
      "epoch": 0.6215841422628273,
      "grad_norm": 12.875,
      "learning_rate": 4.888962327942517e-05,
      "loss": 2.3652,
      "num_input_tokens_seen": 7970848,
      "step": 3810
    },
    {
      "epoch": 0.6223998694836447,
      "grad_norm": 9.0625,
      "learning_rate": 4.8886728071250356e-05,
      "loss": 3.3738,
      "num_input_tokens_seen": 7983216,
      "step": 3815
    },
    {
      "epoch": 0.623215596704462,
      "grad_norm": 7.5625,
      "learning_rate": 4.8883829179441884e-05,
      "loss": 2.6637,
      "num_input_tokens_seen": 7994064,
      "step": 3820
    },
    {
      "epoch": 0.6240313239252794,
      "grad_norm": 7.125,
      "learning_rate": 4.888092660444682e-05,
      "loss": 2.0591,
      "num_input_tokens_seen": 8004656,
      "step": 3825
    },
    {
      "epoch": 0.6248470511460967,
      "grad_norm": 7.09375,
      "learning_rate": 4.887802034671276e-05,
      "loss": 2.8221,
      "num_input_tokens_seen": 8015360,
      "step": 3830
    },
    {
      "epoch": 0.6256627783669141,
      "grad_norm": 5.4375,
      "learning_rate": 4.88751104066879e-05,
      "loss": 1.7193,
      "num_input_tokens_seen": 8025040,
      "step": 3835
    },
    {
      "epoch": 0.6264785055877314,
      "grad_norm": 2.328125,
      "learning_rate": 4.887219678482098e-05,
      "loss": 2.5356,
      "num_input_tokens_seen": 8036576,
      "step": 3840
    },
    {
      "epoch": 0.6272942328085488,
      "grad_norm": 7.46875,
      "learning_rate": 4.8869279481561316e-05,
      "loss": 2.0837,
      "num_input_tokens_seen": 8045728,
      "step": 3845
    },
    {
      "epoch": 0.6281099600293661,
      "grad_norm": 5.75,
      "learning_rate": 4.88663584973588e-05,
      "loss": 2.1651,
      "num_input_tokens_seen": 8056992,
      "step": 3850
    },
    {
      "epoch": 0.6289256872501835,
      "grad_norm": 0.31640625,
      "learning_rate": 4.8863433832663874e-05,
      "loss": 2.1636,
      "num_input_tokens_seen": 8065504,
      "step": 3855
    },
    {
      "epoch": 0.6297414144710008,
      "grad_norm": 5.53125,
      "learning_rate": 4.886050548792757e-05,
      "loss": 1.9006,
      "num_input_tokens_seen": 8076384,
      "step": 3860
    },
    {
      "epoch": 0.6305571416918182,
      "grad_norm": 11.8125,
      "learning_rate": 4.8857573463601465e-05,
      "loss": 2.5184,
      "num_input_tokens_seen": 8087296,
      "step": 3865
    },
    {
      "epoch": 0.6313728689126357,
      "grad_norm": 6.4375,
      "learning_rate": 4.885463776013772e-05,
      "loss": 2.4237,
      "num_input_tokens_seen": 8098368,
      "step": 3870
    },
    {
      "epoch": 0.632188596133453,
      "grad_norm": 12.25,
      "learning_rate": 4.8851698377989056e-05,
      "loss": 2.2896,
      "num_input_tokens_seen": 8109232,
      "step": 3875
    },
    {
      "epoch": 0.6330043233542704,
      "grad_norm": 5.78125,
      "learning_rate": 4.884875531760876e-05,
      "loss": 2.04,
      "num_input_tokens_seen": 8119808,
      "step": 3880
    },
    {
      "epoch": 0.6338200505750877,
      "grad_norm": 9.0625,
      "learning_rate": 4.88458085794507e-05,
      "loss": 2.3342,
      "num_input_tokens_seen": 8129376,
      "step": 3885
    },
    {
      "epoch": 0.6346357777959051,
      "grad_norm": 10.3125,
      "learning_rate": 4.884285816396929e-05,
      "loss": 3.3626,
      "num_input_tokens_seen": 8139168,
      "step": 3890
    },
    {
      "epoch": 0.6354515050167224,
      "grad_norm": 6.46875,
      "learning_rate": 4.8839904071619526e-05,
      "loss": 1.9,
      "num_input_tokens_seen": 8149264,
      "step": 3895
    },
    {
      "epoch": 0.6362672322375398,
      "grad_norm": 3.328125,
      "learning_rate": 4.8836946302856955e-05,
      "loss": 1.9218,
      "num_input_tokens_seen": 8161248,
      "step": 3900
    },
    {
      "epoch": 0.6370829594583571,
      "grad_norm": 10.6875,
      "learning_rate": 4.8833984858137715e-05,
      "loss": 2.6557,
      "num_input_tokens_seen": 8171248,
      "step": 3905
    },
    {
      "epoch": 0.6378986866791745,
      "grad_norm": 3.671875,
      "learning_rate": 4.8831019737918494e-05,
      "loss": 1.7655,
      "num_input_tokens_seen": 8182976,
      "step": 3910
    },
    {
      "epoch": 0.6387144138999918,
      "grad_norm": 2.8125,
      "learning_rate": 4.882805094265655e-05,
      "loss": 1.8309,
      "num_input_tokens_seen": 8192832,
      "step": 3915
    },
    {
      "epoch": 0.6395301411208092,
      "grad_norm": 4.75,
      "learning_rate": 4.8825078472809706e-05,
      "loss": 2.1732,
      "num_input_tokens_seen": 8202992,
      "step": 3920
    },
    {
      "epoch": 0.6403458683416265,
      "grad_norm": 9.25,
      "learning_rate": 4.882210232883635e-05,
      "loss": 2.4585,
      "num_input_tokens_seen": 8214176,
      "step": 3925
    },
    {
      "epoch": 0.6411615955624439,
      "grad_norm": 6.6875,
      "learning_rate": 4.881912251119546e-05,
      "loss": 2.413,
      "num_input_tokens_seen": 8225216,
      "step": 3930
    },
    {
      "epoch": 0.6419773227832613,
      "grad_norm": 9.375,
      "learning_rate": 4.881613902034654e-05,
      "loss": 2.6427,
      "num_input_tokens_seen": 8235680,
      "step": 3935
    },
    {
      "epoch": 0.6427930500040786,
      "grad_norm": 6.90625,
      "learning_rate": 4.88131518567497e-05,
      "loss": 1.1857,
      "num_input_tokens_seen": 8246480,
      "step": 3940
    },
    {
      "epoch": 0.643608777224896,
      "grad_norm": 4.9375,
      "learning_rate": 4.881016102086558e-05,
      "loss": 1.9682,
      "num_input_tokens_seen": 8256928,
      "step": 3945
    },
    {
      "epoch": 0.6444245044457133,
      "grad_norm": 4.90625,
      "learning_rate": 4.8807166513155425e-05,
      "loss": 3.1164,
      "num_input_tokens_seen": 8268096,
      "step": 3950
    },
    {
      "epoch": 0.6452402316665307,
      "grad_norm": 8.5625,
      "learning_rate": 4.8804168334081004e-05,
      "loss": 3.4649,
      "num_input_tokens_seen": 8279440,
      "step": 3955
    },
    {
      "epoch": 0.6460559588873481,
      "grad_norm": 5.09375,
      "learning_rate": 4.880116648410468e-05,
      "loss": 3.0097,
      "num_input_tokens_seen": 8290144,
      "step": 3960
    },
    {
      "epoch": 0.6468716861081655,
      "grad_norm": 0.376953125,
      "learning_rate": 4.879816096368939e-05,
      "loss": 1.2124,
      "num_input_tokens_seen": 8300240,
      "step": 3965
    },
    {
      "epoch": 0.6476874133289828,
      "grad_norm": 7.78125,
      "learning_rate": 4.879515177329861e-05,
      "loss": 2.0724,
      "num_input_tokens_seen": 8309632,
      "step": 3970
    },
    {
      "epoch": 0.6485031405498002,
      "grad_norm": 3.15625,
      "learning_rate": 4.8792138913396394e-05,
      "loss": 2.0452,
      "num_input_tokens_seen": 8320752,
      "step": 3975
    },
    {
      "epoch": 0.6493188677706175,
      "grad_norm": 10.9375,
      "learning_rate": 4.8789122384447374e-05,
      "loss": 2.1118,
      "num_input_tokens_seen": 8331136,
      "step": 3980
    },
    {
      "epoch": 0.6501345949914349,
      "grad_norm": 7.6875,
      "learning_rate": 4.878610218691673e-05,
      "loss": 2.3542,
      "num_input_tokens_seen": 8341552,
      "step": 3985
    },
    {
      "epoch": 0.6509503222122522,
      "grad_norm": 6.5625,
      "learning_rate": 4.87830783212702e-05,
      "loss": 2.6979,
      "num_input_tokens_seen": 8352176,
      "step": 3990
    },
    {
      "epoch": 0.6517660494330696,
      "grad_norm": 9.6875,
      "learning_rate": 4.878005078797413e-05,
      "loss": 3.1937,
      "num_input_tokens_seen": 8361952,
      "step": 3995
    },
    {
      "epoch": 0.652581776653887,
      "grad_norm": 10.75,
      "learning_rate": 4.877701958749539e-05,
      "loss": 2.2672,
      "num_input_tokens_seen": 8372672,
      "step": 4000
    },
    {
      "epoch": 0.652581776653887,
      "eval_loss": 2.5473592281341553,
      "eval_runtime": 135.1225,
      "eval_samples_per_second": 20.167,
      "eval_steps_per_second": 10.087,
      "num_input_tokens_seen": 8372672,
      "step": 4000
    },
    {
      "epoch": 0.6533975038747043,
      "grad_norm": 7.65625,
      "learning_rate": 4.877398472030142e-05,
      "loss": 1.8732,
      "num_input_tokens_seen": 8382752,
      "step": 4005
    },
    {
      "epoch": 0.6542132310955217,
      "grad_norm": 7.0,
      "learning_rate": 4.877094618686024e-05,
      "loss": 1.8587,
      "num_input_tokens_seen": 8394112,
      "step": 4010
    },
    {
      "epoch": 0.655028958316339,
      "grad_norm": 12.25,
      "learning_rate": 4.876790398764045e-05,
      "loss": 2.7755,
      "num_input_tokens_seen": 8404576,
      "step": 4015
    },
    {
      "epoch": 0.6558446855371564,
      "grad_norm": 7.625,
      "learning_rate": 4.8764858123111167e-05,
      "loss": 2.8243,
      "num_input_tokens_seen": 8414592,
      "step": 4020
    },
    {
      "epoch": 0.6566604127579737,
      "grad_norm": 10.125,
      "learning_rate": 4.876180859374212e-05,
      "loss": 3.097,
      "num_input_tokens_seen": 8425168,
      "step": 4025
    },
    {
      "epoch": 0.6574761399787911,
      "grad_norm": 8.5,
      "learning_rate": 4.875875540000357e-05,
      "loss": 2.681,
      "num_input_tokens_seen": 8436768,
      "step": 4030
    },
    {
      "epoch": 0.6582918671996084,
      "grad_norm": 17.5,
      "learning_rate": 4.8755698542366376e-05,
      "loss": 2.1566,
      "num_input_tokens_seen": 8448176,
      "step": 4035
    },
    {
      "epoch": 0.6591075944204258,
      "grad_norm": 6.1875,
      "learning_rate": 4.875263802130193e-05,
      "loss": 2.7086,
      "num_input_tokens_seen": 8460480,
      "step": 4040
    },
    {
      "epoch": 0.6599233216412431,
      "grad_norm": 2.015625,
      "learning_rate": 4.8749573837282207e-05,
      "loss": 2.5483,
      "num_input_tokens_seen": 8470400,
      "step": 4045
    },
    {
      "epoch": 0.6607390488620605,
      "grad_norm": 12.625,
      "learning_rate": 4.874650599077974e-05,
      "loss": 2.7352,
      "num_input_tokens_seen": 8480352,
      "step": 4050
    },
    {
      "epoch": 0.6615547760828779,
      "grad_norm": 6.09375,
      "learning_rate": 4.874343448226764e-05,
      "loss": 2.3886,
      "num_input_tokens_seen": 8491344,
      "step": 4055
    },
    {
      "epoch": 0.6623705033036953,
      "grad_norm": 10.125,
      "learning_rate": 4.874035931221955e-05,
      "loss": 2.279,
      "num_input_tokens_seen": 8502480,
      "step": 4060
    },
    {
      "epoch": 0.6631862305245126,
      "grad_norm": 11.4375,
      "learning_rate": 4.8737280481109724e-05,
      "loss": 1.8528,
      "num_input_tokens_seen": 8512672,
      "step": 4065
    },
    {
      "epoch": 0.66400195774533,
      "grad_norm": 8.1875,
      "learning_rate": 4.873419798941294e-05,
      "loss": 2.2189,
      "num_input_tokens_seen": 8523856,
      "step": 4070
    },
    {
      "epoch": 0.6648176849661473,
      "grad_norm": 3.859375,
      "learning_rate": 4.873111183760458e-05,
      "loss": 1.986,
      "num_input_tokens_seen": 8536192,
      "step": 4075
    },
    {
      "epoch": 0.6656334121869647,
      "grad_norm": 3.765625,
      "learning_rate": 4.8728022026160537e-05,
      "loss": 2.0461,
      "num_input_tokens_seen": 8545136,
      "step": 4080
    },
    {
      "epoch": 0.666449139407782,
      "grad_norm": 8.25,
      "learning_rate": 4.872492855555732e-05,
      "loss": 2.3533,
      "num_input_tokens_seen": 8555456,
      "step": 4085
    },
    {
      "epoch": 0.6672648666285994,
      "grad_norm": 3.25,
      "learning_rate": 4.8721831426271956e-05,
      "loss": 1.2958,
      "num_input_tokens_seen": 8564304,
      "step": 4090
    },
    {
      "epoch": 0.6680805938494168,
      "grad_norm": 0.068359375,
      "learning_rate": 4.87187306387821e-05,
      "loss": 2.3116,
      "num_input_tokens_seen": 8575536,
      "step": 4095
    },
    {
      "epoch": 0.6688963210702341,
      "grad_norm": 1.578125,
      "learning_rate": 4.87156261935659e-05,
      "loss": 2.1941,
      "num_input_tokens_seen": 8587056,
      "step": 4100
    },
    {
      "epoch": 0.6697120482910515,
      "grad_norm": 7.8125,
      "learning_rate": 4.871251809110211e-05,
      "loss": 2.1207,
      "num_input_tokens_seen": 8597808,
      "step": 4105
    },
    {
      "epoch": 0.6705277755118688,
      "grad_norm": 6.03125,
      "learning_rate": 4.8709406331870044e-05,
      "loss": 2.5123,
      "num_input_tokens_seen": 8608592,
      "step": 4110
    },
    {
      "epoch": 0.6713435027326862,
      "grad_norm": 9.5,
      "learning_rate": 4.8706290916349574e-05,
      "loss": 2.4135,
      "num_input_tokens_seen": 8617824,
      "step": 4115
    },
    {
      "epoch": 0.6721592299535035,
      "grad_norm": 9.25,
      "learning_rate": 4.8703171845021134e-05,
      "loss": 1.7784,
      "num_input_tokens_seen": 8628464,
      "step": 4120
    },
    {
      "epoch": 0.6729749571743209,
      "grad_norm": 9.1875,
      "learning_rate": 4.870004911836572e-05,
      "loss": 2.8992,
      "num_input_tokens_seen": 8639168,
      "step": 4125
    },
    {
      "epoch": 0.6737906843951382,
      "grad_norm": 4.1875,
      "learning_rate": 4.869692273686489e-05,
      "loss": 2.2268,
      "num_input_tokens_seen": 8649680,
      "step": 4130
    },
    {
      "epoch": 0.6746064116159556,
      "grad_norm": 9.625,
      "learning_rate": 4.869379270100079e-05,
      "loss": 3.371,
      "num_input_tokens_seen": 8660400,
      "step": 4135
    },
    {
      "epoch": 0.6754221388367729,
      "grad_norm": 8.625,
      "learning_rate": 4.86906590112561e-05,
      "loss": 3.0593,
      "num_input_tokens_seen": 8670080,
      "step": 4140
    },
    {
      "epoch": 0.6762378660575903,
      "grad_norm": 11.375,
      "learning_rate": 4.8687521668114064e-05,
      "loss": 3.0317,
      "num_input_tokens_seen": 8679952,
      "step": 4145
    },
    {
      "epoch": 0.6770535932784078,
      "grad_norm": 8.3125,
      "learning_rate": 4.868438067205853e-05,
      "loss": 2.8708,
      "num_input_tokens_seen": 8689920,
      "step": 4150
    },
    {
      "epoch": 0.6778693204992251,
      "grad_norm": 12.25,
      "learning_rate": 4.8681236023573844e-05,
      "loss": 3.0706,
      "num_input_tokens_seen": 8700432,
      "step": 4155
    },
    {
      "epoch": 0.6786850477200425,
      "grad_norm": 7.875,
      "learning_rate": 4.867808772314497e-05,
      "loss": 3.2197,
      "num_input_tokens_seen": 8711104,
      "step": 4160
    },
    {
      "epoch": 0.6795007749408598,
      "grad_norm": 7.21875,
      "learning_rate": 4.867493577125741e-05,
      "loss": 2.6816,
      "num_input_tokens_seen": 8721264,
      "step": 4165
    },
    {
      "epoch": 0.6803165021616772,
      "grad_norm": 2.046875,
      "learning_rate": 4.867178016839725e-05,
      "loss": 2.1749,
      "num_input_tokens_seen": 8732112,
      "step": 4170
    },
    {
      "epoch": 0.6811322293824945,
      "grad_norm": 9.0,
      "learning_rate": 4.8668620915051094e-05,
      "loss": 1.5776,
      "num_input_tokens_seen": 8742464,
      "step": 4175
    },
    {
      "epoch": 0.6819479566033119,
      "grad_norm": 7.25,
      "learning_rate": 4.866545801170616e-05,
      "loss": 2.6211,
      "num_input_tokens_seen": 8754128,
      "step": 4180
    },
    {
      "epoch": 0.6827636838241292,
      "grad_norm": 5.46875,
      "learning_rate": 4.86622914588502e-05,
      "loss": 1.5238,
      "num_input_tokens_seen": 8765280,
      "step": 4185
    },
    {
      "epoch": 0.6835794110449466,
      "grad_norm": 9.9375,
      "learning_rate": 4.865912125697154e-05,
      "loss": 2.4654,
      "num_input_tokens_seen": 8775712,
      "step": 4190
    },
    {
      "epoch": 0.6843951382657639,
      "grad_norm": 0.07177734375,
      "learning_rate": 4.865594740655907e-05,
      "loss": 1.6411,
      "num_input_tokens_seen": 8785376,
      "step": 4195
    },
    {
      "epoch": 0.6852108654865813,
      "grad_norm": 8.4375,
      "learning_rate": 4.865276990810222e-05,
      "loss": 1.7307,
      "num_input_tokens_seen": 8796048,
      "step": 4200
    },
    {
      "epoch": 0.6852108654865813,
      "eval_loss": 2.558070421218872,
      "eval_runtime": 134.8436,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 8796048,
      "step": 4200
    },
    {
      "epoch": 0.6860265927073986,
      "grad_norm": 9.3125,
      "learning_rate": 4.8649588762091016e-05,
      "loss": 1.0379,
      "num_input_tokens_seen": 8807040,
      "step": 4205
    },
    {
      "epoch": 0.686842319928216,
      "grad_norm": 8.1875,
      "learning_rate": 4.8646403969016016e-05,
      "loss": 1.9028,
      "num_input_tokens_seen": 8817664,
      "step": 4210
    },
    {
      "epoch": 0.6876580471490333,
      "grad_norm": 10.5625,
      "learning_rate": 4.864321552936838e-05,
      "loss": 2.1559,
      "num_input_tokens_seen": 8828880,
      "step": 4215
    },
    {
      "epoch": 0.6884737743698507,
      "grad_norm": 1.5078125,
      "learning_rate": 4.864002344363978e-05,
      "loss": 1.9426,
      "num_input_tokens_seen": 8839840,
      "step": 4220
    },
    {
      "epoch": 0.689289501590668,
      "grad_norm": 10.1875,
      "learning_rate": 4.863682771232248e-05,
      "loss": 1.3108,
      "num_input_tokens_seen": 8850224,
      "step": 4225
    },
    {
      "epoch": 0.6901052288114854,
      "grad_norm": 2.046875,
      "learning_rate": 4.8633628335909324e-05,
      "loss": 1.9218,
      "num_input_tokens_seen": 8862592,
      "step": 4230
    },
    {
      "epoch": 0.6909209560323027,
      "grad_norm": 10.6875,
      "learning_rate": 4.8630425314893676e-05,
      "loss": 1.8785,
      "num_input_tokens_seen": 8872656,
      "step": 4235
    },
    {
      "epoch": 0.6917366832531202,
      "grad_norm": 6.625,
      "learning_rate": 4.862721864976948e-05,
      "loss": 1.9657,
      "num_input_tokens_seen": 8882016,
      "step": 4240
    },
    {
      "epoch": 0.6925524104739376,
      "grad_norm": 3.890625,
      "learning_rate": 4.862400834103125e-05,
      "loss": 1.9744,
      "num_input_tokens_seen": 8891712,
      "step": 4245
    },
    {
      "epoch": 0.6933681376947549,
      "grad_norm": 7.84375,
      "learning_rate": 4.862079438917406e-05,
      "loss": 2.1444,
      "num_input_tokens_seen": 8902000,
      "step": 4250
    },
    {
      "epoch": 0.6941838649155723,
      "grad_norm": 2.203125,
      "learning_rate": 4.8617576794693536e-05,
      "loss": 2.6894,
      "num_input_tokens_seen": 8911040,
      "step": 4255
    },
    {
      "epoch": 0.6949995921363896,
      "grad_norm": 3.84375,
      "learning_rate": 4.8614355558085875e-05,
      "loss": 2.3691,
      "num_input_tokens_seen": 8920912,
      "step": 4260
    },
    {
      "epoch": 0.695815319357207,
      "grad_norm": 7.6875,
      "learning_rate": 4.861113067984783e-05,
      "loss": 2.8364,
      "num_input_tokens_seen": 8931728,
      "step": 4265
    },
    {
      "epoch": 0.6966310465780243,
      "grad_norm": 3.875,
      "learning_rate": 4.860790216047671e-05,
      "loss": 2.4062,
      "num_input_tokens_seen": 8941616,
      "step": 4270
    },
    {
      "epoch": 0.6974467737988417,
      "grad_norm": 10.4375,
      "learning_rate": 4.860467000047041e-05,
      "loss": 1.8364,
      "num_input_tokens_seen": 8951008,
      "step": 4275
    },
    {
      "epoch": 0.698262501019659,
      "grad_norm": 10.875,
      "learning_rate": 4.860143420032737e-05,
      "loss": 1.9861,
      "num_input_tokens_seen": 8961856,
      "step": 4280
    },
    {
      "epoch": 0.6990782282404764,
      "grad_norm": 8.4375,
      "learning_rate": 4.859819476054657e-05,
      "loss": 2.5696,
      "num_input_tokens_seen": 8971792,
      "step": 4285
    },
    {
      "epoch": 0.6998939554612937,
      "grad_norm": 7.65625,
      "learning_rate": 4.859495168162758e-05,
      "loss": 2.2095,
      "num_input_tokens_seen": 8982000,
      "step": 4290
    },
    {
      "epoch": 0.7007096826821111,
      "grad_norm": 11.1875,
      "learning_rate": 4.859170496407054e-05,
      "loss": 4.1751,
      "num_input_tokens_seen": 8990832,
      "step": 4295
    },
    {
      "epoch": 0.7015254099029284,
      "grad_norm": 3.625,
      "learning_rate": 4.8588454608376114e-05,
      "loss": 2.2711,
      "num_input_tokens_seen": 8999904,
      "step": 4300
    },
    {
      "epoch": 0.7023411371237458,
      "grad_norm": 9.0625,
      "learning_rate": 4.8585200615045555e-05,
      "loss": 2.43,
      "num_input_tokens_seen": 9009888,
      "step": 4305
    },
    {
      "epoch": 0.7031568643445631,
      "grad_norm": 7.625,
      "learning_rate": 4.8581942984580674e-05,
      "loss": 1.9469,
      "num_input_tokens_seen": 9020224,
      "step": 4310
    },
    {
      "epoch": 0.7039725915653805,
      "grad_norm": 5.78125,
      "learning_rate": 4.857868171748384e-05,
      "loss": 2.8645,
      "num_input_tokens_seen": 9030176,
      "step": 4315
    },
    {
      "epoch": 0.7047883187861979,
      "grad_norm": 6.25,
      "learning_rate": 4.8575416814257976e-05,
      "loss": 1.8936,
      "num_input_tokens_seen": 9041312,
      "step": 4320
    },
    {
      "epoch": 0.7056040460070152,
      "grad_norm": 12.1875,
      "learning_rate": 4.857214827540657e-05,
      "loss": 2.8018,
      "num_input_tokens_seen": 9052368,
      "step": 4325
    },
    {
      "epoch": 0.7064197732278326,
      "grad_norm": 6.5625,
      "learning_rate": 4.856887610143367e-05,
      "loss": 2.8542,
      "num_input_tokens_seen": 9064096,
      "step": 4330
    },
    {
      "epoch": 0.70723550044865,
      "grad_norm": 14.3125,
      "learning_rate": 4.8565600292843896e-05,
      "loss": 3.1432,
      "num_input_tokens_seen": 9074064,
      "step": 4335
    },
    {
      "epoch": 0.7080512276694674,
      "grad_norm": 9.0,
      "learning_rate": 4.856232085014241e-05,
      "loss": 2.4022,
      "num_input_tokens_seen": 9085728,
      "step": 4340
    },
    {
      "epoch": 0.7088669548902847,
      "grad_norm": 9.0,
      "learning_rate": 4.855903777383495e-05,
      "loss": 2.5207,
      "num_input_tokens_seen": 9097440,
      "step": 4345
    },
    {
      "epoch": 0.7096826821111021,
      "grad_norm": 5.90625,
      "learning_rate": 4.85557510644278e-05,
      "loss": 2.5286,
      "num_input_tokens_seen": 9108320,
      "step": 4350
    },
    {
      "epoch": 0.7104984093319194,
      "grad_norm": 4.5,
      "learning_rate": 4.855246072242782e-05,
      "loss": 1.6211,
      "num_input_tokens_seen": 9119232,
      "step": 4355
    },
    {
      "epoch": 0.7113141365527368,
      "grad_norm": 7.28125,
      "learning_rate": 4.8549166748342414e-05,
      "loss": 2.2292,
      "num_input_tokens_seen": 9128816,
      "step": 4360
    },
    {
      "epoch": 0.7121298637735541,
      "grad_norm": 7.34375,
      "learning_rate": 4.8545869142679556e-05,
      "loss": 3.4955,
      "num_input_tokens_seen": 9137840,
      "step": 4365
    },
    {
      "epoch": 0.7129455909943715,
      "grad_norm": 2.859375,
      "learning_rate": 4.8542567905947776e-05,
      "loss": 2.2395,
      "num_input_tokens_seen": 9148672,
      "step": 4370
    },
    {
      "epoch": 0.7137613182151888,
      "grad_norm": 7.21875,
      "learning_rate": 4.853926303865618e-05,
      "loss": 1.7518,
      "num_input_tokens_seen": 9158720,
      "step": 4375
    },
    {
      "epoch": 0.7145770454360062,
      "grad_norm": 7.0,
      "learning_rate": 4.853595454131441e-05,
      "loss": 2.1249,
      "num_input_tokens_seen": 9168240,
      "step": 4380
    },
    {
      "epoch": 0.7153927726568236,
      "grad_norm": 5.21875,
      "learning_rate": 4.8532642414432674e-05,
      "loss": 2.0307,
      "num_input_tokens_seen": 9178704,
      "step": 4385
    },
    {
      "epoch": 0.7162084998776409,
      "grad_norm": 5.0,
      "learning_rate": 4.8529326658521754e-05,
      "loss": 2.4567,
      "num_input_tokens_seen": 9189648,
      "step": 4390
    },
    {
      "epoch": 0.7170242270984583,
      "grad_norm": 13.0625,
      "learning_rate": 4.8526007274092965e-05,
      "loss": 3.8078,
      "num_input_tokens_seen": 9199696,
      "step": 4395
    },
    {
      "epoch": 0.7178399543192756,
      "grad_norm": 14.75,
      "learning_rate": 4.852268426165822e-05,
      "loss": 2.0188,
      "num_input_tokens_seen": 9210416,
      "step": 4400
    },
    {
      "epoch": 0.7178399543192756,
      "eval_loss": 2.547496795654297,
      "eval_runtime": 135.0287,
      "eval_samples_per_second": 20.181,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 9210416,
      "step": 4400
    },
    {
      "epoch": 0.718655681540093,
      "grad_norm": 6.0,
      "learning_rate": 4.851935762172995e-05,
      "loss": 1.4136,
      "num_input_tokens_seen": 9220016,
      "step": 4405
    },
    {
      "epoch": 0.7194714087609103,
      "grad_norm": 9.0,
      "learning_rate": 4.8516027354821175e-05,
      "loss": 3.8296,
      "num_input_tokens_seen": 9230944,
      "step": 4410
    },
    {
      "epoch": 0.7202871359817277,
      "grad_norm": 8.1875,
      "learning_rate": 4.851269346144546e-05,
      "loss": 3.7847,
      "num_input_tokens_seen": 9240528,
      "step": 4415
    },
    {
      "epoch": 0.721102863202545,
      "grad_norm": 10.5,
      "learning_rate": 4.850935594211693e-05,
      "loss": 2.8351,
      "num_input_tokens_seen": 9250496,
      "step": 4420
    },
    {
      "epoch": 0.7219185904233625,
      "grad_norm": 2.5,
      "learning_rate": 4.850601479735029e-05,
      "loss": 1.8854,
      "num_input_tokens_seen": 9260832,
      "step": 4425
    },
    {
      "epoch": 0.7227343176441798,
      "grad_norm": 4.125,
      "learning_rate": 4.850267002766076e-05,
      "loss": 1.7331,
      "num_input_tokens_seen": 9270352,
      "step": 4430
    },
    {
      "epoch": 0.7235500448649972,
      "grad_norm": 8.875,
      "learning_rate": 4.849932163356417e-05,
      "loss": 2.5217,
      "num_input_tokens_seen": 9281120,
      "step": 4435
    },
    {
      "epoch": 0.7243657720858145,
      "grad_norm": 6.6875,
      "learning_rate": 4.8495969615576864e-05,
      "loss": 2.0177,
      "num_input_tokens_seen": 9292000,
      "step": 4440
    },
    {
      "epoch": 0.7251814993066319,
      "grad_norm": 4.4375,
      "learning_rate": 4.849261397421577e-05,
      "loss": 2.2123,
      "num_input_tokens_seen": 9302480,
      "step": 4445
    },
    {
      "epoch": 0.7259972265274492,
      "grad_norm": 4.28125,
      "learning_rate": 4.848925470999839e-05,
      "loss": 2.4211,
      "num_input_tokens_seen": 9312464,
      "step": 4450
    },
    {
      "epoch": 0.7268129537482666,
      "grad_norm": 6.9375,
      "learning_rate": 4.848589182344273e-05,
      "loss": 2.667,
      "num_input_tokens_seen": 9322224,
      "step": 4455
    },
    {
      "epoch": 0.727628680969084,
      "grad_norm": 11.3125,
      "learning_rate": 4.848252531506742e-05,
      "loss": 4.6915,
      "num_input_tokens_seen": 9332176,
      "step": 4460
    },
    {
      "epoch": 0.7284444081899013,
      "grad_norm": 6.40625,
      "learning_rate": 4.847915518539161e-05,
      "loss": 3.0387,
      "num_input_tokens_seen": 9344176,
      "step": 4465
    },
    {
      "epoch": 0.7292601354107187,
      "grad_norm": 3.5,
      "learning_rate": 4.847578143493501e-05,
      "loss": 1.1769,
      "num_input_tokens_seen": 9355904,
      "step": 4470
    },
    {
      "epoch": 0.730075862631536,
      "grad_norm": 12.6875,
      "learning_rate": 4.847240406421789e-05,
      "loss": 3.601,
      "num_input_tokens_seen": 9366320,
      "step": 4475
    },
    {
      "epoch": 0.7308915898523534,
      "grad_norm": 9.0,
      "learning_rate": 4.84690230737611e-05,
      "loss": 1.9841,
      "num_input_tokens_seen": 9377504,
      "step": 4480
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 3.375,
      "learning_rate": 4.846563846408602e-05,
      "loss": 1.5232,
      "num_input_tokens_seen": 9387936,
      "step": 4485
    },
    {
      "epoch": 0.7325230442939881,
      "grad_norm": 8.3125,
      "learning_rate": 4.84622502357146e-05,
      "loss": 4.1495,
      "num_input_tokens_seen": 9398496,
      "step": 4490
    },
    {
      "epoch": 0.7333387715148054,
      "grad_norm": 15.5625,
      "learning_rate": 4.8458858389169345e-05,
      "loss": 2.6902,
      "num_input_tokens_seen": 9408848,
      "step": 4495
    },
    {
      "epoch": 0.7341544987356228,
      "grad_norm": 4.96875,
      "learning_rate": 4.8455462924973334e-05,
      "loss": 2.1822,
      "num_input_tokens_seen": 9418224,
      "step": 4500
    },
    {
      "epoch": 0.7349702259564401,
      "grad_norm": 1.2421875,
      "learning_rate": 4.845206384365018e-05,
      "loss": 2.3119,
      "num_input_tokens_seen": 9428464,
      "step": 4505
    },
    {
      "epoch": 0.7357859531772575,
      "grad_norm": 6.875,
      "learning_rate": 4.844866114572405e-05,
      "loss": 2.7446,
      "num_input_tokens_seen": 9437792,
      "step": 4510
    },
    {
      "epoch": 0.7366016803980748,
      "grad_norm": 7.5625,
      "learning_rate": 4.8445254831719706e-05,
      "loss": 1.9596,
      "num_input_tokens_seen": 9448144,
      "step": 4515
    },
    {
      "epoch": 0.7374174076188923,
      "grad_norm": 2.265625,
      "learning_rate": 4.8441844902162434e-05,
      "loss": 1.5452,
      "num_input_tokens_seen": 9457312,
      "step": 4520
    },
    {
      "epoch": 0.7382331348397096,
      "grad_norm": 3.453125,
      "learning_rate": 4.843843135757809e-05,
      "loss": 1.126,
      "num_input_tokens_seen": 9469024,
      "step": 4525
    },
    {
      "epoch": 0.739048862060527,
      "grad_norm": 8.0,
      "learning_rate": 4.843501419849308e-05,
      "loss": 3.9736,
      "num_input_tokens_seen": 9480000,
      "step": 4530
    },
    {
      "epoch": 0.7398645892813444,
      "grad_norm": 6.0625,
      "learning_rate": 4.8431593425434386e-05,
      "loss": 2.5313,
      "num_input_tokens_seen": 9491696,
      "step": 4535
    },
    {
      "epoch": 0.7406803165021617,
      "grad_norm": 4.8125,
      "learning_rate": 4.8428169038929526e-05,
      "loss": 2.5195,
      "num_input_tokens_seen": 9502528,
      "step": 4540
    },
    {
      "epoch": 0.7414960437229791,
      "grad_norm": 10.4375,
      "learning_rate": 4.8424741039506575e-05,
      "loss": 2.3166,
      "num_input_tokens_seen": 9512112,
      "step": 4545
    },
    {
      "epoch": 0.7423117709437964,
      "grad_norm": 10.5625,
      "learning_rate": 4.842130942769419e-05,
      "loss": 1.2655,
      "num_input_tokens_seen": 9523568,
      "step": 4550
    },
    {
      "epoch": 0.7431274981646138,
      "grad_norm": 4.0625,
      "learning_rate": 4.841787420402156e-05,
      "loss": 2.5286,
      "num_input_tokens_seen": 9533392,
      "step": 4555
    },
    {
      "epoch": 0.7439432253854311,
      "grad_norm": 5.15625,
      "learning_rate": 4.841443536901844e-05,
      "loss": 3.0285,
      "num_input_tokens_seen": 9544560,
      "step": 4560
    },
    {
      "epoch": 0.7447589526062485,
      "grad_norm": 8.3125,
      "learning_rate": 4.841099292321514e-05,
      "loss": 2.3562,
      "num_input_tokens_seen": 9553984,
      "step": 4565
    },
    {
      "epoch": 0.7455746798270658,
      "grad_norm": 12.5,
      "learning_rate": 4.8407546867142525e-05,
      "loss": 3.4789,
      "num_input_tokens_seen": 9563664,
      "step": 4570
    },
    {
      "epoch": 0.7463904070478832,
      "grad_norm": 11.0625,
      "learning_rate": 4.840409720133203e-05,
      "loss": 1.856,
      "num_input_tokens_seen": 9574256,
      "step": 4575
    },
    {
      "epoch": 0.7472061342687005,
      "grad_norm": 3.703125,
      "learning_rate": 4.8400643926315634e-05,
      "loss": 1.7858,
      "num_input_tokens_seen": 9584800,
      "step": 4580
    },
    {
      "epoch": 0.7480218614895179,
      "grad_norm": 6.3125,
      "learning_rate": 4.839718704262587e-05,
      "loss": 3.5959,
      "num_input_tokens_seen": 9595120,
      "step": 4585
    },
    {
      "epoch": 0.7488375887103352,
      "grad_norm": 4.375,
      "learning_rate": 4.839372655079585e-05,
      "loss": 1.0393,
      "num_input_tokens_seen": 9605024,
      "step": 4590
    },
    {
      "epoch": 0.7496533159311526,
      "grad_norm": 1.921875,
      "learning_rate": 4.83902624513592e-05,
      "loss": 2.8227,
      "num_input_tokens_seen": 9617520,
      "step": 4595
    },
    {
      "epoch": 0.7504690431519699,
      "grad_norm": 5.59375,
      "learning_rate": 4.838679474485014e-05,
      "loss": 2.0703,
      "num_input_tokens_seen": 9628832,
      "step": 4600
    },
    {
      "epoch": 0.7504690431519699,
      "eval_loss": 2.543306827545166,
      "eval_runtime": 135.0617,
      "eval_samples_per_second": 20.176,
      "eval_steps_per_second": 10.092,
      "num_input_tokens_seen": 9628832,
      "step": 4600
    },
    {
      "epoch": 0.7512847703727873,
      "grad_norm": 10.125,
      "learning_rate": 4.838332343180343e-05,
      "loss": 2.2992,
      "num_input_tokens_seen": 9639408,
      "step": 4605
    },
    {
      "epoch": 0.7521004975936046,
      "grad_norm": 7.4375,
      "learning_rate": 4.83798485127544e-05,
      "loss": 1.8661,
      "num_input_tokens_seen": 9650752,
      "step": 4610
    },
    {
      "epoch": 0.7529162248144221,
      "grad_norm": 10.9375,
      "learning_rate": 4.837636998823892e-05,
      "loss": 2.3467,
      "num_input_tokens_seen": 9660672,
      "step": 4615
    },
    {
      "epoch": 0.7537319520352395,
      "grad_norm": 10.125,
      "learning_rate": 4.8372887858793414e-05,
      "loss": 2.0318,
      "num_input_tokens_seen": 9670256,
      "step": 4620
    },
    {
      "epoch": 0.7545476792560568,
      "grad_norm": 8.3125,
      "learning_rate": 4.836940212495489e-05,
      "loss": 4.4494,
      "num_input_tokens_seen": 9680592,
      "step": 4625
    },
    {
      "epoch": 0.7553634064768742,
      "grad_norm": 9.375,
      "learning_rate": 4.836591278726087e-05,
      "loss": 2.0766,
      "num_input_tokens_seen": 9689904,
      "step": 4630
    },
    {
      "epoch": 0.7561791336976915,
      "grad_norm": 8.0625,
      "learning_rate": 4.836241984624947e-05,
      "loss": 2.3191,
      "num_input_tokens_seen": 9699888,
      "step": 4635
    },
    {
      "epoch": 0.7569948609185089,
      "grad_norm": 3.21875,
      "learning_rate": 4.8358923302459336e-05,
      "loss": 2.5817,
      "num_input_tokens_seen": 9710592,
      "step": 4640
    },
    {
      "epoch": 0.7578105881393262,
      "grad_norm": 7.15625,
      "learning_rate": 4.835542315642968e-05,
      "loss": 1.6087,
      "num_input_tokens_seen": 9721184,
      "step": 4645
    },
    {
      "epoch": 0.7586263153601436,
      "grad_norm": 2.71875,
      "learning_rate": 4.8351919408700274e-05,
      "loss": 2.1471,
      "num_input_tokens_seen": 9732384,
      "step": 4650
    },
    {
      "epoch": 0.7594420425809609,
      "grad_norm": 7.65625,
      "learning_rate": 4.834841205981144e-05,
      "loss": 2.3985,
      "num_input_tokens_seen": 9742544,
      "step": 4655
    },
    {
      "epoch": 0.7602577698017783,
      "grad_norm": 3.6875,
      "learning_rate": 4.8344901110304054e-05,
      "loss": 3.0241,
      "num_input_tokens_seen": 9752736,
      "step": 4660
    },
    {
      "epoch": 0.7610734970225956,
      "grad_norm": 5.375,
      "learning_rate": 4.8341386560719534e-05,
      "loss": 3.3465,
      "num_input_tokens_seen": 9763744,
      "step": 4665
    },
    {
      "epoch": 0.761889224243413,
      "grad_norm": 0.34375,
      "learning_rate": 4.833786841159989e-05,
      "loss": 2.0751,
      "num_input_tokens_seen": 9775040,
      "step": 4670
    },
    {
      "epoch": 0.7627049514642303,
      "grad_norm": 15.1875,
      "learning_rate": 4.833434666348765e-05,
      "loss": 3.6634,
      "num_input_tokens_seen": 9785328,
      "step": 4675
    },
    {
      "epoch": 0.7635206786850477,
      "grad_norm": 7.3125,
      "learning_rate": 4.833082131692592e-05,
      "loss": 2.2743,
      "num_input_tokens_seen": 9795328,
      "step": 4680
    },
    {
      "epoch": 0.764336405905865,
      "grad_norm": 4.34375,
      "learning_rate": 4.832729237245835e-05,
      "loss": 3.2662,
      "num_input_tokens_seen": 9805328,
      "step": 4685
    },
    {
      "epoch": 0.7651521331266824,
      "grad_norm": 6.53125,
      "learning_rate": 4.8323759830629145e-05,
      "loss": 3.0139,
      "num_input_tokens_seen": 9814688,
      "step": 4690
    },
    {
      "epoch": 0.7659678603474998,
      "grad_norm": 7.65625,
      "learning_rate": 4.8320223691983066e-05,
      "loss": 2.7113,
      "num_input_tokens_seen": 9825664,
      "step": 4695
    },
    {
      "epoch": 0.7667835875683171,
      "grad_norm": 6.71875,
      "learning_rate": 4.831668395706544e-05,
      "loss": 2.4601,
      "num_input_tokens_seen": 9837760,
      "step": 4700
    },
    {
      "epoch": 0.7675993147891346,
      "grad_norm": 7.34375,
      "learning_rate": 4.8313140626422125e-05,
      "loss": 1.8229,
      "num_input_tokens_seen": 9848512,
      "step": 4705
    },
    {
      "epoch": 0.7684150420099519,
      "grad_norm": 6.1875,
      "learning_rate": 4.830959370059956e-05,
      "loss": 2.0388,
      "num_input_tokens_seen": 9858912,
      "step": 4710
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 6.0625,
      "learning_rate": 4.830604318014472e-05,
      "loss": 1.932,
      "num_input_tokens_seen": 9869024,
      "step": 4715
    },
    {
      "epoch": 0.7700464964515866,
      "grad_norm": 6.5625,
      "learning_rate": 4.830248906560514e-05,
      "loss": 2.619,
      "num_input_tokens_seen": 9878464,
      "step": 4720
    },
    {
      "epoch": 0.770862223672404,
      "grad_norm": 9.1875,
      "learning_rate": 4.829893135752891e-05,
      "loss": 2.4192,
      "num_input_tokens_seen": 9888864,
      "step": 4725
    },
    {
      "epoch": 0.7716779508932213,
      "grad_norm": 2.296875,
      "learning_rate": 4.829537005646466e-05,
      "loss": 2.1911,
      "num_input_tokens_seen": 9899696,
      "step": 4730
    },
    {
      "epoch": 0.7724936781140387,
      "grad_norm": 10.5,
      "learning_rate": 4.8291805162961615e-05,
      "loss": 2.548,
      "num_input_tokens_seen": 9911392,
      "step": 4735
    },
    {
      "epoch": 0.773309405334856,
      "grad_norm": 3.609375,
      "learning_rate": 4.82882366775695e-05,
      "loss": 2.0496,
      "num_input_tokens_seen": 9920304,
      "step": 4740
    },
    {
      "epoch": 0.7741251325556734,
      "grad_norm": 2.71875,
      "learning_rate": 4.828466460083864e-05,
      "loss": 2.6142,
      "num_input_tokens_seen": 9930816,
      "step": 4745
    },
    {
      "epoch": 0.7749408597764907,
      "grad_norm": 1.828125,
      "learning_rate": 4.8281088933319877e-05,
      "loss": 1.6442,
      "num_input_tokens_seen": 9940848,
      "step": 4750
    },
    {
      "epoch": 0.7757565869973081,
      "grad_norm": 6.28125,
      "learning_rate": 4.827750967556464e-05,
      "loss": 3.6581,
      "num_input_tokens_seen": 9951920,
      "step": 4755
    },
    {
      "epoch": 0.7765723142181254,
      "grad_norm": 11.3125,
      "learning_rate": 4.827392682812488e-05,
      "loss": 2.5322,
      "num_input_tokens_seen": 9962176,
      "step": 4760
    },
    {
      "epoch": 0.7773880414389428,
      "grad_norm": 6.125,
      "learning_rate": 4.827034039155312e-05,
      "loss": 2.0869,
      "num_input_tokens_seen": 9973600,
      "step": 4765
    },
    {
      "epoch": 0.7782037686597602,
      "grad_norm": 6.21875,
      "learning_rate": 4.8266750366402445e-05,
      "loss": 2.6236,
      "num_input_tokens_seen": 9984528,
      "step": 4770
    },
    {
      "epoch": 0.7790194958805775,
      "grad_norm": 4.3125,
      "learning_rate": 4.8263156753226476e-05,
      "loss": 2.0488,
      "num_input_tokens_seen": 9993632,
      "step": 4775
    },
    {
      "epoch": 0.7798352231013949,
      "grad_norm": 7.71875,
      "learning_rate": 4.8259559552579394e-05,
      "loss": 1.7693,
      "num_input_tokens_seen": 10004448,
      "step": 4780
    },
    {
      "epoch": 0.7806509503222122,
      "grad_norm": 6.03125,
      "learning_rate": 4.825595876501593e-05,
      "loss": 2.7809,
      "num_input_tokens_seen": 10015664,
      "step": 4785
    },
    {
      "epoch": 0.7814666775430296,
      "grad_norm": 8.75,
      "learning_rate": 4.825235439109137e-05,
      "loss": 1.8827,
      "num_input_tokens_seen": 10026576,
      "step": 4790
    },
    {
      "epoch": 0.7822824047638469,
      "grad_norm": 9.1875,
      "learning_rate": 4.824874643136156e-05,
      "loss": 2.939,
      "num_input_tokens_seen": 10037168,
      "step": 4795
    },
    {
      "epoch": 0.7830981319846644,
      "grad_norm": 6.5,
      "learning_rate": 4.824513488638288e-05,
      "loss": 3.0207,
      "num_input_tokens_seen": 10048144,
      "step": 4800
    },
    {
      "epoch": 0.7830981319846644,
      "eval_loss": 2.536677598953247,
      "eval_runtime": 134.8407,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 10048144,
      "step": 4800
    },
    {
      "epoch": 0.7839138592054817,
      "grad_norm": 10.1875,
      "learning_rate": 4.8241519756712293e-05,
      "loss": 2.8481,
      "num_input_tokens_seen": 10057472,
      "step": 4805
    },
    {
      "epoch": 0.7847295864262991,
      "grad_norm": 13.25,
      "learning_rate": 4.8237901042907285e-05,
      "loss": 4.1903,
      "num_input_tokens_seen": 10068720,
      "step": 4810
    },
    {
      "epoch": 0.7855453136471164,
      "grad_norm": 10.0625,
      "learning_rate": 4.823427874552591e-05,
      "loss": 2.2892,
      "num_input_tokens_seen": 10078336,
      "step": 4815
    },
    {
      "epoch": 0.7863610408679338,
      "grad_norm": 8.0,
      "learning_rate": 4.823065286512677e-05,
      "loss": 1.5361,
      "num_input_tokens_seen": 10088064,
      "step": 4820
    },
    {
      "epoch": 0.7871767680887511,
      "grad_norm": 8.0,
      "learning_rate": 4.8227023402269025e-05,
      "loss": 2.8675,
      "num_input_tokens_seen": 10098928,
      "step": 4825
    },
    {
      "epoch": 0.7879924953095685,
      "grad_norm": 5.875,
      "learning_rate": 4.822339035751239e-05,
      "loss": 2.0661,
      "num_input_tokens_seen": 10108608,
      "step": 4830
    },
    {
      "epoch": 0.7888082225303858,
      "grad_norm": 7.75,
      "learning_rate": 4.8219753731417104e-05,
      "loss": 3.0142,
      "num_input_tokens_seen": 10119168,
      "step": 4835
    },
    {
      "epoch": 0.7896239497512032,
      "grad_norm": 6.59375,
      "learning_rate": 4.821611352454401e-05,
      "loss": 2.4362,
      "num_input_tokens_seen": 10129472,
      "step": 4840
    },
    {
      "epoch": 0.7904396769720206,
      "grad_norm": 9.375,
      "learning_rate": 4.8212469737454444e-05,
      "loss": 2.5736,
      "num_input_tokens_seen": 10138944,
      "step": 4845
    },
    {
      "epoch": 0.7912554041928379,
      "grad_norm": 0.2314453125,
      "learning_rate": 4.820882237071035e-05,
      "loss": 2.6321,
      "num_input_tokens_seen": 10150096,
      "step": 4850
    },
    {
      "epoch": 0.7920711314136553,
      "grad_norm": 13.375,
      "learning_rate": 4.820517142487417e-05,
      "loss": 2.1525,
      "num_input_tokens_seen": 10158512,
      "step": 4855
    },
    {
      "epoch": 0.7928868586344726,
      "grad_norm": 8.375,
      "learning_rate": 4.8201516900508956e-05,
      "loss": 2.8096,
      "num_input_tokens_seen": 10167536,
      "step": 4860
    },
    {
      "epoch": 0.79370258585529,
      "grad_norm": 11.375,
      "learning_rate": 4.819785879817827e-05,
      "loss": 2.8131,
      "num_input_tokens_seen": 10178400,
      "step": 4865
    },
    {
      "epoch": 0.7945183130761073,
      "grad_norm": 5.4375,
      "learning_rate": 4.8194197118446226e-05,
      "loss": 1.5677,
      "num_input_tokens_seen": 10188240,
      "step": 4870
    },
    {
      "epoch": 0.7953340402969247,
      "grad_norm": 6.1875,
      "learning_rate": 4.819053186187752e-05,
      "loss": 3.1932,
      "num_input_tokens_seen": 10198672,
      "step": 4875
    },
    {
      "epoch": 0.796149767517742,
      "grad_norm": 8.125,
      "learning_rate": 4.818686302903736e-05,
      "loss": 3.0827,
      "num_input_tokens_seen": 10208576,
      "step": 4880
    },
    {
      "epoch": 0.7969654947385594,
      "grad_norm": 3.25,
      "learning_rate": 4.818319062049154e-05,
      "loss": 1.9723,
      "num_input_tokens_seen": 10219472,
      "step": 4885
    },
    {
      "epoch": 0.7977812219593767,
      "grad_norm": 3.890625,
      "learning_rate": 4.817951463680639e-05,
      "loss": 2.5122,
      "num_input_tokens_seen": 10230112,
      "step": 4890
    },
    {
      "epoch": 0.7985969491801942,
      "grad_norm": 5.71875,
      "learning_rate": 4.817583507854879e-05,
      "loss": 2.9596,
      "num_input_tokens_seen": 10240992,
      "step": 4895
    },
    {
      "epoch": 0.7994126764010115,
      "grad_norm": 11.25,
      "learning_rate": 4.817215194628617e-05,
      "loss": 1.3114,
      "num_input_tokens_seen": 10250336,
      "step": 4900
    },
    {
      "epoch": 0.8002284036218289,
      "grad_norm": 7.1875,
      "learning_rate": 4.816846524058653e-05,
      "loss": 1.5633,
      "num_input_tokens_seen": 10261600,
      "step": 4905
    },
    {
      "epoch": 0.8010441308426463,
      "grad_norm": 7.46875,
      "learning_rate": 4.816477496201839e-05,
      "loss": 1.7865,
      "num_input_tokens_seen": 10271040,
      "step": 4910
    },
    {
      "epoch": 0.8018598580634636,
      "grad_norm": 4.21875,
      "learning_rate": 4.8161081111150845e-05,
      "loss": 2.0804,
      "num_input_tokens_seen": 10281568,
      "step": 4915
    },
    {
      "epoch": 0.802675585284281,
      "grad_norm": 7.65625,
      "learning_rate": 4.815738368855354e-05,
      "loss": 1.8795,
      "num_input_tokens_seen": 10289952,
      "step": 4920
    },
    {
      "epoch": 0.8034913125050983,
      "grad_norm": 4.46875,
      "learning_rate": 4.815368269479664e-05,
      "loss": 2.6724,
      "num_input_tokens_seen": 10299536,
      "step": 4925
    },
    {
      "epoch": 0.8043070397259157,
      "grad_norm": 4.0625,
      "learning_rate": 4.814997813045092e-05,
      "loss": 3.4789,
      "num_input_tokens_seen": 10309936,
      "step": 4930
    },
    {
      "epoch": 0.805122766946733,
      "grad_norm": 9.4375,
      "learning_rate": 4.814626999608764e-05,
      "loss": 2.5626,
      "num_input_tokens_seen": 10319776,
      "step": 4935
    },
    {
      "epoch": 0.8059384941675504,
      "grad_norm": 3.078125,
      "learning_rate": 4.814255829227865e-05,
      "loss": 1.4127,
      "num_input_tokens_seen": 10330432,
      "step": 4940
    },
    {
      "epoch": 0.8067542213883677,
      "grad_norm": 4.3125,
      "learning_rate": 4.813884301959635e-05,
      "loss": 2.7938,
      "num_input_tokens_seen": 10341936,
      "step": 4945
    },
    {
      "epoch": 0.8075699486091851,
      "grad_norm": 4.1875,
      "learning_rate": 4.813512417861368e-05,
      "loss": 1.6468,
      "num_input_tokens_seen": 10353488,
      "step": 4950
    },
    {
      "epoch": 0.8083856758300024,
      "grad_norm": 5.09375,
      "learning_rate": 4.813140176990411e-05,
      "loss": 1.4526,
      "num_input_tokens_seen": 10364688,
      "step": 4955
    },
    {
      "epoch": 0.8092014030508198,
      "grad_norm": 7.3125,
      "learning_rate": 4.8127675794041714e-05,
      "loss": 1.6773,
      "num_input_tokens_seen": 10373104,
      "step": 4960
    },
    {
      "epoch": 0.8100171302716371,
      "grad_norm": 10.9375,
      "learning_rate": 4.812394625160107e-05,
      "loss": 2.9198,
      "num_input_tokens_seen": 10383776,
      "step": 4965
    },
    {
      "epoch": 0.8108328574924545,
      "grad_norm": 2.25,
      "learning_rate": 4.812021314315732e-05,
      "loss": 1.4405,
      "num_input_tokens_seen": 10395072,
      "step": 4970
    },
    {
      "epoch": 0.8116485847132718,
      "grad_norm": 4.78125,
      "learning_rate": 4.811647646928616e-05,
      "loss": 1.8837,
      "num_input_tokens_seen": 10407264,
      "step": 4975
    },
    {
      "epoch": 0.8124643119340892,
      "grad_norm": 9.5,
      "learning_rate": 4.8112736230563814e-05,
      "loss": 4.0979,
      "num_input_tokens_seen": 10417776,
      "step": 4980
    },
    {
      "epoch": 0.8132800391549067,
      "grad_norm": 7.15625,
      "learning_rate": 4.81089924275671e-05,
      "loss": 1.8326,
      "num_input_tokens_seen": 10428560,
      "step": 4985
    },
    {
      "epoch": 0.814095766375724,
      "grad_norm": 7.15625,
      "learning_rate": 4.810524506087335e-05,
      "loss": 1.9079,
      "num_input_tokens_seen": 10439904,
      "step": 4990
    },
    {
      "epoch": 0.8149114935965414,
      "grad_norm": 6.53125,
      "learning_rate": 4.810149413106044e-05,
      "loss": 2.1585,
      "num_input_tokens_seen": 10449136,
      "step": 4995
    },
    {
      "epoch": 0.8157272208173587,
      "grad_norm": 7.96875,
      "learning_rate": 4.809773963870684e-05,
      "loss": 2.1577,
      "num_input_tokens_seen": 10460144,
      "step": 5000
    },
    {
      "epoch": 0.8157272208173587,
      "eval_loss": 2.5384411811828613,
      "eval_runtime": 135.0685,
      "eval_samples_per_second": 20.175,
      "eval_steps_per_second": 10.091,
      "num_input_tokens_seen": 10460144,
      "step": 5000
    },
    {
      "epoch": 0.8165429480381761,
      "grad_norm": 0.85546875,
      "learning_rate": 4.809398158439151e-05,
      "loss": 1.9975,
      "num_input_tokens_seen": 10470816,
      "step": 5005
    },
    {
      "epoch": 0.8173586752589934,
      "grad_norm": 4.59375,
      "learning_rate": 4.8090219968694005e-05,
      "loss": 2.0518,
      "num_input_tokens_seen": 10482064,
      "step": 5010
    },
    {
      "epoch": 0.8181744024798108,
      "grad_norm": 7.6875,
      "learning_rate": 4.808645479219442e-05,
      "loss": 2.1023,
      "num_input_tokens_seen": 10491408,
      "step": 5015
    },
    {
      "epoch": 0.8189901297006281,
      "grad_norm": 10.3125,
      "learning_rate": 4.8082686055473375e-05,
      "loss": 2.3896,
      "num_input_tokens_seen": 10502896,
      "step": 5020
    },
    {
      "epoch": 0.8198058569214455,
      "grad_norm": 10.25,
      "learning_rate": 4.8078913759112066e-05,
      "loss": 2.1198,
      "num_input_tokens_seen": 10513040,
      "step": 5025
    },
    {
      "epoch": 0.8206215841422628,
      "grad_norm": 3.578125,
      "learning_rate": 4.807513790369223e-05,
      "loss": 2.6971,
      "num_input_tokens_seen": 10523392,
      "step": 5030
    },
    {
      "epoch": 0.8214373113630802,
      "grad_norm": 8.25,
      "learning_rate": 4.8071358489796145e-05,
      "loss": 2.8387,
      "num_input_tokens_seen": 10534592,
      "step": 5035
    },
    {
      "epoch": 0.8222530385838975,
      "grad_norm": 14.75,
      "learning_rate": 4.806757551800665e-05,
      "loss": 2.8668,
      "num_input_tokens_seen": 10544656,
      "step": 5040
    },
    {
      "epoch": 0.8230687658047149,
      "grad_norm": 10.9375,
      "learning_rate": 4.806378898890713e-05,
      "loss": 1.828,
      "num_input_tokens_seen": 10554448,
      "step": 5045
    },
    {
      "epoch": 0.8238844930255322,
      "grad_norm": 8.875,
      "learning_rate": 4.80599989030815e-05,
      "loss": 3.1283,
      "num_input_tokens_seen": 10564864,
      "step": 5050
    },
    {
      "epoch": 0.8247002202463496,
      "grad_norm": 6.21875,
      "learning_rate": 4.805620526111426e-05,
      "loss": 3.3142,
      "num_input_tokens_seen": 10575712,
      "step": 5055
    },
    {
      "epoch": 0.8255159474671669,
      "grad_norm": 9.375,
      "learning_rate": 4.805240806359042e-05,
      "loss": 2.4858,
      "num_input_tokens_seen": 10585792,
      "step": 5060
    },
    {
      "epoch": 0.8263316746879843,
      "grad_norm": 12.6875,
      "learning_rate": 4.804860731109557e-05,
      "loss": 3.2062,
      "num_input_tokens_seen": 10594688,
      "step": 5065
    },
    {
      "epoch": 0.8271474019088016,
      "grad_norm": 5.9375,
      "learning_rate": 4.804480300421581e-05,
      "loss": 2.8882,
      "num_input_tokens_seen": 10603968,
      "step": 5070
    },
    {
      "epoch": 0.827963129129619,
      "grad_norm": 10.375,
      "learning_rate": 4.804099514353784e-05,
      "loss": 2.8833,
      "num_input_tokens_seen": 10614432,
      "step": 5075
    },
    {
      "epoch": 0.8287788563504365,
      "grad_norm": 9.0625,
      "learning_rate": 4.8037183729648867e-05,
      "loss": 2.7341,
      "num_input_tokens_seen": 10624176,
      "step": 5080
    },
    {
      "epoch": 0.8295945835712538,
      "grad_norm": 8.375,
      "learning_rate": 4.803336876313666e-05,
      "loss": 2.8572,
      "num_input_tokens_seen": 10635168,
      "step": 5085
    },
    {
      "epoch": 0.8304103107920712,
      "grad_norm": 4.5,
      "learning_rate": 4.802955024458953e-05,
      "loss": 2.6549,
      "num_input_tokens_seen": 10644336,
      "step": 5090
    },
    {
      "epoch": 0.8312260380128885,
      "grad_norm": 1.9765625,
      "learning_rate": 4.802572817459634e-05,
      "loss": 1.8507,
      "num_input_tokens_seen": 10654656,
      "step": 5095
    },
    {
      "epoch": 0.8320417652337059,
      "grad_norm": 2.671875,
      "learning_rate": 4.802190255374651e-05,
      "loss": 2.2612,
      "num_input_tokens_seen": 10665008,
      "step": 5100
    },
    {
      "epoch": 0.8328574924545232,
      "grad_norm": 8.5,
      "learning_rate": 4.801807338263e-05,
      "loss": 2.7435,
      "num_input_tokens_seen": 10675616,
      "step": 5105
    },
    {
      "epoch": 0.8336732196753406,
      "grad_norm": 9.875,
      "learning_rate": 4.8014240661837306e-05,
      "loss": 2.9749,
      "num_input_tokens_seen": 10686416,
      "step": 5110
    },
    {
      "epoch": 0.8344889468961579,
      "grad_norm": 5.03125,
      "learning_rate": 4.80104043919595e-05,
      "loss": 2.6904,
      "num_input_tokens_seen": 10697056,
      "step": 5115
    },
    {
      "epoch": 0.8353046741169753,
      "grad_norm": 7.40625,
      "learning_rate": 4.800656457358815e-05,
      "loss": 2.4095,
      "num_input_tokens_seen": 10706544,
      "step": 5120
    },
    {
      "epoch": 0.8361204013377926,
      "grad_norm": 0.2431640625,
      "learning_rate": 4.800272120731544e-05,
      "loss": 1.9017,
      "num_input_tokens_seen": 10716752,
      "step": 5125
    },
    {
      "epoch": 0.83693612855861,
      "grad_norm": 7.65625,
      "learning_rate": 4.799887429373404e-05,
      "loss": 2.4891,
      "num_input_tokens_seen": 10727536,
      "step": 5130
    },
    {
      "epoch": 0.8377518557794273,
      "grad_norm": 6.59375,
      "learning_rate": 4.79950238334372e-05,
      "loss": 2.1901,
      "num_input_tokens_seen": 10738000,
      "step": 5135
    },
    {
      "epoch": 0.8385675830002447,
      "grad_norm": 5.6875,
      "learning_rate": 4.799116982701872e-05,
      "loss": 0.7661,
      "num_input_tokens_seen": 10748320,
      "step": 5140
    },
    {
      "epoch": 0.839383310221062,
      "grad_norm": 8.875,
      "learning_rate": 4.7987312275072926e-05,
      "loss": 1.8018,
      "num_input_tokens_seen": 10757120,
      "step": 5145
    },
    {
      "epoch": 0.8401990374418794,
      "grad_norm": 3.375,
      "learning_rate": 4.79834511781947e-05,
      "loss": 2.0503,
      "num_input_tokens_seen": 10766192,
      "step": 5150
    },
    {
      "epoch": 0.8410147646626968,
      "grad_norm": 1.3359375,
      "learning_rate": 4.797958653697947e-05,
      "loss": 1.063,
      "num_input_tokens_seen": 10775088,
      "step": 5155
    },
    {
      "epoch": 0.8418304918835141,
      "grad_norm": 7.71875,
      "learning_rate": 4.7975718352023225e-05,
      "loss": 2.594,
      "num_input_tokens_seen": 10786624,
      "step": 5160
    },
    {
      "epoch": 0.8426462191043315,
      "grad_norm": 8.0625,
      "learning_rate": 4.7971846623922476e-05,
      "loss": 3.0066,
      "num_input_tokens_seen": 10797376,
      "step": 5165
    },
    {
      "epoch": 0.8434619463251488,
      "grad_norm": 6.90625,
      "learning_rate": 4.7967971353274294e-05,
      "loss": 2.6395,
      "num_input_tokens_seen": 10806832,
      "step": 5170
    },
    {
      "epoch": 0.8442776735459663,
      "grad_norm": 3.6875,
      "learning_rate": 4.79640925406763e-05,
      "loss": 1.8819,
      "num_input_tokens_seen": 10817888,
      "step": 5175
    },
    {
      "epoch": 0.8450934007667836,
      "grad_norm": 12.3125,
      "learning_rate": 4.796021018672664e-05,
      "loss": 2.6378,
      "num_input_tokens_seen": 10829120,
      "step": 5180
    },
    {
      "epoch": 0.845909127987601,
      "grad_norm": 9.6875,
      "learning_rate": 4.795632429202405e-05,
      "loss": 2.3422,
      "num_input_tokens_seen": 10839248,
      "step": 5185
    },
    {
      "epoch": 0.8467248552084183,
      "grad_norm": 7.3125,
      "learning_rate": 4.795243485716775e-05,
      "loss": 3.4465,
      "num_input_tokens_seen": 10850624,
      "step": 5190
    },
    {
      "epoch": 0.8475405824292357,
      "grad_norm": 9.75,
      "learning_rate": 4.794854188275757e-05,
      "loss": 2.1124,
      "num_input_tokens_seen": 10861744,
      "step": 5195
    },
    {
      "epoch": 0.848356309650053,
      "grad_norm": 8.75,
      "learning_rate": 4.794464536939384e-05,
      "loss": 1.4574,
      "num_input_tokens_seen": 10871296,
      "step": 5200
    },
    {
      "epoch": 0.848356309650053,
      "eval_loss": 2.5493946075439453,
      "eval_runtime": 135.017,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 10871296,
      "step": 5200
    },
    {
      "epoch": 0.8491720368708704,
      "grad_norm": 7.9375,
      "learning_rate": 4.794074531767745e-05,
      "loss": 3.1758,
      "num_input_tokens_seen": 10881776,
      "step": 5205
    },
    {
      "epoch": 0.8499877640916877,
      "grad_norm": 1.7578125,
      "learning_rate": 4.7936841728209834e-05,
      "loss": 2.5807,
      "num_input_tokens_seen": 10890720,
      "step": 5210
    },
    {
      "epoch": 0.8508034913125051,
      "grad_norm": 5.03125,
      "learning_rate": 4.7932934601593e-05,
      "loss": 2.1814,
      "num_input_tokens_seen": 10900528,
      "step": 5215
    },
    {
      "epoch": 0.8516192185333225,
      "grad_norm": 11.875,
      "learning_rate": 4.792902393842943e-05,
      "loss": 2.2278,
      "num_input_tokens_seen": 10912048,
      "step": 5220
    },
    {
      "epoch": 0.8524349457541398,
      "grad_norm": 6.65625,
      "learning_rate": 4.792510973932225e-05,
      "loss": 2.5504,
      "num_input_tokens_seen": 10921584,
      "step": 5225
    },
    {
      "epoch": 0.8532506729749572,
      "grad_norm": 6.03125,
      "learning_rate": 4.7921192004875036e-05,
      "loss": 3.1665,
      "num_input_tokens_seen": 10932208,
      "step": 5230
    },
    {
      "epoch": 0.8540664001957745,
      "grad_norm": 5.84375,
      "learning_rate": 4.791727073569198e-05,
      "loss": 2.3347,
      "num_input_tokens_seen": 10941984,
      "step": 5235
    },
    {
      "epoch": 0.8548821274165919,
      "grad_norm": 8.125,
      "learning_rate": 4.7913345932377775e-05,
      "loss": 1.7396,
      "num_input_tokens_seen": 10951584,
      "step": 5240
    },
    {
      "epoch": 0.8556978546374092,
      "grad_norm": 5.96875,
      "learning_rate": 4.790941759553769e-05,
      "loss": 2.4622,
      "num_input_tokens_seen": 10961280,
      "step": 5245
    },
    {
      "epoch": 0.8565135818582266,
      "grad_norm": 6.375,
      "learning_rate": 4.79054857257775e-05,
      "loss": 1.7231,
      "num_input_tokens_seen": 10972720,
      "step": 5250
    },
    {
      "epoch": 0.8573293090790439,
      "grad_norm": 4.21875,
      "learning_rate": 4.790155032370357e-05,
      "loss": 2.423,
      "num_input_tokens_seen": 10981936,
      "step": 5255
    },
    {
      "epoch": 0.8581450362998613,
      "grad_norm": 6.5625,
      "learning_rate": 4.789761138992278e-05,
      "loss": 3.4917,
      "num_input_tokens_seen": 10992080,
      "step": 5260
    },
    {
      "epoch": 0.8589607635206787,
      "grad_norm": 4.46875,
      "learning_rate": 4.7893668925042565e-05,
      "loss": 2.1076,
      "num_input_tokens_seen": 11002832,
      "step": 5265
    },
    {
      "epoch": 0.8597764907414961,
      "grad_norm": 5.71875,
      "learning_rate": 4.78897229296709e-05,
      "loss": 1.7452,
      "num_input_tokens_seen": 11013344,
      "step": 5270
    },
    {
      "epoch": 0.8605922179623134,
      "grad_norm": 10.5625,
      "learning_rate": 4.7885773404416315e-05,
      "loss": 3.7345,
      "num_input_tokens_seen": 11022304,
      "step": 5275
    },
    {
      "epoch": 0.8614079451831308,
      "grad_norm": 6.375,
      "learning_rate": 4.788182034988786e-05,
      "loss": 2.2345,
      "num_input_tokens_seen": 11033424,
      "step": 5280
    },
    {
      "epoch": 0.8622236724039481,
      "grad_norm": 7.125,
      "learning_rate": 4.787786376669516e-05,
      "loss": 2.1459,
      "num_input_tokens_seen": 11044112,
      "step": 5285
    },
    {
      "epoch": 0.8630393996247655,
      "grad_norm": 7.90625,
      "learning_rate": 4.787390365544837e-05,
      "loss": 2.3779,
      "num_input_tokens_seen": 11055280,
      "step": 5290
    },
    {
      "epoch": 0.8638551268455829,
      "grad_norm": 4.03125,
      "learning_rate": 4.786994001675818e-05,
      "loss": 2.051,
      "num_input_tokens_seen": 11066416,
      "step": 5295
    },
    {
      "epoch": 0.8646708540664002,
      "grad_norm": 2.6875,
      "learning_rate": 4.786597285123584e-05,
      "loss": 2.3115,
      "num_input_tokens_seen": 11075216,
      "step": 5300
    },
    {
      "epoch": 0.8654865812872176,
      "grad_norm": 6.03125,
      "learning_rate": 4.7862002159493135e-05,
      "loss": 2.6198,
      "num_input_tokens_seen": 11085264,
      "step": 5305
    },
    {
      "epoch": 0.8663023085080349,
      "grad_norm": 11.25,
      "learning_rate": 4.785802794214239e-05,
      "loss": 2.4652,
      "num_input_tokens_seen": 11095536,
      "step": 5310
    },
    {
      "epoch": 0.8671180357288523,
      "grad_norm": 4.28125,
      "learning_rate": 4.7854050199796495e-05,
      "loss": 1.6442,
      "num_input_tokens_seen": 11107360,
      "step": 5315
    },
    {
      "epoch": 0.8679337629496696,
      "grad_norm": 1.796875,
      "learning_rate": 4.7850068933068845e-05,
      "loss": 2.2302,
      "num_input_tokens_seen": 11118336,
      "step": 5320
    },
    {
      "epoch": 0.868749490170487,
      "grad_norm": 8.8125,
      "learning_rate": 4.7846084142573425e-05,
      "loss": 2.4921,
      "num_input_tokens_seen": 11130064,
      "step": 5325
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.1162109375,
      "learning_rate": 4.7842095828924725e-05,
      "loss": 2.4931,
      "num_input_tokens_seen": 11139520,
      "step": 5330
    },
    {
      "epoch": 0.8703809446121217,
      "grad_norm": 6.5,
      "learning_rate": 4.783810399273779e-05,
      "loss": 2.0956,
      "num_input_tokens_seen": 11150224,
      "step": 5335
    },
    {
      "epoch": 0.871196671832939,
      "grad_norm": 4.6875,
      "learning_rate": 4.7834108634628226e-05,
      "loss": 1.9645,
      "num_input_tokens_seen": 11160848,
      "step": 5340
    },
    {
      "epoch": 0.8720123990537564,
      "grad_norm": 5.15625,
      "learning_rate": 4.783010975521216e-05,
      "loss": 2.7703,
      "num_input_tokens_seen": 11171824,
      "step": 5345
    },
    {
      "epoch": 0.8728281262745737,
      "grad_norm": 10.0,
      "learning_rate": 4.782610735510626e-05,
      "loss": 1.8452,
      "num_input_tokens_seen": 11182448,
      "step": 5350
    },
    {
      "epoch": 0.8736438534953911,
      "grad_norm": 7.84375,
      "learning_rate": 4.782210143492776e-05,
      "loss": 2.7839,
      "num_input_tokens_seen": 11192736,
      "step": 5355
    },
    {
      "epoch": 0.8744595807162086,
      "grad_norm": 5.0625,
      "learning_rate": 4.781809199529442e-05,
      "loss": 2.148,
      "num_input_tokens_seen": 11204752,
      "step": 5360
    },
    {
      "epoch": 0.8752753079370259,
      "grad_norm": 4.78125,
      "learning_rate": 4.781407903682454e-05,
      "loss": 1.7689,
      "num_input_tokens_seen": 11216160,
      "step": 5365
    },
    {
      "epoch": 0.8760910351578433,
      "grad_norm": 1.2734375,
      "learning_rate": 4.781006256013698e-05,
      "loss": 1.1763,
      "num_input_tokens_seen": 11227664,
      "step": 5370
    },
    {
      "epoch": 0.8769067623786606,
      "grad_norm": 9.0,
      "learning_rate": 4.7806042565851115e-05,
      "loss": 3.1638,
      "num_input_tokens_seen": 11238160,
      "step": 5375
    },
    {
      "epoch": 0.877722489599478,
      "grad_norm": 11.25,
      "learning_rate": 4.7802019054586895e-05,
      "loss": 3.6178,
      "num_input_tokens_seen": 11248240,
      "step": 5380
    },
    {
      "epoch": 0.8785382168202953,
      "grad_norm": 1.890625,
      "learning_rate": 4.779799202696479e-05,
      "loss": 1.9397,
      "num_input_tokens_seen": 11257760,
      "step": 5385
    },
    {
      "epoch": 0.8793539440411127,
      "grad_norm": 6.4375,
      "learning_rate": 4.779396148360581e-05,
      "loss": 2.6823,
      "num_input_tokens_seen": 11268608,
      "step": 5390
    },
    {
      "epoch": 0.88016967126193,
      "grad_norm": 10.6875,
      "learning_rate": 4.7789927425131517e-05,
      "loss": 2.8279,
      "num_input_tokens_seen": 11278176,
      "step": 5395
    },
    {
      "epoch": 0.8809853984827474,
      "grad_norm": 7.75,
      "learning_rate": 4.778588985216403e-05,
      "loss": 3.7552,
      "num_input_tokens_seen": 11287600,
      "step": 5400
    },
    {
      "epoch": 0.8809853984827474,
      "eval_loss": 2.5324652194976807,
      "eval_runtime": 135.0952,
      "eval_samples_per_second": 20.171,
      "eval_steps_per_second": 10.089,
      "num_input_tokens_seen": 11287600,
      "step": 5400
    },
    {
      "epoch": 0.8818011257035647,
      "grad_norm": 3.828125,
      "learning_rate": 4.778184876532598e-05,
      "loss": 2.5186,
      "num_input_tokens_seen": 11298624,
      "step": 5405
    },
    {
      "epoch": 0.8826168529243821,
      "grad_norm": 6.0625,
      "learning_rate": 4.7777804165240556e-05,
      "loss": 1.6157,
      "num_input_tokens_seen": 11307440,
      "step": 5410
    },
    {
      "epoch": 0.8834325801451994,
      "grad_norm": 9.8125,
      "learning_rate": 4.7773756052531485e-05,
      "loss": 4.2717,
      "num_input_tokens_seen": 11318288,
      "step": 5415
    },
    {
      "epoch": 0.8842483073660168,
      "grad_norm": 4.71875,
      "learning_rate": 4.7769704427823035e-05,
      "loss": 2.1809,
      "num_input_tokens_seen": 11328048,
      "step": 5420
    },
    {
      "epoch": 0.8850640345868341,
      "grad_norm": 12.6875,
      "learning_rate": 4.776564929174003e-05,
      "loss": 2.2513,
      "num_input_tokens_seen": 11338048,
      "step": 5425
    },
    {
      "epoch": 0.8858797618076515,
      "grad_norm": 7.84375,
      "learning_rate": 4.7761590644907806e-05,
      "loss": 1.882,
      "num_input_tokens_seen": 11349600,
      "step": 5430
    },
    {
      "epoch": 0.8866954890284688,
      "grad_norm": 6.8125,
      "learning_rate": 4.7757528487952263e-05,
      "loss": 1.8687,
      "num_input_tokens_seen": 11360000,
      "step": 5435
    },
    {
      "epoch": 0.8875112162492862,
      "grad_norm": 3.359375,
      "learning_rate": 4.7753462821499836e-05,
      "loss": 2.5302,
      "num_input_tokens_seen": 11370624,
      "step": 5440
    },
    {
      "epoch": 0.8883269434701035,
      "grad_norm": 5.0625,
      "learning_rate": 4.774939364617751e-05,
      "loss": 1.869,
      "num_input_tokens_seen": 11381104,
      "step": 5445
    },
    {
      "epoch": 0.889142670690921,
      "grad_norm": 4.09375,
      "learning_rate": 4.7745320962612795e-05,
      "loss": 2.55,
      "num_input_tokens_seen": 11392144,
      "step": 5450
    },
    {
      "epoch": 0.8899583979117384,
      "grad_norm": 4.28125,
      "learning_rate": 4.7741244771433756e-05,
      "loss": 2.9084,
      "num_input_tokens_seen": 11401760,
      "step": 5455
    },
    {
      "epoch": 0.8907741251325557,
      "grad_norm": 5.75,
      "learning_rate": 4.7737165073268985e-05,
      "loss": 1.9265,
      "num_input_tokens_seen": 11412352,
      "step": 5460
    },
    {
      "epoch": 0.8915898523533731,
      "grad_norm": 7.84375,
      "learning_rate": 4.7733081868747626e-05,
      "loss": 1.7198,
      "num_input_tokens_seen": 11421424,
      "step": 5465
    },
    {
      "epoch": 0.8924055795741904,
      "grad_norm": 6.84375,
      "learning_rate": 4.772899515849936e-05,
      "loss": 2.6791,
      "num_input_tokens_seen": 11431680,
      "step": 5470
    },
    {
      "epoch": 0.8932213067950078,
      "grad_norm": 11.5,
      "learning_rate": 4.7724904943154414e-05,
      "loss": 2.3797,
      "num_input_tokens_seen": 11442672,
      "step": 5475
    },
    {
      "epoch": 0.8940370340158251,
      "grad_norm": 9.375,
      "learning_rate": 4.772081122334354e-05,
      "loss": 2.5438,
      "num_input_tokens_seen": 11453712,
      "step": 5480
    },
    {
      "epoch": 0.8948527612366425,
      "grad_norm": 9.6875,
      "learning_rate": 4.771671399969806e-05,
      "loss": 3.078,
      "num_input_tokens_seen": 11464128,
      "step": 5485
    },
    {
      "epoch": 0.8956684884574598,
      "grad_norm": 5.09375,
      "learning_rate": 4.7712613272849794e-05,
      "loss": 3.5604,
      "num_input_tokens_seen": 11474208,
      "step": 5490
    },
    {
      "epoch": 0.8964842156782772,
      "grad_norm": 7.375,
      "learning_rate": 4.770850904343114e-05,
      "loss": 2.2904,
      "num_input_tokens_seen": 11485968,
      "step": 5495
    },
    {
      "epoch": 0.8972999428990945,
      "grad_norm": 8.25,
      "learning_rate": 4.770440131207502e-05,
      "loss": 2.5698,
      "num_input_tokens_seen": 11497104,
      "step": 5500
    },
    {
      "epoch": 0.8981156701199119,
      "grad_norm": 7.65625,
      "learning_rate": 4.7700290079414896e-05,
      "loss": 2.9943,
      "num_input_tokens_seen": 11507440,
      "step": 5505
    },
    {
      "epoch": 0.8989313973407292,
      "grad_norm": 16.5,
      "learning_rate": 4.769617534608477e-05,
      "loss": 3.4936,
      "num_input_tokens_seen": 11518000,
      "step": 5510
    },
    {
      "epoch": 0.8997471245615466,
      "grad_norm": 8.125,
      "learning_rate": 4.7692057112719193e-05,
      "loss": 2.5138,
      "num_input_tokens_seen": 11528592,
      "step": 5515
    },
    {
      "epoch": 0.900562851782364,
      "grad_norm": 7.125,
      "learning_rate": 4.7687935379953234e-05,
      "loss": 2.3693,
      "num_input_tokens_seen": 11538432,
      "step": 5520
    },
    {
      "epoch": 0.9013785790031813,
      "grad_norm": 8.25,
      "learning_rate": 4.7683810148422534e-05,
      "loss": 1.582,
      "num_input_tokens_seen": 11548352,
      "step": 5525
    },
    {
      "epoch": 0.9021943062239987,
      "grad_norm": 0.515625,
      "learning_rate": 4.767968141876324e-05,
      "loss": 3.2872,
      "num_input_tokens_seen": 11558896,
      "step": 5530
    },
    {
      "epoch": 0.903010033444816,
      "grad_norm": 5.21875,
      "learning_rate": 4.767554919161207e-05,
      "loss": 1.6834,
      "num_input_tokens_seen": 11568304,
      "step": 5535
    },
    {
      "epoch": 0.9038257606656334,
      "grad_norm": 7.03125,
      "learning_rate": 4.767141346760624e-05,
      "loss": 1.4889,
      "num_input_tokens_seen": 11579392,
      "step": 5540
    },
    {
      "epoch": 0.9046414878864508,
      "grad_norm": 6.0625,
      "learning_rate": 4.766727424738356e-05,
      "loss": 4.0811,
      "num_input_tokens_seen": 11590048,
      "step": 5545
    },
    {
      "epoch": 0.9054572151072682,
      "grad_norm": 5.0625,
      "learning_rate": 4.7663131531582325e-05,
      "loss": 1.5634,
      "num_input_tokens_seen": 11601216,
      "step": 5550
    },
    {
      "epoch": 0.9062729423280855,
      "grad_norm": 3.890625,
      "learning_rate": 4.765898532084142e-05,
      "loss": 3.1074,
      "num_input_tokens_seen": 11611808,
      "step": 5555
    },
    {
      "epoch": 0.9070886695489029,
      "grad_norm": 6.5,
      "learning_rate": 4.765483561580022e-05,
      "loss": 2.5083,
      "num_input_tokens_seen": 11623408,
      "step": 5560
    },
    {
      "epoch": 0.9079043967697202,
      "grad_norm": 6.3125,
      "learning_rate": 4.7650682417098666e-05,
      "loss": 2.079,
      "num_input_tokens_seen": 11632416,
      "step": 5565
    },
    {
      "epoch": 0.9087201239905376,
      "grad_norm": 9.8125,
      "learning_rate": 4.7646525725377244e-05,
      "loss": 2.4267,
      "num_input_tokens_seen": 11643248,
      "step": 5570
    },
    {
      "epoch": 0.9095358512113549,
      "grad_norm": 4.53125,
      "learning_rate": 4.764236554127696e-05,
      "loss": 3.2853,
      "num_input_tokens_seen": 11654832,
      "step": 5575
    },
    {
      "epoch": 0.9103515784321723,
      "grad_norm": 3.875,
      "learning_rate": 4.7638201865439356e-05,
      "loss": 2.9913,
      "num_input_tokens_seen": 11665696,
      "step": 5580
    },
    {
      "epoch": 0.9111673056529896,
      "grad_norm": 14.75,
      "learning_rate": 4.7634034698506545e-05,
      "loss": 2.9404,
      "num_input_tokens_seen": 11675472,
      "step": 5585
    },
    {
      "epoch": 0.911983032873807,
      "grad_norm": 8.0625,
      "learning_rate": 4.762986404112115e-05,
      "loss": 2.857,
      "num_input_tokens_seen": 11685872,
      "step": 5590
    },
    {
      "epoch": 0.9127987600946244,
      "grad_norm": 7.8125,
      "learning_rate": 4.762568989392633e-05,
      "loss": 2.1901,
      "num_input_tokens_seen": 11696304,
      "step": 5595
    },
    {
      "epoch": 0.9136144873154417,
      "grad_norm": 5.75,
      "learning_rate": 4.76215122575658e-05,
      "loss": 1.8036,
      "num_input_tokens_seen": 11707328,
      "step": 5600
    },
    {
      "epoch": 0.9136144873154417,
      "eval_loss": 2.537649393081665,
      "eval_runtime": 134.8687,
      "eval_samples_per_second": 20.205,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 11707328,
      "step": 5600
    },
    {
      "epoch": 0.9144302145362591,
      "grad_norm": 3.78125,
      "learning_rate": 4.7617331132683795e-05,
      "loss": 3.3887,
      "num_input_tokens_seen": 11717936,
      "step": 5605
    },
    {
      "epoch": 0.9152459417570764,
      "grad_norm": 16.375,
      "learning_rate": 4.7613146519925105e-05,
      "loss": 2.5385,
      "num_input_tokens_seen": 11728688,
      "step": 5610
    },
    {
      "epoch": 0.9160616689778938,
      "grad_norm": 2.21875,
      "learning_rate": 4.7608958419935045e-05,
      "loss": 2.6724,
      "num_input_tokens_seen": 11738016,
      "step": 5615
    },
    {
      "epoch": 0.9168773961987111,
      "grad_norm": 6.96875,
      "learning_rate": 4.760476683335948e-05,
      "loss": 2.8947,
      "num_input_tokens_seen": 11747520,
      "step": 5620
    },
    {
      "epoch": 0.9176931234195285,
      "grad_norm": 9.25,
      "learning_rate": 4.760057176084479e-05,
      "loss": 2.9242,
      "num_input_tokens_seen": 11757344,
      "step": 5625
    },
    {
      "epoch": 0.9185088506403458,
      "grad_norm": 0.2431640625,
      "learning_rate": 4.759637320303793e-05,
      "loss": 3.6776,
      "num_input_tokens_seen": 11768352,
      "step": 5630
    },
    {
      "epoch": 0.9193245778611632,
      "grad_norm": 15.625,
      "learning_rate": 4.759217116058635e-05,
      "loss": 2.0023,
      "num_input_tokens_seen": 11777824,
      "step": 5635
    },
    {
      "epoch": 0.9201403050819806,
      "grad_norm": 3.046875,
      "learning_rate": 4.758796563413807e-05,
      "loss": 2.2695,
      "num_input_tokens_seen": 11788432,
      "step": 5640
    },
    {
      "epoch": 0.920956032302798,
      "grad_norm": 2.015625,
      "learning_rate": 4.758375662434163e-05,
      "loss": 2.4233,
      "num_input_tokens_seen": 11798352,
      "step": 5645
    },
    {
      "epoch": 0.9217717595236153,
      "grad_norm": 4.125,
      "learning_rate": 4.7579544131846114e-05,
      "loss": 2.1305,
      "num_input_tokens_seen": 11807984,
      "step": 5650
    },
    {
      "epoch": 0.9225874867444327,
      "grad_norm": 7.84375,
      "learning_rate": 4.757532815730114e-05,
      "loss": 3.3345,
      "num_input_tokens_seen": 11818448,
      "step": 5655
    },
    {
      "epoch": 0.92340321396525,
      "grad_norm": 8.1875,
      "learning_rate": 4.7571108701356865e-05,
      "loss": 2.4339,
      "num_input_tokens_seen": 11828128,
      "step": 5660
    },
    {
      "epoch": 0.9242189411860674,
      "grad_norm": 2.765625,
      "learning_rate": 4.756688576466398e-05,
      "loss": 1.7845,
      "num_input_tokens_seen": 11839344,
      "step": 5665
    },
    {
      "epoch": 0.9250346684068848,
      "grad_norm": 3.609375,
      "learning_rate": 4.756265934787372e-05,
      "loss": 2.8798,
      "num_input_tokens_seen": 11849280,
      "step": 5670
    },
    {
      "epoch": 0.9258503956277021,
      "grad_norm": 8.375,
      "learning_rate": 4.755842945163785e-05,
      "loss": 2.237,
      "num_input_tokens_seen": 11859296,
      "step": 5675
    },
    {
      "epoch": 0.9266661228485195,
      "grad_norm": 13.375,
      "learning_rate": 4.755419607660867e-05,
      "loss": 2.4382,
      "num_input_tokens_seen": 11870304,
      "step": 5680
    },
    {
      "epoch": 0.9274818500693368,
      "grad_norm": 10.5,
      "learning_rate": 4.7549959223439016e-05,
      "loss": 0.7031,
      "num_input_tokens_seen": 11880736,
      "step": 5685
    },
    {
      "epoch": 0.9282975772901542,
      "grad_norm": 4.0,
      "learning_rate": 4.754571889278228e-05,
      "loss": 1.8747,
      "num_input_tokens_seen": 11891648,
      "step": 5690
    },
    {
      "epoch": 0.9291133045109715,
      "grad_norm": 7.65625,
      "learning_rate": 4.754147508529235e-05,
      "loss": 2.5913,
      "num_input_tokens_seen": 11902480,
      "step": 5695
    },
    {
      "epoch": 0.9299290317317889,
      "grad_norm": 12.5,
      "learning_rate": 4.75372278016237e-05,
      "loss": 3.1828,
      "num_input_tokens_seen": 11912592,
      "step": 5700
    },
    {
      "epoch": 0.9307447589526062,
      "grad_norm": 11.8125,
      "learning_rate": 4.753297704243129e-05,
      "loss": 2.8244,
      "num_input_tokens_seen": 11923024,
      "step": 5705
    },
    {
      "epoch": 0.9315604861734236,
      "grad_norm": 8.375,
      "learning_rate": 4.752872280837066e-05,
      "loss": 3.1457,
      "num_input_tokens_seen": 11933984,
      "step": 5710
    },
    {
      "epoch": 0.9323762133942409,
      "grad_norm": 8.875,
      "learning_rate": 4.752446510009786e-05,
      "loss": 2.2456,
      "num_input_tokens_seen": 11945744,
      "step": 5715
    },
    {
      "epoch": 0.9331919406150583,
      "grad_norm": 3.3125,
      "learning_rate": 4.7520203918269476e-05,
      "loss": 3.4833,
      "num_input_tokens_seen": 11955456,
      "step": 5720
    },
    {
      "epoch": 0.9340076678358756,
      "grad_norm": 5.40625,
      "learning_rate": 4.751593926354265e-05,
      "loss": 3.3501,
      "num_input_tokens_seen": 11965920,
      "step": 5725
    },
    {
      "epoch": 0.9348233950566931,
      "grad_norm": 10.3125,
      "learning_rate": 4.751167113657503e-05,
      "loss": 2.6175,
      "num_input_tokens_seen": 11975232,
      "step": 5730
    },
    {
      "epoch": 0.9356391222775104,
      "grad_norm": 10.5625,
      "learning_rate": 4.7507399538024834e-05,
      "loss": 2.5082,
      "num_input_tokens_seen": 11985664,
      "step": 5735
    },
    {
      "epoch": 0.9364548494983278,
      "grad_norm": 9.4375,
      "learning_rate": 4.750312446855077e-05,
      "loss": 3.6315,
      "num_input_tokens_seen": 11996208,
      "step": 5740
    },
    {
      "epoch": 0.9372705767191452,
      "grad_norm": 9.375,
      "learning_rate": 4.749884592881212e-05,
      "loss": 3.213,
      "num_input_tokens_seen": 12006992,
      "step": 5745
    },
    {
      "epoch": 0.9380863039399625,
      "grad_norm": 5.125,
      "learning_rate": 4.74945639194687e-05,
      "loss": 1.9797,
      "num_input_tokens_seen": 12018064,
      "step": 5750
    },
    {
      "epoch": 0.9389020311607799,
      "grad_norm": 6.25,
      "learning_rate": 4.749027844118083e-05,
      "loss": 2.7685,
      "num_input_tokens_seen": 12028128,
      "step": 5755
    },
    {
      "epoch": 0.9397177583815972,
      "grad_norm": 7.21875,
      "learning_rate": 4.7485989494609395e-05,
      "loss": 2.4822,
      "num_input_tokens_seen": 12040016,
      "step": 5760
    },
    {
      "epoch": 0.9405334856024146,
      "grad_norm": 10.3125,
      "learning_rate": 4.748169708041581e-05,
      "loss": 3.1446,
      "num_input_tokens_seen": 12050608,
      "step": 5765
    },
    {
      "epoch": 0.9413492128232319,
      "grad_norm": 2.09375,
      "learning_rate": 4.7477401199262004e-05,
      "loss": 1.9529,
      "num_input_tokens_seen": 12061296,
      "step": 5770
    },
    {
      "epoch": 0.9421649400440493,
      "grad_norm": 9.6875,
      "learning_rate": 4.747310185181048e-05,
      "loss": 2.2417,
      "num_input_tokens_seen": 12072000,
      "step": 5775
    },
    {
      "epoch": 0.9429806672648666,
      "grad_norm": 11.75,
      "learning_rate": 4.746879903872422e-05,
      "loss": 1.6819,
      "num_input_tokens_seen": 12080992,
      "step": 5780
    },
    {
      "epoch": 0.943796394485684,
      "grad_norm": 1.7265625,
      "learning_rate": 4.746449276066679e-05,
      "loss": 1.9088,
      "num_input_tokens_seen": 12090272,
      "step": 5785
    },
    {
      "epoch": 0.9446121217065013,
      "grad_norm": 4.8125,
      "learning_rate": 4.746018301830227e-05,
      "loss": 2.6754,
      "num_input_tokens_seen": 12100528,
      "step": 5790
    },
    {
      "epoch": 0.9454278489273187,
      "grad_norm": 11.1875,
      "learning_rate": 4.7455869812295275e-05,
      "loss": 4.3559,
      "num_input_tokens_seen": 12110352,
      "step": 5795
    },
    {
      "epoch": 0.946243576148136,
      "grad_norm": 3.5625,
      "learning_rate": 4.7451553143310964e-05,
      "loss": 2.9821,
      "num_input_tokens_seen": 12120624,
      "step": 5800
    },
    {
      "epoch": 0.946243576148136,
      "eval_loss": 2.555554151535034,
      "eval_runtime": 135.0115,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 12120624,
      "step": 5800
    },
    {
      "epoch": 0.9470593033689534,
      "grad_norm": 5.78125,
      "learning_rate": 4.744723301201501e-05,
      "loss": 1.9478,
      "num_input_tokens_seen": 12131440,
      "step": 5805
    },
    {
      "epoch": 0.9478750305897707,
      "grad_norm": 8.625,
      "learning_rate": 4.744290941907364e-05,
      "loss": 2.1477,
      "num_input_tokens_seen": 12141856,
      "step": 5810
    },
    {
      "epoch": 0.9486907578105881,
      "grad_norm": 4.875,
      "learning_rate": 4.7438582365153594e-05,
      "loss": 2.3477,
      "num_input_tokens_seen": 12152512,
      "step": 5815
    },
    {
      "epoch": 0.9495064850314054,
      "grad_norm": 10.375,
      "learning_rate": 4.743425185092217e-05,
      "loss": 2.9303,
      "num_input_tokens_seen": 12163872,
      "step": 5820
    },
    {
      "epoch": 0.9503222122522229,
      "grad_norm": 8.125,
      "learning_rate": 4.742991787704719e-05,
      "loss": 3.2727,
      "num_input_tokens_seen": 12175536,
      "step": 5825
    },
    {
      "epoch": 0.9511379394730403,
      "grad_norm": 8.375,
      "learning_rate": 4.7425580444196994e-05,
      "loss": 2.2856,
      "num_input_tokens_seen": 12186624,
      "step": 5830
    },
    {
      "epoch": 0.9519536666938576,
      "grad_norm": 5.875,
      "learning_rate": 4.742123955304048e-05,
      "loss": 2.6641,
      "num_input_tokens_seen": 12196976,
      "step": 5835
    },
    {
      "epoch": 0.952769393914675,
      "grad_norm": 6.3125,
      "learning_rate": 4.741689520424706e-05,
      "loss": 2.0776,
      "num_input_tokens_seen": 12206784,
      "step": 5840
    },
    {
      "epoch": 0.9535851211354923,
      "grad_norm": 5.65625,
      "learning_rate": 4.741254739848669e-05,
      "loss": 1.2001,
      "num_input_tokens_seen": 12217504,
      "step": 5845
    },
    {
      "epoch": 0.9544008483563097,
      "grad_norm": 8.75,
      "learning_rate": 4.740819613642987e-05,
      "loss": 3.2334,
      "num_input_tokens_seen": 12227360,
      "step": 5850
    },
    {
      "epoch": 0.955216575577127,
      "grad_norm": 4.53125,
      "learning_rate": 4.74038414187476e-05,
      "loss": 2.1254,
      "num_input_tokens_seen": 12237344,
      "step": 5855
    },
    {
      "epoch": 0.9560323027979444,
      "grad_norm": 15.0625,
      "learning_rate": 4.739948324611144e-05,
      "loss": 2.0956,
      "num_input_tokens_seen": 12248400,
      "step": 5860
    },
    {
      "epoch": 0.9568480300187617,
      "grad_norm": 0.73828125,
      "learning_rate": 4.7395121619193465e-05,
      "loss": 2.8317,
      "num_input_tokens_seen": 12258224,
      "step": 5865
    },
    {
      "epoch": 0.9576637572395791,
      "grad_norm": 8.25,
      "learning_rate": 4.7390756538666313e-05,
      "loss": 1.5871,
      "num_input_tokens_seen": 12268736,
      "step": 5870
    },
    {
      "epoch": 0.9584794844603964,
      "grad_norm": 7.40625,
      "learning_rate": 4.738638800520311e-05,
      "loss": 2.5152,
      "num_input_tokens_seen": 12278496,
      "step": 5875
    },
    {
      "epoch": 0.9592952116812138,
      "grad_norm": 7.59375,
      "learning_rate": 4.738201601947757e-05,
      "loss": 2.0659,
      "num_input_tokens_seen": 12289088,
      "step": 5880
    },
    {
      "epoch": 0.9601109389020311,
      "grad_norm": 4.78125,
      "learning_rate": 4.7377640582163876e-05,
      "loss": 3.9642,
      "num_input_tokens_seen": 12299728,
      "step": 5885
    },
    {
      "epoch": 0.9609266661228485,
      "grad_norm": 12.3125,
      "learning_rate": 4.7373261693936786e-05,
      "loss": 3.7051,
      "num_input_tokens_seen": 12309520,
      "step": 5890
    },
    {
      "epoch": 0.9617423933436658,
      "grad_norm": 0.12060546875,
      "learning_rate": 4.7368879355471595e-05,
      "loss": 3.1656,
      "num_input_tokens_seen": 12319776,
      "step": 5895
    },
    {
      "epoch": 0.9625581205644832,
      "grad_norm": 6.59375,
      "learning_rate": 4.736449356744409e-05,
      "loss": 2.453,
      "num_input_tokens_seen": 12329168,
      "step": 5900
    },
    {
      "epoch": 0.9633738477853006,
      "grad_norm": 6.9375,
      "learning_rate": 4.736010433053064e-05,
      "loss": 1.4093,
      "num_input_tokens_seen": 12339696,
      "step": 5905
    },
    {
      "epoch": 0.9641895750061179,
      "grad_norm": 12.0,
      "learning_rate": 4.73557116454081e-05,
      "loss": 3.2938,
      "num_input_tokens_seen": 12349840,
      "step": 5910
    },
    {
      "epoch": 0.9650053022269353,
      "grad_norm": 4.875,
      "learning_rate": 4.735131551275389e-05,
      "loss": 3.278,
      "num_input_tokens_seen": 12361088,
      "step": 5915
    },
    {
      "epoch": 0.9658210294477527,
      "grad_norm": 8.375,
      "learning_rate": 4.734691593324594e-05,
      "loss": 2.0114,
      "num_input_tokens_seen": 12372080,
      "step": 5920
    },
    {
      "epoch": 0.9666367566685701,
      "grad_norm": 6.90625,
      "learning_rate": 4.734251290756272e-05,
      "loss": 1.6471,
      "num_input_tokens_seen": 12382752,
      "step": 5925
    },
    {
      "epoch": 0.9674524838893874,
      "grad_norm": 6.65625,
      "learning_rate": 4.7338106436383246e-05,
      "loss": 3.7974,
      "num_input_tokens_seen": 12392704,
      "step": 5930
    },
    {
      "epoch": 0.9682682111102048,
      "grad_norm": 7.125,
      "learning_rate": 4.733369652038703e-05,
      "loss": 1.705,
      "num_input_tokens_seen": 12404128,
      "step": 5935
    },
    {
      "epoch": 0.9690839383310221,
      "grad_norm": 6.625,
      "learning_rate": 4.7329283160254156e-05,
      "loss": 2.2585,
      "num_input_tokens_seen": 12415936,
      "step": 5940
    },
    {
      "epoch": 0.9698996655518395,
      "grad_norm": 5.28125,
      "learning_rate": 4.732486635666521e-05,
      "loss": 1.4724,
      "num_input_tokens_seen": 12427104,
      "step": 5945
    },
    {
      "epoch": 0.9707153927726568,
      "grad_norm": 1.3515625,
      "learning_rate": 4.732044611030132e-05,
      "loss": 1.7289,
      "num_input_tokens_seen": 12437728,
      "step": 5950
    },
    {
      "epoch": 0.9715311199934742,
      "grad_norm": 3.9375,
      "learning_rate": 4.731602242184414e-05,
      "loss": 2.0069,
      "num_input_tokens_seen": 12447968,
      "step": 5955
    },
    {
      "epoch": 0.9723468472142915,
      "grad_norm": 16.625,
      "learning_rate": 4.7311595291975864e-05,
      "loss": 1.7803,
      "num_input_tokens_seen": 12459072,
      "step": 5960
    },
    {
      "epoch": 0.9731625744351089,
      "grad_norm": 3.46875,
      "learning_rate": 4.7307164721379216e-05,
      "loss": 1.665,
      "num_input_tokens_seen": 12470592,
      "step": 5965
    },
    {
      "epoch": 0.9739783016559262,
      "grad_norm": 6.5625,
      "learning_rate": 4.730273071073743e-05,
      "loss": 2.1494,
      "num_input_tokens_seen": 12480864,
      "step": 5970
    },
    {
      "epoch": 0.9747940288767436,
      "grad_norm": 11.0625,
      "learning_rate": 4.729829326073429e-05,
      "loss": 2.8779,
      "num_input_tokens_seen": 12490672,
      "step": 5975
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 14.125,
      "learning_rate": 4.7293852372054126e-05,
      "loss": 3.0805,
      "num_input_tokens_seen": 12501840,
      "step": 5980
    },
    {
      "epoch": 0.9764254833183783,
      "grad_norm": 5.59375,
      "learning_rate": 4.728940804538176e-05,
      "loss": 2.6568,
      "num_input_tokens_seen": 12511216,
      "step": 5985
    },
    {
      "epoch": 0.9772412105391957,
      "grad_norm": 5.1875,
      "learning_rate": 4.7284960281402556e-05,
      "loss": 1.5319,
      "num_input_tokens_seen": 12521920,
      "step": 5990
    },
    {
      "epoch": 0.978056937760013,
      "grad_norm": 10.625,
      "learning_rate": 4.728050908080244e-05,
      "loss": 3.5632,
      "num_input_tokens_seen": 12533200,
      "step": 5995
    },
    {
      "epoch": 0.9788726649808304,
      "grad_norm": 14.3125,
      "learning_rate": 4.727605444426782e-05,
      "loss": 1.8289,
      "num_input_tokens_seen": 12542416,
      "step": 6000
    },
    {
      "epoch": 0.9788726649808304,
      "eval_loss": 2.549168348312378,
      "eval_runtime": 134.811,
      "eval_samples_per_second": 20.213,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 12542416,
      "step": 6000
    },
    {
      "epoch": 0.9796883922016477,
      "grad_norm": 3.5,
      "learning_rate": 4.727159637248567e-05,
      "loss": 1.621,
      "num_input_tokens_seen": 12553152,
      "step": 6005
    },
    {
      "epoch": 0.9805041194224652,
      "grad_norm": 6.40625,
      "learning_rate": 4.7267134866143474e-05,
      "loss": 3.041,
      "num_input_tokens_seen": 12562960,
      "step": 6010
    },
    {
      "epoch": 0.9813198466432825,
      "grad_norm": 6.4375,
      "learning_rate": 4.726266992592926e-05,
      "loss": 4.6699,
      "num_input_tokens_seen": 12572144,
      "step": 6015
    },
    {
      "epoch": 0.9821355738640999,
      "grad_norm": 5.75,
      "learning_rate": 4.725820155253157e-05,
      "loss": 0.8568,
      "num_input_tokens_seen": 12583120,
      "step": 6020
    },
    {
      "epoch": 0.9829513010849172,
      "grad_norm": 10.25,
      "learning_rate": 4.725372974663948e-05,
      "loss": 3.5086,
      "num_input_tokens_seen": 12594208,
      "step": 6025
    },
    {
      "epoch": 0.9837670283057346,
      "grad_norm": 5.375,
      "learning_rate": 4.724925450894262e-05,
      "loss": 1.9006,
      "num_input_tokens_seen": 12605456,
      "step": 6030
    },
    {
      "epoch": 0.984582755526552,
      "grad_norm": 8.4375,
      "learning_rate": 4.72447758401311e-05,
      "loss": 3.3245,
      "num_input_tokens_seen": 12615872,
      "step": 6035
    },
    {
      "epoch": 0.9853984827473693,
      "grad_norm": 14.375,
      "learning_rate": 4.7240293740895616e-05,
      "loss": 2.5671,
      "num_input_tokens_seen": 12626144,
      "step": 6040
    },
    {
      "epoch": 0.9862142099681867,
      "grad_norm": 7.5,
      "learning_rate": 4.723580821192733e-05,
      "loss": 2.8449,
      "num_input_tokens_seen": 12636112,
      "step": 6045
    },
    {
      "epoch": 0.987029937189004,
      "grad_norm": 8.75,
      "learning_rate": 4.7231319253917996e-05,
      "loss": 2.2511,
      "num_input_tokens_seen": 12645168,
      "step": 6050
    },
    {
      "epoch": 0.9878456644098214,
      "grad_norm": 3.234375,
      "learning_rate": 4.722682686755986e-05,
      "loss": 2.184,
      "num_input_tokens_seen": 12656736,
      "step": 6055
    },
    {
      "epoch": 0.9886613916306387,
      "grad_norm": 4.6875,
      "learning_rate": 4.722233105354569e-05,
      "loss": 2.0117,
      "num_input_tokens_seen": 12667472,
      "step": 6060
    },
    {
      "epoch": 0.9894771188514561,
      "grad_norm": 12.4375,
      "learning_rate": 4.7217831812568815e-05,
      "loss": 2.4944,
      "num_input_tokens_seen": 12678384,
      "step": 6065
    },
    {
      "epoch": 0.9902928460722734,
      "grad_norm": 3.609375,
      "learning_rate": 4.721332914532307e-05,
      "loss": 1.823,
      "num_input_tokens_seen": 12689200,
      "step": 6070
    },
    {
      "epoch": 0.9911085732930908,
      "grad_norm": 10.375,
      "learning_rate": 4.720882305250281e-05,
      "loss": 2.7774,
      "num_input_tokens_seen": 12699568,
      "step": 6075
    },
    {
      "epoch": 0.9919243005139081,
      "grad_norm": 7.78125,
      "learning_rate": 4.720431353480295e-05,
      "loss": 2.2263,
      "num_input_tokens_seen": 12710336,
      "step": 6080
    },
    {
      "epoch": 0.9927400277347255,
      "grad_norm": 8.4375,
      "learning_rate": 4.719980059291891e-05,
      "loss": 3.1793,
      "num_input_tokens_seen": 12719392,
      "step": 6085
    },
    {
      "epoch": 0.9935557549555428,
      "grad_norm": 0.06201171875,
      "learning_rate": 4.7195284227546634e-05,
      "loss": 1.8151,
      "num_input_tokens_seen": 12729424,
      "step": 6090
    },
    {
      "epoch": 0.9943714821763602,
      "grad_norm": 10.125,
      "learning_rate": 4.7190764439382604e-05,
      "loss": 2.6556,
      "num_input_tokens_seen": 12741248,
      "step": 6095
    },
    {
      "epoch": 0.9951872093971775,
      "grad_norm": 9.125,
      "learning_rate": 4.7186241229123826e-05,
      "loss": 2.8441,
      "num_input_tokens_seen": 12751184,
      "step": 6100
    },
    {
      "epoch": 0.996002936617995,
      "grad_norm": 11.875,
      "learning_rate": 4.718171459746785e-05,
      "loss": 1.7279,
      "num_input_tokens_seen": 12760464,
      "step": 6105
    },
    {
      "epoch": 0.9968186638388123,
      "grad_norm": 14.0625,
      "learning_rate": 4.717718454511273e-05,
      "loss": 1.9571,
      "num_input_tokens_seen": 12770672,
      "step": 6110
    },
    {
      "epoch": 0.9976343910596297,
      "grad_norm": 4.46875,
      "learning_rate": 4.7172651072757056e-05,
      "loss": 1.6713,
      "num_input_tokens_seen": 12781456,
      "step": 6115
    },
    {
      "epoch": 0.998450118280447,
      "grad_norm": 5.5625,
      "learning_rate": 4.7168114181099945e-05,
      "loss": 1.3262,
      "num_input_tokens_seen": 12792592,
      "step": 6120
    },
    {
      "epoch": 0.9992658455012644,
      "grad_norm": 6.28125,
      "learning_rate": 4.716357387084105e-05,
      "loss": 2.7664,
      "num_input_tokens_seen": 12803840,
      "step": 6125
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.302734375,
      "learning_rate": 4.715903014268054e-05,
      "loss": 2.6198,
      "num_input_tokens_seen": 12813920,
      "step": 6130
    },
    {
      "epoch": 1.0008157272208174,
      "grad_norm": 11.9375,
      "learning_rate": 4.715448299731911e-05,
      "loss": 2.2941,
      "num_input_tokens_seen": 12824960,
      "step": 6135
    },
    {
      "epoch": 1.0016314544416347,
      "grad_norm": 6.5625,
      "learning_rate": 4.7149932435457986e-05,
      "loss": 1.086,
      "num_input_tokens_seen": 12836112,
      "step": 6140
    },
    {
      "epoch": 1.002447181662452,
      "grad_norm": 4.8125,
      "learning_rate": 4.714537845779894e-05,
      "loss": 1.5446,
      "num_input_tokens_seen": 12846560,
      "step": 6145
    },
    {
      "epoch": 1.0032629088832694,
      "grad_norm": 2.71875,
      "learning_rate": 4.714082106504423e-05,
      "loss": 2.6856,
      "num_input_tokens_seen": 12858208,
      "step": 6150
    },
    {
      "epoch": 1.0040786361040868,
      "grad_norm": 8.75,
      "learning_rate": 4.713626025789667e-05,
      "loss": 1.6021,
      "num_input_tokens_seen": 12868608,
      "step": 6155
    },
    {
      "epoch": 1.0048943633249041,
      "grad_norm": 8.8125,
      "learning_rate": 4.7131696037059606e-05,
      "loss": 0.9059,
      "num_input_tokens_seen": 12878320,
      "step": 6160
    },
    {
      "epoch": 1.0057100905457215,
      "grad_norm": 2.3125,
      "learning_rate": 4.712712840323689e-05,
      "loss": 1.9529,
      "num_input_tokens_seen": 12889248,
      "step": 6165
    },
    {
      "epoch": 1.0065258177665388,
      "grad_norm": 4.59375,
      "learning_rate": 4.71225573571329e-05,
      "loss": 1.3517,
      "num_input_tokens_seen": 12900432,
      "step": 6170
    },
    {
      "epoch": 1.0073415449873562,
      "grad_norm": 2.921875,
      "learning_rate": 4.711798289945256e-05,
      "loss": 2.3083,
      "num_input_tokens_seen": 12910800,
      "step": 6175
    },
    {
      "epoch": 1.0081572722081735,
      "grad_norm": 6.4375,
      "learning_rate": 4.71134050309013e-05,
      "loss": 2.4419,
      "num_input_tokens_seen": 12921872,
      "step": 6180
    },
    {
      "epoch": 1.0089729994289909,
      "grad_norm": 7.75,
      "learning_rate": 4.710882375218509e-05,
      "loss": 2.385,
      "num_input_tokens_seen": 12933024,
      "step": 6185
    },
    {
      "epoch": 1.0097887266498082,
      "grad_norm": 0.515625,
      "learning_rate": 4.7104239064010424e-05,
      "loss": 2.1937,
      "num_input_tokens_seen": 12943536,
      "step": 6190
    },
    {
      "epoch": 1.0106044538706256,
      "grad_norm": 10.125,
      "learning_rate": 4.709965096708432e-05,
      "loss": 2.363,
      "num_input_tokens_seen": 12952080,
      "step": 6195
    },
    {
      "epoch": 1.011420181091443,
      "grad_norm": 5.875,
      "learning_rate": 4.709505946211431e-05,
      "loss": 1.8497,
      "num_input_tokens_seen": 12963008,
      "step": 6200
    },
    {
      "epoch": 1.011420181091443,
      "eval_loss": 2.5441558361053467,
      "eval_runtime": 134.972,
      "eval_samples_per_second": 20.189,
      "eval_steps_per_second": 10.098,
      "num_input_tokens_seen": 12963008,
      "step": 6200
    },
    {
      "epoch": 1.0122359083122603,
      "grad_norm": 2.453125,
      "learning_rate": 4.709046454980846e-05,
      "loss": 1.761,
      "num_input_tokens_seen": 12974864,
      "step": 6205
    },
    {
      "epoch": 1.0130516355330776,
      "grad_norm": 5.21875,
      "learning_rate": 4.708586623087538e-05,
      "loss": 1.9405,
      "num_input_tokens_seen": 12985216,
      "step": 6210
    },
    {
      "epoch": 1.013867362753895,
      "grad_norm": 4.625,
      "learning_rate": 4.708126450602418e-05,
      "loss": 3.8876,
      "num_input_tokens_seen": 12996976,
      "step": 6215
    },
    {
      "epoch": 1.0146830899747123,
      "grad_norm": 6.03125,
      "learning_rate": 4.7076659375964495e-05,
      "loss": 2.7066,
      "num_input_tokens_seen": 13007952,
      "step": 6220
    },
    {
      "epoch": 1.01549881719553,
      "grad_norm": 7.09375,
      "learning_rate": 4.707205084140651e-05,
      "loss": 2.507,
      "num_input_tokens_seen": 13016672,
      "step": 6225
    },
    {
      "epoch": 1.0163145444163473,
      "grad_norm": 11.6875,
      "learning_rate": 4.7067438903060904e-05,
      "loss": 2.4329,
      "num_input_tokens_seen": 13027520,
      "step": 6230
    },
    {
      "epoch": 1.0171302716371646,
      "grad_norm": 10.5,
      "learning_rate": 4.70628235616389e-05,
      "loss": 2.6443,
      "num_input_tokens_seen": 13037552,
      "step": 6235
    },
    {
      "epoch": 1.017945998857982,
      "grad_norm": 8.625,
      "learning_rate": 4.7058204817852256e-05,
      "loss": 2.5943,
      "num_input_tokens_seen": 13047728,
      "step": 6240
    },
    {
      "epoch": 1.0187617260787993,
      "grad_norm": 8.4375,
      "learning_rate": 4.705358267241322e-05,
      "loss": 2.1596,
      "num_input_tokens_seen": 13058880,
      "step": 6245
    },
    {
      "epoch": 1.0195774532996167,
      "grad_norm": 11.125,
      "learning_rate": 4.704895712603459e-05,
      "loss": 2.7262,
      "num_input_tokens_seen": 13069856,
      "step": 6250
    },
    {
      "epoch": 1.020393180520434,
      "grad_norm": 7.5,
      "learning_rate": 4.704432817942969e-05,
      "loss": 3.3593,
      "num_input_tokens_seen": 13080160,
      "step": 6255
    },
    {
      "epoch": 1.0212089077412514,
      "grad_norm": 4.59375,
      "learning_rate": 4.703969583331236e-05,
      "loss": 1.729,
      "num_input_tokens_seen": 13090832,
      "step": 6260
    },
    {
      "epoch": 1.0220246349620687,
      "grad_norm": 4.0,
      "learning_rate": 4.7035060088396965e-05,
      "loss": 1.9941,
      "num_input_tokens_seen": 13100976,
      "step": 6265
    },
    {
      "epoch": 1.022840362182886,
      "grad_norm": 11.9375,
      "learning_rate": 4.703042094539839e-05,
      "loss": 3.7058,
      "num_input_tokens_seen": 13109904,
      "step": 6270
    },
    {
      "epoch": 1.0236560894037035,
      "grad_norm": 2.859375,
      "learning_rate": 4.702577840503206e-05,
      "loss": 1.8261,
      "num_input_tokens_seen": 13120032,
      "step": 6275
    },
    {
      "epoch": 1.0244718166245208,
      "grad_norm": 17.5,
      "learning_rate": 4.70211324680139e-05,
      "loss": 2.3398,
      "num_input_tokens_seen": 13130448,
      "step": 6280
    },
    {
      "epoch": 1.0252875438453382,
      "grad_norm": 1.71875,
      "learning_rate": 4.7016483135060386e-05,
      "loss": 1.472,
      "num_input_tokens_seen": 13142448,
      "step": 6285
    },
    {
      "epoch": 1.0261032710661555,
      "grad_norm": 2.171875,
      "learning_rate": 4.701183040688849e-05,
      "loss": 1.0879,
      "num_input_tokens_seen": 13152592,
      "step": 6290
    },
    {
      "epoch": 1.0269189982869729,
      "grad_norm": 5.09375,
      "learning_rate": 4.700717428421573e-05,
      "loss": 1.8312,
      "num_input_tokens_seen": 13163904,
      "step": 6295
    },
    {
      "epoch": 1.0277347255077902,
      "grad_norm": 3.71875,
      "learning_rate": 4.700251476776014e-05,
      "loss": 2.3739,
      "num_input_tokens_seen": 13174064,
      "step": 6300
    },
    {
      "epoch": 1.0285504527286076,
      "grad_norm": 9.0,
      "learning_rate": 4.699785185824026e-05,
      "loss": 2.4082,
      "num_input_tokens_seen": 13184528,
      "step": 6305
    },
    {
      "epoch": 1.029366179949425,
      "grad_norm": 4.4375,
      "learning_rate": 4.699318555637519e-05,
      "loss": 2.3398,
      "num_input_tokens_seen": 13194464,
      "step": 6310
    },
    {
      "epoch": 1.0301819071702423,
      "grad_norm": 10.1875,
      "learning_rate": 4.6988515862884525e-05,
      "loss": 1.8384,
      "num_input_tokens_seen": 13206272,
      "step": 6315
    },
    {
      "epoch": 1.0309976343910596,
      "grad_norm": 5.3125,
      "learning_rate": 4.698384277848838e-05,
      "loss": 3.2116,
      "num_input_tokens_seen": 13216944,
      "step": 6320
    },
    {
      "epoch": 1.031813361611877,
      "grad_norm": 4.46875,
      "learning_rate": 4.6979166303907425e-05,
      "loss": 2.2847,
      "num_input_tokens_seen": 13227664,
      "step": 6325
    },
    {
      "epoch": 1.0326290888326943,
      "grad_norm": 7.96875,
      "learning_rate": 4.697448643986281e-05,
      "loss": 2.4183,
      "num_input_tokens_seen": 13238832,
      "step": 6330
    },
    {
      "epoch": 1.0334448160535117,
      "grad_norm": 1.2109375,
      "learning_rate": 4.696980318707624e-05,
      "loss": 1.7859,
      "num_input_tokens_seen": 13248880,
      "step": 6335
    },
    {
      "epoch": 1.034260543274329,
      "grad_norm": 6.625,
      "learning_rate": 4.6965116546269924e-05,
      "loss": 1.8692,
      "num_input_tokens_seen": 13259408,
      "step": 6340
    },
    {
      "epoch": 1.0350762704951464,
      "grad_norm": 6.9375,
      "learning_rate": 4.6960426518166615e-05,
      "loss": 3.024,
      "num_input_tokens_seen": 13270432,
      "step": 6345
    },
    {
      "epoch": 1.0358919977159637,
      "grad_norm": 4.09375,
      "learning_rate": 4.6955733103489556e-05,
      "loss": 1.4392,
      "num_input_tokens_seen": 13282112,
      "step": 6350
    },
    {
      "epoch": 1.036707724936781,
      "grad_norm": 6.0625,
      "learning_rate": 4.695103630296255e-05,
      "loss": 2.0951,
      "num_input_tokens_seen": 13293248,
      "step": 6355
    },
    {
      "epoch": 1.0375234521575984,
      "grad_norm": 8.0625,
      "learning_rate": 4.694633611730988e-05,
      "loss": 2.2979,
      "num_input_tokens_seen": 13303488,
      "step": 6360
    },
    {
      "epoch": 1.0383391793784158,
      "grad_norm": 9.5,
      "learning_rate": 4.694163254725639e-05,
      "loss": 3.425,
      "num_input_tokens_seen": 13314848,
      "step": 6365
    },
    {
      "epoch": 1.0391549065992332,
      "grad_norm": 7.03125,
      "learning_rate": 4.693692559352743e-05,
      "loss": 4.1539,
      "num_input_tokens_seen": 13325712,
      "step": 6370
    },
    {
      "epoch": 1.0399706338200505,
      "grad_norm": 7.21875,
      "learning_rate": 4.693221525684886e-05,
      "loss": 2.5029,
      "num_input_tokens_seen": 13335664,
      "step": 6375
    },
    {
      "epoch": 1.0407863610408679,
      "grad_norm": 8.25,
      "learning_rate": 4.6927501537947084e-05,
      "loss": 1.6925,
      "num_input_tokens_seen": 13345760,
      "step": 6380
    },
    {
      "epoch": 1.0416020882616852,
      "grad_norm": 4.375,
      "learning_rate": 4.692278443754901e-05,
      "loss": 1.8803,
      "num_input_tokens_seen": 13356144,
      "step": 6385
    },
    {
      "epoch": 1.0424178154825026,
      "grad_norm": 3.0625,
      "learning_rate": 4.691806395638208e-05,
      "loss": 2.0043,
      "num_input_tokens_seen": 13366432,
      "step": 6390
    },
    {
      "epoch": 1.04323354270332,
      "grad_norm": 8.875,
      "learning_rate": 4.6913340095174255e-05,
      "loss": 1.9346,
      "num_input_tokens_seen": 13378800,
      "step": 6395
    },
    {
      "epoch": 1.0440492699241373,
      "grad_norm": 4.15625,
      "learning_rate": 4.690861285465399e-05,
      "loss": 1.2982,
      "num_input_tokens_seen": 13388336,
      "step": 6400
    },
    {
      "epoch": 1.0440492699241373,
      "eval_loss": 2.5316240787506104,
      "eval_runtime": 134.8601,
      "eval_samples_per_second": 20.206,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 13388336,
      "step": 6400
    },
    {
      "epoch": 1.0448649971449546,
      "grad_norm": 5.75,
      "learning_rate": 4.690388223555031e-05,
      "loss": 1.554,
      "num_input_tokens_seen": 13399088,
      "step": 6405
    },
    {
      "epoch": 1.0456807243657722,
      "grad_norm": 5.96875,
      "learning_rate": 4.689914823859273e-05,
      "loss": 1.4387,
      "num_input_tokens_seen": 13410384,
      "step": 6410
    },
    {
      "epoch": 1.0464964515865895,
      "grad_norm": 8.6875,
      "learning_rate": 4.689441086451129e-05,
      "loss": 2.3521,
      "num_input_tokens_seen": 13419216,
      "step": 6415
    },
    {
      "epoch": 1.047312178807407,
      "grad_norm": 9.875,
      "learning_rate": 4.688967011403655e-05,
      "loss": 4.2419,
      "num_input_tokens_seen": 13430224,
      "step": 6420
    },
    {
      "epoch": 1.0481279060282243,
      "grad_norm": 5.5,
      "learning_rate": 4.68849259878996e-05,
      "loss": 1.707,
      "num_input_tokens_seen": 13441424,
      "step": 6425
    },
    {
      "epoch": 1.0489436332490416,
      "grad_norm": 7.53125,
      "learning_rate": 4.6880178486832036e-05,
      "loss": 2.474,
      "num_input_tokens_seen": 13451760,
      "step": 6430
    },
    {
      "epoch": 1.049759360469859,
      "grad_norm": 4.78125,
      "learning_rate": 4.687542761156598e-05,
      "loss": 2.1315,
      "num_input_tokens_seen": 13462464,
      "step": 6435
    },
    {
      "epoch": 1.0505750876906763,
      "grad_norm": 8.3125,
      "learning_rate": 4.6870673362834096e-05,
      "loss": 3.3358,
      "num_input_tokens_seen": 13471968,
      "step": 6440
    },
    {
      "epoch": 1.0513908149114937,
      "grad_norm": 5.25,
      "learning_rate": 4.6865915741369526e-05,
      "loss": 3.8159,
      "num_input_tokens_seen": 13483856,
      "step": 6445
    },
    {
      "epoch": 1.052206542132311,
      "grad_norm": 8.3125,
      "learning_rate": 4.686115474790597e-05,
      "loss": 1.8372,
      "num_input_tokens_seen": 13494464,
      "step": 6450
    },
    {
      "epoch": 1.0530222693531284,
      "grad_norm": 12.25,
      "learning_rate": 4.685639038317762e-05,
      "loss": 3.3756,
      "num_input_tokens_seen": 13503664,
      "step": 6455
    },
    {
      "epoch": 1.0538379965739457,
      "grad_norm": 6.96875,
      "learning_rate": 4.685162264791921e-05,
      "loss": 2.6894,
      "num_input_tokens_seen": 13515824,
      "step": 6460
    },
    {
      "epoch": 1.054653723794763,
      "grad_norm": 4.34375,
      "learning_rate": 4.684685154286599e-05,
      "loss": 1.2832,
      "num_input_tokens_seen": 13525712,
      "step": 6465
    },
    {
      "epoch": 1.0554694510155804,
      "grad_norm": 8.125,
      "learning_rate": 4.684207706875371e-05,
      "loss": 2.6077,
      "num_input_tokens_seen": 13536176,
      "step": 6470
    },
    {
      "epoch": 1.0562851782363978,
      "grad_norm": 10.8125,
      "learning_rate": 4.683729922631866e-05,
      "loss": 4.0819,
      "num_input_tokens_seen": 13546704,
      "step": 6475
    },
    {
      "epoch": 1.0571009054572151,
      "grad_norm": 9.75,
      "learning_rate": 4.683251801629765e-05,
      "loss": 2.8638,
      "num_input_tokens_seen": 13557520,
      "step": 6480
    },
    {
      "epoch": 1.0579166326780325,
      "grad_norm": 8.3125,
      "learning_rate": 4.6827733439428e-05,
      "loss": 2.2073,
      "num_input_tokens_seen": 13568640,
      "step": 6485
    },
    {
      "epoch": 1.0587323598988498,
      "grad_norm": 11.875,
      "learning_rate": 4.682294549644754e-05,
      "loss": 2.9975,
      "num_input_tokens_seen": 13580016,
      "step": 6490
    },
    {
      "epoch": 1.0595480871196672,
      "grad_norm": 6.59375,
      "learning_rate": 4.681815418809464e-05,
      "loss": 1.5164,
      "num_input_tokens_seen": 13589744,
      "step": 6495
    },
    {
      "epoch": 1.0603638143404845,
      "grad_norm": 5.875,
      "learning_rate": 4.681335951510819e-05,
      "loss": 1.446,
      "num_input_tokens_seen": 13598896,
      "step": 6500
    },
    {
      "epoch": 1.061179541561302,
      "grad_norm": 11.1875,
      "learning_rate": 4.6808561478227576e-05,
      "loss": 3.2373,
      "num_input_tokens_seen": 13610352,
      "step": 6505
    },
    {
      "epoch": 1.0619952687821193,
      "grad_norm": 2.65625,
      "learning_rate": 4.680376007819271e-05,
      "loss": 1.783,
      "num_input_tokens_seen": 13621536,
      "step": 6510
    },
    {
      "epoch": 1.0628109960029366,
      "grad_norm": 10.375,
      "learning_rate": 4.679895531574405e-05,
      "loss": 1.3744,
      "num_input_tokens_seen": 13632704,
      "step": 6515
    },
    {
      "epoch": 1.063626723223754,
      "grad_norm": 8.4375,
      "learning_rate": 4.679414719162253e-05,
      "loss": 3.1847,
      "num_input_tokens_seen": 13642656,
      "step": 6520
    },
    {
      "epoch": 1.0644424504445713,
      "grad_norm": 4.90625,
      "learning_rate": 4.6789335706569635e-05,
      "loss": 2.4636,
      "num_input_tokens_seen": 13652800,
      "step": 6525
    },
    {
      "epoch": 1.0652581776653887,
      "grad_norm": 4.15625,
      "learning_rate": 4.678452086132734e-05,
      "loss": 2.388,
      "num_input_tokens_seen": 13664160,
      "step": 6530
    },
    {
      "epoch": 1.066073904886206,
      "grad_norm": 5.875,
      "learning_rate": 4.677970265663818e-05,
      "loss": 3.9124,
      "num_input_tokens_seen": 13675824,
      "step": 6535
    },
    {
      "epoch": 1.0668896321070234,
      "grad_norm": 4.4375,
      "learning_rate": 4.677488109324517e-05,
      "loss": 1.9941,
      "num_input_tokens_seen": 13684960,
      "step": 6540
    },
    {
      "epoch": 1.0677053593278407,
      "grad_norm": 18.875,
      "learning_rate": 4.6770056171891846e-05,
      "loss": 2.2253,
      "num_input_tokens_seen": 13696688,
      "step": 6545
    },
    {
      "epoch": 1.068521086548658,
      "grad_norm": 5.25,
      "learning_rate": 4.6765227893322286e-05,
      "loss": 3.2079,
      "num_input_tokens_seen": 13707440,
      "step": 6550
    },
    {
      "epoch": 1.0693368137694754,
      "grad_norm": 9.1875,
      "learning_rate": 4.676039625828107e-05,
      "loss": 2.862,
      "num_input_tokens_seen": 13718368,
      "step": 6555
    },
    {
      "epoch": 1.0701525409902928,
      "grad_norm": 4.59375,
      "learning_rate": 4.675556126751328e-05,
      "loss": 2.6244,
      "num_input_tokens_seen": 13729152,
      "step": 6560
    },
    {
      "epoch": 1.0709682682111101,
      "grad_norm": 7.75,
      "learning_rate": 4.6750722921764556e-05,
      "loss": 2.242,
      "num_input_tokens_seen": 13740672,
      "step": 6565
    },
    {
      "epoch": 1.0717839954319275,
      "grad_norm": 7.65625,
      "learning_rate": 4.674588122178102e-05,
      "loss": 2.2849,
      "num_input_tokens_seen": 13751360,
      "step": 6570
    },
    {
      "epoch": 1.0725997226527448,
      "grad_norm": 2.390625,
      "learning_rate": 4.674103616830931e-05,
      "loss": 1.8315,
      "num_input_tokens_seen": 13762896,
      "step": 6575
    },
    {
      "epoch": 1.0734154498735622,
      "grad_norm": 8.125,
      "learning_rate": 4.673618776209663e-05,
      "loss": 2.4437,
      "num_input_tokens_seen": 13774192,
      "step": 6580
    },
    {
      "epoch": 1.0742311770943795,
      "grad_norm": 7.75,
      "learning_rate": 4.673133600389063e-05,
      "loss": 3.4912,
      "num_input_tokens_seen": 13784304,
      "step": 6585
    },
    {
      "epoch": 1.075046904315197,
      "grad_norm": 7.5625,
      "learning_rate": 4.672648089443953e-05,
      "loss": 2.4513,
      "num_input_tokens_seen": 13794432,
      "step": 6590
    },
    {
      "epoch": 1.0758626315360145,
      "grad_norm": 8.5625,
      "learning_rate": 4.672162243449204e-05,
      "loss": 1.3882,
      "num_input_tokens_seen": 13805872,
      "step": 6595
    },
    {
      "epoch": 1.0766783587568316,
      "grad_norm": 15.75,
      "learning_rate": 4.67167606247974e-05,
      "loss": 2.5333,
      "num_input_tokens_seen": 13816224,
      "step": 6600
    },
    {
      "epoch": 1.0766783587568316,
      "eval_loss": 2.5479660034179688,
      "eval_runtime": 134.926,
      "eval_samples_per_second": 20.196,
      "eval_steps_per_second": 10.102,
      "num_input_tokens_seen": 13816224,
      "step": 6600
    },
    {
      "epoch": 1.0774940859776492,
      "grad_norm": 6.84375,
      "learning_rate": 4.671189546610536e-05,
      "loss": 1.9081,
      "num_input_tokens_seen": 13827264,
      "step": 6605
    },
    {
      "epoch": 1.0783098131984665,
      "grad_norm": 7.03125,
      "learning_rate": 4.67070269591662e-05,
      "loss": 1.6404,
      "num_input_tokens_seen": 13837344,
      "step": 6610
    },
    {
      "epoch": 1.0791255404192839,
      "grad_norm": 9.0,
      "learning_rate": 4.670215510473068e-05,
      "loss": 1.2935,
      "num_input_tokens_seen": 13848368,
      "step": 6615
    },
    {
      "epoch": 1.0799412676401012,
      "grad_norm": 11.0625,
      "learning_rate": 4.669727990355013e-05,
      "loss": 3.1689,
      "num_input_tokens_seen": 13858688,
      "step": 6620
    },
    {
      "epoch": 1.0807569948609186,
      "grad_norm": 3.75,
      "learning_rate": 4.669240135637635e-05,
      "loss": 2.5636,
      "num_input_tokens_seen": 13867728,
      "step": 6625
    },
    {
      "epoch": 1.081572722081736,
      "grad_norm": 8.6875,
      "learning_rate": 4.6687519463961675e-05,
      "loss": 3.2371,
      "num_input_tokens_seen": 13879216,
      "step": 6630
    },
    {
      "epoch": 1.0823884493025533,
      "grad_norm": 2.421875,
      "learning_rate": 4.668263422705896e-05,
      "loss": 2.1644,
      "num_input_tokens_seen": 13888976,
      "step": 6635
    },
    {
      "epoch": 1.0832041765233706,
      "grad_norm": 10.9375,
      "learning_rate": 4.667774564642156e-05,
      "loss": 3.4243,
      "num_input_tokens_seen": 13898064,
      "step": 6640
    },
    {
      "epoch": 1.084019903744188,
      "grad_norm": 7.21875,
      "learning_rate": 4.6672853722803365e-05,
      "loss": 1.8537,
      "num_input_tokens_seen": 13909136,
      "step": 6645
    },
    {
      "epoch": 1.0848356309650053,
      "grad_norm": 2.6875,
      "learning_rate": 4.666795845695877e-05,
      "loss": 1.807,
      "num_input_tokens_seen": 13919632,
      "step": 6650
    },
    {
      "epoch": 1.0856513581858227,
      "grad_norm": 6.96875,
      "learning_rate": 4.666305984964269e-05,
      "loss": 3.7924,
      "num_input_tokens_seen": 13929552,
      "step": 6655
    },
    {
      "epoch": 1.08646708540664,
      "grad_norm": 3.984375,
      "learning_rate": 4.6658157901610535e-05,
      "loss": 2.2932,
      "num_input_tokens_seen": 13941232,
      "step": 6660
    },
    {
      "epoch": 1.0872828126274574,
      "grad_norm": 12.75,
      "learning_rate": 4.665325261361826e-05,
      "loss": 2.8005,
      "num_input_tokens_seen": 13953264,
      "step": 6665
    },
    {
      "epoch": 1.0880985398482748,
      "grad_norm": 11.3125,
      "learning_rate": 4.664834398642232e-05,
      "loss": 2.9229,
      "num_input_tokens_seen": 13963248,
      "step": 6670
    },
    {
      "epoch": 1.0889142670690921,
      "grad_norm": 6.1875,
      "learning_rate": 4.6643432020779686e-05,
      "loss": 2.7717,
      "num_input_tokens_seen": 13973680,
      "step": 6675
    },
    {
      "epoch": 1.0897299942899095,
      "grad_norm": 9.5625,
      "learning_rate": 4.663851671744786e-05,
      "loss": 2.5093,
      "num_input_tokens_seen": 13984528,
      "step": 6680
    },
    {
      "epoch": 1.0905457215107268,
      "grad_norm": 6.90625,
      "learning_rate": 4.6633598077184815e-05,
      "loss": 0.7262,
      "num_input_tokens_seen": 13994032,
      "step": 6685
    },
    {
      "epoch": 1.0913614487315442,
      "grad_norm": 10.8125,
      "learning_rate": 4.662867610074908e-05,
      "loss": 2.7007,
      "num_input_tokens_seen": 14004976,
      "step": 6690
    },
    {
      "epoch": 1.0921771759523615,
      "grad_norm": 5.53125,
      "learning_rate": 4.6623750788899696e-05,
      "loss": 0.8619,
      "num_input_tokens_seen": 14014240,
      "step": 6695
    },
    {
      "epoch": 1.0929929031731789,
      "grad_norm": 5.65625,
      "learning_rate": 4.6618822142396195e-05,
      "loss": 2.3617,
      "num_input_tokens_seen": 14025152,
      "step": 6700
    },
    {
      "epoch": 1.0938086303939962,
      "grad_norm": 7.90625,
      "learning_rate": 4.661389016199864e-05,
      "loss": 1.7394,
      "num_input_tokens_seen": 14036528,
      "step": 6705
    },
    {
      "epoch": 1.0946243576148136,
      "grad_norm": 5.1875,
      "learning_rate": 4.660895484846761e-05,
      "loss": 3.3268,
      "num_input_tokens_seen": 14047568,
      "step": 6710
    },
    {
      "epoch": 1.095440084835631,
      "grad_norm": 11.4375,
      "learning_rate": 4.660401620256418e-05,
      "loss": 3.3054,
      "num_input_tokens_seen": 14057760,
      "step": 6715
    },
    {
      "epoch": 1.0962558120564483,
      "grad_norm": 4.3125,
      "learning_rate": 4.659907422504997e-05,
      "loss": 2.9211,
      "num_input_tokens_seen": 14067936,
      "step": 6720
    },
    {
      "epoch": 1.0970715392772656,
      "grad_norm": 6.5625,
      "learning_rate": 4.6594128916687074e-05,
      "loss": 2.6511,
      "num_input_tokens_seen": 14077616,
      "step": 6725
    },
    {
      "epoch": 1.097887266498083,
      "grad_norm": 10.8125,
      "learning_rate": 4.658918027823813e-05,
      "loss": 2.1828,
      "num_input_tokens_seen": 14087440,
      "step": 6730
    },
    {
      "epoch": 1.0987029937189003,
      "grad_norm": 9.4375,
      "learning_rate": 4.658422831046628e-05,
      "loss": 1.69,
      "num_input_tokens_seen": 14096880,
      "step": 6735
    },
    {
      "epoch": 1.0995187209397177,
      "grad_norm": 5.84375,
      "learning_rate": 4.657927301413518e-05,
      "loss": 2.5988,
      "num_input_tokens_seen": 14108320,
      "step": 6740
    },
    {
      "epoch": 1.100334448160535,
      "grad_norm": 3.265625,
      "learning_rate": 4.657431439000901e-05,
      "loss": 2.5977,
      "num_input_tokens_seen": 14116368,
      "step": 6745
    },
    {
      "epoch": 1.1011501753813524,
      "grad_norm": 1.75,
      "learning_rate": 4.656935243885243e-05,
      "loss": 1.6304,
      "num_input_tokens_seen": 14126352,
      "step": 6750
    },
    {
      "epoch": 1.1019659026021698,
      "grad_norm": 1.6875,
      "learning_rate": 4.656438716143066e-05,
      "loss": 1.296,
      "num_input_tokens_seen": 14137920,
      "step": 6755
    },
    {
      "epoch": 1.102781629822987,
      "grad_norm": 8.1875,
      "learning_rate": 4.6559418558509384e-05,
      "loss": 2.2682,
      "num_input_tokens_seen": 14148448,
      "step": 6760
    },
    {
      "epoch": 1.1035973570438045,
      "grad_norm": 3.828125,
      "learning_rate": 4.6554446630854833e-05,
      "loss": 1.7127,
      "num_input_tokens_seen": 14157248,
      "step": 6765
    },
    {
      "epoch": 1.1044130842646218,
      "grad_norm": 8.6875,
      "learning_rate": 4.654947137923374e-05,
      "loss": 2.3281,
      "num_input_tokens_seen": 14167216,
      "step": 6770
    },
    {
      "epoch": 1.1052288114854392,
      "grad_norm": 7.78125,
      "learning_rate": 4.654449280441335e-05,
      "loss": 2.1477,
      "num_input_tokens_seen": 14176352,
      "step": 6775
    },
    {
      "epoch": 1.1060445387062567,
      "grad_norm": 6.53125,
      "learning_rate": 4.653951090716143e-05,
      "loss": 2.6341,
      "num_input_tokens_seen": 14185840,
      "step": 6780
    },
    {
      "epoch": 1.1068602659270739,
      "grad_norm": 4.09375,
      "learning_rate": 4.653452568824625e-05,
      "loss": 2.1709,
      "num_input_tokens_seen": 14196064,
      "step": 6785
    },
    {
      "epoch": 1.1076759931478914,
      "grad_norm": 8.125,
      "learning_rate": 4.6529537148436585e-05,
      "loss": 1.4094,
      "num_input_tokens_seen": 14207152,
      "step": 6790
    },
    {
      "epoch": 1.1084917203687088,
      "grad_norm": 7.3125,
      "learning_rate": 4.6524545288501734e-05,
      "loss": 3.4547,
      "num_input_tokens_seen": 14217392,
      "step": 6795
    },
    {
      "epoch": 1.1093074475895262,
      "grad_norm": 5.625,
      "learning_rate": 4.6519550109211506e-05,
      "loss": 2.6627,
      "num_input_tokens_seen": 14228240,
      "step": 6800
    },
    {
      "epoch": 1.1093074475895262,
      "eval_loss": 2.5380496978759766,
      "eval_runtime": 134.8409,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 14228240,
      "step": 6800
    },
    {
      "epoch": 1.1101231748103435,
      "grad_norm": 6.3125,
      "learning_rate": 4.651455161133622e-05,
      "loss": 2.7996,
      "num_input_tokens_seen": 14237536,
      "step": 6805
    },
    {
      "epoch": 1.1109389020311609,
      "grad_norm": 5.625,
      "learning_rate": 4.6509549795646704e-05,
      "loss": 3.0776,
      "num_input_tokens_seen": 14249440,
      "step": 6810
    },
    {
      "epoch": 1.1117546292519782,
      "grad_norm": 5.78125,
      "learning_rate": 4.6504544662914306e-05,
      "loss": 3.4046,
      "num_input_tokens_seen": 14259952,
      "step": 6815
    },
    {
      "epoch": 1.1125703564727956,
      "grad_norm": 11.8125,
      "learning_rate": 4.6499536213910876e-05,
      "loss": 3.9139,
      "num_input_tokens_seen": 14270032,
      "step": 6820
    },
    {
      "epoch": 1.113386083693613,
      "grad_norm": 7.4375,
      "learning_rate": 4.6494524449408786e-05,
      "loss": 3.3805,
      "num_input_tokens_seen": 14279776,
      "step": 6825
    },
    {
      "epoch": 1.1142018109144303,
      "grad_norm": 7.5,
      "learning_rate": 4.6489509370180903e-05,
      "loss": 1.6634,
      "num_input_tokens_seen": 14289776,
      "step": 6830
    },
    {
      "epoch": 1.1150175381352476,
      "grad_norm": 6.0625,
      "learning_rate": 4.648449097700063e-05,
      "loss": 2.532,
      "num_input_tokens_seen": 14298544,
      "step": 6835
    },
    {
      "epoch": 1.115833265356065,
      "grad_norm": 6.8125,
      "learning_rate": 4.647946927064185e-05,
      "loss": 1.6581,
      "num_input_tokens_seen": 14308624,
      "step": 6840
    },
    {
      "epoch": 1.1166489925768823,
      "grad_norm": 9.125,
      "learning_rate": 4.647444425187898e-05,
      "loss": 2.2007,
      "num_input_tokens_seen": 14318512,
      "step": 6845
    },
    {
      "epoch": 1.1174647197976997,
      "grad_norm": 2.46875,
      "learning_rate": 4.646941592148695e-05,
      "loss": 2.7297,
      "num_input_tokens_seen": 14329520,
      "step": 6850
    },
    {
      "epoch": 1.118280447018517,
      "grad_norm": 12.5625,
      "learning_rate": 4.646438428024117e-05,
      "loss": 3.3047,
      "num_input_tokens_seen": 14339536,
      "step": 6855
    },
    {
      "epoch": 1.1190961742393344,
      "grad_norm": 4.46875,
      "learning_rate": 4.64593493289176e-05,
      "loss": 2.8425,
      "num_input_tokens_seen": 14350128,
      "step": 6860
    },
    {
      "epoch": 1.1199119014601517,
      "grad_norm": 7.71875,
      "learning_rate": 4.64543110682927e-05,
      "loss": 1.9209,
      "num_input_tokens_seen": 14360832,
      "step": 6865
    },
    {
      "epoch": 1.120727628680969,
      "grad_norm": 6.59375,
      "learning_rate": 4.644926949914341e-05,
      "loss": 2.7893,
      "num_input_tokens_seen": 14372192,
      "step": 6870
    },
    {
      "epoch": 1.1215433559017864,
      "grad_norm": 11.5,
      "learning_rate": 4.644422462224722e-05,
      "loss": 3.2243,
      "num_input_tokens_seen": 14382064,
      "step": 6875
    },
    {
      "epoch": 1.1223590831226038,
      "grad_norm": 4.65625,
      "learning_rate": 4.643917643838211e-05,
      "loss": 1.9737,
      "num_input_tokens_seen": 14391312,
      "step": 6880
    },
    {
      "epoch": 1.1231748103434211,
      "grad_norm": 7.59375,
      "learning_rate": 4.6434124948326564e-05,
      "loss": 3.3349,
      "num_input_tokens_seen": 14401424,
      "step": 6885
    },
    {
      "epoch": 1.1239905375642385,
      "grad_norm": 7.90625,
      "learning_rate": 4.6429070152859594e-05,
      "loss": 2.5061,
      "num_input_tokens_seen": 14410832,
      "step": 6890
    },
    {
      "epoch": 1.1248062647850559,
      "grad_norm": 4.21875,
      "learning_rate": 4.6424012052760714e-05,
      "loss": 2.3073,
      "num_input_tokens_seen": 14421888,
      "step": 6895
    },
    {
      "epoch": 1.1256219920058732,
      "grad_norm": 9.0625,
      "learning_rate": 4.6418950648809945e-05,
      "loss": 2.4475,
      "num_input_tokens_seen": 14432848,
      "step": 6900
    },
    {
      "epoch": 1.1264377192266906,
      "grad_norm": 2.84375,
      "learning_rate": 4.641388594178782e-05,
      "loss": 2.1236,
      "num_input_tokens_seen": 14443728,
      "step": 6905
    },
    {
      "epoch": 1.127253446447508,
      "grad_norm": 3.1875,
      "learning_rate": 4.640881793247538e-05,
      "loss": 2.3005,
      "num_input_tokens_seen": 14455296,
      "step": 6910
    },
    {
      "epoch": 1.1280691736683253,
      "grad_norm": 5.09375,
      "learning_rate": 4.6403746621654173e-05,
      "loss": 1.8599,
      "num_input_tokens_seen": 14466016,
      "step": 6915
    },
    {
      "epoch": 1.1288849008891426,
      "grad_norm": 4.34375,
      "learning_rate": 4.639867201010626e-05,
      "loss": 3.3185,
      "num_input_tokens_seen": 14475296,
      "step": 6920
    },
    {
      "epoch": 1.12970062810996,
      "grad_norm": 6.59375,
      "learning_rate": 4.6393594098614204e-05,
      "loss": 2.6306,
      "num_input_tokens_seen": 14485280,
      "step": 6925
    },
    {
      "epoch": 1.1305163553307773,
      "grad_norm": 8.0625,
      "learning_rate": 4.63885128879611e-05,
      "loss": 1.8298,
      "num_input_tokens_seen": 14494960,
      "step": 6930
    },
    {
      "epoch": 1.1313320825515947,
      "grad_norm": 11.625,
      "learning_rate": 4.638342837893052e-05,
      "loss": 3.1982,
      "num_input_tokens_seen": 14506048,
      "step": 6935
    },
    {
      "epoch": 1.132147809772412,
      "grad_norm": 7.125,
      "learning_rate": 4.6378340572306565e-05,
      "loss": 3.4628,
      "num_input_tokens_seen": 14515264,
      "step": 6940
    },
    {
      "epoch": 1.1329635369932294,
      "grad_norm": 9.8125,
      "learning_rate": 4.6373249468873833e-05,
      "loss": 2.4782,
      "num_input_tokens_seen": 14524304,
      "step": 6945
    },
    {
      "epoch": 1.1337792642140467,
      "grad_norm": 7.5625,
      "learning_rate": 4.636815506941744e-05,
      "loss": 2.3501,
      "num_input_tokens_seen": 14535152,
      "step": 6950
    },
    {
      "epoch": 1.134594991434864,
      "grad_norm": 6.09375,
      "learning_rate": 4.6363057374723004e-05,
      "loss": 2.3922,
      "num_input_tokens_seen": 14545488,
      "step": 6955
    },
    {
      "epoch": 1.1354107186556814,
      "grad_norm": 4.3125,
      "learning_rate": 4.635795638557666e-05,
      "loss": 3.8938,
      "num_input_tokens_seen": 14556304,
      "step": 6960
    },
    {
      "epoch": 1.136226445876499,
      "grad_norm": 7.40625,
      "learning_rate": 4.635285210276504e-05,
      "loss": 2.2939,
      "num_input_tokens_seen": 14567424,
      "step": 6965
    },
    {
      "epoch": 1.1370421730973161,
      "grad_norm": 5.125,
      "learning_rate": 4.6347744527075295e-05,
      "loss": 2.8028,
      "num_input_tokens_seen": 14576928,
      "step": 6970
    },
    {
      "epoch": 1.1378579003181337,
      "grad_norm": 7.09375,
      "learning_rate": 4.634263365929506e-05,
      "loss": 2.8515,
      "num_input_tokens_seen": 14585392,
      "step": 6975
    },
    {
      "epoch": 1.1386736275389508,
      "grad_norm": 11.25,
      "learning_rate": 4.6337519500212515e-05,
      "loss": 2.3309,
      "num_input_tokens_seen": 14595792,
      "step": 6980
    },
    {
      "epoch": 1.1394893547597684,
      "grad_norm": 8.25,
      "learning_rate": 4.633240205061632e-05,
      "loss": 2.8317,
      "num_input_tokens_seen": 14606464,
      "step": 6985
    },
    {
      "epoch": 1.1403050819805858,
      "grad_norm": 3.359375,
      "learning_rate": 4.632728131129565e-05,
      "loss": 2.4585,
      "num_input_tokens_seen": 14616576,
      "step": 6990
    },
    {
      "epoch": 1.1411208092014031,
      "grad_norm": 5.3125,
      "learning_rate": 4.632215728304018e-05,
      "loss": 1.8745,
      "num_input_tokens_seen": 14626944,
      "step": 6995
    },
    {
      "epoch": 1.1419365364222205,
      "grad_norm": 8.1875,
      "learning_rate": 4.63170299666401e-05,
      "loss": 3.0394,
      "num_input_tokens_seen": 14637984,
      "step": 7000
    },
    {
      "epoch": 1.1419365364222205,
      "eval_loss": 2.5363268852233887,
      "eval_runtime": 135.0809,
      "eval_samples_per_second": 20.173,
      "eval_steps_per_second": 10.09,
      "num_input_tokens_seen": 14637984,
      "step": 7000
    },
    {
      "epoch": 1.1427522636430378,
      "grad_norm": 1.8828125,
      "learning_rate": 4.631189936288612e-05,
      "loss": 2.1908,
      "num_input_tokens_seen": 14648064,
      "step": 7005
    },
    {
      "epoch": 1.1435679908638552,
      "grad_norm": 0.83203125,
      "learning_rate": 4.630676547256944e-05,
      "loss": 1.7766,
      "num_input_tokens_seen": 14658048,
      "step": 7010
    },
    {
      "epoch": 1.1443837180846725,
      "grad_norm": 4.25,
      "learning_rate": 4.630162829648176e-05,
      "loss": 1.0917,
      "num_input_tokens_seen": 14668256,
      "step": 7015
    },
    {
      "epoch": 1.14519944530549,
      "grad_norm": 5.6875,
      "learning_rate": 4.629648783541531e-05,
      "loss": 1.7112,
      "num_input_tokens_seen": 14679280,
      "step": 7020
    },
    {
      "epoch": 1.1460151725263072,
      "grad_norm": 4.5,
      "learning_rate": 4.6291344090162804e-05,
      "loss": 2.1924,
      "num_input_tokens_seen": 14688992,
      "step": 7025
    },
    {
      "epoch": 1.1468308997471246,
      "grad_norm": 2.640625,
      "learning_rate": 4.628619706151748e-05,
      "loss": 2.4936,
      "num_input_tokens_seen": 14699328,
      "step": 7030
    },
    {
      "epoch": 1.147646626967942,
      "grad_norm": 1.3046875,
      "learning_rate": 4.628104675027306e-05,
      "loss": 2.275,
      "num_input_tokens_seen": 14711264,
      "step": 7035
    },
    {
      "epoch": 1.1484623541887593,
      "grad_norm": 8.5625,
      "learning_rate": 4.6275893157223805e-05,
      "loss": 2.5831,
      "num_input_tokens_seen": 14721504,
      "step": 7040
    },
    {
      "epoch": 1.1492780814095767,
      "grad_norm": 8.75,
      "learning_rate": 4.627073628316445e-05,
      "loss": 1.6562,
      "num_input_tokens_seen": 14730896,
      "step": 7045
    },
    {
      "epoch": 1.150093808630394,
      "grad_norm": 7.0,
      "learning_rate": 4.626557612889026e-05,
      "loss": 2.058,
      "num_input_tokens_seen": 14740448,
      "step": 7050
    },
    {
      "epoch": 1.1509095358512114,
      "grad_norm": 4.875,
      "learning_rate": 4.626041269519699e-05,
      "loss": 2.8781,
      "num_input_tokens_seen": 14751216,
      "step": 7055
    },
    {
      "epoch": 1.1517252630720287,
      "grad_norm": 1.984375,
      "learning_rate": 4.6255245982880905e-05,
      "loss": 2.3806,
      "num_input_tokens_seen": 14760720,
      "step": 7060
    },
    {
      "epoch": 1.152540990292846,
      "grad_norm": 11.1875,
      "learning_rate": 4.625007599273879e-05,
      "loss": 1.8402,
      "num_input_tokens_seen": 14772848,
      "step": 7065
    },
    {
      "epoch": 1.1533567175136634,
      "grad_norm": 4.375,
      "learning_rate": 4.6244902725567895e-05,
      "loss": 1.56,
      "num_input_tokens_seen": 14782528,
      "step": 7070
    },
    {
      "epoch": 1.1541724447344808,
      "grad_norm": 3.5,
      "learning_rate": 4.6239726182166024e-05,
      "loss": 1.9506,
      "num_input_tokens_seen": 14791968,
      "step": 7075
    },
    {
      "epoch": 1.1549881719552981,
      "grad_norm": 11.9375,
      "learning_rate": 4.623454636333147e-05,
      "loss": 1.659,
      "num_input_tokens_seen": 14801616,
      "step": 7080
    },
    {
      "epoch": 1.1558038991761155,
      "grad_norm": 5.625,
      "learning_rate": 4.622936326986301e-05,
      "loss": 2.0274,
      "num_input_tokens_seen": 14812272,
      "step": 7085
    },
    {
      "epoch": 1.1566196263969328,
      "grad_norm": 3.046875,
      "learning_rate": 4.6224176902559946e-05,
      "loss": 2.1539,
      "num_input_tokens_seen": 14823552,
      "step": 7090
    },
    {
      "epoch": 1.1574353536177502,
      "grad_norm": 9.5625,
      "learning_rate": 4.621898726222209e-05,
      "loss": 1.9231,
      "num_input_tokens_seen": 14833920,
      "step": 7095
    },
    {
      "epoch": 1.1582510808385675,
      "grad_norm": 4.6875,
      "learning_rate": 4.6213794349649744e-05,
      "loss": 1.0234,
      "num_input_tokens_seen": 14842816,
      "step": 7100
    },
    {
      "epoch": 1.159066808059385,
      "grad_norm": 3.53125,
      "learning_rate": 4.6208598165643715e-05,
      "loss": 2.2755,
      "num_input_tokens_seen": 14853104,
      "step": 7105
    },
    {
      "epoch": 1.1598825352802022,
      "grad_norm": 10.5625,
      "learning_rate": 4.620339871100533e-05,
      "loss": 1.922,
      "num_input_tokens_seen": 14862176,
      "step": 7110
    },
    {
      "epoch": 1.1606982625010196,
      "grad_norm": 7.625,
      "learning_rate": 4.6198195986536394e-05,
      "loss": 1.8391,
      "num_input_tokens_seen": 14872160,
      "step": 7115
    },
    {
      "epoch": 1.161513989721837,
      "grad_norm": 7.34375,
      "learning_rate": 4.619298999303926e-05,
      "loss": 1.513,
      "num_input_tokens_seen": 14882576,
      "step": 7120
    },
    {
      "epoch": 1.1623297169426543,
      "grad_norm": 4.0625,
      "learning_rate": 4.618778073131673e-05,
      "loss": 3.0759,
      "num_input_tokens_seen": 14893040,
      "step": 7125
    },
    {
      "epoch": 1.1631454441634717,
      "grad_norm": 6.5,
      "learning_rate": 4.618256820217215e-05,
      "loss": 2.3181,
      "num_input_tokens_seen": 14904336,
      "step": 7130
    },
    {
      "epoch": 1.163961171384289,
      "grad_norm": 6.9375,
      "learning_rate": 4.617735240640936e-05,
      "loss": 2.516,
      "num_input_tokens_seen": 14915840,
      "step": 7135
    },
    {
      "epoch": 1.1647768986051064,
      "grad_norm": 8.875,
      "learning_rate": 4.6172133344832705e-05,
      "loss": 2.8026,
      "num_input_tokens_seen": 14925776,
      "step": 7140
    },
    {
      "epoch": 1.1655926258259237,
      "grad_norm": 7.4375,
      "learning_rate": 4.6166911018247004e-05,
      "loss": 2.2109,
      "num_input_tokens_seen": 14936208,
      "step": 7145
    },
    {
      "epoch": 1.1664083530467413,
      "grad_norm": 4.53125,
      "learning_rate": 4.616168542745764e-05,
      "loss": 2.641,
      "num_input_tokens_seen": 14947456,
      "step": 7150
    },
    {
      "epoch": 1.1672240802675584,
      "grad_norm": 6.15625,
      "learning_rate": 4.6156456573270446e-05,
      "loss": 2.2243,
      "num_input_tokens_seen": 14956112,
      "step": 7155
    },
    {
      "epoch": 1.168039807488376,
      "grad_norm": 5.5,
      "learning_rate": 4.615122445649177e-05,
      "loss": 1.6195,
      "num_input_tokens_seen": 14965344,
      "step": 7160
    },
    {
      "epoch": 1.1688555347091931,
      "grad_norm": 7.84375,
      "learning_rate": 4.6145989077928486e-05,
      "loss": 1.9419,
      "num_input_tokens_seen": 14975888,
      "step": 7165
    },
    {
      "epoch": 1.1696712619300107,
      "grad_norm": 8.0625,
      "learning_rate": 4.6140750438387953e-05,
      "loss": 1.5907,
      "num_input_tokens_seen": 14986160,
      "step": 7170
    },
    {
      "epoch": 1.170486989150828,
      "grad_norm": 11.375,
      "learning_rate": 4.613550853867803e-05,
      "loss": 2.863,
      "num_input_tokens_seen": 14996544,
      "step": 7175
    },
    {
      "epoch": 1.1713027163716454,
      "grad_norm": 5.625,
      "learning_rate": 4.613026337960708e-05,
      "loss": 2.1381,
      "num_input_tokens_seen": 15007712,
      "step": 7180
    },
    {
      "epoch": 1.1721184435924628,
      "grad_norm": 7.5625,
      "learning_rate": 4.612501496198398e-05,
      "loss": 3.8136,
      "num_input_tokens_seen": 15017056,
      "step": 7185
    },
    {
      "epoch": 1.17293417081328,
      "grad_norm": 13.625,
      "learning_rate": 4.61197632866181e-05,
      "loss": 4.3741,
      "num_input_tokens_seen": 15028048,
      "step": 7190
    },
    {
      "epoch": 1.1737498980340975,
      "grad_norm": 12.75,
      "learning_rate": 4.611450835431931e-05,
      "loss": 1.6193,
      "num_input_tokens_seen": 15038096,
      "step": 7195
    },
    {
      "epoch": 1.1745656252549148,
      "grad_norm": 19.5,
      "learning_rate": 4.6109250165898e-05,
      "loss": 2.7862,
      "num_input_tokens_seen": 15049216,
      "step": 7200
    },
    {
      "epoch": 1.1745656252549148,
      "eval_loss": 2.54175066947937,
      "eval_runtime": 134.8401,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 15049216,
      "step": 7200
    },
    {
      "epoch": 1.1753813524757322,
      "grad_norm": 6.09375,
      "learning_rate": 4.610398872216503e-05,
      "loss": 2.2944,
      "num_input_tokens_seen": 15060080,
      "step": 7205
    },
    {
      "epoch": 1.1761970796965495,
      "grad_norm": 9.0625,
      "learning_rate": 4.6098724023931796e-05,
      "loss": 2.0494,
      "num_input_tokens_seen": 15071408,
      "step": 7210
    },
    {
      "epoch": 1.1770128069173669,
      "grad_norm": 9.625,
      "learning_rate": 4.609345607201017e-05,
      "loss": 2.5433,
      "num_input_tokens_seen": 15080496,
      "step": 7215
    },
    {
      "epoch": 1.1778285341381842,
      "grad_norm": 0.142578125,
      "learning_rate": 4.608818486721254e-05,
      "loss": 1.5975,
      "num_input_tokens_seen": 15090592,
      "step": 7220
    },
    {
      "epoch": 1.1786442613590016,
      "grad_norm": 4.34375,
      "learning_rate": 4.608291041035179e-05,
      "loss": 1.0553,
      "num_input_tokens_seen": 15100832,
      "step": 7225
    },
    {
      "epoch": 1.179459988579819,
      "grad_norm": 7.53125,
      "learning_rate": 4.607763270224132e-05,
      "loss": 2.1474,
      "num_input_tokens_seen": 15111104,
      "step": 7230
    },
    {
      "epoch": 1.1802757158006363,
      "grad_norm": 6.46875,
      "learning_rate": 4.6072351743695e-05,
      "loss": 2.0628,
      "num_input_tokens_seen": 15121152,
      "step": 7235
    },
    {
      "epoch": 1.1810914430214536,
      "grad_norm": 11.25,
      "learning_rate": 4.606706753552723e-05,
      "loss": 3.0488,
      "num_input_tokens_seen": 15132800,
      "step": 7240
    },
    {
      "epoch": 1.181907170242271,
      "grad_norm": 7.375,
      "learning_rate": 4.6061780078552906e-05,
      "loss": 2.26,
      "num_input_tokens_seen": 15143632,
      "step": 7245
    },
    {
      "epoch": 1.1827228974630883,
      "grad_norm": 4.53125,
      "learning_rate": 4.605648937358742e-05,
      "loss": 1.4248,
      "num_input_tokens_seen": 15154688,
      "step": 7250
    },
    {
      "epoch": 1.1835386246839057,
      "grad_norm": 10.8125,
      "learning_rate": 4.605119542144665e-05,
      "loss": 3.4472,
      "num_input_tokens_seen": 15165136,
      "step": 7255
    },
    {
      "epoch": 1.184354351904723,
      "grad_norm": 7.0625,
      "learning_rate": 4.604589822294701e-05,
      "loss": 1.8409,
      "num_input_tokens_seen": 15173968,
      "step": 7260
    },
    {
      "epoch": 1.1851700791255404,
      "grad_norm": 11.125,
      "learning_rate": 4.604059777890537e-05,
      "loss": 2.4529,
      "num_input_tokens_seen": 15184720,
      "step": 7265
    },
    {
      "epoch": 1.1859858063463578,
      "grad_norm": 9.0625,
      "learning_rate": 4.6035294090139145e-05,
      "loss": 2.0046,
      "num_input_tokens_seen": 15195024,
      "step": 7270
    },
    {
      "epoch": 1.186801533567175,
      "grad_norm": 8.875,
      "learning_rate": 4.6029987157466226e-05,
      "loss": 3.3057,
      "num_input_tokens_seen": 15206256,
      "step": 7275
    },
    {
      "epoch": 1.1876172607879925,
      "grad_norm": 7.125,
      "learning_rate": 4.602467698170502e-05,
      "loss": 2.1189,
      "num_input_tokens_seen": 15218832,
      "step": 7280
    },
    {
      "epoch": 1.1884329880088098,
      "grad_norm": 7.8125,
      "learning_rate": 4.601936356367439e-05,
      "loss": 2.8011,
      "num_input_tokens_seen": 15229536,
      "step": 7285
    },
    {
      "epoch": 1.1892487152296272,
      "grad_norm": 8.0,
      "learning_rate": 4.601404690419377e-05,
      "loss": 2.4708,
      "num_input_tokens_seen": 15240512,
      "step": 7290
    },
    {
      "epoch": 1.1900644424504445,
      "grad_norm": 6.5625,
      "learning_rate": 4.600872700408303e-05,
      "loss": 2.636,
      "num_input_tokens_seen": 15250080,
      "step": 7295
    },
    {
      "epoch": 1.1908801696712619,
      "grad_norm": 6.9375,
      "learning_rate": 4.600340386416258e-05,
      "loss": 3.0933,
      "num_input_tokens_seen": 15260704,
      "step": 7300
    },
    {
      "epoch": 1.1916958968920792,
      "grad_norm": 7.15625,
      "learning_rate": 4.5998077485253296e-05,
      "loss": 2.0272,
      "num_input_tokens_seen": 15269504,
      "step": 7305
    },
    {
      "epoch": 1.1925116241128966,
      "grad_norm": 4.78125,
      "learning_rate": 4.59927478681766e-05,
      "loss": 2.2788,
      "num_input_tokens_seen": 15280064,
      "step": 7310
    },
    {
      "epoch": 1.193327351333714,
      "grad_norm": 8.25,
      "learning_rate": 4.5987415013754366e-05,
      "loss": 2.0007,
      "num_input_tokens_seen": 15291456,
      "step": 7315
    },
    {
      "epoch": 1.1941430785545313,
      "grad_norm": 4.125,
      "learning_rate": 4.598207892280899e-05,
      "loss": 2.2253,
      "num_input_tokens_seen": 15303968,
      "step": 7320
    },
    {
      "epoch": 1.1949588057753486,
      "grad_norm": 4.65625,
      "learning_rate": 4.597673959616337e-05,
      "loss": 1.6033,
      "num_input_tokens_seen": 15313760,
      "step": 7325
    },
    {
      "epoch": 1.195774532996166,
      "grad_norm": 4.15625,
      "learning_rate": 4.597139703464089e-05,
      "loss": 1.1099,
      "num_input_tokens_seen": 15324192,
      "step": 7330
    },
    {
      "epoch": 1.1965902602169836,
      "grad_norm": 3.515625,
      "learning_rate": 4.596605123906545e-05,
      "loss": 1.6314,
      "num_input_tokens_seen": 15335456,
      "step": 7335
    },
    {
      "epoch": 1.1974059874378007,
      "grad_norm": 3.46875,
      "learning_rate": 4.596070221026143e-05,
      "loss": 1.4956,
      "num_input_tokens_seen": 15345600,
      "step": 7340
    },
    {
      "epoch": 1.1982217146586183,
      "grad_norm": 7.03125,
      "learning_rate": 4.595534994905372e-05,
      "loss": 2.2547,
      "num_input_tokens_seen": 15356544,
      "step": 7345
    },
    {
      "epoch": 1.1990374418794354,
      "grad_norm": 8.4375,
      "learning_rate": 4.594999445626771e-05,
      "loss": 1.4879,
      "num_input_tokens_seen": 15367232,
      "step": 7350
    },
    {
      "epoch": 1.199853169100253,
      "grad_norm": 6.5625,
      "learning_rate": 4.5944635732729276e-05,
      "loss": 3.7935,
      "num_input_tokens_seen": 15377168,
      "step": 7355
    },
    {
      "epoch": 1.2006688963210703,
      "grad_norm": 12.5625,
      "learning_rate": 4.5939273779264804e-05,
      "loss": 2.3969,
      "num_input_tokens_seen": 15387088,
      "step": 7360
    },
    {
      "epoch": 1.2014846235418877,
      "grad_norm": 7.25,
      "learning_rate": 4.593390859670118e-05,
      "loss": 2.2434,
      "num_input_tokens_seen": 15397584,
      "step": 7365
    },
    {
      "epoch": 1.202300350762705,
      "grad_norm": 14.9375,
      "learning_rate": 4.5928540185865776e-05,
      "loss": 2.43,
      "num_input_tokens_seen": 15407264,
      "step": 7370
    },
    {
      "epoch": 1.2031160779835224,
      "grad_norm": 10.625,
      "learning_rate": 4.592316854758648e-05,
      "loss": 3.6633,
      "num_input_tokens_seen": 15418080,
      "step": 7375
    },
    {
      "epoch": 1.2039318052043397,
      "grad_norm": 6.15625,
      "learning_rate": 4.5917793682691646e-05,
      "loss": 3.0852,
      "num_input_tokens_seen": 15428464,
      "step": 7380
    },
    {
      "epoch": 1.204747532425157,
      "grad_norm": 3.0625,
      "learning_rate": 4.5912415592010164e-05,
      "loss": 2.6455,
      "num_input_tokens_seen": 15438320,
      "step": 7385
    },
    {
      "epoch": 1.2055632596459744,
      "grad_norm": 16.25,
      "learning_rate": 4.5907034276371386e-05,
      "loss": 2.7999,
      "num_input_tokens_seen": 15449360,
      "step": 7390
    },
    {
      "epoch": 1.2063789868667918,
      "grad_norm": 7.1875,
      "learning_rate": 4.5901649736605196e-05,
      "loss": 1.8303,
      "num_input_tokens_seen": 15460496,
      "step": 7395
    },
    {
      "epoch": 1.2071947140876091,
      "grad_norm": 8.3125,
      "learning_rate": 4.589626197354195e-05,
      "loss": 2.3596,
      "num_input_tokens_seen": 15471984,
      "step": 7400
    },
    {
      "epoch": 1.2071947140876091,
      "eval_loss": 2.558281660079956,
      "eval_runtime": 135.0649,
      "eval_samples_per_second": 20.175,
      "eval_steps_per_second": 10.091,
      "num_input_tokens_seen": 15471984,
      "step": 7400
    },
    {
      "epoch": 1.2080104413084265,
      "grad_norm": 1.59375,
      "learning_rate": 4.5890870988012504e-05,
      "loss": 2.091,
      "num_input_tokens_seen": 15481472,
      "step": 7405
    },
    {
      "epoch": 1.2088261685292438,
      "grad_norm": 6.25,
      "learning_rate": 4.5885476780848226e-05,
      "loss": 2.1034,
      "num_input_tokens_seen": 15492544,
      "step": 7410
    },
    {
      "epoch": 1.2096418957500612,
      "grad_norm": 11.9375,
      "learning_rate": 4.5880079352880964e-05,
      "loss": 1.8024,
      "num_input_tokens_seen": 15503408,
      "step": 7415
    },
    {
      "epoch": 1.2104576229708786,
      "grad_norm": 4.15625,
      "learning_rate": 4.5874678704943065e-05,
      "loss": 2.0026,
      "num_input_tokens_seen": 15514256,
      "step": 7420
    },
    {
      "epoch": 1.211273350191696,
      "grad_norm": 9.5625,
      "learning_rate": 4.5869274837867394e-05,
      "loss": 2.8967,
      "num_input_tokens_seen": 15525456,
      "step": 7425
    },
    {
      "epoch": 1.2120890774125133,
      "grad_norm": 8.3125,
      "learning_rate": 4.5863867752487275e-05,
      "loss": 3.706,
      "num_input_tokens_seen": 15535600,
      "step": 7430
    },
    {
      "epoch": 1.2129048046333306,
      "grad_norm": 5.625,
      "learning_rate": 4.5858457449636554e-05,
      "loss": 2.11,
      "num_input_tokens_seen": 15546976,
      "step": 7435
    },
    {
      "epoch": 1.213720531854148,
      "grad_norm": 12.625,
      "learning_rate": 4.5853043930149574e-05,
      "loss": 2.3336,
      "num_input_tokens_seen": 15556928,
      "step": 7440
    },
    {
      "epoch": 1.2145362590749653,
      "grad_norm": 6.78125,
      "learning_rate": 4.584762719486117e-05,
      "loss": 1.7933,
      "num_input_tokens_seen": 15568192,
      "step": 7445
    },
    {
      "epoch": 1.2153519862957827,
      "grad_norm": 5.09375,
      "learning_rate": 4.584220724460665e-05,
      "loss": 2.4251,
      "num_input_tokens_seen": 15579360,
      "step": 7450
    },
    {
      "epoch": 1.2161677135166,
      "grad_norm": 7.9375,
      "learning_rate": 4.5836784080221865e-05,
      "loss": 3.3301,
      "num_input_tokens_seen": 15589680,
      "step": 7455
    },
    {
      "epoch": 1.2169834407374174,
      "grad_norm": 6.03125,
      "learning_rate": 4.583135770254312e-05,
      "loss": 2.9671,
      "num_input_tokens_seen": 15600704,
      "step": 7460
    },
    {
      "epoch": 1.2177991679582347,
      "grad_norm": 5.125,
      "learning_rate": 4.5825928112407236e-05,
      "loss": 2.569,
      "num_input_tokens_seen": 15611024,
      "step": 7465
    },
    {
      "epoch": 1.218614895179052,
      "grad_norm": 2.90625,
      "learning_rate": 4.582049531065152e-05,
      "loss": 2.7915,
      "num_input_tokens_seen": 15621232,
      "step": 7470
    },
    {
      "epoch": 1.2194306223998694,
      "grad_norm": 2.71875,
      "learning_rate": 4.5815059298113783e-05,
      "loss": 1.4746,
      "num_input_tokens_seen": 15630976,
      "step": 7475
    },
    {
      "epoch": 1.2202463496206868,
      "grad_norm": 6.875,
      "learning_rate": 4.580962007563232e-05,
      "loss": 2.3339,
      "num_input_tokens_seen": 15640880,
      "step": 7480
    },
    {
      "epoch": 1.2210620768415041,
      "grad_norm": 2.265625,
      "learning_rate": 4.5804177644045935e-05,
      "loss": 2.695,
      "num_input_tokens_seen": 15651040,
      "step": 7485
    },
    {
      "epoch": 1.2218778040623215,
      "grad_norm": 4.21875,
      "learning_rate": 4.579873200419391e-05,
      "loss": 2.2661,
      "num_input_tokens_seen": 15661248,
      "step": 7490
    },
    {
      "epoch": 1.2226935312831388,
      "grad_norm": 4.90625,
      "learning_rate": 4.5793283156916046e-05,
      "loss": 3.0956,
      "num_input_tokens_seen": 15670560,
      "step": 7495
    },
    {
      "epoch": 1.2235092585039562,
      "grad_norm": 6.9375,
      "learning_rate": 4.578783110305261e-05,
      "loss": 2.3361,
      "num_input_tokens_seen": 15681296,
      "step": 7500
    },
    {
      "epoch": 1.2243249857247736,
      "grad_norm": 8.25,
      "learning_rate": 4.578237584344438e-05,
      "loss": 3.1864,
      "num_input_tokens_seen": 15692880,
      "step": 7505
    },
    {
      "epoch": 1.225140712945591,
      "grad_norm": 2.328125,
      "learning_rate": 4.577691737893263e-05,
      "loss": 2.6673,
      "num_input_tokens_seen": 15703168,
      "step": 7510
    },
    {
      "epoch": 1.2259564401664083,
      "grad_norm": 8.0,
      "learning_rate": 4.577145571035912e-05,
      "loss": 2.9582,
      "num_input_tokens_seen": 15713568,
      "step": 7515
    },
    {
      "epoch": 1.2267721673872258,
      "grad_norm": 5.25,
      "learning_rate": 4.576599083856611e-05,
      "loss": 2.0738,
      "num_input_tokens_seen": 15723872,
      "step": 7520
    },
    {
      "epoch": 1.227587894608043,
      "grad_norm": 7.03125,
      "learning_rate": 4.576052276439635e-05,
      "loss": 2.1149,
      "num_input_tokens_seen": 15733472,
      "step": 7525
    },
    {
      "epoch": 1.2284036218288605,
      "grad_norm": 2.828125,
      "learning_rate": 4.575505148869308e-05,
      "loss": 3.2454,
      "num_input_tokens_seen": 15743920,
      "step": 7530
    },
    {
      "epoch": 1.2292193490496777,
      "grad_norm": 9.625,
      "learning_rate": 4.574957701230006e-05,
      "loss": 3.307,
      "num_input_tokens_seen": 15753648,
      "step": 7535
    },
    {
      "epoch": 1.2300350762704952,
      "grad_norm": 4.59375,
      "learning_rate": 4.57440993360615e-05,
      "loss": 1.7898,
      "num_input_tokens_seen": 15764160,
      "step": 7540
    },
    {
      "epoch": 1.2308508034913126,
      "grad_norm": 11.5625,
      "learning_rate": 4.5738618460822134e-05,
      "loss": 2.3753,
      "num_input_tokens_seen": 15774512,
      "step": 7545
    },
    {
      "epoch": 1.23166653071213,
      "grad_norm": 9.0625,
      "learning_rate": 4.573313438742719e-05,
      "loss": 2.4045,
      "num_input_tokens_seen": 15784944,
      "step": 7550
    },
    {
      "epoch": 1.2324822579329473,
      "grad_norm": 6.1875,
      "learning_rate": 4.5727647116722374e-05,
      "loss": 2.7328,
      "num_input_tokens_seen": 15795776,
      "step": 7555
    },
    {
      "epoch": 1.2332979851537647,
      "grad_norm": 8.25,
      "learning_rate": 4.5722156649553884e-05,
      "loss": 2.9951,
      "num_input_tokens_seen": 15806304,
      "step": 7560
    },
    {
      "epoch": 1.234113712374582,
      "grad_norm": 4.9375,
      "learning_rate": 4.571666298676843e-05,
      "loss": 2.2625,
      "num_input_tokens_seen": 15817376,
      "step": 7565
    },
    {
      "epoch": 1.2349294395953994,
      "grad_norm": 13.25,
      "learning_rate": 4.571116612921321e-05,
      "loss": 2.7347,
      "num_input_tokens_seen": 15827072,
      "step": 7570
    },
    {
      "epoch": 1.2357451668162167,
      "grad_norm": 2.3125,
      "learning_rate": 4.57056660777359e-05,
      "loss": 1.1264,
      "num_input_tokens_seen": 15839040,
      "step": 7575
    },
    {
      "epoch": 1.236560894037034,
      "grad_norm": 8.125,
      "learning_rate": 4.5700162833184666e-05,
      "loss": 2.3227,
      "num_input_tokens_seen": 15849296,
      "step": 7580
    },
    {
      "epoch": 1.2373766212578514,
      "grad_norm": 3.953125,
      "learning_rate": 4.5694656396408195e-05,
      "loss": 1.3727,
      "num_input_tokens_seen": 15859424,
      "step": 7585
    },
    {
      "epoch": 1.2381923484786688,
      "grad_norm": 7.5,
      "learning_rate": 4.5689146768255646e-05,
      "loss": 2.17,
      "num_input_tokens_seen": 15870864,
      "step": 7590
    },
    {
      "epoch": 1.2390080756994861,
      "grad_norm": 8.75,
      "learning_rate": 4.568363394957667e-05,
      "loss": 2.8341,
      "num_input_tokens_seen": 15881056,
      "step": 7595
    },
    {
      "epoch": 1.2398238029203035,
      "grad_norm": 9.0625,
      "learning_rate": 4.567811794122141e-05,
      "loss": 2.6311,
      "num_input_tokens_seen": 15891152,
      "step": 7600
    },
    {
      "epoch": 1.2398238029203035,
      "eval_loss": 2.5384037494659424,
      "eval_runtime": 134.8844,
      "eval_samples_per_second": 20.202,
      "eval_steps_per_second": 10.105,
      "num_input_tokens_seen": 15891152,
      "step": 7600
    },
    {
      "epoch": 1.2406395301411208,
      "grad_norm": 10.4375,
      "learning_rate": 4.56725987440405e-05,
      "loss": 2.9682,
      "num_input_tokens_seen": 15901296,
      "step": 7605
    },
    {
      "epoch": 1.2414552573619382,
      "grad_norm": 5.625,
      "learning_rate": 4.566707635888508e-05,
      "loss": 1.8171,
      "num_input_tokens_seen": 15911280,
      "step": 7610
    },
    {
      "epoch": 1.2422709845827555,
      "grad_norm": 6.53125,
      "learning_rate": 4.566155078660677e-05,
      "loss": 2.4598,
      "num_input_tokens_seen": 15920128,
      "step": 7615
    },
    {
      "epoch": 1.2430867118035729,
      "grad_norm": 11.375,
      "learning_rate": 4.565602202805768e-05,
      "loss": 2.0089,
      "num_input_tokens_seen": 15930768,
      "step": 7620
    },
    {
      "epoch": 1.2439024390243902,
      "grad_norm": 6.34375,
      "learning_rate": 4.56504900840904e-05,
      "loss": 1.2776,
      "num_input_tokens_seen": 15940592,
      "step": 7625
    },
    {
      "epoch": 1.2447181662452076,
      "grad_norm": 4.3125,
      "learning_rate": 4.564495495555805e-05,
      "loss": 1.3889,
      "num_input_tokens_seen": 15950160,
      "step": 7630
    },
    {
      "epoch": 1.245533893466025,
      "grad_norm": 4.75,
      "learning_rate": 4.5639416643314204e-05,
      "loss": 1.2862,
      "num_input_tokens_seen": 15960880,
      "step": 7635
    },
    {
      "epoch": 1.2463496206868423,
      "grad_norm": 7.3125,
      "learning_rate": 4.5633875148212946e-05,
      "loss": 1.912,
      "num_input_tokens_seen": 15972064,
      "step": 7640
    },
    {
      "epoch": 1.2471653479076596,
      "grad_norm": 5.6875,
      "learning_rate": 4.562833047110883e-05,
      "loss": 2.5649,
      "num_input_tokens_seen": 15983184,
      "step": 7645
    },
    {
      "epoch": 1.247981075128477,
      "grad_norm": 5.59375,
      "learning_rate": 4.5622782612856923e-05,
      "loss": 2.2973,
      "num_input_tokens_seen": 15994624,
      "step": 7650
    },
    {
      "epoch": 1.2487968023492944,
      "grad_norm": 1.7890625,
      "learning_rate": 4.561723157431278e-05,
      "loss": 1.5844,
      "num_input_tokens_seen": 16004336,
      "step": 7655
    },
    {
      "epoch": 1.2496125295701117,
      "grad_norm": 8.25,
      "learning_rate": 4.5611677356332435e-05,
      "loss": 2.2688,
      "num_input_tokens_seen": 16014400,
      "step": 7660
    },
    {
      "epoch": 1.250428256790929,
      "grad_norm": 10.4375,
      "learning_rate": 4.560611995977242e-05,
      "loss": 2.6278,
      "num_input_tokens_seen": 16024720,
      "step": 7665
    },
    {
      "epoch": 1.2512439840117464,
      "grad_norm": 8.5,
      "learning_rate": 4.560055938548975e-05,
      "loss": 2.6172,
      "num_input_tokens_seen": 16036624,
      "step": 7670
    },
    {
      "epoch": 1.2520597112325638,
      "grad_norm": 11.625,
      "learning_rate": 4.5594995634341944e-05,
      "loss": 2.5714,
      "num_input_tokens_seen": 16046240,
      "step": 7675
    },
    {
      "epoch": 1.2528754384533811,
      "grad_norm": 9.5,
      "learning_rate": 4.5589428707187e-05,
      "loss": 2.9079,
      "num_input_tokens_seen": 16057232,
      "step": 7680
    },
    {
      "epoch": 1.2536911656741985,
      "grad_norm": 10.375,
      "learning_rate": 4.55838586048834e-05,
      "loss": 1.9369,
      "num_input_tokens_seen": 16067296,
      "step": 7685
    },
    {
      "epoch": 1.2545068928950158,
      "grad_norm": 16.875,
      "learning_rate": 4.557828532829013e-05,
      "loss": 3.1759,
      "num_input_tokens_seen": 16079312,
      "step": 7690
    },
    {
      "epoch": 1.2553226201158334,
      "grad_norm": 0.265625,
      "learning_rate": 4.557270887826667e-05,
      "loss": 2.2967,
      "num_input_tokens_seen": 16089776,
      "step": 7695
    },
    {
      "epoch": 1.2561383473366505,
      "grad_norm": 11.375,
      "learning_rate": 4.556712925567296e-05,
      "loss": 2.3469,
      "num_input_tokens_seen": 16100544,
      "step": 7700
    },
    {
      "epoch": 1.256954074557468,
      "grad_norm": 5.21875,
      "learning_rate": 4.5561546461369454e-05,
      "loss": 2.4295,
      "num_input_tokens_seen": 16111424,
      "step": 7705
    },
    {
      "epoch": 1.2577698017782852,
      "grad_norm": 10.5625,
      "learning_rate": 4.55559604962171e-05,
      "loss": 3.0378,
      "num_input_tokens_seen": 16121056,
      "step": 7710
    },
    {
      "epoch": 1.2585855289991028,
      "grad_norm": 1.9609375,
      "learning_rate": 4.55503713610773e-05,
      "loss": 2.651,
      "num_input_tokens_seen": 16130304,
      "step": 7715
    },
    {
      "epoch": 1.25940125621992,
      "grad_norm": 17.25,
      "learning_rate": 4.5544779056812e-05,
      "loss": 2.5304,
      "num_input_tokens_seen": 16141936,
      "step": 7720
    },
    {
      "epoch": 1.2602169834407375,
      "grad_norm": 10.875,
      "learning_rate": 4.553918358428358e-05,
      "loss": 2.5715,
      "num_input_tokens_seen": 16151664,
      "step": 7725
    },
    {
      "epoch": 1.2610327106615546,
      "grad_norm": 7.0625,
      "learning_rate": 4.553358494435494e-05,
      "loss": 2.8047,
      "num_input_tokens_seen": 16161936,
      "step": 7730
    },
    {
      "epoch": 1.2618484378823722,
      "grad_norm": 8.9375,
      "learning_rate": 4.5527983137889464e-05,
      "loss": 2.9944,
      "num_input_tokens_seen": 16172944,
      "step": 7735
    },
    {
      "epoch": 1.2626641651031896,
      "grad_norm": 2.5625,
      "learning_rate": 4.5522378165751015e-05,
      "loss": 1.7051,
      "num_input_tokens_seen": 16183136,
      "step": 7740
    },
    {
      "epoch": 1.263479892324007,
      "grad_norm": 5.03125,
      "learning_rate": 4.5516770028803954e-05,
      "loss": 2.0574,
      "num_input_tokens_seen": 16194592,
      "step": 7745
    },
    {
      "epoch": 1.2642956195448243,
      "grad_norm": 5.0,
      "learning_rate": 4.5511158727913116e-05,
      "loss": 3.056,
      "num_input_tokens_seen": 16203296,
      "step": 7750
    },
    {
      "epoch": 1.2651113467656416,
      "grad_norm": 10.3125,
      "learning_rate": 4.5505544263943856e-05,
      "loss": 2.6448,
      "num_input_tokens_seen": 16214096,
      "step": 7755
    },
    {
      "epoch": 1.265927073986459,
      "grad_norm": 3.5625,
      "learning_rate": 4.549992663776197e-05,
      "loss": 1.4155,
      "num_input_tokens_seen": 16224400,
      "step": 7760
    },
    {
      "epoch": 1.2667428012072763,
      "grad_norm": 7.9375,
      "learning_rate": 4.5494305850233786e-05,
      "loss": 2.6791,
      "num_input_tokens_seen": 16235152,
      "step": 7765
    },
    {
      "epoch": 1.2675585284280937,
      "grad_norm": 5.75,
      "learning_rate": 4.5488681902226094e-05,
      "loss": 2.8537,
      "num_input_tokens_seen": 16245552,
      "step": 7770
    },
    {
      "epoch": 1.268374255648911,
      "grad_norm": 6.5625,
      "learning_rate": 4.5483054794606174e-05,
      "loss": 2.5883,
      "num_input_tokens_seen": 16258272,
      "step": 7775
    },
    {
      "epoch": 1.2691899828697284,
      "grad_norm": 7.21875,
      "learning_rate": 4.547742452824179e-05,
      "loss": 1.6048,
      "num_input_tokens_seen": 16268096,
      "step": 7780
    },
    {
      "epoch": 1.2700057100905457,
      "grad_norm": 1.796875,
      "learning_rate": 4.5471791104001215e-05,
      "loss": 2.0426,
      "num_input_tokens_seen": 16278752,
      "step": 7785
    },
    {
      "epoch": 1.270821437311363,
      "grad_norm": 11.1875,
      "learning_rate": 4.546615452275319e-05,
      "loss": 2.8123,
      "num_input_tokens_seen": 16289328,
      "step": 7790
    },
    {
      "epoch": 1.2716371645321805,
      "grad_norm": 6.5625,
      "learning_rate": 4.5460514785366944e-05,
      "loss": 2.1361,
      "num_input_tokens_seen": 16299248,
      "step": 7795
    },
    {
      "epoch": 1.2724528917529978,
      "grad_norm": 6.375,
      "learning_rate": 4.545487189271219e-05,
      "loss": 1.9245,
      "num_input_tokens_seen": 16309376,
      "step": 7800
    },
    {
      "epoch": 1.2724528917529978,
      "eval_loss": 2.5406174659729004,
      "eval_runtime": 134.9886,
      "eval_samples_per_second": 20.187,
      "eval_steps_per_second": 10.097,
      "num_input_tokens_seen": 16309376,
      "step": 7800
    },
    {
      "epoch": 1.2732686189738152,
      "grad_norm": 5.625,
      "learning_rate": 4.544922584565914e-05,
      "loss": 2.609,
      "num_input_tokens_seen": 16319664,
      "step": 7805
    },
    {
      "epoch": 1.2740843461946325,
      "grad_norm": 8.5,
      "learning_rate": 4.544357664507848e-05,
      "loss": 1.9248,
      "num_input_tokens_seen": 16329776,
      "step": 7810
    },
    {
      "epoch": 1.2749000734154499,
      "grad_norm": 7.03125,
      "learning_rate": 4.54379242918414e-05,
      "loss": 3.1428,
      "num_input_tokens_seen": 16339520,
      "step": 7815
    },
    {
      "epoch": 1.2757158006362672,
      "grad_norm": 5.125,
      "learning_rate": 4.543226878681955e-05,
      "loss": 2.4886,
      "num_input_tokens_seen": 16351040,
      "step": 7820
    },
    {
      "epoch": 1.2765315278570846,
      "grad_norm": 1.8671875,
      "learning_rate": 4.5426610130885087e-05,
      "loss": 1.0436,
      "num_input_tokens_seen": 16360848,
      "step": 7825
    },
    {
      "epoch": 1.277347255077902,
      "grad_norm": 6.28125,
      "learning_rate": 4.542094832491064e-05,
      "loss": 3.4212,
      "num_input_tokens_seen": 16372256,
      "step": 7830
    },
    {
      "epoch": 1.2781629822987193,
      "grad_norm": 7.875,
      "learning_rate": 4.541528336976934e-05,
      "loss": 2.5651,
      "num_input_tokens_seen": 16384016,
      "step": 7835
    },
    {
      "epoch": 1.2789787095195366,
      "grad_norm": 6.0625,
      "learning_rate": 4.540961526633479e-05,
      "loss": 2.6668,
      "num_input_tokens_seen": 16395104,
      "step": 7840
    },
    {
      "epoch": 1.279794436740354,
      "grad_norm": 5.5625,
      "learning_rate": 4.540394401548108e-05,
      "loss": 2.9675,
      "num_input_tokens_seen": 16405280,
      "step": 7845
    },
    {
      "epoch": 1.2806101639611713,
      "grad_norm": 10.0625,
      "learning_rate": 4.539826961808279e-05,
      "loss": 1.8778,
      "num_input_tokens_seen": 16415264,
      "step": 7850
    },
    {
      "epoch": 1.2814258911819887,
      "grad_norm": 3.15625,
      "learning_rate": 4.5392592075014994e-05,
      "loss": 1.9129,
      "num_input_tokens_seen": 16424928,
      "step": 7855
    },
    {
      "epoch": 1.282241618402806,
      "grad_norm": 2.25,
      "learning_rate": 4.538691138715322e-05,
      "loss": 1.4572,
      "num_input_tokens_seen": 16434864,
      "step": 7860
    },
    {
      "epoch": 1.2830573456236234,
      "grad_norm": 10.5625,
      "learning_rate": 4.5381227555373516e-05,
      "loss": 2.9183,
      "num_input_tokens_seen": 16445584,
      "step": 7865
    },
    {
      "epoch": 1.2838730728444407,
      "grad_norm": 2.984375,
      "learning_rate": 4.537554058055239e-05,
      "loss": 1.5132,
      "num_input_tokens_seen": 16455440,
      "step": 7870
    },
    {
      "epoch": 1.284688800065258,
      "grad_norm": 8.1875,
      "learning_rate": 4.5369850463566865e-05,
      "loss": 2.5803,
      "num_input_tokens_seen": 16465024,
      "step": 7875
    },
    {
      "epoch": 1.2855045272860757,
      "grad_norm": 8.375,
      "learning_rate": 4.5364157205294404e-05,
      "loss": 1.9098,
      "num_input_tokens_seen": 16476128,
      "step": 7880
    },
    {
      "epoch": 1.2863202545068928,
      "grad_norm": 3.625,
      "learning_rate": 4.5358460806612996e-05,
      "loss": 2.3376,
      "num_input_tokens_seen": 16486704,
      "step": 7885
    },
    {
      "epoch": 1.2871359817277104,
      "grad_norm": 3.109375,
      "learning_rate": 4.535276126840109e-05,
      "loss": 2.509,
      "num_input_tokens_seen": 16497984,
      "step": 7890
    },
    {
      "epoch": 1.2879517089485275,
      "grad_norm": 2.953125,
      "learning_rate": 4.5347058591537626e-05,
      "loss": 2.7248,
      "num_input_tokens_seen": 16508848,
      "step": 7895
    },
    {
      "epoch": 1.288767436169345,
      "grad_norm": 7.34375,
      "learning_rate": 4.534135277690203e-05,
      "loss": 2.1849,
      "num_input_tokens_seen": 16519008,
      "step": 7900
    },
    {
      "epoch": 1.2895831633901622,
      "grad_norm": 6.53125,
      "learning_rate": 4.533564382537421e-05,
      "loss": 2.814,
      "num_input_tokens_seen": 16529120,
      "step": 7905
    },
    {
      "epoch": 1.2903988906109798,
      "grad_norm": 6.6875,
      "learning_rate": 4.532993173783456e-05,
      "loss": 2.3283,
      "num_input_tokens_seen": 16540032,
      "step": 7910
    },
    {
      "epoch": 1.291214617831797,
      "grad_norm": 4.75,
      "learning_rate": 4.5324216515163954e-05,
      "loss": 2.3421,
      "num_input_tokens_seen": 16551664,
      "step": 7915
    },
    {
      "epoch": 1.2920303450526145,
      "grad_norm": 6.34375,
      "learning_rate": 4.531849815824375e-05,
      "loss": 2.5789,
      "num_input_tokens_seen": 16561856,
      "step": 7920
    },
    {
      "epoch": 1.2928460722734318,
      "grad_norm": 6.28125,
      "learning_rate": 4.5312776667955795e-05,
      "loss": 2.5213,
      "num_input_tokens_seen": 16572688,
      "step": 7925
    },
    {
      "epoch": 1.2936617994942492,
      "grad_norm": 2.96875,
      "learning_rate": 4.5307052045182405e-05,
      "loss": 2.3967,
      "num_input_tokens_seen": 16583024,
      "step": 7930
    },
    {
      "epoch": 1.2944775267150666,
      "grad_norm": 11.0625,
      "learning_rate": 4.53013242908064e-05,
      "loss": 2.7529,
      "num_input_tokens_seen": 16593936,
      "step": 7935
    },
    {
      "epoch": 1.295293253935884,
      "grad_norm": 8.1875,
      "learning_rate": 4.529559340571107e-05,
      "loss": 1.5142,
      "num_input_tokens_seen": 16604016,
      "step": 7940
    },
    {
      "epoch": 1.2961089811567013,
      "grad_norm": 6.1875,
      "learning_rate": 4.528985939078018e-05,
      "loss": 1.5558,
      "num_input_tokens_seen": 16613744,
      "step": 7945
    },
    {
      "epoch": 1.2969247083775186,
      "grad_norm": 8.125,
      "learning_rate": 4.5284122246898e-05,
      "loss": 3.0316,
      "num_input_tokens_seen": 16625168,
      "step": 7950
    },
    {
      "epoch": 1.297740435598336,
      "grad_norm": 7.625,
      "learning_rate": 4.527838197494926e-05,
      "loss": 2.9055,
      "num_input_tokens_seen": 16635664,
      "step": 7955
    },
    {
      "epoch": 1.2985561628191533,
      "grad_norm": 8.625,
      "learning_rate": 4.527263857581918e-05,
      "loss": 1.9755,
      "num_input_tokens_seen": 16646736,
      "step": 7960
    },
    {
      "epoch": 1.2993718900399707,
      "grad_norm": 4.9375,
      "learning_rate": 4.526689205039347e-05,
      "loss": 3.1641,
      "num_input_tokens_seen": 16658432,
      "step": 7965
    },
    {
      "epoch": 1.300187617260788,
      "grad_norm": 4.96875,
      "learning_rate": 4.5261142399558324e-05,
      "loss": 1.9703,
      "num_input_tokens_seen": 16668608,
      "step": 7970
    },
    {
      "epoch": 1.3010033444816054,
      "grad_norm": 10.1875,
      "learning_rate": 4.525538962420041e-05,
      "loss": 2.7511,
      "num_input_tokens_seen": 16678224,
      "step": 7975
    },
    {
      "epoch": 1.3018190717024227,
      "grad_norm": 8.4375,
      "learning_rate": 4.524963372520685e-05,
      "loss": 2.2828,
      "num_input_tokens_seen": 16688960,
      "step": 7980
    },
    {
      "epoch": 1.30263479892324,
      "grad_norm": 3.84375,
      "learning_rate": 4.524387470346531e-05,
      "loss": 1.4005,
      "num_input_tokens_seen": 16699552,
      "step": 7985
    },
    {
      "epoch": 1.3034505261440574,
      "grad_norm": 3.421875,
      "learning_rate": 4.5238112559863885e-05,
      "loss": 2.1994,
      "num_input_tokens_seen": 16710720,
      "step": 7990
    },
    {
      "epoch": 1.3042662533648748,
      "grad_norm": 8.25,
      "learning_rate": 4.5232347295291175e-05,
      "loss": 1.922,
      "num_input_tokens_seen": 16719968,
      "step": 7995
    },
    {
      "epoch": 1.3050819805856921,
      "grad_norm": 6.5,
      "learning_rate": 4.522657891063626e-05,
      "loss": 1.3522,
      "num_input_tokens_seen": 16729632,
      "step": 8000
    },
    {
      "epoch": 1.3050819805856921,
      "eval_loss": 2.5443472862243652,
      "eval_runtime": 134.6824,
      "eval_samples_per_second": 20.233,
      "eval_steps_per_second": 10.12,
      "num_input_tokens_seen": 16729632,
      "step": 8000
    },
    {
      "epoch": 1.3058977078065095,
      "grad_norm": 8.75,
      "learning_rate": 4.52208074067887e-05,
      "loss": 2.08,
      "num_input_tokens_seen": 16740000,
      "step": 8005
    },
    {
      "epoch": 1.3067134350273268,
      "grad_norm": 11.9375,
      "learning_rate": 4.5215032784638516e-05,
      "loss": 4.4784,
      "num_input_tokens_seen": 16748928,
      "step": 8010
    },
    {
      "epoch": 1.3075291622481442,
      "grad_norm": 9.3125,
      "learning_rate": 4.5209255045076245e-05,
      "loss": 2.2037,
      "num_input_tokens_seen": 16759728,
      "step": 8015
    },
    {
      "epoch": 1.3083448894689615,
      "grad_norm": 7.125,
      "learning_rate": 4.5203474188992875e-05,
      "loss": 1.6067,
      "num_input_tokens_seen": 16770672,
      "step": 8020
    },
    {
      "epoch": 1.309160616689779,
      "grad_norm": 8.4375,
      "learning_rate": 4.51976902172799e-05,
      "loss": 2.3676,
      "num_input_tokens_seen": 16781008,
      "step": 8025
    },
    {
      "epoch": 1.3099763439105963,
      "grad_norm": 5.59375,
      "learning_rate": 4.519190313082927e-05,
      "loss": 1.862,
      "num_input_tokens_seen": 16791008,
      "step": 8030
    },
    {
      "epoch": 1.3107920711314136,
      "grad_norm": 5.90625,
      "learning_rate": 4.518611293053343e-05,
      "loss": 1.9986,
      "num_input_tokens_seen": 16800896,
      "step": 8035
    },
    {
      "epoch": 1.311607798352231,
      "grad_norm": 2.171875,
      "learning_rate": 4.51803196172853e-05,
      "loss": 3.0175,
      "num_input_tokens_seen": 16809072,
      "step": 8040
    },
    {
      "epoch": 1.3124235255730483,
      "grad_norm": 2.828125,
      "learning_rate": 4.517452319197828e-05,
      "loss": 3.0101,
      "num_input_tokens_seen": 16818976,
      "step": 8045
    },
    {
      "epoch": 1.3132392527938657,
      "grad_norm": 5.90625,
      "learning_rate": 4.5168723655506265e-05,
      "loss": 2.6953,
      "num_input_tokens_seen": 16828176,
      "step": 8050
    },
    {
      "epoch": 1.314054980014683,
      "grad_norm": 6.09375,
      "learning_rate": 4.51629210087636e-05,
      "loss": 2.3099,
      "num_input_tokens_seen": 16837680,
      "step": 8055
    },
    {
      "epoch": 1.3148707072355004,
      "grad_norm": 8.375,
      "learning_rate": 4.515711525264513e-05,
      "loss": 3.1837,
      "num_input_tokens_seen": 16847344,
      "step": 8060
    },
    {
      "epoch": 1.315686434456318,
      "grad_norm": 4.75,
      "learning_rate": 4.5151306388046175e-05,
      "loss": 1.6865,
      "num_input_tokens_seen": 16858832,
      "step": 8065
    },
    {
      "epoch": 1.316502161677135,
      "grad_norm": 6.15625,
      "learning_rate": 4.514549441586255e-05,
      "loss": 2.0347,
      "num_input_tokens_seen": 16868672,
      "step": 8070
    },
    {
      "epoch": 1.3173178888979526,
      "grad_norm": 12.0625,
      "learning_rate": 4.513967933699051e-05,
      "loss": 3.2597,
      "num_input_tokens_seen": 16879392,
      "step": 8075
    },
    {
      "epoch": 1.3181336161187698,
      "grad_norm": 1.96875,
      "learning_rate": 4.513386115232684e-05,
      "loss": 1.7723,
      "num_input_tokens_seen": 16889568,
      "step": 8080
    },
    {
      "epoch": 1.3189493433395874,
      "grad_norm": 15.0625,
      "learning_rate": 4.5128039862768745e-05,
      "loss": 3.5898,
      "num_input_tokens_seen": 16899392,
      "step": 8085
    },
    {
      "epoch": 1.3197650705604045,
      "grad_norm": 3.3125,
      "learning_rate": 4.512221546921397e-05,
      "loss": 2.6839,
      "num_input_tokens_seen": 16910448,
      "step": 8090
    },
    {
      "epoch": 1.320580797781222,
      "grad_norm": 3.046875,
      "learning_rate": 4.5116387972560694e-05,
      "loss": 1.4084,
      "num_input_tokens_seen": 16919808,
      "step": 8095
    },
    {
      "epoch": 1.3213965250020392,
      "grad_norm": 3.53125,
      "learning_rate": 4.511055737370759e-05,
      "loss": 1.3778,
      "num_input_tokens_seen": 16930288,
      "step": 8100
    },
    {
      "epoch": 1.3222122522228568,
      "grad_norm": 3.390625,
      "learning_rate": 4.510472367355383e-05,
      "loss": 1.8917,
      "num_input_tokens_seen": 16941200,
      "step": 8105
    },
    {
      "epoch": 1.3230279794436741,
      "grad_norm": 14.75,
      "learning_rate": 4.509888687299901e-05,
      "loss": 2.8396,
      "num_input_tokens_seen": 16950624,
      "step": 8110
    },
    {
      "epoch": 1.3238437066644915,
      "grad_norm": 16.25,
      "learning_rate": 4.5093046972943266e-05,
      "loss": 3.5705,
      "num_input_tokens_seen": 16960960,
      "step": 8115
    },
    {
      "epoch": 1.3246594338853088,
      "grad_norm": 0.85546875,
      "learning_rate": 4.508720397428717e-05,
      "loss": 2.7155,
      "num_input_tokens_seen": 16970224,
      "step": 8120
    },
    {
      "epoch": 1.3254751611061262,
      "grad_norm": 3.640625,
      "learning_rate": 4.508135787793178e-05,
      "loss": 1.5689,
      "num_input_tokens_seen": 16981600,
      "step": 8125
    },
    {
      "epoch": 1.3262908883269435,
      "grad_norm": 4.5,
      "learning_rate": 4.5075508684778664e-05,
      "loss": 2.3816,
      "num_input_tokens_seen": 16991264,
      "step": 8130
    },
    {
      "epoch": 1.3271066155477609,
      "grad_norm": 12.3125,
      "learning_rate": 4.506965639572982e-05,
      "loss": 2.723,
      "num_input_tokens_seen": 17003248,
      "step": 8135
    },
    {
      "epoch": 1.3279223427685782,
      "grad_norm": 10.5,
      "learning_rate": 4.506380101168774e-05,
      "loss": 2.6515,
      "num_input_tokens_seen": 17014016,
      "step": 8140
    },
    {
      "epoch": 1.3287380699893956,
      "grad_norm": 10.5,
      "learning_rate": 4.505794253355542e-05,
      "loss": 2.1346,
      "num_input_tokens_seen": 17023840,
      "step": 8145
    },
    {
      "epoch": 1.329553797210213,
      "grad_norm": 8.5,
      "learning_rate": 4.5052080962236286e-05,
      "loss": 2.7023,
      "num_input_tokens_seen": 17034176,
      "step": 8150
    },
    {
      "epoch": 1.3303695244310303,
      "grad_norm": 8.9375,
      "learning_rate": 4.504621629863428e-05,
      "loss": 2.8469,
      "num_input_tokens_seen": 17045424,
      "step": 8155
    },
    {
      "epoch": 1.3311852516518476,
      "grad_norm": 3.171875,
      "learning_rate": 4.504034854365381e-05,
      "loss": 3.282,
      "num_input_tokens_seen": 17056080,
      "step": 8160
    },
    {
      "epoch": 1.332000978872665,
      "grad_norm": 8.25,
      "learning_rate": 4.503447769819974e-05,
      "loss": 3.1554,
      "num_input_tokens_seen": 17065776,
      "step": 8165
    },
    {
      "epoch": 1.3328167060934824,
      "grad_norm": 7.90625,
      "learning_rate": 4.502860376317745e-05,
      "loss": 2.1478,
      "num_input_tokens_seen": 17076320,
      "step": 8170
    },
    {
      "epoch": 1.3336324333142997,
      "grad_norm": 7.78125,
      "learning_rate": 4.502272673949276e-05,
      "loss": 2.6394,
      "num_input_tokens_seen": 17086304,
      "step": 8175
    },
    {
      "epoch": 1.334448160535117,
      "grad_norm": 8.0625,
      "learning_rate": 4.501684662805199e-05,
      "loss": 1.7187,
      "num_input_tokens_seen": 17096976,
      "step": 8180
    },
    {
      "epoch": 1.3352638877559344,
      "grad_norm": 7.09375,
      "learning_rate": 4.5010963429761924e-05,
      "loss": 1.8292,
      "num_input_tokens_seen": 17108064,
      "step": 8185
    },
    {
      "epoch": 1.3360796149767518,
      "grad_norm": 9.875,
      "learning_rate": 4.500507714552982e-05,
      "loss": 2.5231,
      "num_input_tokens_seen": 17118912,
      "step": 8190
    },
    {
      "epoch": 1.3368953421975691,
      "grad_norm": 8.25,
      "learning_rate": 4.499918777626342e-05,
      "loss": 2.7606,
      "num_input_tokens_seen": 17129168,
      "step": 8195
    },
    {
      "epoch": 1.3377110694183865,
      "grad_norm": 8.3125,
      "learning_rate": 4.499329532287093e-05,
      "loss": 2.9042,
      "num_input_tokens_seen": 17139952,
      "step": 8200
    },
    {
      "epoch": 1.3377110694183865,
      "eval_loss": 2.5448949337005615,
      "eval_runtime": 134.8864,
      "eval_samples_per_second": 20.202,
      "eval_steps_per_second": 10.105,
      "num_input_tokens_seen": 17139952,
      "step": 8200
    },
    {
      "epoch": 1.3385267966392038,
      "grad_norm": 12.75,
      "learning_rate": 4.4987399786261064e-05,
      "loss": 1.9021,
      "num_input_tokens_seen": 17149664,
      "step": 8205
    },
    {
      "epoch": 1.3393425238600212,
      "grad_norm": 9.4375,
      "learning_rate": 4.498150116734297e-05,
      "loss": 3.6334,
      "num_input_tokens_seen": 17159040,
      "step": 8210
    },
    {
      "epoch": 1.3401582510808385,
      "grad_norm": 7.59375,
      "learning_rate": 4.4975599467026294e-05,
      "loss": 2.2201,
      "num_input_tokens_seen": 17169360,
      "step": 8215
    },
    {
      "epoch": 1.3409739783016559,
      "grad_norm": 10.0,
      "learning_rate": 4.496969468622114e-05,
      "loss": 3.4071,
      "num_input_tokens_seen": 17181920,
      "step": 8220
    },
    {
      "epoch": 1.3417897055224732,
      "grad_norm": 7.28125,
      "learning_rate": 4.496378682583813e-05,
      "loss": 1.0302,
      "num_input_tokens_seen": 17191600,
      "step": 8225
    },
    {
      "epoch": 1.3426054327432906,
      "grad_norm": 10.4375,
      "learning_rate": 4.495787588678829e-05,
      "loss": 3.1304,
      "num_input_tokens_seen": 17201776,
      "step": 8230
    },
    {
      "epoch": 1.343421159964108,
      "grad_norm": 9.9375,
      "learning_rate": 4.4951961869983196e-05,
      "loss": 2.6814,
      "num_input_tokens_seen": 17212864,
      "step": 8235
    },
    {
      "epoch": 1.3442368871849253,
      "grad_norm": 9.25,
      "learning_rate": 4.494604477633485e-05,
      "loss": 1.916,
      "num_input_tokens_seen": 17223888,
      "step": 8240
    },
    {
      "epoch": 1.3450526144057426,
      "grad_norm": 15.1875,
      "learning_rate": 4.4940124606755734e-05,
      "loss": 1.2973,
      "num_input_tokens_seen": 17234064,
      "step": 8245
    },
    {
      "epoch": 1.3458683416265602,
      "grad_norm": 4.4375,
      "learning_rate": 4.493420136215882e-05,
      "loss": 1.6788,
      "num_input_tokens_seen": 17243296,
      "step": 8250
    },
    {
      "epoch": 1.3466840688473773,
      "grad_norm": 11.375,
      "learning_rate": 4.492827504345756e-05,
      "loss": 1.9523,
      "num_input_tokens_seen": 17253520,
      "step": 8255
    },
    {
      "epoch": 1.347499796068195,
      "grad_norm": 4.40625,
      "learning_rate": 4.492234565156584e-05,
      "loss": 2.1785,
      "num_input_tokens_seen": 17264208,
      "step": 8260
    },
    {
      "epoch": 1.348315523289012,
      "grad_norm": 8.3125,
      "learning_rate": 4.491641318739807e-05,
      "loss": 2.0787,
      "num_input_tokens_seen": 17273296,
      "step": 8265
    },
    {
      "epoch": 1.3491312505098296,
      "grad_norm": 8.75,
      "learning_rate": 4.4910477651869096e-05,
      "loss": 2.9158,
      "num_input_tokens_seen": 17284848,
      "step": 8270
    },
    {
      "epoch": 1.3499469777306468,
      "grad_norm": 6.15625,
      "learning_rate": 4.4904539045894254e-05,
      "loss": 1.9896,
      "num_input_tokens_seen": 17296592,
      "step": 8275
    },
    {
      "epoch": 1.3507627049514643,
      "grad_norm": 7.3125,
      "learning_rate": 4.4898597370389364e-05,
      "loss": 2.6489,
      "num_input_tokens_seen": 17307264,
      "step": 8280
    },
    {
      "epoch": 1.3515784321722815,
      "grad_norm": 10.1875,
      "learning_rate": 4.489265262627069e-05,
      "loss": 2.1398,
      "num_input_tokens_seen": 17317040,
      "step": 8285
    },
    {
      "epoch": 1.352394159393099,
      "grad_norm": 1.375,
      "learning_rate": 4.488670481445499e-05,
      "loss": 1.9753,
      "num_input_tokens_seen": 17327072,
      "step": 8290
    },
    {
      "epoch": 1.3532098866139162,
      "grad_norm": 2.203125,
      "learning_rate": 4.488075393585951e-05,
      "loss": 1.4988,
      "num_input_tokens_seen": 17337360,
      "step": 8295
    },
    {
      "epoch": 1.3540256138347337,
      "grad_norm": 13.375,
      "learning_rate": 4.487479999140193e-05,
      "loss": 2.4538,
      "num_input_tokens_seen": 17347232,
      "step": 8300
    },
    {
      "epoch": 1.354841341055551,
      "grad_norm": 8.0625,
      "learning_rate": 4.4868842982000425e-05,
      "loss": 2.9812,
      "num_input_tokens_seen": 17357424,
      "step": 8305
    },
    {
      "epoch": 1.3556570682763684,
      "grad_norm": 7.46875,
      "learning_rate": 4.486288290857365e-05,
      "loss": 3.7324,
      "num_input_tokens_seen": 17367408,
      "step": 8310
    },
    {
      "epoch": 1.3564727954971858,
      "grad_norm": 4.84375,
      "learning_rate": 4.4856919772040715e-05,
      "loss": 2.0072,
      "num_input_tokens_seen": 17378480,
      "step": 8315
    },
    {
      "epoch": 1.3572885227180032,
      "grad_norm": 4.09375,
      "learning_rate": 4.485095357332122e-05,
      "loss": 1.7019,
      "num_input_tokens_seen": 17388784,
      "step": 8320
    },
    {
      "epoch": 1.3581042499388205,
      "grad_norm": 9.125,
      "learning_rate": 4.484498431333521e-05,
      "loss": 2.2157,
      "num_input_tokens_seen": 17399552,
      "step": 8325
    },
    {
      "epoch": 1.3589199771596379,
      "grad_norm": 2.28125,
      "learning_rate": 4.4839011993003245e-05,
      "loss": 2.6732,
      "num_input_tokens_seen": 17411168,
      "step": 8330
    },
    {
      "epoch": 1.3597357043804552,
      "grad_norm": 8.4375,
      "learning_rate": 4.4833036613246305e-05,
      "loss": 2.7798,
      "num_input_tokens_seen": 17422224,
      "step": 8335
    },
    {
      "epoch": 1.3605514316012726,
      "grad_norm": 9.5,
      "learning_rate": 4.482705817498589e-05,
      "loss": 2.9483,
      "num_input_tokens_seen": 17433792,
      "step": 8340
    },
    {
      "epoch": 1.36136715882209,
      "grad_norm": 11.5,
      "learning_rate": 4.4821076679143934e-05,
      "loss": 4.2509,
      "num_input_tokens_seen": 17444688,
      "step": 8345
    },
    {
      "epoch": 1.3621828860429073,
      "grad_norm": 2.828125,
      "learning_rate": 4.481509212664288e-05,
      "loss": 3.326,
      "num_input_tokens_seen": 17454720,
      "step": 8350
    },
    {
      "epoch": 1.3629986132637246,
      "grad_norm": 12.9375,
      "learning_rate": 4.480910451840559e-05,
      "loss": 2.1183,
      "num_input_tokens_seen": 17463424,
      "step": 8355
    },
    {
      "epoch": 1.363814340484542,
      "grad_norm": 5.78125,
      "learning_rate": 4.480311385535546e-05,
      "loss": 2.4097,
      "num_input_tokens_seen": 17474416,
      "step": 8360
    },
    {
      "epoch": 1.3646300677053593,
      "grad_norm": 5.875,
      "learning_rate": 4.47971201384163e-05,
      "loss": 2.4651,
      "num_input_tokens_seen": 17484496,
      "step": 8365
    },
    {
      "epoch": 1.3654457949261767,
      "grad_norm": 6.75,
      "learning_rate": 4.4791123368512446e-05,
      "loss": 2.473,
      "num_input_tokens_seen": 17495136,
      "step": 8370
    },
    {
      "epoch": 1.366261522146994,
      "grad_norm": 6.40625,
      "learning_rate": 4.478512354656864e-05,
      "loss": 2.3176,
      "num_input_tokens_seen": 17504848,
      "step": 8375
    },
    {
      "epoch": 1.3670772493678114,
      "grad_norm": 7.3125,
      "learning_rate": 4.477912067351016e-05,
      "loss": 1.6594,
      "num_input_tokens_seen": 17515072,
      "step": 8380
    },
    {
      "epoch": 1.3678929765886287,
      "grad_norm": 3.828125,
      "learning_rate": 4.477311475026271e-05,
      "loss": 3.1007,
      "num_input_tokens_seen": 17524672,
      "step": 8385
    },
    {
      "epoch": 1.368708703809446,
      "grad_norm": 10.375,
      "learning_rate": 4.476710577775248e-05,
      "loss": 2.9775,
      "num_input_tokens_seen": 17534976,
      "step": 8390
    },
    {
      "epoch": 1.3695244310302634,
      "grad_norm": 4.09375,
      "learning_rate": 4.476109375690612e-05,
      "loss": 1.6781,
      "num_input_tokens_seen": 17545248,
      "step": 8395
    },
    {
      "epoch": 1.3703401582510808,
      "grad_norm": 6.4375,
      "learning_rate": 4.4755078688650784e-05,
      "loss": 2.6583,
      "num_input_tokens_seen": 17557136,
      "step": 8400
    },
    {
      "epoch": 1.3703401582510808,
      "eval_loss": 2.540050983428955,
      "eval_runtime": 135.0136,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 17557136,
      "step": 8400
    },
    {
      "epoch": 1.3711558854718982,
      "grad_norm": 13.0625,
      "learning_rate": 4.474906057391406e-05,
      "loss": 3.358,
      "num_input_tokens_seen": 17567696,
      "step": 8405
    },
    {
      "epoch": 1.3719716126927155,
      "grad_norm": 10.25,
      "learning_rate": 4.4743039413624e-05,
      "loss": 2.8831,
      "num_input_tokens_seen": 17577008,
      "step": 8410
    },
    {
      "epoch": 1.3727873399135329,
      "grad_norm": 2.96875,
      "learning_rate": 4.473701520870916e-05,
      "loss": 2.505,
      "num_input_tokens_seen": 17587344,
      "step": 8415
    },
    {
      "epoch": 1.3736030671343502,
      "grad_norm": 4.90625,
      "learning_rate": 4.4730987960098544e-05,
      "loss": 2.9895,
      "num_input_tokens_seen": 17598096,
      "step": 8420
    },
    {
      "epoch": 1.3744187943551676,
      "grad_norm": 5.40625,
      "learning_rate": 4.4724957668721635e-05,
      "loss": 2.0436,
      "num_input_tokens_seen": 17609488,
      "step": 8425
    },
    {
      "epoch": 1.375234521575985,
      "grad_norm": 2.90625,
      "learning_rate": 4.471892433550836e-05,
      "loss": 1.4558,
      "num_input_tokens_seen": 17619920,
      "step": 8430
    },
    {
      "epoch": 1.3760502487968023,
      "grad_norm": 5.09375,
      "learning_rate": 4.471288796138916e-05,
      "loss": 3.1056,
      "num_input_tokens_seen": 17629600,
      "step": 8435
    },
    {
      "epoch": 1.3768659760176196,
      "grad_norm": 8.5625,
      "learning_rate": 4.470684854729491e-05,
      "loss": 2.015,
      "num_input_tokens_seen": 17640752,
      "step": 8440
    },
    {
      "epoch": 1.3776817032384372,
      "grad_norm": 9.1875,
      "learning_rate": 4.4700806094156955e-05,
      "loss": 2.4412,
      "num_input_tokens_seen": 17650432,
      "step": 8445
    },
    {
      "epoch": 1.3784974304592543,
      "grad_norm": 18.0,
      "learning_rate": 4.469476060290713e-05,
      "loss": 2.296,
      "num_input_tokens_seen": 17661600,
      "step": 8450
    },
    {
      "epoch": 1.379313157680072,
      "grad_norm": 2.078125,
      "learning_rate": 4.468871207447772e-05,
      "loss": 1.805,
      "num_input_tokens_seen": 17671792,
      "step": 8455
    },
    {
      "epoch": 1.380128884900889,
      "grad_norm": 8.1875,
      "learning_rate": 4.4682660509801486e-05,
      "loss": 2.446,
      "num_input_tokens_seen": 17682128,
      "step": 8460
    },
    {
      "epoch": 1.3809446121217066,
      "grad_norm": 6.65625,
      "learning_rate": 4.467660590981165e-05,
      "loss": 3.3865,
      "num_input_tokens_seen": 17694128,
      "step": 8465
    },
    {
      "epoch": 1.3817603393425237,
      "grad_norm": 13.0,
      "learning_rate": 4.467054827544191e-05,
      "loss": 2.6766,
      "num_input_tokens_seen": 17704128,
      "step": 8470
    },
    {
      "epoch": 1.3825760665633413,
      "grad_norm": 1.9296875,
      "learning_rate": 4.4664487607626434e-05,
      "loss": 1.8885,
      "num_input_tokens_seen": 17712960,
      "step": 8475
    },
    {
      "epoch": 1.3833917937841584,
      "grad_norm": 14.625,
      "learning_rate": 4.4658423907299845e-05,
      "loss": 2.7156,
      "num_input_tokens_seen": 17724096,
      "step": 8480
    },
    {
      "epoch": 1.384207521004976,
      "grad_norm": 7.84375,
      "learning_rate": 4.465235717539725e-05,
      "loss": 2.1476,
      "num_input_tokens_seen": 17735728,
      "step": 8485
    },
    {
      "epoch": 1.3850232482257934,
      "grad_norm": 4.34375,
      "learning_rate": 4.464628741285421e-05,
      "loss": 1.8076,
      "num_input_tokens_seen": 17745968,
      "step": 8490
    },
    {
      "epoch": 1.3858389754466107,
      "grad_norm": 2.984375,
      "learning_rate": 4.4640214620606754e-05,
      "loss": 2.437,
      "num_input_tokens_seen": 17757744,
      "step": 8495
    },
    {
      "epoch": 1.386654702667428,
      "grad_norm": 7.75,
      "learning_rate": 4.46341387995914e-05,
      "loss": 2.4694,
      "num_input_tokens_seen": 17767824,
      "step": 8500
    },
    {
      "epoch": 1.3874704298882454,
      "grad_norm": 6.75,
      "learning_rate": 4.4628059950745106e-05,
      "loss": 2.5645,
      "num_input_tokens_seen": 17778128,
      "step": 8505
    },
    {
      "epoch": 1.3882861571090628,
      "grad_norm": 9.5625,
      "learning_rate": 4.4621978075005297e-05,
      "loss": 2.0868,
      "num_input_tokens_seen": 17787888,
      "step": 8510
    },
    {
      "epoch": 1.3891018843298801,
      "grad_norm": 9.1875,
      "learning_rate": 4.461589317330989e-05,
      "loss": 2.4031,
      "num_input_tokens_seen": 17799664,
      "step": 8515
    },
    {
      "epoch": 1.3899176115506975,
      "grad_norm": 8.5,
      "learning_rate": 4.460980524659724e-05,
      "loss": 1.7137,
      "num_input_tokens_seen": 17809360,
      "step": 8520
    },
    {
      "epoch": 1.3907333387715148,
      "grad_norm": 10.0,
      "learning_rate": 4.46037142958062e-05,
      "loss": 2.7602,
      "num_input_tokens_seen": 17819184,
      "step": 8525
    },
    {
      "epoch": 1.3915490659923322,
      "grad_norm": 10.4375,
      "learning_rate": 4.4597620321876046e-05,
      "loss": 2.6755,
      "num_input_tokens_seen": 17830320,
      "step": 8530
    },
    {
      "epoch": 1.3923647932131495,
      "grad_norm": 3.59375,
      "learning_rate": 4.459152332574656e-05,
      "loss": 2.0528,
      "num_input_tokens_seen": 17838880,
      "step": 8535
    },
    {
      "epoch": 1.393180520433967,
      "grad_norm": 7.03125,
      "learning_rate": 4.4585423308357985e-05,
      "loss": 1.4005,
      "num_input_tokens_seen": 17848848,
      "step": 8540
    },
    {
      "epoch": 1.3939962476547842,
      "grad_norm": 7.53125,
      "learning_rate": 4.457932027065102e-05,
      "loss": 2.3557,
      "num_input_tokens_seen": 17859792,
      "step": 8545
    },
    {
      "epoch": 1.3948119748756016,
      "grad_norm": 5.71875,
      "learning_rate": 4.45732142135668e-05,
      "loss": 1.867,
      "num_input_tokens_seen": 17869872,
      "step": 8550
    },
    {
      "epoch": 1.395627702096419,
      "grad_norm": 5.0,
      "learning_rate": 4.4567105138046986e-05,
      "loss": 3.0756,
      "num_input_tokens_seen": 17881392,
      "step": 8555
    },
    {
      "epoch": 1.3964434293172363,
      "grad_norm": 5.15625,
      "learning_rate": 4.456099304503365e-05,
      "loss": 2.0689,
      "num_input_tokens_seen": 17892064,
      "step": 8560
    },
    {
      "epoch": 1.3972591565380537,
      "grad_norm": 7.0,
      "learning_rate": 4.455487793546939e-05,
      "loss": 2.1211,
      "num_input_tokens_seen": 17903280,
      "step": 8565
    },
    {
      "epoch": 1.398074883758871,
      "grad_norm": 9.8125,
      "learning_rate": 4.454875981029719e-05,
      "loss": 2.5391,
      "num_input_tokens_seen": 17913584,
      "step": 8570
    },
    {
      "epoch": 1.3988906109796884,
      "grad_norm": 4.0,
      "learning_rate": 4.454263867046057e-05,
      "loss": 0.876,
      "num_input_tokens_seen": 17923328,
      "step": 8575
    },
    {
      "epoch": 1.3997063382005057,
      "grad_norm": 4.34375,
      "learning_rate": 4.4536514516903484e-05,
      "loss": 1.562,
      "num_input_tokens_seen": 17933728,
      "step": 8580
    },
    {
      "epoch": 1.400522065421323,
      "grad_norm": 12.4375,
      "learning_rate": 4.453038735057034e-05,
      "loss": 3.1081,
      "num_input_tokens_seen": 17943376,
      "step": 8585
    },
    {
      "epoch": 1.4013377926421404,
      "grad_norm": 6.65625,
      "learning_rate": 4.4524257172406034e-05,
      "loss": 1.9284,
      "num_input_tokens_seen": 17954160,
      "step": 8590
    },
    {
      "epoch": 1.4021535198629578,
      "grad_norm": 8.0,
      "learning_rate": 4.451812398335592e-05,
      "loss": 3.2021,
      "num_input_tokens_seen": 17964384,
      "step": 8595
    },
    {
      "epoch": 1.4029692470837751,
      "grad_norm": 13.5,
      "learning_rate": 4.4511987784365805e-05,
      "loss": 2.7435,
      "num_input_tokens_seen": 17974864,
      "step": 8600
    },
    {
      "epoch": 1.4029692470837751,
      "eval_loss": 2.5591657161712646,
      "eval_runtime": 135.0229,
      "eval_samples_per_second": 20.182,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 17974864,
      "step": 8600
    },
    {
      "epoch": 1.4037849743045925,
      "grad_norm": 8.5,
      "learning_rate": 4.450584857638197e-05,
      "loss": 3.0168,
      "num_input_tokens_seen": 17983792,
      "step": 8605
    },
    {
      "epoch": 1.4046007015254098,
      "grad_norm": 3.984375,
      "learning_rate": 4.449970636035116e-05,
      "loss": 1.9743,
      "num_input_tokens_seen": 17994688,
      "step": 8610
    },
    {
      "epoch": 1.4054164287462272,
      "grad_norm": 6.28125,
      "learning_rate": 4.4493561137220574e-05,
      "loss": 2.567,
      "num_input_tokens_seen": 18005024,
      "step": 8615
    },
    {
      "epoch": 1.4062321559670445,
      "grad_norm": 5.09375,
      "learning_rate": 4.44874129079379e-05,
      "loss": 1.2527,
      "num_input_tokens_seen": 18015104,
      "step": 8620
    },
    {
      "epoch": 1.407047883187862,
      "grad_norm": 6.28125,
      "learning_rate": 4.4481261673451255e-05,
      "loss": 3.0014,
      "num_input_tokens_seen": 18024976,
      "step": 8625
    },
    {
      "epoch": 1.4078636104086795,
      "grad_norm": 9.3125,
      "learning_rate": 4.4475107434709245e-05,
      "loss": 2.8707,
      "num_input_tokens_seen": 18035520,
      "step": 8630
    },
    {
      "epoch": 1.4086793376294966,
      "grad_norm": 4.96875,
      "learning_rate": 4.446895019266093e-05,
      "loss": 0.6995,
      "num_input_tokens_seen": 18045312,
      "step": 8635
    },
    {
      "epoch": 1.4094950648503142,
      "grad_norm": 11.625,
      "learning_rate": 4.446278994825583e-05,
      "loss": 2.6908,
      "num_input_tokens_seen": 18055872,
      "step": 8640
    },
    {
      "epoch": 1.4103107920711313,
      "grad_norm": 6.59375,
      "learning_rate": 4.445662670244394e-05,
      "loss": 1.8127,
      "num_input_tokens_seen": 18066208,
      "step": 8645
    },
    {
      "epoch": 1.4111265192919489,
      "grad_norm": 5.625,
      "learning_rate": 4.44504604561757e-05,
      "loss": 2.8751,
      "num_input_tokens_seen": 18075968,
      "step": 8650
    },
    {
      "epoch": 1.411942246512766,
      "grad_norm": 11.4375,
      "learning_rate": 4.4444291210402035e-05,
      "loss": 4.1479,
      "num_input_tokens_seen": 18085728,
      "step": 8655
    },
    {
      "epoch": 1.4127579737335836,
      "grad_norm": 9.9375,
      "learning_rate": 4.443811896607431e-05,
      "loss": 2.9089,
      "num_input_tokens_seen": 18096144,
      "step": 8660
    },
    {
      "epoch": 1.4135737009544007,
      "grad_norm": 11.375,
      "learning_rate": 4.443194372414436e-05,
      "loss": 2.2833,
      "num_input_tokens_seen": 18107840,
      "step": 8665
    },
    {
      "epoch": 1.4143894281752183,
      "grad_norm": 7.59375,
      "learning_rate": 4.442576548556449e-05,
      "loss": 4.3797,
      "num_input_tokens_seen": 18118064,
      "step": 8670
    },
    {
      "epoch": 1.4152051553960356,
      "grad_norm": 7.90625,
      "learning_rate": 4.441958425128747e-05,
      "loss": 1.4752,
      "num_input_tokens_seen": 18128656,
      "step": 8675
    },
    {
      "epoch": 1.416020882616853,
      "grad_norm": 8.5,
      "learning_rate": 4.4413400022266515e-05,
      "loss": 2.6552,
      "num_input_tokens_seen": 18140560,
      "step": 8680
    },
    {
      "epoch": 1.4168366098376703,
      "grad_norm": 3.6875,
      "learning_rate": 4.4407212799455313e-05,
      "loss": 1.3233,
      "num_input_tokens_seen": 18152048,
      "step": 8685
    },
    {
      "epoch": 1.4176523370584877,
      "grad_norm": 3.046875,
      "learning_rate": 4.4401022583808003e-05,
      "loss": 1.2866,
      "num_input_tokens_seen": 18161808,
      "step": 8690
    },
    {
      "epoch": 1.418468064279305,
      "grad_norm": 3.734375,
      "learning_rate": 4.439482937627921e-05,
      "loss": 1.6412,
      "num_input_tokens_seen": 18173488,
      "step": 8695
    },
    {
      "epoch": 1.4192837915001224,
      "grad_norm": 9.875,
      "learning_rate": 4.4388633177824004e-05,
      "loss": 3.2706,
      "num_input_tokens_seen": 18182448,
      "step": 8700
    },
    {
      "epoch": 1.4200995187209398,
      "grad_norm": 5.375,
      "learning_rate": 4.4382433989397895e-05,
      "loss": 3.5,
      "num_input_tokens_seen": 18191296,
      "step": 8705
    },
    {
      "epoch": 1.420915245941757,
      "grad_norm": 5.90625,
      "learning_rate": 4.4376231811956895e-05,
      "loss": 3.1701,
      "num_input_tokens_seen": 18202128,
      "step": 8710
    },
    {
      "epoch": 1.4217309731625745,
      "grad_norm": 4.21875,
      "learning_rate": 4.437002664645745e-05,
      "loss": 1.6259,
      "num_input_tokens_seen": 18213280,
      "step": 8715
    },
    {
      "epoch": 1.4225467003833918,
      "grad_norm": 5.5,
      "learning_rate": 4.436381849385649e-05,
      "loss": 1.9465,
      "num_input_tokens_seen": 18223936,
      "step": 8720
    },
    {
      "epoch": 1.4233624276042092,
      "grad_norm": 7.71875,
      "learning_rate": 4.435760735511136e-05,
      "loss": 3.6822,
      "num_input_tokens_seen": 18234800,
      "step": 8725
    },
    {
      "epoch": 1.4241781548250265,
      "grad_norm": 2.890625,
      "learning_rate": 4.435139323117992e-05,
      "loss": 1.7686,
      "num_input_tokens_seen": 18244240,
      "step": 8730
    },
    {
      "epoch": 1.4249938820458439,
      "grad_norm": 2.109375,
      "learning_rate": 4.434517612302046e-05,
      "loss": 2.1102,
      "num_input_tokens_seen": 18256240,
      "step": 8735
    },
    {
      "epoch": 1.4258096092666612,
      "grad_norm": 8.8125,
      "learning_rate": 4.433895603159174e-05,
      "loss": 2.2475,
      "num_input_tokens_seen": 18268608,
      "step": 8740
    },
    {
      "epoch": 1.4266253364874786,
      "grad_norm": 3.84375,
      "learning_rate": 4.433273295785296e-05,
      "loss": 2.1408,
      "num_input_tokens_seen": 18279936,
      "step": 8745
    },
    {
      "epoch": 1.427441063708296,
      "grad_norm": 11.0625,
      "learning_rate": 4.432650690276382e-05,
      "loss": 2.4811,
      "num_input_tokens_seen": 18289552,
      "step": 8750
    },
    {
      "epoch": 1.4282567909291133,
      "grad_norm": 6.09375,
      "learning_rate": 4.4320277867284435e-05,
      "loss": 2.2092,
      "num_input_tokens_seen": 18301600,
      "step": 8755
    },
    {
      "epoch": 1.4290725181499306,
      "grad_norm": 6.65625,
      "learning_rate": 4.431404585237541e-05,
      "loss": 1.6207,
      "num_input_tokens_seen": 18311328,
      "step": 8760
    },
    {
      "epoch": 1.429888245370748,
      "grad_norm": 4.5625,
      "learning_rate": 4.43078108589978e-05,
      "loss": 1.8095,
      "num_input_tokens_seen": 18321408,
      "step": 8765
    },
    {
      "epoch": 1.4307039725915653,
      "grad_norm": 8.1875,
      "learning_rate": 4.4301572888113116e-05,
      "loss": 2.3585,
      "num_input_tokens_seen": 18332720,
      "step": 8770
    },
    {
      "epoch": 1.4315196998123827,
      "grad_norm": 6.34375,
      "learning_rate": 4.4295331940683337e-05,
      "loss": 2.205,
      "num_input_tokens_seen": 18342576,
      "step": 8775
    },
    {
      "epoch": 1.4323354270332,
      "grad_norm": 6.75,
      "learning_rate": 4.428908801767089e-05,
      "loss": 1.7936,
      "num_input_tokens_seen": 18353360,
      "step": 8780
    },
    {
      "epoch": 1.4331511542540174,
      "grad_norm": 5.8125,
      "learning_rate": 4.428284112003868e-05,
      "loss": 3.0113,
      "num_input_tokens_seen": 18363200,
      "step": 8785
    },
    {
      "epoch": 1.4339668814748348,
      "grad_norm": 0.259765625,
      "learning_rate": 4.4276591248750033e-05,
      "loss": 2.5336,
      "num_input_tokens_seen": 18373456,
      "step": 8790
    },
    {
      "epoch": 1.434782608695652,
      "grad_norm": 3.546875,
      "learning_rate": 4.4270338404768774e-05,
      "loss": 2.7362,
      "num_input_tokens_seen": 18384432,
      "step": 8795
    },
    {
      "epoch": 1.4355983359164695,
      "grad_norm": 3.953125,
      "learning_rate": 4.426408258905917e-05,
      "loss": 2.7448,
      "num_input_tokens_seen": 18394320,
      "step": 8800
    },
    {
      "epoch": 1.4355983359164695,
      "eval_loss": 2.5488646030426025,
      "eval_runtime": 134.9944,
      "eval_samples_per_second": 20.186,
      "eval_steps_per_second": 10.097,
      "num_input_tokens_seen": 18394320,
      "step": 8800
    },
    {
      "epoch": 1.4364140631372868,
      "grad_norm": 12.125,
      "learning_rate": 4.425782380258594e-05,
      "loss": 2.5851,
      "num_input_tokens_seen": 18404992,
      "step": 8805
    },
    {
      "epoch": 1.4372297903581042,
      "grad_norm": 4.09375,
      "learning_rate": 4.425156204631427e-05,
      "loss": 2.4947,
      "num_input_tokens_seen": 18416096,
      "step": 8810
    },
    {
      "epoch": 1.4380455175789217,
      "grad_norm": 10.1875,
      "learning_rate": 4.424529732120981e-05,
      "loss": 2.6177,
      "num_input_tokens_seen": 18426832,
      "step": 8815
    },
    {
      "epoch": 1.4388612447997389,
      "grad_norm": 3.59375,
      "learning_rate": 4.423902962823864e-05,
      "loss": 1.1538,
      "num_input_tokens_seen": 18437584,
      "step": 8820
    },
    {
      "epoch": 1.4396769720205564,
      "grad_norm": 13.25,
      "learning_rate": 4.423275896836733e-05,
      "loss": 3.9667,
      "num_input_tokens_seen": 18447024,
      "step": 8825
    },
    {
      "epoch": 1.4404926992413736,
      "grad_norm": 10.1875,
      "learning_rate": 4.42264853425629e-05,
      "loss": 3.4546,
      "num_input_tokens_seen": 18458512,
      "step": 8830
    },
    {
      "epoch": 1.4413084264621912,
      "grad_norm": 5.0625,
      "learning_rate": 4.4220208751792816e-05,
      "loss": 3.0126,
      "num_input_tokens_seen": 18469824,
      "step": 8835
    },
    {
      "epoch": 1.4421241536830083,
      "grad_norm": 5.625,
      "learning_rate": 4.421392919702499e-05,
      "loss": 2.1316,
      "num_input_tokens_seen": 18479936,
      "step": 8840
    },
    {
      "epoch": 1.4429398809038259,
      "grad_norm": 2.46875,
      "learning_rate": 4.4207646679227846e-05,
      "loss": 2.1269,
      "num_input_tokens_seen": 18489072,
      "step": 8845
    },
    {
      "epoch": 1.443755608124643,
      "grad_norm": 2.8125,
      "learning_rate": 4.42013611993702e-05,
      "loss": 2.1949,
      "num_input_tokens_seen": 18499200,
      "step": 8850
    },
    {
      "epoch": 1.4445713353454606,
      "grad_norm": 4.40625,
      "learning_rate": 4.419507275842135e-05,
      "loss": 2.6193,
      "num_input_tokens_seen": 18510080,
      "step": 8855
    },
    {
      "epoch": 1.445387062566278,
      "grad_norm": 5.53125,
      "learning_rate": 4.418878135735106e-05,
      "loss": 1.9718,
      "num_input_tokens_seen": 18517264,
      "step": 8860
    },
    {
      "epoch": 1.4462027897870953,
      "grad_norm": 8.5625,
      "learning_rate": 4.418248699712955e-05,
      "loss": 1.5938,
      "num_input_tokens_seen": 18528544,
      "step": 8865
    },
    {
      "epoch": 1.4470185170079126,
      "grad_norm": 6.0625,
      "learning_rate": 4.417618967872748e-05,
      "loss": 2.6605,
      "num_input_tokens_seen": 18540048,
      "step": 8870
    },
    {
      "epoch": 1.44783424422873,
      "grad_norm": 7.4375,
      "learning_rate": 4.4169889403115985e-05,
      "loss": 2.1911,
      "num_input_tokens_seen": 18550688,
      "step": 8875
    },
    {
      "epoch": 1.4486499714495473,
      "grad_norm": 2.671875,
      "learning_rate": 4.4163586171266627e-05,
      "loss": 2.6802,
      "num_input_tokens_seen": 18560960,
      "step": 8880
    },
    {
      "epoch": 1.4494656986703647,
      "grad_norm": 12.0,
      "learning_rate": 4.415727998415147e-05,
      "loss": 3.3054,
      "num_input_tokens_seen": 18572224,
      "step": 8885
    },
    {
      "epoch": 1.450281425891182,
      "grad_norm": 6.4375,
      "learning_rate": 4.4150970842742985e-05,
      "loss": 2.3937,
      "num_input_tokens_seen": 18581872,
      "step": 8890
    },
    {
      "epoch": 1.4510971531119994,
      "grad_norm": 5.09375,
      "learning_rate": 4.4144658748014134e-05,
      "loss": 1.1845,
      "num_input_tokens_seen": 18592736,
      "step": 8895
    },
    {
      "epoch": 1.4519128803328167,
      "grad_norm": 5.21875,
      "learning_rate": 4.413834370093831e-05,
      "loss": 1.9351,
      "num_input_tokens_seen": 18603232,
      "step": 8900
    },
    {
      "epoch": 1.452728607553634,
      "grad_norm": 8.8125,
      "learning_rate": 4.413202570248939e-05,
      "loss": 2.4058,
      "num_input_tokens_seen": 18614256,
      "step": 8905
    },
    {
      "epoch": 1.4535443347744514,
      "grad_norm": 2.1875,
      "learning_rate": 4.412570475364167e-05,
      "loss": 1.8693,
      "num_input_tokens_seen": 18625968,
      "step": 8910
    },
    {
      "epoch": 1.4543600619952688,
      "grad_norm": 10.125,
      "learning_rate": 4.411938085536994e-05,
      "loss": 1.8954,
      "num_input_tokens_seen": 18636048,
      "step": 8915
    },
    {
      "epoch": 1.4551757892160861,
      "grad_norm": 5.3125,
      "learning_rate": 4.41130540086494e-05,
      "loss": 1.2701,
      "num_input_tokens_seen": 18646144,
      "step": 8920
    },
    {
      "epoch": 1.4559915164369035,
      "grad_norm": 3.265625,
      "learning_rate": 4.4106724214455754e-05,
      "loss": 2.0906,
      "num_input_tokens_seen": 18657312,
      "step": 8925
    },
    {
      "epoch": 1.4568072436577209,
      "grad_norm": 6.625,
      "learning_rate": 4.4100391473765115e-05,
      "loss": 2.4887,
      "num_input_tokens_seen": 18668976,
      "step": 8930
    },
    {
      "epoch": 1.4576229708785382,
      "grad_norm": 12.9375,
      "learning_rate": 4.409405578755408e-05,
      "loss": 3.8476,
      "num_input_tokens_seen": 18679728,
      "step": 8935
    },
    {
      "epoch": 1.4584386980993556,
      "grad_norm": 4.125,
      "learning_rate": 4.4087717156799705e-05,
      "loss": 1.5907,
      "num_input_tokens_seen": 18691296,
      "step": 8940
    },
    {
      "epoch": 1.459254425320173,
      "grad_norm": 5.84375,
      "learning_rate": 4.408137558247946e-05,
      "loss": 1.7532,
      "num_input_tokens_seen": 18702080,
      "step": 8945
    },
    {
      "epoch": 1.4600701525409903,
      "grad_norm": 16.375,
      "learning_rate": 4.4075031065571306e-05,
      "loss": 2.6004,
      "num_input_tokens_seen": 18713120,
      "step": 8950
    },
    {
      "epoch": 1.4608858797618076,
      "grad_norm": 7.25,
      "learning_rate": 4.406868360705366e-05,
      "loss": 2.9629,
      "num_input_tokens_seen": 18723488,
      "step": 8955
    },
    {
      "epoch": 1.461701606982625,
      "grad_norm": 9.625,
      "learning_rate": 4.406233320790536e-05,
      "loss": 2.617,
      "num_input_tokens_seen": 18733968,
      "step": 8960
    },
    {
      "epoch": 1.4625173342034423,
      "grad_norm": 3.625,
      "learning_rate": 4.4055979869105734e-05,
      "loss": 1.9915,
      "num_input_tokens_seen": 18745664,
      "step": 8965
    },
    {
      "epoch": 1.4633330614242597,
      "grad_norm": 4.53125,
      "learning_rate": 4.404962359163454e-05,
      "loss": 2.7435,
      "num_input_tokens_seen": 18755680,
      "step": 8970
    },
    {
      "epoch": 1.464148788645077,
      "grad_norm": 9.125,
      "learning_rate": 4.404326437647199e-05,
      "loss": 3.6306,
      "num_input_tokens_seen": 18765920,
      "step": 8975
    },
    {
      "epoch": 1.4649645158658944,
      "grad_norm": 16.75,
      "learning_rate": 4.403690222459877e-05,
      "loss": 1.5588,
      "num_input_tokens_seen": 18776480,
      "step": 8980
    },
    {
      "epoch": 1.4657802430867117,
      "grad_norm": 8.6875,
      "learning_rate": 4.4030537136995984e-05,
      "loss": 3.4508,
      "num_input_tokens_seen": 18787632,
      "step": 8985
    },
    {
      "epoch": 1.466595970307529,
      "grad_norm": 11.0,
      "learning_rate": 4.402416911464523e-05,
      "loss": 2.5324,
      "num_input_tokens_seen": 18797488,
      "step": 8990
    },
    {
      "epoch": 1.4674116975283464,
      "grad_norm": 12.5,
      "learning_rate": 4.4017798158528516e-05,
      "loss": 2.3699,
      "num_input_tokens_seen": 18808880,
      "step": 8995
    },
    {
      "epoch": 1.468227424749164,
      "grad_norm": 8.875,
      "learning_rate": 4.401142426962834e-05,
      "loss": 3.6386,
      "num_input_tokens_seen": 18820208,
      "step": 9000
    },
    {
      "epoch": 1.468227424749164,
      "eval_loss": 2.5527656078338623,
      "eval_runtime": 134.8717,
      "eval_samples_per_second": 20.204,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 18820208,
      "step": 9000
    },
    {
      "epoch": 1.4690431519699811,
      "grad_norm": 7.6875,
      "learning_rate": 4.400504744892763e-05,
      "loss": 2.4906,
      "num_input_tokens_seen": 18830160,
      "step": 9005
    },
    {
      "epoch": 1.4698588791907987,
      "grad_norm": 2.484375,
      "learning_rate": 4.399866769740975e-05,
      "loss": 3.1367,
      "num_input_tokens_seen": 18839408,
      "step": 9010
    },
    {
      "epoch": 1.4706746064116158,
      "grad_norm": 15.0625,
      "learning_rate": 4.399228501605859e-05,
      "loss": 2.3759,
      "num_input_tokens_seen": 18847520,
      "step": 9015
    },
    {
      "epoch": 1.4714903336324334,
      "grad_norm": 8.625,
      "learning_rate": 4.398589940585839e-05,
      "loss": 2.9692,
      "num_input_tokens_seen": 18858608,
      "step": 9020
    },
    {
      "epoch": 1.4723060608532506,
      "grad_norm": 8.8125,
      "learning_rate": 4.3979510867793917e-05,
      "loss": 2.3692,
      "num_input_tokens_seen": 18868416,
      "step": 9025
    },
    {
      "epoch": 1.4731217880740681,
      "grad_norm": 2.671875,
      "learning_rate": 4.3973119402850346e-05,
      "loss": 2.5802,
      "num_input_tokens_seen": 18877408,
      "step": 9030
    },
    {
      "epoch": 1.4739375152948853,
      "grad_norm": 3.40625,
      "learning_rate": 4.396672501201334e-05,
      "loss": 1.6646,
      "num_input_tokens_seen": 18887552,
      "step": 9035
    },
    {
      "epoch": 1.4747532425157028,
      "grad_norm": 1.5078125,
      "learning_rate": 4.396032769626899e-05,
      "loss": 1.3979,
      "num_input_tokens_seen": 18896832,
      "step": 9040
    },
    {
      "epoch": 1.4755689697365202,
      "grad_norm": 7.9375,
      "learning_rate": 4.395392745660384e-05,
      "loss": 2.4439,
      "num_input_tokens_seen": 18907856,
      "step": 9045
    },
    {
      "epoch": 1.4763846969573375,
      "grad_norm": 11.0625,
      "learning_rate": 4.394752429400488e-05,
      "loss": 3.273,
      "num_input_tokens_seen": 18919008,
      "step": 9050
    },
    {
      "epoch": 1.477200424178155,
      "grad_norm": 1.6015625,
      "learning_rate": 4.394111820945957e-05,
      "loss": 2.0073,
      "num_input_tokens_seen": 18931344,
      "step": 9055
    },
    {
      "epoch": 1.4780161513989722,
      "grad_norm": 5.6875,
      "learning_rate": 4.393470920395579e-05,
      "loss": 2.8412,
      "num_input_tokens_seen": 18940528,
      "step": 9060
    },
    {
      "epoch": 1.4788318786197896,
      "grad_norm": 10.0625,
      "learning_rate": 4.392829727848192e-05,
      "loss": 2.9761,
      "num_input_tokens_seen": 18950784,
      "step": 9065
    },
    {
      "epoch": 1.479647605840607,
      "grad_norm": 5.96875,
      "learning_rate": 4.392188243402673e-05,
      "loss": 2.8897,
      "num_input_tokens_seen": 18961616,
      "step": 9070
    },
    {
      "epoch": 1.4804633330614243,
      "grad_norm": 4.5625,
      "learning_rate": 4.391546467157949e-05,
      "loss": 2.3154,
      "num_input_tokens_seen": 18972880,
      "step": 9075
    },
    {
      "epoch": 1.4812790602822417,
      "grad_norm": 6.9375,
      "learning_rate": 4.390904399212988e-05,
      "loss": 1.7477,
      "num_input_tokens_seen": 18984704,
      "step": 9080
    },
    {
      "epoch": 1.482094787503059,
      "grad_norm": 3.296875,
      "learning_rate": 4.390262039666807e-05,
      "loss": 2.0176,
      "num_input_tokens_seen": 18996560,
      "step": 9085
    },
    {
      "epoch": 1.4829105147238764,
      "grad_norm": 2.28125,
      "learning_rate": 4.389619388618464e-05,
      "loss": 2.1006,
      "num_input_tokens_seen": 19006176,
      "step": 9090
    },
    {
      "epoch": 1.4837262419446937,
      "grad_norm": 0.71484375,
      "learning_rate": 4.3889764461670655e-05,
      "loss": 1.9793,
      "num_input_tokens_seen": 19017120,
      "step": 9095
    },
    {
      "epoch": 1.484541969165511,
      "grad_norm": 6.125,
      "learning_rate": 4.38833321241176e-05,
      "loss": 2.067,
      "num_input_tokens_seen": 19028512,
      "step": 9100
    },
    {
      "epoch": 1.4853576963863284,
      "grad_norm": 4.28125,
      "learning_rate": 4.3876896874517434e-05,
      "loss": 1.9383,
      "num_input_tokens_seen": 19038992,
      "step": 9105
    },
    {
      "epoch": 1.4861734236071458,
      "grad_norm": 11.375,
      "learning_rate": 4.3870458713862554e-05,
      "loss": 2.7034,
      "num_input_tokens_seen": 19049728,
      "step": 9110
    },
    {
      "epoch": 1.4869891508279631,
      "grad_norm": 15.375,
      "learning_rate": 4.386401764314579e-05,
      "loss": 2.3843,
      "num_input_tokens_seen": 19059520,
      "step": 9115
    },
    {
      "epoch": 1.4878048780487805,
      "grad_norm": 4.6875,
      "learning_rate": 4.385757366336045e-05,
      "loss": 2.4358,
      "num_input_tokens_seen": 19070944,
      "step": 9120
    },
    {
      "epoch": 1.4886206052695978,
      "grad_norm": 8.75,
      "learning_rate": 4.385112677550027e-05,
      "loss": 3.5552,
      "num_input_tokens_seen": 19082672,
      "step": 9125
    },
    {
      "epoch": 1.4894363324904152,
      "grad_norm": 7.09375,
      "learning_rate": 4.384467698055945e-05,
      "loss": 3.0621,
      "num_input_tokens_seen": 19093536,
      "step": 9130
    },
    {
      "epoch": 1.4902520597112325,
      "grad_norm": 8.5625,
      "learning_rate": 4.383822427953261e-05,
      "loss": 2.1582,
      "num_input_tokens_seen": 19104848,
      "step": 9135
    },
    {
      "epoch": 1.4910677869320499,
      "grad_norm": 10.125,
      "learning_rate": 4.3831768673414864e-05,
      "loss": 2.286,
      "num_input_tokens_seen": 19116704,
      "step": 9140
    },
    {
      "epoch": 1.4918835141528672,
      "grad_norm": 8.5625,
      "learning_rate": 4.382531016320173e-05,
      "loss": 1.3295,
      "num_input_tokens_seen": 19128688,
      "step": 9145
    },
    {
      "epoch": 1.4926992413736846,
      "grad_norm": 1.28125,
      "learning_rate": 4.3818848749889184e-05,
      "loss": 1.5414,
      "num_input_tokens_seen": 19141664,
      "step": 9150
    },
    {
      "epoch": 1.493514968594502,
      "grad_norm": 5.1875,
      "learning_rate": 4.381238443447368e-05,
      "loss": 2.55,
      "num_input_tokens_seen": 19151408,
      "step": 9155
    },
    {
      "epoch": 1.4943306958153193,
      "grad_norm": 5.5625,
      "learning_rate": 4.380591721795208e-05,
      "loss": 2.8566,
      "num_input_tokens_seen": 19162016,
      "step": 9160
    },
    {
      "epoch": 1.4951464230361367,
      "grad_norm": 6.09375,
      "learning_rate": 4.3799447101321723e-05,
      "loss": 3.038,
      "num_input_tokens_seen": 19172064,
      "step": 9165
    },
    {
      "epoch": 1.495962150256954,
      "grad_norm": 7.53125,
      "learning_rate": 4.379297408558036e-05,
      "loss": 3.4502,
      "num_input_tokens_seen": 19181456,
      "step": 9170
    },
    {
      "epoch": 1.4967778774777714,
      "grad_norm": 15.9375,
      "learning_rate": 4.378649817172624e-05,
      "loss": 3.5764,
      "num_input_tokens_seen": 19191440,
      "step": 9175
    },
    {
      "epoch": 1.4975936046985887,
      "grad_norm": 5.0625,
      "learning_rate": 4.378001936075801e-05,
      "loss": 2.8938,
      "num_input_tokens_seen": 19203504,
      "step": 9180
    },
    {
      "epoch": 1.4984093319194063,
      "grad_norm": 7.40625,
      "learning_rate": 4.377353765367479e-05,
      "loss": 4.5858,
      "num_input_tokens_seen": 19214064,
      "step": 9185
    },
    {
      "epoch": 1.4992250591402234,
      "grad_norm": 8.3125,
      "learning_rate": 4.376705305147614e-05,
      "loss": 3.2526,
      "num_input_tokens_seen": 19223792,
      "step": 9190
    },
    {
      "epoch": 1.500040786361041,
      "grad_norm": 5.84375,
      "learning_rate": 4.376056555516206e-05,
      "loss": 2.5012,
      "num_input_tokens_seen": 19233312,
      "step": 9195
    },
    {
      "epoch": 1.5008565135818581,
      "grad_norm": 13.375,
      "learning_rate": 4.375407516573302e-05,
      "loss": 1.1522,
      "num_input_tokens_seen": 19244192,
      "step": 9200
    },
    {
      "epoch": 1.5008565135818581,
      "eval_loss": 2.5400924682617188,
      "eval_runtime": 135.0531,
      "eval_samples_per_second": 20.177,
      "eval_steps_per_second": 10.092,
      "num_input_tokens_seen": 19244192,
      "step": 9200
    },
    {
      "epoch": 1.5016722408026757,
      "grad_norm": 7.84375,
      "learning_rate": 4.3747581884189913e-05,
      "loss": 3.1154,
      "num_input_tokens_seen": 19254640,
      "step": 9205
    },
    {
      "epoch": 1.5024879680234928,
      "grad_norm": 8.1875,
      "learning_rate": 4.374108571153408e-05,
      "loss": 2.1562,
      "num_input_tokens_seen": 19265568,
      "step": 9210
    },
    {
      "epoch": 1.5033036952443104,
      "grad_norm": 12.8125,
      "learning_rate": 4.3734586648767316e-05,
      "loss": 2.4074,
      "num_input_tokens_seen": 19275360,
      "step": 9215
    },
    {
      "epoch": 1.5041194224651275,
      "grad_norm": 6.625,
      "learning_rate": 4.372808469689186e-05,
      "loss": 1.8542,
      "num_input_tokens_seen": 19286800,
      "step": 9220
    },
    {
      "epoch": 1.504935149685945,
      "grad_norm": 3.25,
      "learning_rate": 4.372157985691039e-05,
      "loss": 0.8788,
      "num_input_tokens_seen": 19296192,
      "step": 9225
    },
    {
      "epoch": 1.5057508769067622,
      "grad_norm": 5.5625,
      "learning_rate": 4.371507212982603e-05,
      "loss": 1.4798,
      "num_input_tokens_seen": 19306864,
      "step": 9230
    },
    {
      "epoch": 1.5065666041275798,
      "grad_norm": 11.125,
      "learning_rate": 4.370856151664236e-05,
      "loss": 2.9215,
      "num_input_tokens_seen": 19318288,
      "step": 9235
    },
    {
      "epoch": 1.507382331348397,
      "grad_norm": 4.53125,
      "learning_rate": 4.3702048018363404e-05,
      "loss": 1.2685,
      "num_input_tokens_seen": 19328112,
      "step": 9240
    },
    {
      "epoch": 1.5081980585692145,
      "grad_norm": 8.125,
      "learning_rate": 4.369553163599362e-05,
      "loss": 2.4029,
      "num_input_tokens_seen": 19339872,
      "step": 9245
    },
    {
      "epoch": 1.5090137857900316,
      "grad_norm": 10.5,
      "learning_rate": 4.3689012370537904e-05,
      "loss": 3.1361,
      "num_input_tokens_seen": 19349424,
      "step": 9250
    },
    {
      "epoch": 1.5098295130108492,
      "grad_norm": 6.6875,
      "learning_rate": 4.368249022300164e-05,
      "loss": 2.5712,
      "num_input_tokens_seen": 19360560,
      "step": 9255
    },
    {
      "epoch": 1.5106452402316666,
      "grad_norm": 6.25,
      "learning_rate": 4.367596519439059e-05,
      "loss": 1.9279,
      "num_input_tokens_seen": 19369968,
      "step": 9260
    },
    {
      "epoch": 1.511460967452484,
      "grad_norm": 2.875,
      "learning_rate": 4.366943728571101e-05,
      "loss": 3.1247,
      "num_input_tokens_seen": 19378624,
      "step": 9265
    },
    {
      "epoch": 1.5122766946733013,
      "grad_norm": 2.453125,
      "learning_rate": 4.366290649796959e-05,
      "loss": 1.5744,
      "num_input_tokens_seen": 19388224,
      "step": 9270
    },
    {
      "epoch": 1.5130924218941186,
      "grad_norm": 8.6875,
      "learning_rate": 4.3656372832173456e-05,
      "loss": 2.543,
      "num_input_tokens_seen": 19399120,
      "step": 9275
    },
    {
      "epoch": 1.513908149114936,
      "grad_norm": 12.0625,
      "learning_rate": 4.364983628933017e-05,
      "loss": 4.1608,
      "num_input_tokens_seen": 19409584,
      "step": 9280
    },
    {
      "epoch": 1.5147238763357533,
      "grad_norm": 14.6875,
      "learning_rate": 4.364329687044777e-05,
      "loss": 2.9229,
      "num_input_tokens_seen": 19419360,
      "step": 9285
    },
    {
      "epoch": 1.5155396035565707,
      "grad_norm": 5.0625,
      "learning_rate": 4.36367545765347e-05,
      "loss": 2.0604,
      "num_input_tokens_seen": 19428368,
      "step": 9290
    },
    {
      "epoch": 1.516355330777388,
      "grad_norm": 9.625,
      "learning_rate": 4.363020940859988e-05,
      "loss": 3.5126,
      "num_input_tokens_seen": 19438240,
      "step": 9295
    },
    {
      "epoch": 1.5171710579982054,
      "grad_norm": 5.875,
      "learning_rate": 4.362366136765263e-05,
      "loss": 2.4625,
      "num_input_tokens_seen": 19448784,
      "step": 9300
    },
    {
      "epoch": 1.5179867852190227,
      "grad_norm": 8.75,
      "learning_rate": 4.361711045470278e-05,
      "loss": 2.9133,
      "num_input_tokens_seen": 19458752,
      "step": 9305
    },
    {
      "epoch": 1.51880251243984,
      "grad_norm": 3.921875,
      "learning_rate": 4.3610556670760524e-05,
      "loss": 1.2429,
      "num_input_tokens_seen": 19469888,
      "step": 9310
    },
    {
      "epoch": 1.5196182396606575,
      "grad_norm": 5.78125,
      "learning_rate": 4.360400001683657e-05,
      "loss": 1.3817,
      "num_input_tokens_seen": 19480848,
      "step": 9315
    },
    {
      "epoch": 1.5204339668814748,
      "grad_norm": 12.1875,
      "learning_rate": 4.3597440493942e-05,
      "loss": 2.3602,
      "num_input_tokens_seen": 19491280,
      "step": 9320
    },
    {
      "epoch": 1.5212496941022922,
      "grad_norm": 5.8125,
      "learning_rate": 4.3590878103088405e-05,
      "loss": 1.841,
      "num_input_tokens_seen": 19501904,
      "step": 9325
    },
    {
      "epoch": 1.5220654213231095,
      "grad_norm": 7.53125,
      "learning_rate": 4.358431284528779e-05,
      "loss": 2.2008,
      "num_input_tokens_seen": 19511984,
      "step": 9330
    },
    {
      "epoch": 1.5228811485439269,
      "grad_norm": 10.75,
      "learning_rate": 4.357774472155257e-05,
      "loss": 3.1643,
      "num_input_tokens_seen": 19521952,
      "step": 9335
    },
    {
      "epoch": 1.5236968757647442,
      "grad_norm": 3.515625,
      "learning_rate": 4.3571173732895664e-05,
      "loss": 1.7009,
      "num_input_tokens_seen": 19531648,
      "step": 9340
    },
    {
      "epoch": 1.5245126029855616,
      "grad_norm": 8.6875,
      "learning_rate": 4.356459988033039e-05,
      "loss": 2.5876,
      "num_input_tokens_seen": 19540672,
      "step": 9345
    },
    {
      "epoch": 1.5253283302063791,
      "grad_norm": 10.625,
      "learning_rate": 4.355802316487051e-05,
      "loss": 3.7905,
      "num_input_tokens_seen": 19549328,
      "step": 9350
    },
    {
      "epoch": 1.5261440574271963,
      "grad_norm": 7.375,
      "learning_rate": 4.355144358753025e-05,
      "loss": 2.6895,
      "num_input_tokens_seen": 19559872,
      "step": 9355
    },
    {
      "epoch": 1.5269597846480139,
      "grad_norm": 8.5625,
      "learning_rate": 4.354486114932425e-05,
      "loss": 2.4303,
      "num_input_tokens_seen": 19570880,
      "step": 9360
    },
    {
      "epoch": 1.527775511868831,
      "grad_norm": 4.03125,
      "learning_rate": 4.353827585126762e-05,
      "loss": 1.2927,
      "num_input_tokens_seen": 19581536,
      "step": 9365
    },
    {
      "epoch": 1.5285912390896486,
      "grad_norm": 10.1875,
      "learning_rate": 4.353168769437588e-05,
      "loss": 3.0233,
      "num_input_tokens_seen": 19591344,
      "step": 9370
    },
    {
      "epoch": 1.5294069663104657,
      "grad_norm": 8.5625,
      "learning_rate": 4.3525096679665014e-05,
      "loss": 2.2789,
      "num_input_tokens_seen": 19601664,
      "step": 9375
    },
    {
      "epoch": 1.5302226935312833,
      "grad_norm": 0.123046875,
      "learning_rate": 4.351850280815144e-05,
      "loss": 1.5994,
      "num_input_tokens_seen": 19612192,
      "step": 9380
    },
    {
      "epoch": 1.5310384207521004,
      "grad_norm": 6.84375,
      "learning_rate": 4.3511906080852014e-05,
      "loss": 1.8388,
      "num_input_tokens_seen": 19622256,
      "step": 9385
    },
    {
      "epoch": 1.531854147972918,
      "grad_norm": 6.46875,
      "learning_rate": 4.350530649878404e-05,
      "loss": 1.4125,
      "num_input_tokens_seen": 19633424,
      "step": 9390
    },
    {
      "epoch": 1.532669875193735,
      "grad_norm": 5.8125,
      "learning_rate": 4.3498704062965246e-05,
      "loss": 2.4417,
      "num_input_tokens_seen": 19643584,
      "step": 9395
    },
    {
      "epoch": 1.5334856024145527,
      "grad_norm": 8.0625,
      "learning_rate": 4.3492098774413815e-05,
      "loss": 2.7058,
      "num_input_tokens_seen": 19654192,
      "step": 9400
    },
    {
      "epoch": 1.5334856024145527,
      "eval_loss": 2.5398964881896973,
      "eval_runtime": 134.9947,
      "eval_samples_per_second": 20.186,
      "eval_steps_per_second": 10.097,
      "num_input_tokens_seen": 19654192,
      "step": 9400
    },
    {
      "epoch": 1.5343013296353698,
      "grad_norm": 7.53125,
      "learning_rate": 4.3485490634148375e-05,
      "loss": 2.8161,
      "num_input_tokens_seen": 19665248,
      "step": 9405
    },
    {
      "epoch": 1.5351170568561874,
      "grad_norm": 6.90625,
      "learning_rate": 4.347887964318797e-05,
      "loss": 2.0126,
      "num_input_tokens_seen": 19674784,
      "step": 9410
    },
    {
      "epoch": 1.5359327840770045,
      "grad_norm": 1.7265625,
      "learning_rate": 4.34722658025521e-05,
      "loss": 1.6498,
      "num_input_tokens_seen": 19685632,
      "step": 9415
    },
    {
      "epoch": 1.536748511297822,
      "grad_norm": 4.8125,
      "learning_rate": 4.346564911326071e-05,
      "loss": 3.838,
      "num_input_tokens_seen": 19697184,
      "step": 9420
    },
    {
      "epoch": 1.5375642385186392,
      "grad_norm": 4.40625,
      "learning_rate": 4.345902957633418e-05,
      "loss": 1.0547,
      "num_input_tokens_seen": 19708064,
      "step": 9425
    },
    {
      "epoch": 1.5383799657394568,
      "grad_norm": 4.125,
      "learning_rate": 4.345240719279331e-05,
      "loss": 2.7169,
      "num_input_tokens_seen": 19717664,
      "step": 9430
    },
    {
      "epoch": 1.539195692960274,
      "grad_norm": 4.46875,
      "learning_rate": 4.3445781963659374e-05,
      "loss": 1.9193,
      "num_input_tokens_seen": 19728736,
      "step": 9435
    },
    {
      "epoch": 1.5400114201810915,
      "grad_norm": 3.96875,
      "learning_rate": 4.3439153889954045e-05,
      "loss": 3.4239,
      "num_input_tokens_seen": 19740080,
      "step": 9440
    },
    {
      "epoch": 1.5408271474019088,
      "grad_norm": 12.1875,
      "learning_rate": 4.343252297269946e-05,
      "loss": 2.7399,
      "num_input_tokens_seen": 19750688,
      "step": 9445
    },
    {
      "epoch": 1.5416428746227262,
      "grad_norm": 2.046875,
      "learning_rate": 4.342588921291821e-05,
      "loss": 1.8026,
      "num_input_tokens_seen": 19762176,
      "step": 9450
    },
    {
      "epoch": 1.5424586018435436,
      "grad_norm": 7.71875,
      "learning_rate": 4.341925261163328e-05,
      "loss": 3.1393,
      "num_input_tokens_seen": 19771920,
      "step": 9455
    },
    {
      "epoch": 1.543274329064361,
      "grad_norm": 7.78125,
      "learning_rate": 4.341261316986813e-05,
      "loss": 1.768,
      "num_input_tokens_seen": 19782784,
      "step": 9460
    },
    {
      "epoch": 1.5440900562851783,
      "grad_norm": 5.15625,
      "learning_rate": 4.340597088864664e-05,
      "loss": 2.0342,
      "num_input_tokens_seen": 19794256,
      "step": 9465
    },
    {
      "epoch": 1.5449057835059956,
      "grad_norm": 7.0625,
      "learning_rate": 4.339932576899313e-05,
      "loss": 2.4258,
      "num_input_tokens_seen": 19805168,
      "step": 9470
    },
    {
      "epoch": 1.545721510726813,
      "grad_norm": 5.84375,
      "learning_rate": 4.3392677811932375e-05,
      "loss": 2.2318,
      "num_input_tokens_seen": 19816288,
      "step": 9475
    },
    {
      "epoch": 1.5465372379476303,
      "grad_norm": 2.5625,
      "learning_rate": 4.338602701848956e-05,
      "loss": 1.4446,
      "num_input_tokens_seen": 19826144,
      "step": 9480
    },
    {
      "epoch": 1.5473529651684477,
      "grad_norm": 4.90625,
      "learning_rate": 4.337937338969033e-05,
      "loss": 2.8775,
      "num_input_tokens_seen": 19835840,
      "step": 9485
    },
    {
      "epoch": 1.548168692389265,
      "grad_norm": 13.3125,
      "learning_rate": 4.337271692656075e-05,
      "loss": 3.0846,
      "num_input_tokens_seen": 19848000,
      "step": 9490
    },
    {
      "epoch": 1.5489844196100824,
      "grad_norm": 6.15625,
      "learning_rate": 4.336605763012733e-05,
      "loss": 2.3266,
      "num_input_tokens_seen": 19858496,
      "step": 9495
    },
    {
      "epoch": 1.5498001468308997,
      "grad_norm": 6.65625,
      "learning_rate": 4.3359395501417026e-05,
      "loss": 1.5661,
      "num_input_tokens_seen": 19870304,
      "step": 9500
    },
    {
      "epoch": 1.550615874051717,
      "grad_norm": 4.28125,
      "learning_rate": 4.335273054145722e-05,
      "loss": 1.3735,
      "num_input_tokens_seen": 19881904,
      "step": 9505
    },
    {
      "epoch": 1.5514316012725344,
      "grad_norm": 5.4375,
      "learning_rate": 4.334606275127572e-05,
      "loss": 2.4925,
      "num_input_tokens_seen": 19892720,
      "step": 9510
    },
    {
      "epoch": 1.5522473284933518,
      "grad_norm": 5.21875,
      "learning_rate": 4.33393921319008e-05,
      "loss": 2.0904,
      "num_input_tokens_seen": 19902800,
      "step": 9515
    },
    {
      "epoch": 1.5530630557141691,
      "grad_norm": 8.5,
      "learning_rate": 4.3332718684361146e-05,
      "loss": 2.5532,
      "num_input_tokens_seen": 19913424,
      "step": 9520
    },
    {
      "epoch": 1.5538787829349865,
      "grad_norm": 8.0625,
      "learning_rate": 4.332604240968588e-05,
      "loss": 1.8244,
      "num_input_tokens_seen": 19922976,
      "step": 9525
    },
    {
      "epoch": 1.5546945101558038,
      "grad_norm": 5.90625,
      "learning_rate": 4.331936330890459e-05,
      "loss": 2.9644,
      "num_input_tokens_seen": 19932992,
      "step": 9530
    },
    {
      "epoch": 1.5555102373766214,
      "grad_norm": 18.125,
      "learning_rate": 4.331268138304725e-05,
      "loss": 1.7415,
      "num_input_tokens_seen": 19944160,
      "step": 9535
    },
    {
      "epoch": 1.5563259645974385,
      "grad_norm": 11.4375,
      "learning_rate": 4.330599663314431e-05,
      "loss": 1.007,
      "num_input_tokens_seen": 19955216,
      "step": 9540
    },
    {
      "epoch": 1.5571416918182561,
      "grad_norm": 5.84375,
      "learning_rate": 4.329930906022665e-05,
      "loss": 3.0016,
      "num_input_tokens_seen": 19965440,
      "step": 9545
    },
    {
      "epoch": 1.5579574190390733,
      "grad_norm": 6.5625,
      "learning_rate": 4.3292618665325564e-05,
      "loss": 1.9259,
      "num_input_tokens_seen": 19973936,
      "step": 9550
    },
    {
      "epoch": 1.5587731462598908,
      "grad_norm": 14.1875,
      "learning_rate": 4.3285925449472796e-05,
      "loss": 3.0655,
      "num_input_tokens_seen": 19984752,
      "step": 9555
    },
    {
      "epoch": 1.559588873480708,
      "grad_norm": 4.375,
      "learning_rate": 4.327922941370054e-05,
      "loss": 1.6832,
      "num_input_tokens_seen": 19995056,
      "step": 9560
    },
    {
      "epoch": 1.5604046007015255,
      "grad_norm": 6.875,
      "learning_rate": 4.3272530559041384e-05,
      "loss": 1.3606,
      "num_input_tokens_seen": 20005888,
      "step": 9565
    },
    {
      "epoch": 1.5612203279223427,
      "grad_norm": 2.75,
      "learning_rate": 4.32658288865284e-05,
      "loss": 2.2324,
      "num_input_tokens_seen": 20016688,
      "step": 9570
    },
    {
      "epoch": 1.5620360551431602,
      "grad_norm": 16.125,
      "learning_rate": 4.325912439719505e-05,
      "loss": 2.4209,
      "num_input_tokens_seen": 20027808,
      "step": 9575
    },
    {
      "epoch": 1.5628517823639774,
      "grad_norm": 2.6875,
      "learning_rate": 4.3252417092075266e-05,
      "loss": 1.0929,
      "num_input_tokens_seen": 20038144,
      "step": 9580
    },
    {
      "epoch": 1.563667509584795,
      "grad_norm": 3.765625,
      "learning_rate": 4.3245706972203385e-05,
      "loss": 3.2779,
      "num_input_tokens_seen": 20049552,
      "step": 9585
    },
    {
      "epoch": 1.564483236805612,
      "grad_norm": 7.59375,
      "learning_rate": 4.323899403861421e-05,
      "loss": 2.8241,
      "num_input_tokens_seen": 20057744,
      "step": 9590
    },
    {
      "epoch": 1.5652989640264297,
      "grad_norm": 2.640625,
      "learning_rate": 4.3232278292342935e-05,
      "loss": 1.6569,
      "num_input_tokens_seen": 20067536,
      "step": 9595
    },
    {
      "epoch": 1.5661146912472468,
      "grad_norm": 6.46875,
      "learning_rate": 4.322555973442524e-05,
      "loss": 1.7284,
      "num_input_tokens_seen": 20077520,
      "step": 9600
    },
    {
      "epoch": 1.5661146912472468,
      "eval_loss": 2.555157423019409,
      "eval_runtime": 134.7593,
      "eval_samples_per_second": 20.221,
      "eval_steps_per_second": 10.114,
      "num_input_tokens_seen": 20077520,
      "step": 9600
    },
    {
      "epoch": 1.5669304184680644,
      "grad_norm": 3.75,
      "learning_rate": 4.3218838365897184e-05,
      "loss": 1.7063,
      "num_input_tokens_seen": 20086640,
      "step": 9605
    },
    {
      "epoch": 1.5677461456888815,
      "grad_norm": 7.125,
      "learning_rate": 4.3212114187795306e-05,
      "loss": 2.1828,
      "num_input_tokens_seen": 20097472,
      "step": 9610
    },
    {
      "epoch": 1.568561872909699,
      "grad_norm": 12.4375,
      "learning_rate": 4.320538720115656e-05,
      "loss": 1.602,
      "num_input_tokens_seen": 20107104,
      "step": 9615
    },
    {
      "epoch": 1.5693776001305162,
      "grad_norm": 9.6875,
      "learning_rate": 4.319865740701831e-05,
      "loss": 1.4984,
      "num_input_tokens_seen": 20117280,
      "step": 9620
    },
    {
      "epoch": 1.5701933273513338,
      "grad_norm": 12.4375,
      "learning_rate": 4.3191924806418396e-05,
      "loss": 2.1557,
      "num_input_tokens_seen": 20129120,
      "step": 9625
    },
    {
      "epoch": 1.5710090545721511,
      "grad_norm": 5.65625,
      "learning_rate": 4.318518940039507e-05,
      "loss": 1.5604,
      "num_input_tokens_seen": 20138672,
      "step": 9630
    },
    {
      "epoch": 1.5718247817929685,
      "grad_norm": 16.125,
      "learning_rate": 4.3178451189987e-05,
      "loss": 2.5694,
      "num_input_tokens_seen": 20149888,
      "step": 9635
    },
    {
      "epoch": 1.5726405090137858,
      "grad_norm": 4.625,
      "learning_rate": 4.3171710176233315e-05,
      "loss": 1.3876,
      "num_input_tokens_seen": 20160192,
      "step": 9640
    },
    {
      "epoch": 1.5734562362346032,
      "grad_norm": 10.3125,
      "learning_rate": 4.316496636017355e-05,
      "loss": 3.0902,
      "num_input_tokens_seen": 20170608,
      "step": 9645
    },
    {
      "epoch": 1.5742719634554205,
      "grad_norm": 5.59375,
      "learning_rate": 4.315821974284771e-05,
      "loss": 2.1185,
      "num_input_tokens_seen": 20181936,
      "step": 9650
    },
    {
      "epoch": 1.5750876906762379,
      "grad_norm": 5.90625,
      "learning_rate": 4.315147032529619e-05,
      "loss": 2.2076,
      "num_input_tokens_seen": 20191008,
      "step": 9655
    },
    {
      "epoch": 1.5759034178970552,
      "grad_norm": 15.3125,
      "learning_rate": 4.3144718108559845e-05,
      "loss": 2.3942,
      "num_input_tokens_seen": 20199840,
      "step": 9660
    },
    {
      "epoch": 1.5767191451178726,
      "grad_norm": 7.875,
      "learning_rate": 4.3137963093679945e-05,
      "loss": 2.6206,
      "num_input_tokens_seen": 20211152,
      "step": 9665
    },
    {
      "epoch": 1.57753487233869,
      "grad_norm": 4.46875,
      "learning_rate": 4.31312052816982e-05,
      "loss": 1.6106,
      "num_input_tokens_seen": 20220304,
      "step": 9670
    },
    {
      "epoch": 1.5783505995595073,
      "grad_norm": 6.21875,
      "learning_rate": 4.312444467365675e-05,
      "loss": 1.7438,
      "num_input_tokens_seen": 20230432,
      "step": 9675
    },
    {
      "epoch": 1.5791663267803246,
      "grad_norm": 6.03125,
      "learning_rate": 4.311768127059816e-05,
      "loss": 2.0414,
      "num_input_tokens_seen": 20240896,
      "step": 9680
    },
    {
      "epoch": 1.579982054001142,
      "grad_norm": 4.875,
      "learning_rate": 4.3110915073565444e-05,
      "loss": 2.1504,
      "num_input_tokens_seen": 20252448,
      "step": 9685
    },
    {
      "epoch": 1.5807977812219594,
      "grad_norm": 1.75,
      "learning_rate": 4.310414608360203e-05,
      "loss": 2.5433,
      "num_input_tokens_seen": 20262768,
      "step": 9690
    },
    {
      "epoch": 1.5816135084427767,
      "grad_norm": 5.3125,
      "learning_rate": 4.309737430175177e-05,
      "loss": 2.1019,
      "num_input_tokens_seen": 20273584,
      "step": 9695
    },
    {
      "epoch": 1.582429235663594,
      "grad_norm": 8.4375,
      "learning_rate": 4.309059972905897e-05,
      "loss": 2.0451,
      "num_input_tokens_seen": 20285056,
      "step": 9700
    },
    {
      "epoch": 1.5832449628844114,
      "grad_norm": 9.25,
      "learning_rate": 4.308382236656836e-05,
      "loss": 1.8828,
      "num_input_tokens_seen": 20296000,
      "step": 9705
    },
    {
      "epoch": 1.5840606901052288,
      "grad_norm": 4.5,
      "learning_rate": 4.307704221532507e-05,
      "loss": 2.2228,
      "num_input_tokens_seen": 20306464,
      "step": 9710
    },
    {
      "epoch": 1.5848764173260461,
      "grad_norm": 9.9375,
      "learning_rate": 4.307025927637471e-05,
      "loss": 2.4342,
      "num_input_tokens_seen": 20316320,
      "step": 9715
    },
    {
      "epoch": 1.5856921445468637,
      "grad_norm": 8.875,
      "learning_rate": 4.306347355076328e-05,
      "loss": 2.5944,
      "num_input_tokens_seen": 20327488,
      "step": 9720
    },
    {
      "epoch": 1.5865078717676808,
      "grad_norm": 4.34375,
      "learning_rate": 4.305668503953724e-05,
      "loss": 4.4652,
      "num_input_tokens_seen": 20336832,
      "step": 9725
    },
    {
      "epoch": 1.5873235989884984,
      "grad_norm": 5.0,
      "learning_rate": 4.3049893743743436e-05,
      "loss": 2.4719,
      "num_input_tokens_seen": 20346288,
      "step": 9730
    },
    {
      "epoch": 1.5881393262093155,
      "grad_norm": 9.0625,
      "learning_rate": 4.304309966442919e-05,
      "loss": 2.1083,
      "num_input_tokens_seen": 20356016,
      "step": 9735
    },
    {
      "epoch": 1.588955053430133,
      "grad_norm": 12.9375,
      "learning_rate": 4.303630280264224e-05,
      "loss": 3.3368,
      "num_input_tokens_seen": 20368208,
      "step": 9740
    },
    {
      "epoch": 1.5897707806509502,
      "grad_norm": 3.515625,
      "learning_rate": 4.302950315943074e-05,
      "loss": 2.7414,
      "num_input_tokens_seen": 20379408,
      "step": 9745
    },
    {
      "epoch": 1.5905865078717678,
      "grad_norm": 7.0,
      "learning_rate": 4.3022700735843275e-05,
      "loss": 2.9612,
      "num_input_tokens_seen": 20388864,
      "step": 9750
    },
    {
      "epoch": 1.591402235092585,
      "grad_norm": 2.8125,
      "learning_rate": 4.301589553292887e-05,
      "loss": 2.1642,
      "num_input_tokens_seen": 20399120,
      "step": 9755
    },
    {
      "epoch": 1.5922179623134025,
      "grad_norm": 7.21875,
      "learning_rate": 4.300908755173697e-05,
      "loss": 1.8425,
      "num_input_tokens_seen": 20410464,
      "step": 9760
    },
    {
      "epoch": 1.5930336895342196,
      "grad_norm": 1.4921875,
      "learning_rate": 4.300227679331745e-05,
      "loss": 2.1656,
      "num_input_tokens_seen": 20420400,
      "step": 9765
    },
    {
      "epoch": 1.5938494167550372,
      "grad_norm": 10.625,
      "learning_rate": 4.299546325872063e-05,
      "loss": 2.9693,
      "num_input_tokens_seen": 20430512,
      "step": 9770
    },
    {
      "epoch": 1.5946651439758543,
      "grad_norm": 13.0625,
      "learning_rate": 4.2988646948997225e-05,
      "loss": 2.0745,
      "num_input_tokens_seen": 20440848,
      "step": 9775
    },
    {
      "epoch": 1.595480871196672,
      "grad_norm": 6.625,
      "learning_rate": 4.29818278651984e-05,
      "loss": 2.458,
      "num_input_tokens_seen": 20451520,
      "step": 9780
    },
    {
      "epoch": 1.596296598417489,
      "grad_norm": 7.34375,
      "learning_rate": 4.297500600837574e-05,
      "loss": 3.2988,
      "num_input_tokens_seen": 20461856,
      "step": 9785
    },
    {
      "epoch": 1.5971123256383066,
      "grad_norm": 8.25,
      "learning_rate": 4.2968181379581276e-05,
      "loss": 3.4894,
      "num_input_tokens_seen": 20473648,
      "step": 9790
    },
    {
      "epoch": 1.5979280528591238,
      "grad_norm": 9.1875,
      "learning_rate": 4.296135397986743e-05,
      "loss": 2.5406,
      "num_input_tokens_seen": 20482752,
      "step": 9795
    },
    {
      "epoch": 1.5987437800799413,
      "grad_norm": 6.1875,
      "learning_rate": 4.295452381028709e-05,
      "loss": 2.7139,
      "num_input_tokens_seen": 20493344,
      "step": 9800
    },
    {
      "epoch": 1.5987437800799413,
      "eval_loss": 2.5283308029174805,
      "eval_runtime": 134.8676,
      "eval_samples_per_second": 20.205,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 20493344,
      "step": 9800
    },
    {
      "epoch": 1.5995595073007585,
      "grad_norm": 16.0,
      "learning_rate": 4.294769087189354e-05,
      "loss": 3.8364,
      "num_input_tokens_seen": 20505264,
      "step": 9805
    },
    {
      "epoch": 1.600375234521576,
      "grad_norm": 6.9375,
      "learning_rate": 4.294085516574052e-05,
      "loss": 1.7072,
      "num_input_tokens_seen": 20515632,
      "step": 9810
    },
    {
      "epoch": 1.6011909617423934,
      "grad_norm": 5.09375,
      "learning_rate": 4.2934016692882176e-05,
      "loss": 1.2228,
      "num_input_tokens_seen": 20526848,
      "step": 9815
    },
    {
      "epoch": 1.6020066889632107,
      "grad_norm": 7.875,
      "learning_rate": 4.292717545437308e-05,
      "loss": 2.6823,
      "num_input_tokens_seen": 20536416,
      "step": 9820
    },
    {
      "epoch": 1.602822416184028,
      "grad_norm": 7.75,
      "learning_rate": 4.292033145126825e-05,
      "loss": 2.4265,
      "num_input_tokens_seen": 20547008,
      "step": 9825
    },
    {
      "epoch": 1.6036381434048455,
      "grad_norm": 9.3125,
      "learning_rate": 4.29134846846231e-05,
      "loss": 1.145,
      "num_input_tokens_seen": 20558064,
      "step": 9830
    },
    {
      "epoch": 1.6044538706256628,
      "grad_norm": 4.25,
      "learning_rate": 4.29066351554935e-05,
      "loss": 1.7239,
      "num_input_tokens_seen": 20568512,
      "step": 9835
    },
    {
      "epoch": 1.6052695978464802,
      "grad_norm": 7.59375,
      "learning_rate": 4.289978286493574e-05,
      "loss": 2.7111,
      "num_input_tokens_seen": 20578256,
      "step": 9840
    },
    {
      "epoch": 1.6060853250672975,
      "grad_norm": 12.6875,
      "learning_rate": 4.28929278140065e-05,
      "loss": 3.2203,
      "num_input_tokens_seen": 20589344,
      "step": 9845
    },
    {
      "epoch": 1.6069010522881149,
      "grad_norm": 4.3125,
      "learning_rate": 4.288607000376295e-05,
      "loss": 2.4628,
      "num_input_tokens_seen": 20600960,
      "step": 9850
    },
    {
      "epoch": 1.6077167795089322,
      "grad_norm": 7.5625,
      "learning_rate": 4.2879209435262624e-05,
      "loss": 2.3288,
      "num_input_tokens_seen": 20611424,
      "step": 9855
    },
    {
      "epoch": 1.6085325067297496,
      "grad_norm": 6.84375,
      "learning_rate": 4.287234610956353e-05,
      "loss": 3.4665,
      "num_input_tokens_seen": 20621824,
      "step": 9860
    },
    {
      "epoch": 1.609348233950567,
      "grad_norm": 12.5625,
      "learning_rate": 4.2865480027724056e-05,
      "loss": 2.9214,
      "num_input_tokens_seen": 20631264,
      "step": 9865
    },
    {
      "epoch": 1.6101639611713843,
      "grad_norm": 8.5625,
      "learning_rate": 4.285861119080306e-05,
      "loss": 2.6535,
      "num_input_tokens_seen": 20642096,
      "step": 9870
    },
    {
      "epoch": 1.6109796883922016,
      "grad_norm": 9.875,
      "learning_rate": 4.2851739599859784e-05,
      "loss": 2.9815,
      "num_input_tokens_seen": 20652976,
      "step": 9875
    },
    {
      "epoch": 1.611795415613019,
      "grad_norm": 10.3125,
      "learning_rate": 4.2844865255953934e-05,
      "loss": 3.4443,
      "num_input_tokens_seen": 20661920,
      "step": 9880
    },
    {
      "epoch": 1.6126111428338363,
      "grad_norm": 11.4375,
      "learning_rate": 4.2837988160145605e-05,
      "loss": 2.2692,
      "num_input_tokens_seen": 20672336,
      "step": 9885
    },
    {
      "epoch": 1.6134268700546537,
      "grad_norm": 4.0,
      "learning_rate": 4.2831108313495336e-05,
      "loss": 1.8791,
      "num_input_tokens_seen": 20684448,
      "step": 9890
    },
    {
      "epoch": 1.614242597275471,
      "grad_norm": 3.421875,
      "learning_rate": 4.282422571706408e-05,
      "loss": 2.0034,
      "num_input_tokens_seen": 20695248,
      "step": 9895
    },
    {
      "epoch": 1.6150583244962884,
      "grad_norm": 2.28125,
      "learning_rate": 4.281734037191323e-05,
      "loss": 2.2174,
      "num_input_tokens_seen": 20705488,
      "step": 9900
    },
    {
      "epoch": 1.615874051717106,
      "grad_norm": 5.5,
      "learning_rate": 4.281045227910459e-05,
      "loss": 1.578,
      "num_input_tokens_seen": 20715120,
      "step": 9905
    },
    {
      "epoch": 1.616689778937923,
      "grad_norm": 3.703125,
      "learning_rate": 4.280356143970038e-05,
      "loss": 1.6327,
      "num_input_tokens_seen": 20724624,
      "step": 9910
    },
    {
      "epoch": 1.6175055061587407,
      "grad_norm": 11.3125,
      "learning_rate": 4.279666785476327e-05,
      "loss": 2.7766,
      "num_input_tokens_seen": 20735664,
      "step": 9915
    },
    {
      "epoch": 1.6183212333795578,
      "grad_norm": 8.5,
      "learning_rate": 4.2789771525356325e-05,
      "loss": 2.478,
      "num_input_tokens_seen": 20745104,
      "step": 9920
    },
    {
      "epoch": 1.6191369606003754,
      "grad_norm": 8.5,
      "learning_rate": 4.2782872452543056e-05,
      "loss": 2.5877,
      "num_input_tokens_seen": 20755632,
      "step": 9925
    },
    {
      "epoch": 1.6199526878211925,
      "grad_norm": 5.8125,
      "learning_rate": 4.2775970637387376e-05,
      "loss": 1.6479,
      "num_input_tokens_seen": 20766304,
      "step": 9930
    },
    {
      "epoch": 1.62076841504201,
      "grad_norm": 5.875,
      "learning_rate": 4.276906608095363e-05,
      "loss": 3.4129,
      "num_input_tokens_seen": 20777712,
      "step": 9935
    },
    {
      "epoch": 1.6215841422628272,
      "grad_norm": 5.59375,
      "learning_rate": 4.276215878430661e-05,
      "loss": 2.6432,
      "num_input_tokens_seen": 20788000,
      "step": 9940
    },
    {
      "epoch": 1.6223998694836448,
      "grad_norm": 4.96875,
      "learning_rate": 4.275524874851149e-05,
      "loss": 1.8947,
      "num_input_tokens_seen": 20798720,
      "step": 9945
    },
    {
      "epoch": 1.623215596704462,
      "grad_norm": 6.6875,
      "learning_rate": 4.274833597463388e-05,
      "loss": 1.7858,
      "num_input_tokens_seen": 20808704,
      "step": 9950
    },
    {
      "epoch": 1.6240313239252795,
      "grad_norm": 10.875,
      "learning_rate": 4.2741420463739824e-05,
      "loss": 2.6123,
      "num_input_tokens_seen": 20818848,
      "step": 9955
    },
    {
      "epoch": 1.6248470511460966,
      "grad_norm": 9.0625,
      "learning_rate": 4.273450221689578e-05,
      "loss": 3.578,
      "num_input_tokens_seen": 20828304,
      "step": 9960
    },
    {
      "epoch": 1.6256627783669142,
      "grad_norm": 8.4375,
      "learning_rate": 4.272758123516863e-05,
      "loss": 2.9122,
      "num_input_tokens_seen": 20837840,
      "step": 9965
    },
    {
      "epoch": 1.6264785055877313,
      "grad_norm": 6.78125,
      "learning_rate": 4.272065751962567e-05,
      "loss": 1.7519,
      "num_input_tokens_seen": 20848608,
      "step": 9970
    },
    {
      "epoch": 1.627294232808549,
      "grad_norm": 11.5,
      "learning_rate": 4.271373107133464e-05,
      "loss": 2.5725,
      "num_input_tokens_seen": 20859616,
      "step": 9975
    },
    {
      "epoch": 1.628109960029366,
      "grad_norm": 20.25,
      "learning_rate": 4.270680189136366e-05,
      "loss": 3.8603,
      "num_input_tokens_seen": 20869984,
      "step": 9980
    },
    {
      "epoch": 1.6289256872501836,
      "grad_norm": 12.125,
      "learning_rate": 4.269986998078132e-05,
      "loss": 3.1739,
      "num_input_tokens_seen": 20880448,
      "step": 9985
    },
    {
      "epoch": 1.6297414144710007,
      "grad_norm": 12.0625,
      "learning_rate": 4.2692935340656595e-05,
      "loss": 2.2646,
      "num_input_tokens_seen": 20890480,
      "step": 9990
    },
    {
      "epoch": 1.6305571416918183,
      "grad_norm": 5.4375,
      "learning_rate": 4.26859979720589e-05,
      "loss": 2.554,
      "num_input_tokens_seen": 20902416,
      "step": 9995
    },
    {
      "epoch": 1.6313728689126357,
      "grad_norm": 5.75,
      "learning_rate": 4.267905787605806e-05,
      "loss": 2.3272,
      "num_input_tokens_seen": 20912896,
      "step": 10000
    },
    {
      "epoch": 1.6313728689126357,
      "eval_loss": 2.542886972427368,
      "eval_runtime": 135.0431,
      "eval_samples_per_second": 20.179,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 20912896,
      "step": 10000
    },
    {
      "epoch": 1.632188596133453,
      "grad_norm": 8.9375,
      "learning_rate": 4.267211505372433e-05,
      "loss": 1.6226,
      "num_input_tokens_seen": 20923056,
      "step": 10005
    },
    {
      "epoch": 1.6330043233542704,
      "grad_norm": 7.96875,
      "learning_rate": 4.266516950612837e-05,
      "loss": 2.0991,
      "num_input_tokens_seen": 20934336,
      "step": 10010
    },
    {
      "epoch": 1.6338200505750877,
      "grad_norm": 4.25,
      "learning_rate": 4.265822123434128e-05,
      "loss": 2.9489,
      "num_input_tokens_seen": 20945440,
      "step": 10015
    },
    {
      "epoch": 1.634635777795905,
      "grad_norm": 3.640625,
      "learning_rate": 4.265127023943457e-05,
      "loss": 1.4088,
      "num_input_tokens_seen": 20956544,
      "step": 10020
    },
    {
      "epoch": 1.6354515050167224,
      "grad_norm": 5.6875,
      "learning_rate": 4.2644316522480176e-05,
      "loss": 3.0212,
      "num_input_tokens_seen": 20966560,
      "step": 10025
    },
    {
      "epoch": 1.6362672322375398,
      "grad_norm": 8.4375,
      "learning_rate": 4.263736008455044e-05,
      "loss": 1.8325,
      "num_input_tokens_seen": 20976784,
      "step": 10030
    },
    {
      "epoch": 1.6370829594583571,
      "grad_norm": 4.59375,
      "learning_rate": 4.2630400926718125e-05,
      "loss": 2.3791,
      "num_input_tokens_seen": 20988256,
      "step": 10035
    },
    {
      "epoch": 1.6378986866791745,
      "grad_norm": 6.96875,
      "learning_rate": 4.262343905005644e-05,
      "loss": 1.1471,
      "num_input_tokens_seen": 20997648,
      "step": 10040
    },
    {
      "epoch": 1.6387144138999918,
      "grad_norm": 1.5625,
      "learning_rate": 4.261647445563897e-05,
      "loss": 2.0964,
      "num_input_tokens_seen": 21008480,
      "step": 10045
    },
    {
      "epoch": 1.6395301411208092,
      "grad_norm": 6.96875,
      "learning_rate": 4.260950714453976e-05,
      "loss": 3.2704,
      "num_input_tokens_seen": 21018384,
      "step": 10050
    },
    {
      "epoch": 1.6403458683416265,
      "grad_norm": 7.375,
      "learning_rate": 4.2602537117833266e-05,
      "loss": 1.6717,
      "num_input_tokens_seen": 21029424,
      "step": 10055
    },
    {
      "epoch": 1.641161595562444,
      "grad_norm": 13.4375,
      "learning_rate": 4.259556437659433e-05,
      "loss": 1.8708,
      "num_input_tokens_seen": 21039936,
      "step": 10060
    },
    {
      "epoch": 1.6419773227832613,
      "grad_norm": 4.9375,
      "learning_rate": 4.258858892189825e-05,
      "loss": 1.7369,
      "num_input_tokens_seen": 21049904,
      "step": 10065
    },
    {
      "epoch": 1.6427930500040786,
      "grad_norm": 6.5,
      "learning_rate": 4.2581610754820725e-05,
      "loss": 1.122,
      "num_input_tokens_seen": 21059312,
      "step": 10070
    },
    {
      "epoch": 1.643608777224896,
      "grad_norm": 6.0,
      "learning_rate": 4.2574629876437876e-05,
      "loss": 0.9097,
      "num_input_tokens_seen": 21068816,
      "step": 10075
    },
    {
      "epoch": 1.6444245044457133,
      "grad_norm": 14.5625,
      "learning_rate": 4.256764628782625e-05,
      "loss": 2.4086,
      "num_input_tokens_seen": 21080128,
      "step": 10080
    },
    {
      "epoch": 1.6452402316665307,
      "grad_norm": 12.125,
      "learning_rate": 4.256065999006279e-05,
      "loss": 2.7302,
      "num_input_tokens_seen": 21090800,
      "step": 10085
    },
    {
      "epoch": 1.6460559588873482,
      "grad_norm": 5.0625,
      "learning_rate": 4.2553670984224885e-05,
      "loss": 2.3471,
      "num_input_tokens_seen": 21101744,
      "step": 10090
    },
    {
      "epoch": 1.6468716861081654,
      "grad_norm": 6.75,
      "learning_rate": 4.254667927139032e-05,
      "loss": 3.115,
      "num_input_tokens_seen": 21112320,
      "step": 10095
    },
    {
      "epoch": 1.647687413328983,
      "grad_norm": 8.0,
      "learning_rate": 4.2539684852637295e-05,
      "loss": 2.7829,
      "num_input_tokens_seen": 21123376,
      "step": 10100
    },
    {
      "epoch": 1.6485031405498,
      "grad_norm": 5.59375,
      "learning_rate": 4.253268772904446e-05,
      "loss": 2.4823,
      "num_input_tokens_seen": 21135104,
      "step": 10105
    },
    {
      "epoch": 1.6493188677706176,
      "grad_norm": 4.03125,
      "learning_rate": 4.252568790169085e-05,
      "loss": 3.1325,
      "num_input_tokens_seen": 21145744,
      "step": 10110
    },
    {
      "epoch": 1.6501345949914348,
      "grad_norm": 5.5625,
      "learning_rate": 4.251868537165592e-05,
      "loss": 1.9709,
      "num_input_tokens_seen": 21157488,
      "step": 10115
    },
    {
      "epoch": 1.6509503222122524,
      "grad_norm": 5.15625,
      "learning_rate": 4.251168014001955e-05,
      "loss": 1.5757,
      "num_input_tokens_seen": 21167120,
      "step": 10120
    },
    {
      "epoch": 1.6517660494330695,
      "grad_norm": 14.0625,
      "learning_rate": 4.250467220786204e-05,
      "loss": 2.7046,
      "num_input_tokens_seen": 21178704,
      "step": 10125
    },
    {
      "epoch": 1.652581776653887,
      "grad_norm": 7.46875,
      "learning_rate": 4.249766157626409e-05,
      "loss": 2.9901,
      "num_input_tokens_seen": 21188640,
      "step": 10130
    },
    {
      "epoch": 1.6533975038747042,
      "grad_norm": 11.5,
      "learning_rate": 4.249064824630684e-05,
      "loss": 3.1327,
      "num_input_tokens_seen": 21199312,
      "step": 10135
    },
    {
      "epoch": 1.6542132310955218,
      "grad_norm": 7.21875,
      "learning_rate": 4.248363221907183e-05,
      "loss": 2.0713,
      "num_input_tokens_seen": 21209824,
      "step": 10140
    },
    {
      "epoch": 1.655028958316339,
      "grad_norm": 6.59375,
      "learning_rate": 4.2476613495641026e-05,
      "loss": 2.7075,
      "num_input_tokens_seen": 21220832,
      "step": 10145
    },
    {
      "epoch": 1.6558446855371565,
      "grad_norm": 8.125,
      "learning_rate": 4.246959207709679e-05,
      "loss": 3.6813,
      "num_input_tokens_seen": 21230928,
      "step": 10150
    },
    {
      "epoch": 1.6566604127579736,
      "grad_norm": 7.0625,
      "learning_rate": 4.246256796452192e-05,
      "loss": 1.5339,
      "num_input_tokens_seen": 21240928,
      "step": 10155
    },
    {
      "epoch": 1.6574761399787912,
      "grad_norm": 5.15625,
      "learning_rate": 4.245554115899962e-05,
      "loss": 2.1236,
      "num_input_tokens_seen": 21251280,
      "step": 10160
    },
    {
      "epoch": 1.6582918671996083,
      "grad_norm": 9.8125,
      "learning_rate": 4.2448511661613514e-05,
      "loss": 2.2097,
      "num_input_tokens_seen": 21261072,
      "step": 10165
    },
    {
      "epoch": 1.6591075944204259,
      "grad_norm": 4.65625,
      "learning_rate": 4.2441479473447635e-05,
      "loss": 1.7586,
      "num_input_tokens_seen": 21270416,
      "step": 10170
    },
    {
      "epoch": 1.659923321641243,
      "grad_norm": 4.4375,
      "learning_rate": 4.243444459558644e-05,
      "loss": 2.6622,
      "num_input_tokens_seen": 21280976,
      "step": 10175
    },
    {
      "epoch": 1.6607390488620606,
      "grad_norm": 13.375,
      "learning_rate": 4.24274070291148e-05,
      "loss": 2.9632,
      "num_input_tokens_seen": 21290304,
      "step": 10180
    },
    {
      "epoch": 1.661554776082878,
      "grad_norm": 2.515625,
      "learning_rate": 4.242036677511798e-05,
      "loss": 1.6508,
      "num_input_tokens_seen": 21299504,
      "step": 10185
    },
    {
      "epoch": 1.6623705033036953,
      "grad_norm": 11.3125,
      "learning_rate": 4.241332383468169e-05,
      "loss": 2.7934,
      "num_input_tokens_seen": 21308288,
      "step": 10190
    },
    {
      "epoch": 1.6631862305245126,
      "grad_norm": 10.9375,
      "learning_rate": 4.2406278208892034e-05,
      "loss": 3.0532,
      "num_input_tokens_seen": 21318704,
      "step": 10195
    },
    {
      "epoch": 1.66400195774533,
      "grad_norm": 4.90625,
      "learning_rate": 4.2399229898835536e-05,
      "loss": 2.7251,
      "num_input_tokens_seen": 21328976,
      "step": 10200
    },
    {
      "epoch": 1.66400195774533,
      "eval_loss": 2.5554885864257812,
      "eval_runtime": 135.1264,
      "eval_samples_per_second": 20.166,
      "eval_steps_per_second": 10.087,
      "num_input_tokens_seen": 21328976,
      "step": 10200
    },
    {
      "epoch": 1.6648176849661473,
      "grad_norm": 9.25,
      "learning_rate": 4.239217890559914e-05,
      "loss": 2.7855,
      "num_input_tokens_seen": 21339840,
      "step": 10205
    },
    {
      "epoch": 1.6656334121869647,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.238512523027019e-05,
      "loss": 0.9088,
      "num_input_tokens_seen": 21350368,
      "step": 10210
    },
    {
      "epoch": 1.666449139407782,
      "grad_norm": 7.1875,
      "learning_rate": 4.237806887393645e-05,
      "loss": 2.8211,
      "num_input_tokens_seen": 21360736,
      "step": 10215
    },
    {
      "epoch": 1.6672648666285994,
      "grad_norm": 9.25,
      "learning_rate": 4.237100983768611e-05,
      "loss": 1.4705,
      "num_input_tokens_seen": 21372096,
      "step": 10220
    },
    {
      "epoch": 1.6680805938494168,
      "grad_norm": 2.09375,
      "learning_rate": 4.2363948122607756e-05,
      "loss": 1.5441,
      "num_input_tokens_seen": 21382240,
      "step": 10225
    },
    {
      "epoch": 1.6688963210702341,
      "grad_norm": 10.375,
      "learning_rate": 4.235688372979039e-05,
      "loss": 2.7682,
      "num_input_tokens_seen": 21392864,
      "step": 10230
    },
    {
      "epoch": 1.6697120482910515,
      "grad_norm": 7.40625,
      "learning_rate": 4.234981666032343e-05,
      "loss": 3.0931,
      "num_input_tokens_seen": 21403008,
      "step": 10235
    },
    {
      "epoch": 1.6705277755118688,
      "grad_norm": 9.625,
      "learning_rate": 4.2342746915296704e-05,
      "loss": 2.3487,
      "num_input_tokens_seen": 21414048,
      "step": 10240
    },
    {
      "epoch": 1.6713435027326862,
      "grad_norm": 7.25,
      "learning_rate": 4.233567449580047e-05,
      "loss": 2.3831,
      "num_input_tokens_seen": 21425792,
      "step": 10245
    },
    {
      "epoch": 1.6721592299535035,
      "grad_norm": 8.0625,
      "learning_rate": 4.232859940292537e-05,
      "loss": 3.5676,
      "num_input_tokens_seen": 21434736,
      "step": 10250
    },
    {
      "epoch": 1.6729749571743209,
      "grad_norm": 1.4375,
      "learning_rate": 4.232152163776248e-05,
      "loss": 2.0049,
      "num_input_tokens_seen": 21445984,
      "step": 10255
    },
    {
      "epoch": 1.6737906843951382,
      "grad_norm": 0.53125,
      "learning_rate": 4.231444120140328e-05,
      "loss": 1.6876,
      "num_input_tokens_seen": 21457440,
      "step": 10260
    },
    {
      "epoch": 1.6746064116159556,
      "grad_norm": 7.03125,
      "learning_rate": 4.230735809493967e-05,
      "loss": 2.5528,
      "num_input_tokens_seen": 21468112,
      "step": 10265
    },
    {
      "epoch": 1.675422138836773,
      "grad_norm": 9.5,
      "learning_rate": 4.2300272319463926e-05,
      "loss": 3.9087,
      "num_input_tokens_seen": 21478912,
      "step": 10270
    },
    {
      "epoch": 1.6762378660575903,
      "grad_norm": 2.765625,
      "learning_rate": 4.2293183876068786e-05,
      "loss": 1.9837,
      "num_input_tokens_seen": 21489264,
      "step": 10275
    },
    {
      "epoch": 1.6770535932784076,
      "grad_norm": 9.25,
      "learning_rate": 4.228609276584737e-05,
      "loss": 2.3761,
      "num_input_tokens_seen": 21500944,
      "step": 10280
    },
    {
      "epoch": 1.6778693204992252,
      "grad_norm": 7.5,
      "learning_rate": 4.227899898989323e-05,
      "loss": 2.1788,
      "num_input_tokens_seen": 21510848,
      "step": 10285
    },
    {
      "epoch": 1.6786850477200423,
      "grad_norm": 5.53125,
      "learning_rate": 4.2271902549300293e-05,
      "loss": 2.2827,
      "num_input_tokens_seen": 21520384,
      "step": 10290
    },
    {
      "epoch": 1.67950077494086,
      "grad_norm": 5.03125,
      "learning_rate": 4.226480344516294e-05,
      "loss": 1.9437,
      "num_input_tokens_seen": 21531600,
      "step": 10295
    },
    {
      "epoch": 1.680316502161677,
      "grad_norm": 4.5,
      "learning_rate": 4.2257701678575925e-05,
      "loss": 2.1368,
      "num_input_tokens_seen": 21541760,
      "step": 10300
    },
    {
      "epoch": 1.6811322293824946,
      "grad_norm": 7.0625,
      "learning_rate": 4.225059725063444e-05,
      "loss": 1.9154,
      "num_input_tokens_seen": 21552000,
      "step": 10305
    },
    {
      "epoch": 1.6819479566033118,
      "grad_norm": 5.4375,
      "learning_rate": 4.2243490162434074e-05,
      "loss": 2.0984,
      "num_input_tokens_seen": 21562416,
      "step": 10310
    },
    {
      "epoch": 1.6827636838241293,
      "grad_norm": 8.25,
      "learning_rate": 4.223638041507083e-05,
      "loss": 2.6947,
      "num_input_tokens_seen": 21572656,
      "step": 10315
    },
    {
      "epoch": 1.6835794110449465,
      "grad_norm": 3.078125,
      "learning_rate": 4.2229268009641124e-05,
      "loss": 2.6856,
      "num_input_tokens_seen": 21583456,
      "step": 10320
    },
    {
      "epoch": 1.684395138265764,
      "grad_norm": 3.203125,
      "learning_rate": 4.222215294724177e-05,
      "loss": 2.5418,
      "num_input_tokens_seen": 21593696,
      "step": 10325
    },
    {
      "epoch": 1.6852108654865812,
      "grad_norm": 10.5625,
      "learning_rate": 4.2215035228970005e-05,
      "loss": 1.5621,
      "num_input_tokens_seen": 21602832,
      "step": 10330
    },
    {
      "epoch": 1.6860265927073987,
      "grad_norm": 10.125,
      "learning_rate": 4.2207914855923464e-05,
      "loss": 1.9968,
      "num_input_tokens_seen": 21613760,
      "step": 10335
    },
    {
      "epoch": 1.6868423199282159,
      "grad_norm": 3.1875,
      "learning_rate": 4.220079182920021e-05,
      "loss": 1.9081,
      "num_input_tokens_seen": 21622576,
      "step": 10340
    },
    {
      "epoch": 1.6876580471490334,
      "grad_norm": 7.90625,
      "learning_rate": 4.2193666149898705e-05,
      "loss": 2.6379,
      "num_input_tokens_seen": 21634032,
      "step": 10345
    },
    {
      "epoch": 1.6884737743698506,
      "grad_norm": 4.96875,
      "learning_rate": 4.21865378191178e-05,
      "loss": 2.3657,
      "num_input_tokens_seen": 21644592,
      "step": 10350
    },
    {
      "epoch": 1.6892895015906682,
      "grad_norm": 8.25,
      "learning_rate": 4.217940683795678e-05,
      "loss": 2.3895,
      "num_input_tokens_seen": 21655552,
      "step": 10355
    },
    {
      "epoch": 1.6901052288114853,
      "grad_norm": 10.375,
      "learning_rate": 4.217227320751534e-05,
      "loss": 2.3325,
      "num_input_tokens_seen": 21666960,
      "step": 10360
    },
    {
      "epoch": 1.6909209560323029,
      "grad_norm": 3.828125,
      "learning_rate": 4.216513692889358e-05,
      "loss": 2.191,
      "num_input_tokens_seen": 21676608,
      "step": 10365
    },
    {
      "epoch": 1.6917366832531202,
      "grad_norm": 7.8125,
      "learning_rate": 4.215799800319199e-05,
      "loss": 2.014,
      "num_input_tokens_seen": 21686576,
      "step": 10370
    },
    {
      "epoch": 1.6925524104739376,
      "grad_norm": 2.40625,
      "learning_rate": 4.2150856431511485e-05,
      "loss": 2.4317,
      "num_input_tokens_seen": 21697104,
      "step": 10375
    },
    {
      "epoch": 1.693368137694755,
      "grad_norm": 8.75,
      "learning_rate": 4.214371221495339e-05,
      "loss": 3.8414,
      "num_input_tokens_seen": 21707376,
      "step": 10380
    },
    {
      "epoch": 1.6941838649155723,
      "grad_norm": 4.21875,
      "learning_rate": 4.213656535461942e-05,
      "loss": 2.6699,
      "num_input_tokens_seen": 21718848,
      "step": 10385
    },
    {
      "epoch": 1.6949995921363896,
      "grad_norm": 6.375,
      "learning_rate": 4.2129415851611734e-05,
      "loss": 2.3662,
      "num_input_tokens_seen": 21729968,
      "step": 10390
    },
    {
      "epoch": 1.695815319357207,
      "grad_norm": 6.0,
      "learning_rate": 4.2122263707032855e-05,
      "loss": 1.4607,
      "num_input_tokens_seen": 21740832,
      "step": 10395
    },
    {
      "epoch": 1.6966310465780243,
      "grad_norm": 8.25,
      "learning_rate": 4.211510892198574e-05,
      "loss": 3.511,
      "num_input_tokens_seen": 21752192,
      "step": 10400
    },
    {
      "epoch": 1.6966310465780243,
      "eval_loss": 2.530885696411133,
      "eval_runtime": 134.8787,
      "eval_samples_per_second": 20.203,
      "eval_steps_per_second": 10.105,
      "num_input_tokens_seen": 21752192,
      "step": 10400
    },
    {
      "epoch": 1.6974467737988417,
      "grad_norm": 1.1640625,
      "learning_rate": 4.210795149757375e-05,
      "loss": 1.8275,
      "num_input_tokens_seen": 21761344,
      "step": 10405
    },
    {
      "epoch": 1.698262501019659,
      "grad_norm": 7.3125,
      "learning_rate": 4.210079143490065e-05,
      "loss": 2.6952,
      "num_input_tokens_seen": 21770720,
      "step": 10410
    },
    {
      "epoch": 1.6990782282404764,
      "grad_norm": 6.875,
      "learning_rate": 4.2093628735070604e-05,
      "loss": 1.688,
      "num_input_tokens_seen": 21780208,
      "step": 10415
    },
    {
      "epoch": 1.6998939554612937,
      "grad_norm": 8.8125,
      "learning_rate": 4.208646339918819e-05,
      "loss": 2.0777,
      "num_input_tokens_seen": 21790624,
      "step": 10420
    },
    {
      "epoch": 1.700709682682111,
      "grad_norm": 7.8125,
      "learning_rate": 4.2079295428358414e-05,
      "loss": 2.3737,
      "num_input_tokens_seen": 21801008,
      "step": 10425
    },
    {
      "epoch": 1.7015254099029284,
      "grad_norm": 8.1875,
      "learning_rate": 4.207212482368664e-05,
      "loss": 2.0515,
      "num_input_tokens_seen": 21812656,
      "step": 10430
    },
    {
      "epoch": 1.7023411371237458,
      "grad_norm": 3.90625,
      "learning_rate": 4.206495158627867e-05,
      "loss": 3.2055,
      "num_input_tokens_seen": 21824000,
      "step": 10435
    },
    {
      "epoch": 1.7031568643445631,
      "grad_norm": 9.9375,
      "learning_rate": 4.205777571724073e-05,
      "loss": 2.3469,
      "num_input_tokens_seen": 21834976,
      "step": 10440
    },
    {
      "epoch": 1.7039725915653805,
      "grad_norm": 15.5625,
      "learning_rate": 4.20505972176794e-05,
      "loss": 1.9091,
      "num_input_tokens_seen": 21845344,
      "step": 10445
    },
    {
      "epoch": 1.7047883187861979,
      "grad_norm": 4.3125,
      "learning_rate": 4.204341608870171e-05,
      "loss": 2.3963,
      "num_input_tokens_seen": 21856944,
      "step": 10450
    },
    {
      "epoch": 1.7056040460070152,
      "grad_norm": 5.03125,
      "learning_rate": 4.203623233141508e-05,
      "loss": 1.7123,
      "num_input_tokens_seen": 21867856,
      "step": 10455
    },
    {
      "epoch": 1.7064197732278326,
      "grad_norm": 9.875,
      "learning_rate": 4.2029045946927334e-05,
      "loss": 2.5352,
      "num_input_tokens_seen": 21877504,
      "step": 10460
    },
    {
      "epoch": 1.70723550044865,
      "grad_norm": 0.061767578125,
      "learning_rate": 4.20218569363467e-05,
      "loss": 1.4753,
      "num_input_tokens_seen": 21886352,
      "step": 10465
    },
    {
      "epoch": 1.7080512276694675,
      "grad_norm": 4.15625,
      "learning_rate": 4.2014665300781834e-05,
      "loss": 1.2279,
      "num_input_tokens_seen": 21896976,
      "step": 10470
    },
    {
      "epoch": 1.7088669548902846,
      "grad_norm": 2.125,
      "learning_rate": 4.200747104134174e-05,
      "loss": 1.6798,
      "num_input_tokens_seen": 21906160,
      "step": 10475
    },
    {
      "epoch": 1.7096826821111022,
      "grad_norm": 6.59375,
      "learning_rate": 4.200027415913588e-05,
      "loss": 1.8652,
      "num_input_tokens_seen": 21915792,
      "step": 10480
    },
    {
      "epoch": 1.7104984093319193,
      "grad_norm": 6.96875,
      "learning_rate": 4.1993074655274126e-05,
      "loss": 4.3753,
      "num_input_tokens_seen": 21926080,
      "step": 10485
    },
    {
      "epoch": 1.711314136552737,
      "grad_norm": 11.5625,
      "learning_rate": 4.198587253086669e-05,
      "loss": 2.8847,
      "num_input_tokens_seen": 21936656,
      "step": 10490
    },
    {
      "epoch": 1.712129863773554,
      "grad_norm": 14.3125,
      "learning_rate": 4.197866778702426e-05,
      "loss": 2.3488,
      "num_input_tokens_seen": 21947488,
      "step": 10495
    },
    {
      "epoch": 1.7129455909943716,
      "grad_norm": 10.125,
      "learning_rate": 4.197146042485789e-05,
      "loss": 2.4031,
      "num_input_tokens_seen": 21957760,
      "step": 10500
    },
    {
      "epoch": 1.7137613182151887,
      "grad_norm": 5.1875,
      "learning_rate": 4.1964250445479046e-05,
      "loss": 1.9205,
      "num_input_tokens_seen": 21967472,
      "step": 10505
    },
    {
      "epoch": 1.7145770454360063,
      "grad_norm": 10.25,
      "learning_rate": 4.19570378499996e-05,
      "loss": 3.2976,
      "num_input_tokens_seen": 21977568,
      "step": 10510
    },
    {
      "epoch": 1.7153927726568234,
      "grad_norm": 15.6875,
      "learning_rate": 4.194982263953182e-05,
      "loss": 4.1467,
      "num_input_tokens_seen": 21988224,
      "step": 10515
    },
    {
      "epoch": 1.716208499877641,
      "grad_norm": 2.9375,
      "learning_rate": 4.194260481518838e-05,
      "loss": 2.7116,
      "num_input_tokens_seen": 21998032,
      "step": 10520
    },
    {
      "epoch": 1.7170242270984581,
      "grad_norm": 14.25,
      "learning_rate": 4.1935384378082366e-05,
      "loss": 3.4925,
      "num_input_tokens_seen": 22008800,
      "step": 10525
    },
    {
      "epoch": 1.7178399543192757,
      "grad_norm": 10.625,
      "learning_rate": 4.1928161329327267e-05,
      "loss": 4.0193,
      "num_input_tokens_seen": 22020000,
      "step": 10530
    },
    {
      "epoch": 1.7186556815400928,
      "grad_norm": 6.28125,
      "learning_rate": 4.1920935670036945e-05,
      "loss": 2.767,
      "num_input_tokens_seen": 22031712,
      "step": 10535
    },
    {
      "epoch": 1.7194714087609104,
      "grad_norm": 7.90625,
      "learning_rate": 4.1913707401325705e-05,
      "loss": 1.5344,
      "num_input_tokens_seen": 22042352,
      "step": 10540
    },
    {
      "epoch": 1.7202871359817276,
      "grad_norm": 8.6875,
      "learning_rate": 4.1906476524308235e-05,
      "loss": 2.9611,
      "num_input_tokens_seen": 22054064,
      "step": 10545
    },
    {
      "epoch": 1.7211028632025451,
      "grad_norm": 2.78125,
      "learning_rate": 4.189924304009962e-05,
      "loss": 1.8644,
      "num_input_tokens_seen": 22064368,
      "step": 10550
    },
    {
      "epoch": 1.7219185904233625,
      "grad_norm": 9.1875,
      "learning_rate": 4.189200694981537e-05,
      "loss": 2.8982,
      "num_input_tokens_seen": 22074064,
      "step": 10555
    },
    {
      "epoch": 1.7227343176441798,
      "grad_norm": 14.375,
      "learning_rate": 4.188476825457136e-05,
      "loss": 2.3784,
      "num_input_tokens_seen": 22084864,
      "step": 10560
    },
    {
      "epoch": 1.7235500448649972,
      "grad_norm": 2.65625,
      "learning_rate": 4.18775269554839e-05,
      "loss": 1.3378,
      "num_input_tokens_seen": 22094544,
      "step": 10565
    },
    {
      "epoch": 1.7243657720858145,
      "grad_norm": 6.84375,
      "learning_rate": 4.187028305366969e-05,
      "loss": 2.3798,
      "num_input_tokens_seen": 22102656,
      "step": 10570
    },
    {
      "epoch": 1.725181499306632,
      "grad_norm": 7.59375,
      "learning_rate": 4.1863036550245824e-05,
      "loss": 1.953,
      "num_input_tokens_seen": 22112976,
      "step": 10575
    },
    {
      "epoch": 1.7259972265274492,
      "grad_norm": 11.5625,
      "learning_rate": 4.1855787446329806e-05,
      "loss": 3.6301,
      "num_input_tokens_seen": 22123936,
      "step": 10580
    },
    {
      "epoch": 1.7268129537482666,
      "grad_norm": 7.9375,
      "learning_rate": 4.184853574303955e-05,
      "loss": 2.9563,
      "num_input_tokens_seen": 22134736,
      "step": 10585
    },
    {
      "epoch": 1.727628680969084,
      "grad_norm": 1.6875,
      "learning_rate": 4.184128144149334e-05,
      "loss": 2.2153,
      "num_input_tokens_seen": 22144736,
      "step": 10590
    },
    {
      "epoch": 1.7284444081899013,
      "grad_norm": 12.5,
      "learning_rate": 4.1834024542809896e-05,
      "loss": 3.638,
      "num_input_tokens_seen": 22154656,
      "step": 10595
    },
    {
      "epoch": 1.7292601354107187,
      "grad_norm": 6.03125,
      "learning_rate": 4.1826765048108315e-05,
      "loss": 2.7897,
      "num_input_tokens_seen": 22164912,
      "step": 10600
    },
    {
      "epoch": 1.7292601354107187,
      "eval_loss": 2.539090394973755,
      "eval_runtime": 134.9487,
      "eval_samples_per_second": 20.193,
      "eval_steps_per_second": 10.1,
      "num_input_tokens_seen": 22164912,
      "step": 10600
    },
    {
      "epoch": 1.730075862631536,
      "grad_norm": 7.4375,
      "learning_rate": 4.181950295850811e-05,
      "loss": 2.4834,
      "num_input_tokens_seen": 22174560,
      "step": 10605
    },
    {
      "epoch": 1.7308915898523534,
      "grad_norm": 3.21875,
      "learning_rate": 4.181223827512918e-05,
      "loss": 3.7932,
      "num_input_tokens_seen": 22185552,
      "step": 10610
    },
    {
      "epoch": 1.7317073170731707,
      "grad_norm": 14.4375,
      "learning_rate": 4.180497099909183e-05,
      "loss": 1.6313,
      "num_input_tokens_seen": 22194352,
      "step": 10615
    },
    {
      "epoch": 1.732523044293988,
      "grad_norm": 7.8125,
      "learning_rate": 4.179770113151677e-05,
      "loss": 1.7586,
      "num_input_tokens_seen": 22203600,
      "step": 10620
    },
    {
      "epoch": 1.7333387715148054,
      "grad_norm": 0.265625,
      "learning_rate": 4.179042867352511e-05,
      "loss": 2.8869,
      "num_input_tokens_seen": 22214720,
      "step": 10625
    },
    {
      "epoch": 1.7341544987356228,
      "grad_norm": 6.71875,
      "learning_rate": 4.1783153626238334e-05,
      "loss": 2.9075,
      "num_input_tokens_seen": 22227568,
      "step": 10630
    },
    {
      "epoch": 1.7349702259564401,
      "grad_norm": 5.6875,
      "learning_rate": 4.177587599077836e-05,
      "loss": 3.3877,
      "num_input_tokens_seen": 22239392,
      "step": 10635
    },
    {
      "epoch": 1.7357859531772575,
      "grad_norm": 2.59375,
      "learning_rate": 4.1768595768267494e-05,
      "loss": 1.8827,
      "num_input_tokens_seen": 22249808,
      "step": 10640
    },
    {
      "epoch": 1.7366016803980748,
      "grad_norm": 7.375,
      "learning_rate": 4.176131295982843e-05,
      "loss": 2.1767,
      "num_input_tokens_seen": 22260512,
      "step": 10645
    },
    {
      "epoch": 1.7374174076188922,
      "grad_norm": 6.96875,
      "learning_rate": 4.1754027566584276e-05,
      "loss": 2.4503,
      "num_input_tokens_seen": 22270144,
      "step": 10650
    },
    {
      "epoch": 1.7382331348397098,
      "grad_norm": 10.625,
      "learning_rate": 4.174673958965852e-05,
      "loss": 3.9162,
      "num_input_tokens_seen": 22280688,
      "step": 10655
    },
    {
      "epoch": 1.739048862060527,
      "grad_norm": 6.03125,
      "learning_rate": 4.173944903017507e-05,
      "loss": 3.0249,
      "num_input_tokens_seen": 22291568,
      "step": 10660
    },
    {
      "epoch": 1.7398645892813445,
      "grad_norm": 10.5,
      "learning_rate": 4.173215588925822e-05,
      "loss": 2.3746,
      "num_input_tokens_seen": 22300912,
      "step": 10665
    },
    {
      "epoch": 1.7406803165021616,
      "grad_norm": 11.5625,
      "learning_rate": 4.172486016803266e-05,
      "loss": 2.6148,
      "num_input_tokens_seen": 22310688,
      "step": 10670
    },
    {
      "epoch": 1.7414960437229792,
      "grad_norm": 5.4375,
      "learning_rate": 4.171756186762349e-05,
      "loss": 2.9161,
      "num_input_tokens_seen": 22320720,
      "step": 10675
    },
    {
      "epoch": 1.7423117709437963,
      "grad_norm": 9.5,
      "learning_rate": 4.171026098915619e-05,
      "loss": 1.9717,
      "num_input_tokens_seen": 22331072,
      "step": 10680
    },
    {
      "epoch": 1.7431274981646139,
      "grad_norm": 10.625,
      "learning_rate": 4.170295753375665e-05,
      "loss": 4.0071,
      "num_input_tokens_seen": 22341072,
      "step": 10685
    },
    {
      "epoch": 1.743943225385431,
      "grad_norm": 11.75,
      "learning_rate": 4.169565150255117e-05,
      "loss": 2.3715,
      "num_input_tokens_seen": 22350624,
      "step": 10690
    },
    {
      "epoch": 1.7447589526062486,
      "grad_norm": 5.46875,
      "learning_rate": 4.16883428966664e-05,
      "loss": 1.7984,
      "num_input_tokens_seen": 22361200,
      "step": 10695
    },
    {
      "epoch": 1.7455746798270657,
      "grad_norm": 5.46875,
      "learning_rate": 4.168103171722944e-05,
      "loss": 1.6843,
      "num_input_tokens_seen": 22372464,
      "step": 10700
    },
    {
      "epoch": 1.7463904070478833,
      "grad_norm": 9.4375,
      "learning_rate": 4.167371796536777e-05,
      "loss": 2.9907,
      "num_input_tokens_seen": 22383936,
      "step": 10705
    },
    {
      "epoch": 1.7472061342687004,
      "grad_norm": 4.4375,
      "learning_rate": 4.166640164220924e-05,
      "loss": 2.9674,
      "num_input_tokens_seen": 22394144,
      "step": 10710
    },
    {
      "epoch": 1.748021861489518,
      "grad_norm": 9.625,
      "learning_rate": 4.1659082748882144e-05,
      "loss": 1.4832,
      "num_input_tokens_seen": 22404000,
      "step": 10715
    },
    {
      "epoch": 1.7488375887103351,
      "grad_norm": 4.28125,
      "learning_rate": 4.1651761286515135e-05,
      "loss": 2.2355,
      "num_input_tokens_seen": 22415232,
      "step": 10720
    },
    {
      "epoch": 1.7496533159311527,
      "grad_norm": 4.25,
      "learning_rate": 4.164443725623728e-05,
      "loss": 1.6445,
      "num_input_tokens_seen": 22424672,
      "step": 10725
    },
    {
      "epoch": 1.7504690431519698,
      "grad_norm": 5.0,
      "learning_rate": 4.163711065917802e-05,
      "loss": 1.9583,
      "num_input_tokens_seen": 22434256,
      "step": 10730
    },
    {
      "epoch": 1.7512847703727874,
      "grad_norm": 2.3125,
      "learning_rate": 4.1629781496467234e-05,
      "loss": 2.0795,
      "num_input_tokens_seen": 22444288,
      "step": 10735
    },
    {
      "epoch": 1.7521004975936045,
      "grad_norm": 6.46875,
      "learning_rate": 4.1622449769235164e-05,
      "loss": 3.4827,
      "num_input_tokens_seen": 22454880,
      "step": 10740
    },
    {
      "epoch": 1.752916224814422,
      "grad_norm": 7.09375,
      "learning_rate": 4.161511547861243e-05,
      "loss": 3.4251,
      "num_input_tokens_seen": 22466000,
      "step": 10745
    },
    {
      "epoch": 1.7537319520352395,
      "grad_norm": 4.40625,
      "learning_rate": 4.1607778625730104e-05,
      "loss": 2.075,
      "num_input_tokens_seen": 22476240,
      "step": 10750
    },
    {
      "epoch": 1.7545476792560568,
      "grad_norm": 8.9375,
      "learning_rate": 4.160043921171961e-05,
      "loss": 2.5045,
      "num_input_tokens_seen": 22485968,
      "step": 10755
    },
    {
      "epoch": 1.7553634064768742,
      "grad_norm": 8.8125,
      "learning_rate": 4.159309723771276e-05,
      "loss": 1.776,
      "num_input_tokens_seen": 22497824,
      "step": 10760
    },
    {
      "epoch": 1.7561791336976915,
      "grad_norm": 9.0625,
      "learning_rate": 4.158575270484181e-05,
      "loss": 2.9436,
      "num_input_tokens_seen": 22509344,
      "step": 10765
    },
    {
      "epoch": 1.7569948609185089,
      "grad_norm": 3.71875,
      "learning_rate": 4.157840561423936e-05,
      "loss": 2.382,
      "num_input_tokens_seen": 22521760,
      "step": 10770
    },
    {
      "epoch": 1.7578105881393262,
      "grad_norm": 12.25,
      "learning_rate": 4.1571055967038416e-05,
      "loss": 1.3928,
      "num_input_tokens_seen": 22532000,
      "step": 10775
    },
    {
      "epoch": 1.7586263153601436,
      "grad_norm": 10.375,
      "learning_rate": 4.156370376437241e-05,
      "loss": 3.461,
      "num_input_tokens_seen": 22542112,
      "step": 10780
    },
    {
      "epoch": 1.759442042580961,
      "grad_norm": 6.25,
      "learning_rate": 4.155634900737513e-05,
      "loss": 2.9181,
      "num_input_tokens_seen": 22553504,
      "step": 10785
    },
    {
      "epoch": 1.7602577698017783,
      "grad_norm": 7.90625,
      "learning_rate": 4.1548991697180764e-05,
      "loss": 1.8568,
      "num_input_tokens_seen": 22563872,
      "step": 10790
    },
    {
      "epoch": 1.7610734970225956,
      "grad_norm": 7.71875,
      "learning_rate": 4.1541631834923914e-05,
      "loss": 1.5312,
      "num_input_tokens_seen": 22575024,
      "step": 10795
    },
    {
      "epoch": 1.761889224243413,
      "grad_norm": 6.6875,
      "learning_rate": 4.153426942173956e-05,
      "loss": 2.4284,
      "num_input_tokens_seen": 22585216,
      "step": 10800
    },
    {
      "epoch": 1.761889224243413,
      "eval_loss": 2.5454680919647217,
      "eval_runtime": 135.0717,
      "eval_samples_per_second": 20.174,
      "eval_steps_per_second": 10.091,
      "num_input_tokens_seen": 22585216,
      "step": 10800
    },
    {
      "epoch": 1.7627049514642303,
      "grad_norm": 14.6875,
      "learning_rate": 4.152690445876308e-05,
      "loss": 2.6657,
      "num_input_tokens_seen": 22595088,
      "step": 10805
    },
    {
      "epoch": 1.7635206786850477,
      "grad_norm": 3.1875,
      "learning_rate": 4.1519536947130245e-05,
      "loss": 2.1949,
      "num_input_tokens_seen": 22604784,
      "step": 10810
    },
    {
      "epoch": 1.764336405905865,
      "grad_norm": 6.03125,
      "learning_rate": 4.151216688797722e-05,
      "loss": 2.6813,
      "num_input_tokens_seen": 22615040,
      "step": 10815
    },
    {
      "epoch": 1.7651521331266824,
      "grad_norm": 4.46875,
      "learning_rate": 4.150479428244054e-05,
      "loss": 2.1004,
      "num_input_tokens_seen": 22626928,
      "step": 10820
    },
    {
      "epoch": 1.7659678603474998,
      "grad_norm": 9.0,
      "learning_rate": 4.1497419131657176e-05,
      "loss": 1.895,
      "num_input_tokens_seen": 22636896,
      "step": 10825
    },
    {
      "epoch": 1.766783587568317,
      "grad_norm": 3.90625,
      "learning_rate": 4.149004143676447e-05,
      "loss": 1.5149,
      "num_input_tokens_seen": 22649296,
      "step": 10830
    },
    {
      "epoch": 1.7675993147891345,
      "grad_norm": 3.25,
      "learning_rate": 4.148266119890015e-05,
      "loss": 1.8836,
      "num_input_tokens_seen": 22658160,
      "step": 10835
    },
    {
      "epoch": 1.768415042009952,
      "grad_norm": 12.125,
      "learning_rate": 4.1475278419202324e-05,
      "loss": 2.2586,
      "num_input_tokens_seen": 22666944,
      "step": 10840
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.859375,
      "learning_rate": 4.146789309880953e-05,
      "loss": 3.2418,
      "num_input_tokens_seen": 22678032,
      "step": 10845
    },
    {
      "epoch": 1.7700464964515867,
      "grad_norm": 4.96875,
      "learning_rate": 4.146050523886068e-05,
      "loss": 1.3837,
      "num_input_tokens_seen": 22687536,
      "step": 10850
    },
    {
      "epoch": 1.7708622236724039,
      "grad_norm": 9.9375,
      "learning_rate": 4.1453114840495055e-05,
      "loss": 3.2864,
      "num_input_tokens_seen": 22698784,
      "step": 10855
    },
    {
      "epoch": 1.7716779508932214,
      "grad_norm": 7.3125,
      "learning_rate": 4.1445721904852364e-05,
      "loss": 1.8127,
      "num_input_tokens_seen": 22709536,
      "step": 10860
    },
    {
      "epoch": 1.7724936781140386,
      "grad_norm": 1.75,
      "learning_rate": 4.143832643307269e-05,
      "loss": 1.3252,
      "num_input_tokens_seen": 22718960,
      "step": 10865
    },
    {
      "epoch": 1.7733094053348561,
      "grad_norm": 10.625,
      "learning_rate": 4.1430928426296503e-05,
      "loss": 3.6312,
      "num_input_tokens_seen": 22729696,
      "step": 10870
    },
    {
      "epoch": 1.7741251325556733,
      "grad_norm": 2.515625,
      "learning_rate": 4.142352788566466e-05,
      "loss": 1.9633,
      "num_input_tokens_seen": 22739808,
      "step": 10875
    },
    {
      "epoch": 1.7749408597764909,
      "grad_norm": 11.0,
      "learning_rate": 4.1416124812318424e-05,
      "loss": 2.199,
      "num_input_tokens_seen": 22752112,
      "step": 10880
    },
    {
      "epoch": 1.775756586997308,
      "grad_norm": 10.6875,
      "learning_rate": 4.1408719207399453e-05,
      "loss": 2.5672,
      "num_input_tokens_seen": 22763392,
      "step": 10885
    },
    {
      "epoch": 1.7765723142181256,
      "grad_norm": 6.53125,
      "learning_rate": 4.140131107204978e-05,
      "loss": 2.0084,
      "num_input_tokens_seen": 22773408,
      "step": 10890
    },
    {
      "epoch": 1.7773880414389427,
      "grad_norm": 5.46875,
      "learning_rate": 4.139390040741182e-05,
      "loss": 3.0286,
      "num_input_tokens_seen": 22783936,
      "step": 10895
    },
    {
      "epoch": 1.7782037686597603,
      "grad_norm": 7.96875,
      "learning_rate": 4.1386487214628396e-05,
      "loss": 2.7703,
      "num_input_tokens_seen": 22795152,
      "step": 10900
    },
    {
      "epoch": 1.7790194958805774,
      "grad_norm": 7.0625,
      "learning_rate": 4.137907149484272e-05,
      "loss": 2.2327,
      "num_input_tokens_seen": 22805104,
      "step": 10905
    },
    {
      "epoch": 1.779835223101395,
      "grad_norm": 9.125,
      "learning_rate": 4.137165324919839e-05,
      "loss": 3.2997,
      "num_input_tokens_seen": 22815520,
      "step": 10910
    },
    {
      "epoch": 1.780650950322212,
      "grad_norm": 10.5,
      "learning_rate": 4.136423247883939e-05,
      "loss": 1.4504,
      "num_input_tokens_seen": 22825424,
      "step": 10915
    },
    {
      "epoch": 1.7814666775430297,
      "grad_norm": 8.375,
      "learning_rate": 4.135680918491009e-05,
      "loss": 2.0813,
      "num_input_tokens_seen": 22835968,
      "step": 10920
    },
    {
      "epoch": 1.7822824047638468,
      "grad_norm": 7.34375,
      "learning_rate": 4.1349383368555265e-05,
      "loss": 2.9522,
      "num_input_tokens_seen": 22846432,
      "step": 10925
    },
    {
      "epoch": 1.7830981319846644,
      "grad_norm": 5.4375,
      "learning_rate": 4.1341955030920065e-05,
      "loss": 1.9342,
      "num_input_tokens_seen": 22856160,
      "step": 10930
    },
    {
      "epoch": 1.7839138592054817,
      "grad_norm": 3.6875,
      "learning_rate": 4.1334524173150036e-05,
      "loss": 2.4893,
      "num_input_tokens_seen": 22866512,
      "step": 10935
    },
    {
      "epoch": 1.784729586426299,
      "grad_norm": 6.78125,
      "learning_rate": 4.13270907963911e-05,
      "loss": 3.3057,
      "num_input_tokens_seen": 22876272,
      "step": 10940
    },
    {
      "epoch": 1.7855453136471164,
      "grad_norm": 3.734375,
      "learning_rate": 4.131965490178959e-05,
      "loss": 1.7925,
      "num_input_tokens_seen": 22886592,
      "step": 10945
    },
    {
      "epoch": 1.7863610408679338,
      "grad_norm": 5.90625,
      "learning_rate": 4.131221649049222e-05,
      "loss": 2.0263,
      "num_input_tokens_seen": 22897968,
      "step": 10950
    },
    {
      "epoch": 1.7871767680887511,
      "grad_norm": 4.5625,
      "learning_rate": 4.130477556364606e-05,
      "loss": 2.4973,
      "num_input_tokens_seen": 22907792,
      "step": 10955
    },
    {
      "epoch": 1.7879924953095685,
      "grad_norm": 8.875,
      "learning_rate": 4.129733212239861e-05,
      "loss": 2.3772,
      "num_input_tokens_seen": 22917264,
      "step": 10960
    },
    {
      "epoch": 1.7888082225303858,
      "grad_norm": 8.4375,
      "learning_rate": 4.128988616789774e-05,
      "loss": 1.9849,
      "num_input_tokens_seen": 22928608,
      "step": 10965
    },
    {
      "epoch": 1.7896239497512032,
      "grad_norm": 9.0,
      "learning_rate": 4.1282437701291724e-05,
      "loss": 2.6225,
      "num_input_tokens_seen": 22939888,
      "step": 10970
    },
    {
      "epoch": 1.7904396769720206,
      "grad_norm": 0.1923828125,
      "learning_rate": 4.1274986723729184e-05,
      "loss": 2.9913,
      "num_input_tokens_seen": 22950864,
      "step": 10975
    },
    {
      "epoch": 1.791255404192838,
      "grad_norm": 6.9375,
      "learning_rate": 4.126753323635917e-05,
      "loss": 2.0562,
      "num_input_tokens_seen": 22962656,
      "step": 10980
    },
    {
      "epoch": 1.7920711314136553,
      "grad_norm": 2.96875,
      "learning_rate": 4.12600772403311e-05,
      "loss": 3.0361,
      "num_input_tokens_seen": 22974112,
      "step": 10985
    },
    {
      "epoch": 1.7928868586344726,
      "grad_norm": 8.6875,
      "learning_rate": 4.125261873679479e-05,
      "loss": 1.8447,
      "num_input_tokens_seen": 22985232,
      "step": 10990
    },
    {
      "epoch": 1.79370258585529,
      "grad_norm": 12.75,
      "learning_rate": 4.124515772690042e-05,
      "loss": 2.978,
      "num_input_tokens_seen": 22995312,
      "step": 10995
    },
    {
      "epoch": 1.7945183130761073,
      "grad_norm": 0.1826171875,
      "learning_rate": 4.123769421179858e-05,
      "loss": 2.028,
      "num_input_tokens_seen": 23005600,
      "step": 11000
    },
    {
      "epoch": 1.7945183130761073,
      "eval_loss": 2.5415890216827393,
      "eval_runtime": 135.1545,
      "eval_samples_per_second": 20.162,
      "eval_steps_per_second": 10.085,
      "num_input_tokens_seen": 23005600,
      "step": 11000
    },
    {
      "epoch": 1.7953340402969247,
      "grad_norm": 13.375,
      "learning_rate": 4.1230228192640236e-05,
      "loss": 2.9247,
      "num_input_tokens_seen": 23014720,
      "step": 11005
    },
    {
      "epoch": 1.796149767517742,
      "grad_norm": 5.1875,
      "learning_rate": 4.122275967057675e-05,
      "loss": 2.1675,
      "num_input_tokens_seen": 23024944,
      "step": 11010
    },
    {
      "epoch": 1.7969654947385594,
      "grad_norm": 4.1875,
      "learning_rate": 4.1215288646759846e-05,
      "loss": 1.4726,
      "num_input_tokens_seen": 23035856,
      "step": 11015
    },
    {
      "epoch": 1.7977812219593767,
      "grad_norm": 4.53125,
      "learning_rate": 4.120781512234166e-05,
      "loss": 2.8311,
      "num_input_tokens_seen": 23045360,
      "step": 11020
    },
    {
      "epoch": 1.7985969491801943,
      "grad_norm": 10.4375,
      "learning_rate": 4.120033909847471e-05,
      "loss": 2.0828,
      "num_input_tokens_seen": 23056368,
      "step": 11025
    },
    {
      "epoch": 1.7994126764010114,
      "grad_norm": 3.171875,
      "learning_rate": 4.119286057631187e-05,
      "loss": 2.2712,
      "num_input_tokens_seen": 23066736,
      "step": 11030
    },
    {
      "epoch": 1.800228403621829,
      "grad_norm": 9.0,
      "learning_rate": 4.118537955700646e-05,
      "loss": 2.0837,
      "num_input_tokens_seen": 23078992,
      "step": 11035
    },
    {
      "epoch": 1.8010441308426461,
      "grad_norm": 7.25,
      "learning_rate": 4.11778960417121e-05,
      "loss": 1.7473,
      "num_input_tokens_seen": 23089984,
      "step": 11040
    },
    {
      "epoch": 1.8018598580634637,
      "grad_norm": 13.5625,
      "learning_rate": 4.117041003158288e-05,
      "loss": 2.8248,
      "num_input_tokens_seen": 23099904,
      "step": 11045
    },
    {
      "epoch": 1.8026755852842808,
      "grad_norm": 5.03125,
      "learning_rate": 4.1162921527773215e-05,
      "loss": 2.6457,
      "num_input_tokens_seen": 23110208,
      "step": 11050
    },
    {
      "epoch": 1.8034913125050984,
      "grad_norm": 10.1875,
      "learning_rate": 4.115543053143794e-05,
      "loss": 3.4696,
      "num_input_tokens_seen": 23120544,
      "step": 11055
    },
    {
      "epoch": 1.8043070397259156,
      "grad_norm": 7.6875,
      "learning_rate": 4.114793704373226e-05,
      "loss": 2.8976,
      "num_input_tokens_seen": 23131472,
      "step": 11060
    },
    {
      "epoch": 1.8051227669467331,
      "grad_norm": 6.25,
      "learning_rate": 4.114044106581175e-05,
      "loss": 1.779,
      "num_input_tokens_seen": 23140640,
      "step": 11065
    },
    {
      "epoch": 1.8059384941675503,
      "grad_norm": 7.78125,
      "learning_rate": 4.11329425988324e-05,
      "loss": 3.3835,
      "num_input_tokens_seen": 23150800,
      "step": 11070
    },
    {
      "epoch": 1.8067542213883678,
      "grad_norm": 6.71875,
      "learning_rate": 4.112544164395056e-05,
      "loss": 2.7285,
      "num_input_tokens_seen": 23161360,
      "step": 11075
    },
    {
      "epoch": 1.807569948609185,
      "grad_norm": 4.90625,
      "learning_rate": 4.111793820232297e-05,
      "loss": 1.0915,
      "num_input_tokens_seen": 23171776,
      "step": 11080
    },
    {
      "epoch": 1.8083856758300025,
      "grad_norm": 8.5625,
      "learning_rate": 4.1110432275106767e-05,
      "loss": 3.2355,
      "num_input_tokens_seen": 23181344,
      "step": 11085
    },
    {
      "epoch": 1.8092014030508197,
      "grad_norm": 8.3125,
      "learning_rate": 4.110292386345944e-05,
      "loss": 2.6463,
      "num_input_tokens_seen": 23191808,
      "step": 11090
    },
    {
      "epoch": 1.8100171302716372,
      "grad_norm": 3.9375,
      "learning_rate": 4.109541296853891e-05,
      "loss": 1.9661,
      "num_input_tokens_seen": 23201312,
      "step": 11095
    },
    {
      "epoch": 1.8108328574924544,
      "grad_norm": 10.8125,
      "learning_rate": 4.108789959150341e-05,
      "loss": 1.6202,
      "num_input_tokens_seen": 23212416,
      "step": 11100
    },
    {
      "epoch": 1.811648584713272,
      "grad_norm": 14.1875,
      "learning_rate": 4.108038373351163e-05,
      "loss": 2.6634,
      "num_input_tokens_seen": 23221024,
      "step": 11105
    },
    {
      "epoch": 1.812464311934089,
      "grad_norm": 0.142578125,
      "learning_rate": 4.10728653957226e-05,
      "loss": 1.2823,
      "num_input_tokens_seen": 23230416,
      "step": 11110
    },
    {
      "epoch": 1.8132800391549067,
      "grad_norm": 8.25,
      "learning_rate": 4.106534457929575e-05,
      "loss": 3.5251,
      "num_input_tokens_seen": 23240192,
      "step": 11115
    },
    {
      "epoch": 1.814095766375724,
      "grad_norm": 5.875,
      "learning_rate": 4.105782128539086e-05,
      "loss": 2.2713,
      "num_input_tokens_seen": 23248992,
      "step": 11120
    },
    {
      "epoch": 1.8149114935965414,
      "grad_norm": 1.59375,
      "learning_rate": 4.1050295515168144e-05,
      "loss": 1.1684,
      "num_input_tokens_seen": 23258688,
      "step": 11125
    },
    {
      "epoch": 1.8157272208173587,
      "grad_norm": 7.84375,
      "learning_rate": 4.1042767269788155e-05,
      "loss": 1.9391,
      "num_input_tokens_seen": 23268448,
      "step": 11130
    },
    {
      "epoch": 1.816542948038176,
      "grad_norm": 9.3125,
      "learning_rate": 4.103523655041185e-05,
      "loss": 3.8295,
      "num_input_tokens_seen": 23277616,
      "step": 11135
    },
    {
      "epoch": 1.8173586752589934,
      "grad_norm": 3.6875,
      "learning_rate": 4.102770335820055e-05,
      "loss": 3.4805,
      "num_input_tokens_seen": 23287136,
      "step": 11140
    },
    {
      "epoch": 1.8181744024798108,
      "grad_norm": 6.9375,
      "learning_rate": 4.1020167694315984e-05,
      "loss": 1.8822,
      "num_input_tokens_seen": 23296880,
      "step": 11145
    },
    {
      "epoch": 1.8189901297006281,
      "grad_norm": 8.1875,
      "learning_rate": 4.101262955992023e-05,
      "loss": 2.8875,
      "num_input_tokens_seen": 23305856,
      "step": 11150
    },
    {
      "epoch": 1.8198058569214455,
      "grad_norm": 5.5,
      "learning_rate": 4.100508895617578e-05,
      "loss": 2.3623,
      "num_input_tokens_seen": 23315936,
      "step": 11155
    },
    {
      "epoch": 1.8206215841422628,
      "grad_norm": 4.90625,
      "learning_rate": 4.099754588424547e-05,
      "loss": 2.2908,
      "num_input_tokens_seen": 23326752,
      "step": 11160
    },
    {
      "epoch": 1.8214373113630802,
      "grad_norm": 5.40625,
      "learning_rate": 4.0990000345292546e-05,
      "loss": 3.5846,
      "num_input_tokens_seen": 23338544,
      "step": 11165
    },
    {
      "epoch": 1.8222530385838975,
      "grad_norm": 9.5625,
      "learning_rate": 4.098245234048064e-05,
      "loss": 2.2862,
      "num_input_tokens_seen": 23348800,
      "step": 11170
    },
    {
      "epoch": 1.8230687658047149,
      "grad_norm": 5.875,
      "learning_rate": 4.0974901870973726e-05,
      "loss": 2.0122,
      "num_input_tokens_seen": 23359824,
      "step": 11175
    },
    {
      "epoch": 1.8238844930255322,
      "grad_norm": 6.28125,
      "learning_rate": 4.096734893793619e-05,
      "loss": 2.8398,
      "num_input_tokens_seen": 23369920,
      "step": 11180
    },
    {
      "epoch": 1.8247002202463496,
      "grad_norm": 6.15625,
      "learning_rate": 4.095979354253279e-05,
      "loss": 1.9091,
      "num_input_tokens_seen": 23380864,
      "step": 11185
    },
    {
      "epoch": 1.825515947467167,
      "grad_norm": 3.59375,
      "learning_rate": 4.0952235685928656e-05,
      "loss": 2.5836,
      "num_input_tokens_seen": 23391888,
      "step": 11190
    },
    {
      "epoch": 1.8263316746879843,
      "grad_norm": 6.15625,
      "learning_rate": 4.094467536928932e-05,
      "loss": 1.8714,
      "num_input_tokens_seen": 23403312,
      "step": 11195
    },
    {
      "epoch": 1.8271474019088016,
      "grad_norm": 5.1875,
      "learning_rate": 4.093711259378067e-05,
      "loss": 2.9135,
      "num_input_tokens_seen": 23413712,
      "step": 11200
    },
    {
      "epoch": 1.8271474019088016,
      "eval_loss": 2.548494338989258,
      "eval_runtime": 135.0508,
      "eval_samples_per_second": 20.178,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 23413712,
      "step": 11200
    },
    {
      "epoch": 1.827963129129619,
      "grad_norm": 6.46875,
      "learning_rate": 4.092954736056897e-05,
      "loss": 2.3802,
      "num_input_tokens_seen": 23424544,
      "step": 11205
    },
    {
      "epoch": 1.8287788563504366,
      "grad_norm": 1.7734375,
      "learning_rate": 4.09219796708209e-05,
      "loss": 2.1915,
      "num_input_tokens_seen": 23435200,
      "step": 11210
    },
    {
      "epoch": 1.8295945835712537,
      "grad_norm": 8.0,
      "learning_rate": 4.0914409525703464e-05,
      "loss": 3.4955,
      "num_input_tokens_seen": 23444192,
      "step": 11215
    },
    {
      "epoch": 1.8304103107920713,
      "grad_norm": 5.40625,
      "learning_rate": 4.090683692638408e-05,
      "loss": 3.2005,
      "num_input_tokens_seen": 23454144,
      "step": 11220
    },
    {
      "epoch": 1.8312260380128884,
      "grad_norm": 0.60546875,
      "learning_rate": 4.089926187403056e-05,
      "loss": 2.6453,
      "num_input_tokens_seen": 23464656,
      "step": 11225
    },
    {
      "epoch": 1.832041765233706,
      "grad_norm": 7.28125,
      "learning_rate": 4.0891684369811044e-05,
      "loss": 3.4868,
      "num_input_tokens_seen": 23474112,
      "step": 11230
    },
    {
      "epoch": 1.8328574924545231,
      "grad_norm": 7.34375,
      "learning_rate": 4.0884104414894107e-05,
      "loss": 2.1139,
      "num_input_tokens_seen": 23485456,
      "step": 11235
    },
    {
      "epoch": 1.8336732196753407,
      "grad_norm": 5.5,
      "learning_rate": 4.087652201044864e-05,
      "loss": 2.7731,
      "num_input_tokens_seen": 23496576,
      "step": 11240
    },
    {
      "epoch": 1.8344889468961578,
      "grad_norm": 5.46875,
      "learning_rate": 4.086893715764397e-05,
      "loss": 3.1156,
      "num_input_tokens_seen": 23506976,
      "step": 11245
    },
    {
      "epoch": 1.8353046741169754,
      "grad_norm": 8.1875,
      "learning_rate": 4.086134985764977e-05,
      "loss": 2.0548,
      "num_input_tokens_seen": 23517776,
      "step": 11250
    },
    {
      "epoch": 1.8361204013377925,
      "grad_norm": 1.609375,
      "learning_rate": 4.0853760111636085e-05,
      "loss": 2.7464,
      "num_input_tokens_seen": 23528096,
      "step": 11255
    },
    {
      "epoch": 1.83693612855861,
      "grad_norm": 8.875,
      "learning_rate": 4.084616792077337e-05,
      "loss": 2.9937,
      "num_input_tokens_seen": 23538816,
      "step": 11260
    },
    {
      "epoch": 1.8377518557794272,
      "grad_norm": 2.640625,
      "learning_rate": 4.083857328623243e-05,
      "loss": 1.7838,
      "num_input_tokens_seen": 23548688,
      "step": 11265
    },
    {
      "epoch": 1.8385675830002448,
      "grad_norm": 8.625,
      "learning_rate": 4.083097620918444e-05,
      "loss": 3.5507,
      "num_input_tokens_seen": 23559440,
      "step": 11270
    },
    {
      "epoch": 1.839383310221062,
      "grad_norm": 8.25,
      "learning_rate": 4.082337669080097e-05,
      "loss": 0.9322,
      "num_input_tokens_seen": 23569488,
      "step": 11275
    },
    {
      "epoch": 1.8401990374418795,
      "grad_norm": 4.78125,
      "learning_rate": 4.081577473225398e-05,
      "loss": 1.7583,
      "num_input_tokens_seen": 23581088,
      "step": 11280
    },
    {
      "epoch": 1.8410147646626966,
      "grad_norm": 17.625,
      "learning_rate": 4.080817033471577e-05,
      "loss": 2.5401,
      "num_input_tokens_seen": 23590048,
      "step": 11285
    },
    {
      "epoch": 1.8418304918835142,
      "grad_norm": 9.0625,
      "learning_rate": 4.080056349935903e-05,
      "loss": 2.7579,
      "num_input_tokens_seen": 23600544,
      "step": 11290
    },
    {
      "epoch": 1.8426462191043314,
      "grad_norm": 9.125,
      "learning_rate": 4.079295422735684e-05,
      "loss": 2.7838,
      "num_input_tokens_seen": 23609984,
      "step": 11295
    },
    {
      "epoch": 1.843461946325149,
      "grad_norm": 14.375,
      "learning_rate": 4.078534251988264e-05,
      "loss": 2.6622,
      "num_input_tokens_seen": 23620400,
      "step": 11300
    },
    {
      "epoch": 1.8442776735459663,
      "grad_norm": 8.0625,
      "learning_rate": 4.077772837811025e-05,
      "loss": 2.0743,
      "num_input_tokens_seen": 23629840,
      "step": 11305
    },
    {
      "epoch": 1.8450934007667836,
      "grad_norm": 7.90625,
      "learning_rate": 4.0770111803213874e-05,
      "loss": 3.0851,
      "num_input_tokens_seen": 23641696,
      "step": 11310
    },
    {
      "epoch": 1.845909127987601,
      "grad_norm": 3.515625,
      "learning_rate": 4.076249279636807e-05,
      "loss": 1.4606,
      "num_input_tokens_seen": 23651680,
      "step": 11315
    },
    {
      "epoch": 1.8467248552084183,
      "grad_norm": 9.6875,
      "learning_rate": 4.075487135874781e-05,
      "loss": 2.8207,
      "num_input_tokens_seen": 23662528,
      "step": 11320
    },
    {
      "epoch": 1.8475405824292357,
      "grad_norm": 10.5625,
      "learning_rate": 4.074724749152837e-05,
      "loss": 2.2815,
      "num_input_tokens_seen": 23673328,
      "step": 11325
    },
    {
      "epoch": 1.848356309650053,
      "grad_norm": 6.28125,
      "learning_rate": 4.07396211958855e-05,
      "loss": 2.1915,
      "num_input_tokens_seen": 23683872,
      "step": 11330
    },
    {
      "epoch": 1.8491720368708704,
      "grad_norm": 4.21875,
      "learning_rate": 4.073199247299523e-05,
      "loss": 2.6668,
      "num_input_tokens_seen": 23693360,
      "step": 11335
    },
    {
      "epoch": 1.8499877640916877,
      "grad_norm": 11.0,
      "learning_rate": 4.072436132403403e-05,
      "loss": 2.3273,
      "num_input_tokens_seen": 23704112,
      "step": 11340
    },
    {
      "epoch": 1.850803491312505,
      "grad_norm": 1.4140625,
      "learning_rate": 4.0716727750178704e-05,
      "loss": 2.0404,
      "num_input_tokens_seen": 23713984,
      "step": 11345
    },
    {
      "epoch": 1.8516192185333225,
      "grad_norm": 7.59375,
      "learning_rate": 4.0709091752606455e-05,
      "loss": 3.2624,
      "num_input_tokens_seen": 23724608,
      "step": 11350
    },
    {
      "epoch": 1.8524349457541398,
      "grad_norm": 12.9375,
      "learning_rate": 4.070145333249484e-05,
      "loss": 2.4581,
      "num_input_tokens_seen": 23735472,
      "step": 11355
    },
    {
      "epoch": 1.8532506729749572,
      "grad_norm": 11.125,
      "learning_rate": 4.069381249102181e-05,
      "loss": 2.4519,
      "num_input_tokens_seen": 23746592,
      "step": 11360
    },
    {
      "epoch": 1.8540664001957745,
      "grad_norm": 8.3125,
      "learning_rate": 4.0686169229365665e-05,
      "loss": 2.8099,
      "num_input_tokens_seen": 23757536,
      "step": 11365
    },
    {
      "epoch": 1.8548821274165919,
      "grad_norm": 11.5,
      "learning_rate": 4.067852354870511e-05,
      "loss": 3.5504,
      "num_input_tokens_seen": 23767584,
      "step": 11370
    },
    {
      "epoch": 1.8556978546374092,
      "grad_norm": 5.46875,
      "learning_rate": 4.067087545021919e-05,
      "loss": 2.2058,
      "num_input_tokens_seen": 23777136,
      "step": 11375
    },
    {
      "epoch": 1.8565135818582266,
      "grad_norm": 8.6875,
      "learning_rate": 4.066322493508734e-05,
      "loss": 3.6205,
      "num_input_tokens_seen": 23787872,
      "step": 11380
    },
    {
      "epoch": 1.857329309079044,
      "grad_norm": 7.375,
      "learning_rate": 4.065557200448937e-05,
      "loss": 3.8011,
      "num_input_tokens_seen": 23796464,
      "step": 11385
    },
    {
      "epoch": 1.8581450362998613,
      "grad_norm": 6.28125,
      "learning_rate": 4.064791665960546e-05,
      "loss": 3.8066,
      "num_input_tokens_seen": 23807360,
      "step": 11390
    },
    {
      "epoch": 1.8589607635206788,
      "grad_norm": 4.6875,
      "learning_rate": 4.064025890161615e-05,
      "loss": 1.6115,
      "num_input_tokens_seen": 23816768,
      "step": 11395
    },
    {
      "epoch": 1.859776490741496,
      "grad_norm": 7.5625,
      "learning_rate": 4.0632598731702373e-05,
      "loss": 1.8818,
      "num_input_tokens_seen": 23827536,
      "step": 11400
    },
    {
      "epoch": 1.859776490741496,
      "eval_loss": 2.5369606018066406,
      "eval_runtime": 134.9998,
      "eval_samples_per_second": 20.185,
      "eval_steps_per_second": 10.096,
      "num_input_tokens_seen": 23827536,
      "step": 11400
    },
    {
      "epoch": 1.8605922179623136,
      "grad_norm": 7.8125,
      "learning_rate": 4.0624936151045426e-05,
      "loss": 2.2367,
      "num_input_tokens_seen": 23837824,
      "step": 11405
    },
    {
      "epoch": 1.8614079451831307,
      "grad_norm": 8.6875,
      "learning_rate": 4.061727116082696e-05,
      "loss": 3.2004,
      "num_input_tokens_seen": 23847904,
      "step": 11410
    },
    {
      "epoch": 1.8622236724039483,
      "grad_norm": 6.96875,
      "learning_rate": 4.060960376222903e-05,
      "loss": 2.4678,
      "num_input_tokens_seen": 23856816,
      "step": 11415
    },
    {
      "epoch": 1.8630393996247654,
      "grad_norm": 4.21875,
      "learning_rate": 4.0601933956434034e-05,
      "loss": 2.1978,
      "num_input_tokens_seen": 23865392,
      "step": 11420
    },
    {
      "epoch": 1.863855126845583,
      "grad_norm": 2.625,
      "learning_rate": 4.059426174462476e-05,
      "loss": 2.297,
      "num_input_tokens_seen": 23876384,
      "step": 11425
    },
    {
      "epoch": 1.8646708540664,
      "grad_norm": 4.8125,
      "learning_rate": 4.058658712798435e-05,
      "loss": 1.8264,
      "num_input_tokens_seen": 23887792,
      "step": 11430
    },
    {
      "epoch": 1.8654865812872177,
      "grad_norm": 6.75,
      "learning_rate": 4.0578910107696336e-05,
      "loss": 2.5089,
      "num_input_tokens_seen": 23897968,
      "step": 11435
    },
    {
      "epoch": 1.8663023085080348,
      "grad_norm": 8.5625,
      "learning_rate": 4.05712306849446e-05,
      "loss": 2.3993,
      "num_input_tokens_seen": 23907632,
      "step": 11440
    },
    {
      "epoch": 1.8671180357288524,
      "grad_norm": 12.875,
      "learning_rate": 4.0563548860913415e-05,
      "loss": 1.976,
      "num_input_tokens_seen": 23917440,
      "step": 11445
    },
    {
      "epoch": 1.8679337629496695,
      "grad_norm": 8.6875,
      "learning_rate": 4.0555864636787414e-05,
      "loss": 2.2438,
      "num_input_tokens_seen": 23928144,
      "step": 11450
    },
    {
      "epoch": 1.868749490170487,
      "grad_norm": 6.90625,
      "learning_rate": 4.054817801375159e-05,
      "loss": 2.6947,
      "num_input_tokens_seen": 23938480,
      "step": 11455
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 4.4375,
      "learning_rate": 4.054048899299134e-05,
      "loss": 2.8081,
      "num_input_tokens_seen": 23948192,
      "step": 11460
    },
    {
      "epoch": 1.8703809446121218,
      "grad_norm": 5.71875,
      "learning_rate": 4.0532797575692385e-05,
      "loss": 3.7468,
      "num_input_tokens_seen": 23958784,
      "step": 11465
    },
    {
      "epoch": 1.871196671832939,
      "grad_norm": 5.96875,
      "learning_rate": 4.052510376304085e-05,
      "loss": 2.9304,
      "num_input_tokens_seen": 23969008,
      "step": 11470
    },
    {
      "epoch": 1.8720123990537565,
      "grad_norm": 9.125,
      "learning_rate": 4.051740755622321e-05,
      "loss": 2.3511,
      "num_input_tokens_seen": 23978688,
      "step": 11475
    },
    {
      "epoch": 1.8728281262745736,
      "grad_norm": 14.625,
      "learning_rate": 4.050970895642632e-05,
      "loss": 4.4255,
      "num_input_tokens_seen": 23987408,
      "step": 11480
    },
    {
      "epoch": 1.8736438534953912,
      "grad_norm": 5.75,
      "learning_rate": 4.050200796483741e-05,
      "loss": 1.6992,
      "num_input_tokens_seen": 23997328,
      "step": 11485
    },
    {
      "epoch": 1.8744595807162086,
      "grad_norm": 6.375,
      "learning_rate": 4.049430458264405e-05,
      "loss": 3.2928,
      "num_input_tokens_seen": 24006304,
      "step": 11490
    },
    {
      "epoch": 1.875275307937026,
      "grad_norm": 6.5,
      "learning_rate": 4.048659881103422e-05,
      "loss": 2.3,
      "num_input_tokens_seen": 24017856,
      "step": 11495
    },
    {
      "epoch": 1.8760910351578433,
      "grad_norm": 6.96875,
      "learning_rate": 4.0478890651196235e-05,
      "loss": 2.3548,
      "num_input_tokens_seen": 24029760,
      "step": 11500
    },
    {
      "epoch": 1.8769067623786606,
      "grad_norm": 3.640625,
      "learning_rate": 4.047118010431879e-05,
      "loss": 3.428,
      "num_input_tokens_seen": 24040336,
      "step": 11505
    },
    {
      "epoch": 1.877722489599478,
      "grad_norm": 4.3125,
      "learning_rate": 4.046346717159094e-05,
      "loss": 2.5567,
      "num_input_tokens_seen": 24051472,
      "step": 11510
    },
    {
      "epoch": 1.8785382168202953,
      "grad_norm": 10.875,
      "learning_rate": 4.045575185420214e-05,
      "loss": 2.718,
      "num_input_tokens_seen": 24061920,
      "step": 11515
    },
    {
      "epoch": 1.8793539440411127,
      "grad_norm": 3.515625,
      "learning_rate": 4.0448034153342165e-05,
      "loss": 1.6905,
      "num_input_tokens_seen": 24071424,
      "step": 11520
    },
    {
      "epoch": 1.88016967126193,
      "grad_norm": 9.1875,
      "learning_rate": 4.0440314070201194e-05,
      "loss": 3.4315,
      "num_input_tokens_seen": 24081712,
      "step": 11525
    },
    {
      "epoch": 1.8809853984827474,
      "grad_norm": 9.9375,
      "learning_rate": 4.043259160596976e-05,
      "loss": 2.6037,
      "num_input_tokens_seen": 24091248,
      "step": 11530
    },
    {
      "epoch": 1.8818011257035647,
      "grad_norm": 4.9375,
      "learning_rate": 4.0424866761838767e-05,
      "loss": 2.7409,
      "num_input_tokens_seen": 24103200,
      "step": 11535
    },
    {
      "epoch": 1.882616852924382,
      "grad_norm": 5.8125,
      "learning_rate": 4.041713953899948e-05,
      "loss": 2.4161,
      "num_input_tokens_seen": 24113200,
      "step": 11540
    },
    {
      "epoch": 1.8834325801451994,
      "grad_norm": 12.125,
      "learning_rate": 4.0409409938643515e-05,
      "loss": 3.3316,
      "num_input_tokens_seen": 24121952,
      "step": 11545
    },
    {
      "epoch": 1.8842483073660168,
      "grad_norm": 6.15625,
      "learning_rate": 4.0401677961962904e-05,
      "loss": 1.7687,
      "num_input_tokens_seen": 24131840,
      "step": 11550
    },
    {
      "epoch": 1.8850640345868341,
      "grad_norm": 7.96875,
      "learning_rate": 4.039394361015001e-05,
      "loss": 2.2257,
      "num_input_tokens_seen": 24143936,
      "step": 11555
    },
    {
      "epoch": 1.8858797618076515,
      "grad_norm": 5.53125,
      "learning_rate": 4.038620688439755e-05,
      "loss": 1.8754,
      "num_input_tokens_seen": 24153168,
      "step": 11560
    },
    {
      "epoch": 1.8866954890284688,
      "grad_norm": 1.515625,
      "learning_rate": 4.037846778589862e-05,
      "loss": 1.0507,
      "num_input_tokens_seen": 24164576,
      "step": 11565
    },
    {
      "epoch": 1.8875112162492862,
      "grad_norm": 10.125,
      "learning_rate": 4.0370726315846715e-05,
      "loss": 2.1941,
      "num_input_tokens_seen": 24176608,
      "step": 11570
    },
    {
      "epoch": 1.8883269434701035,
      "grad_norm": 7.46875,
      "learning_rate": 4.036298247543565e-05,
      "loss": 2.4753,
      "num_input_tokens_seen": 24187696,
      "step": 11575
    },
    {
      "epoch": 1.8891426706909211,
      "grad_norm": 10.4375,
      "learning_rate": 4.035523626585962e-05,
      "loss": 1.8263,
      "num_input_tokens_seen": 24199280,
      "step": 11580
    },
    {
      "epoch": 1.8899583979117383,
      "grad_norm": 16.875,
      "learning_rate": 4.0347487688313194e-05,
      "loss": 3.9414,
      "num_input_tokens_seen": 24209504,
      "step": 11585
    },
    {
      "epoch": 1.8907741251325558,
      "grad_norm": 8.125,
      "learning_rate": 4.0339736743991296e-05,
      "loss": 2.0571,
      "num_input_tokens_seen": 24220576,
      "step": 11590
    },
    {
      "epoch": 1.891589852353373,
      "grad_norm": 5.75,
      "learning_rate": 4.0331983434089227e-05,
      "loss": 3.0148,
      "num_input_tokens_seen": 24230752,
      "step": 11595
    },
    {
      "epoch": 1.8924055795741905,
      "grad_norm": 3.859375,
      "learning_rate": 4.032422775980264e-05,
      "loss": 1.3497,
      "num_input_tokens_seen": 24242256,
      "step": 11600
    },
    {
      "epoch": 1.8924055795741905,
      "eval_loss": 2.549673080444336,
      "eval_runtime": 135.0674,
      "eval_samples_per_second": 20.175,
      "eval_steps_per_second": 10.091,
      "num_input_tokens_seen": 24242256,
      "step": 11600
    },
    {
      "epoch": 1.8932213067950077,
      "grad_norm": 3.640625,
      "learning_rate": 4.031646972232754e-05,
      "loss": 1.3969,
      "num_input_tokens_seen": 24252608,
      "step": 11605
    },
    {
      "epoch": 1.8940370340158252,
      "grad_norm": 7.0625,
      "learning_rate": 4.0308709322860344e-05,
      "loss": 1.9765,
      "num_input_tokens_seen": 24262416,
      "step": 11610
    },
    {
      "epoch": 1.8948527612366424,
      "grad_norm": 11.1875,
      "learning_rate": 4.0300946562597784e-05,
      "loss": 2.9223,
      "num_input_tokens_seen": 24273952,
      "step": 11615
    },
    {
      "epoch": 1.89566848845746,
      "grad_norm": 6.4375,
      "learning_rate": 4.029318144273698e-05,
      "loss": 2.3679,
      "num_input_tokens_seen": 24282352,
      "step": 11620
    },
    {
      "epoch": 1.896484215678277,
      "grad_norm": 7.90625,
      "learning_rate": 4.0285413964475415e-05,
      "loss": 2.5045,
      "num_input_tokens_seen": 24293440,
      "step": 11625
    },
    {
      "epoch": 1.8972999428990946,
      "grad_norm": 7.46875,
      "learning_rate": 4.0277644129010927e-05,
      "loss": 1.8866,
      "num_input_tokens_seen": 24304640,
      "step": 11630
    },
    {
      "epoch": 1.8981156701199118,
      "grad_norm": 1.203125,
      "learning_rate": 4.0269871937541724e-05,
      "loss": 2.1028,
      "num_input_tokens_seen": 24314496,
      "step": 11635
    },
    {
      "epoch": 1.8989313973407294,
      "grad_norm": 7.9375,
      "learning_rate": 4.026209739126637e-05,
      "loss": 2.0925,
      "num_input_tokens_seen": 24324096,
      "step": 11640
    },
    {
      "epoch": 1.8997471245615465,
      "grad_norm": 7.59375,
      "learning_rate": 4.025432049138381e-05,
      "loss": 3.0681,
      "num_input_tokens_seen": 24334672,
      "step": 11645
    },
    {
      "epoch": 1.900562851782364,
      "grad_norm": 4.59375,
      "learning_rate": 4.0246541239093325e-05,
      "loss": 2.6697,
      "num_input_tokens_seen": 24344512,
      "step": 11650
    },
    {
      "epoch": 1.9013785790031812,
      "grad_norm": 11.5625,
      "learning_rate": 4.023875963559459e-05,
      "loss": 2.952,
      "num_input_tokens_seen": 24354208,
      "step": 11655
    },
    {
      "epoch": 1.9021943062239988,
      "grad_norm": 8.5625,
      "learning_rate": 4.023097568208761e-05,
      "loss": 2.4238,
      "num_input_tokens_seen": 24364928,
      "step": 11660
    },
    {
      "epoch": 1.903010033444816,
      "grad_norm": 6.96875,
      "learning_rate": 4.022318937977277e-05,
      "loss": 2.7698,
      "num_input_tokens_seen": 24376576,
      "step": 11665
    },
    {
      "epoch": 1.9038257606656335,
      "grad_norm": 1.4375,
      "learning_rate": 4.021540072985084e-05,
      "loss": 1.5888,
      "num_input_tokens_seen": 24387504,
      "step": 11670
    },
    {
      "epoch": 1.9046414878864508,
      "grad_norm": 2.53125,
      "learning_rate": 4.020760973352289e-05,
      "loss": 3.0971,
      "num_input_tokens_seen": 24398400,
      "step": 11675
    },
    {
      "epoch": 1.9054572151072682,
      "grad_norm": 8.8125,
      "learning_rate": 4.019981639199042e-05,
      "loss": 2.9334,
      "num_input_tokens_seen": 24408896,
      "step": 11680
    },
    {
      "epoch": 1.9062729423280855,
      "grad_norm": 7.75,
      "learning_rate": 4.0192020706455245e-05,
      "loss": 1.9956,
      "num_input_tokens_seen": 24418928,
      "step": 11685
    },
    {
      "epoch": 1.9070886695489029,
      "grad_norm": 8.125,
      "learning_rate": 4.018422267811956e-05,
      "loss": 2.9171,
      "num_input_tokens_seen": 24429792,
      "step": 11690
    },
    {
      "epoch": 1.9079043967697202,
      "grad_norm": 7.09375,
      "learning_rate": 4.017642230818592e-05,
      "loss": 2.2483,
      "num_input_tokens_seen": 24439808,
      "step": 11695
    },
    {
      "epoch": 1.9087201239905376,
      "grad_norm": 14.125,
      "learning_rate": 4.0168619597857246e-05,
      "loss": 2.5742,
      "num_input_tokens_seen": 24449200,
      "step": 11700
    },
    {
      "epoch": 1.909535851211355,
      "grad_norm": 11.5625,
      "learning_rate": 4.016081454833681e-05,
      "loss": 3.0056,
      "num_input_tokens_seen": 24459552,
      "step": 11705
    },
    {
      "epoch": 1.9103515784321723,
      "grad_norm": 4.59375,
      "learning_rate": 4.0153007160828245e-05,
      "loss": 2.7527,
      "num_input_tokens_seen": 24469008,
      "step": 11710
    },
    {
      "epoch": 1.9111673056529896,
      "grad_norm": 6.625,
      "learning_rate": 4.0145197436535555e-05,
      "loss": 2.0328,
      "num_input_tokens_seen": 24480576,
      "step": 11715
    },
    {
      "epoch": 1.911983032873807,
      "grad_norm": 5.96875,
      "learning_rate": 4.0137385376663095e-05,
      "loss": 2.4346,
      "num_input_tokens_seen": 24489856,
      "step": 11720
    },
    {
      "epoch": 1.9127987600946244,
      "grad_norm": 8.3125,
      "learning_rate": 4.012957098241558e-05,
      "loss": 2.8905,
      "num_input_tokens_seen": 24500080,
      "step": 11725
    },
    {
      "epoch": 1.9136144873154417,
      "grad_norm": 3.21875,
      "learning_rate": 4.0121754254998076e-05,
      "loss": 1.3164,
      "num_input_tokens_seen": 24510480,
      "step": 11730
    },
    {
      "epoch": 1.914430214536259,
      "grad_norm": 5.90625,
      "learning_rate": 4.011393519561606e-05,
      "loss": 1.4809,
      "num_input_tokens_seen": 24519168,
      "step": 11735
    },
    {
      "epoch": 1.9152459417570764,
      "grad_norm": 5.46875,
      "learning_rate": 4.010611380547529e-05,
      "loss": 1.9885,
      "num_input_tokens_seen": 24528720,
      "step": 11740
    },
    {
      "epoch": 1.9160616689778938,
      "grad_norm": 3.4375,
      "learning_rate": 4.009829008578192e-05,
      "loss": 3.1988,
      "num_input_tokens_seen": 24539408,
      "step": 11745
    },
    {
      "epoch": 1.9168773961987111,
      "grad_norm": 6.40625,
      "learning_rate": 4.00904640377425e-05,
      "loss": 1.5636,
      "num_input_tokens_seen": 24549280,
      "step": 11750
    },
    {
      "epoch": 1.9176931234195285,
      "grad_norm": 5.40625,
      "learning_rate": 4.0082635662563886e-05,
      "loss": 2.5178,
      "num_input_tokens_seen": 24559536,
      "step": 11755
    },
    {
      "epoch": 1.9185088506403458,
      "grad_norm": 7.78125,
      "learning_rate": 4.007480496145331e-05,
      "loss": 3.2025,
      "num_input_tokens_seen": 24571120,
      "step": 11760
    },
    {
      "epoch": 1.9193245778611632,
      "grad_norm": 8.0625,
      "learning_rate": 4.006697193561837e-05,
      "loss": 1.9617,
      "num_input_tokens_seen": 24582864,
      "step": 11765
    },
    {
      "epoch": 1.9201403050819805,
      "grad_norm": 5.6875,
      "learning_rate": 4.005913658626701e-05,
      "loss": 2.34,
      "num_input_tokens_seen": 24592016,
      "step": 11770
    },
    {
      "epoch": 1.920956032302798,
      "grad_norm": 6.40625,
      "learning_rate": 4.005129891460754e-05,
      "loss": 2.3168,
      "num_input_tokens_seen": 24603728,
      "step": 11775
    },
    {
      "epoch": 1.9217717595236152,
      "grad_norm": 3.140625,
      "learning_rate": 4.004345892184864e-05,
      "loss": 1.9279,
      "num_input_tokens_seen": 24612416,
      "step": 11780
    },
    {
      "epoch": 1.9225874867444328,
      "grad_norm": 7.0625,
      "learning_rate": 4.003561660919932e-05,
      "loss": 1.7097,
      "num_input_tokens_seen": 24624080,
      "step": 11785
    },
    {
      "epoch": 1.92340321396525,
      "grad_norm": 5.78125,
      "learning_rate": 4.002777197786897e-05,
      "loss": 1.1837,
      "num_input_tokens_seen": 24633392,
      "step": 11790
    },
    {
      "epoch": 1.9242189411860675,
      "grad_norm": 11.375,
      "learning_rate": 4.0019925029067326e-05,
      "loss": 1.5724,
      "num_input_tokens_seen": 24643712,
      "step": 11795
    },
    {
      "epoch": 1.9250346684068846,
      "grad_norm": 6.40625,
      "learning_rate": 4.0012075764004495e-05,
      "loss": 2.077,
      "num_input_tokens_seen": 24655408,
      "step": 11800
    },
    {
      "epoch": 1.9250346684068846,
      "eval_loss": 2.532294511795044,
      "eval_runtime": 134.8739,
      "eval_samples_per_second": 20.204,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 24655408,
      "step": 11800
    },
    {
      "epoch": 1.9258503956277022,
      "grad_norm": 18.25,
      "learning_rate": 4.000422418389094e-05,
      "loss": 2.2146,
      "num_input_tokens_seen": 24666144,
      "step": 11805
    },
    {
      "epoch": 1.9266661228485193,
      "grad_norm": 6.46875,
      "learning_rate": 3.999637028993744e-05,
      "loss": 3.2053,
      "num_input_tokens_seen": 24677264,
      "step": 11810
    },
    {
      "epoch": 1.927481850069337,
      "grad_norm": 7.6875,
      "learning_rate": 3.99885140833552e-05,
      "loss": 2.0437,
      "num_input_tokens_seen": 24688864,
      "step": 11815
    },
    {
      "epoch": 1.928297577290154,
      "grad_norm": 12.25,
      "learning_rate": 3.998065556535572e-05,
      "loss": 2.5752,
      "num_input_tokens_seen": 24697600,
      "step": 11820
    },
    {
      "epoch": 1.9291133045109716,
      "grad_norm": 7.90625,
      "learning_rate": 3.9972794737150895e-05,
      "loss": 2.7432,
      "num_input_tokens_seen": 24708608,
      "step": 11825
    },
    {
      "epoch": 1.9299290317317888,
      "grad_norm": 6.9375,
      "learning_rate": 3.996493159995297e-05,
      "loss": 3.82,
      "num_input_tokens_seen": 24718944,
      "step": 11830
    },
    {
      "epoch": 1.9307447589526063,
      "grad_norm": 5.125,
      "learning_rate": 3.995706615497453e-05,
      "loss": 2.5464,
      "num_input_tokens_seen": 24729152,
      "step": 11835
    },
    {
      "epoch": 1.9315604861734235,
      "grad_norm": 10.1875,
      "learning_rate": 3.994919840342852e-05,
      "loss": 2.9565,
      "num_input_tokens_seen": 24739152,
      "step": 11840
    },
    {
      "epoch": 1.932376213394241,
      "grad_norm": 2.71875,
      "learning_rate": 3.994132834652825e-05,
      "loss": 2.8477,
      "num_input_tokens_seen": 24748384,
      "step": 11845
    },
    {
      "epoch": 1.9331919406150582,
      "grad_norm": 6.5625,
      "learning_rate": 3.99334559854874e-05,
      "loss": 2.235,
      "num_input_tokens_seen": 24760544,
      "step": 11850
    },
    {
      "epoch": 1.9340076678358757,
      "grad_norm": 5.9375,
      "learning_rate": 3.9925581321519955e-05,
      "loss": 2.7372,
      "num_input_tokens_seen": 24769376,
      "step": 11855
    },
    {
      "epoch": 1.934823395056693,
      "grad_norm": 8.8125,
      "learning_rate": 3.991770435584031e-05,
      "loss": 3.5214,
      "num_input_tokens_seen": 24778816,
      "step": 11860
    },
    {
      "epoch": 1.9356391222775104,
      "grad_norm": 3.734375,
      "learning_rate": 3.990982508966319e-05,
      "loss": 1.3603,
      "num_input_tokens_seen": 24789696,
      "step": 11865
    },
    {
      "epoch": 1.9364548494983278,
      "grad_norm": 10.875,
      "learning_rate": 3.990194352420367e-05,
      "loss": 2.0686,
      "num_input_tokens_seen": 24799536,
      "step": 11870
    },
    {
      "epoch": 1.9372705767191452,
      "grad_norm": 6.3125,
      "learning_rate": 3.9894059660677184e-05,
      "loss": 2.3669,
      "num_input_tokens_seen": 24809232,
      "step": 11875
    },
    {
      "epoch": 1.9380863039399625,
      "grad_norm": 11.25,
      "learning_rate": 3.9886173500299526e-05,
      "loss": 3.0545,
      "num_input_tokens_seen": 24819872,
      "step": 11880
    },
    {
      "epoch": 1.9389020311607799,
      "grad_norm": 10.1875,
      "learning_rate": 3.987828504428685e-05,
      "loss": 2.841,
      "num_input_tokens_seen": 24831072,
      "step": 11885
    },
    {
      "epoch": 1.9397177583815972,
      "grad_norm": 6.71875,
      "learning_rate": 3.987039429385565e-05,
      "loss": 2.1879,
      "num_input_tokens_seen": 24842384,
      "step": 11890
    },
    {
      "epoch": 1.9405334856024146,
      "grad_norm": 4.71875,
      "learning_rate": 3.986250125022277e-05,
      "loss": 2.2361,
      "num_input_tokens_seen": 24852016,
      "step": 11895
    },
    {
      "epoch": 1.941349212823232,
      "grad_norm": 5.875,
      "learning_rate": 3.985460591460544e-05,
      "loss": 2.6951,
      "num_input_tokens_seen": 24862448,
      "step": 11900
    },
    {
      "epoch": 1.9421649400440493,
      "grad_norm": 8.8125,
      "learning_rate": 3.984670828822118e-05,
      "loss": 2.2953,
      "num_input_tokens_seen": 24871696,
      "step": 11905
    },
    {
      "epoch": 1.9429806672648666,
      "grad_norm": 4.21875,
      "learning_rate": 3.983880837228794e-05,
      "loss": 2.6165,
      "num_input_tokens_seen": 24882912,
      "step": 11910
    },
    {
      "epoch": 1.943796394485684,
      "grad_norm": 2.578125,
      "learning_rate": 3.983090616802396e-05,
      "loss": 2.0177,
      "num_input_tokens_seen": 24892256,
      "step": 11915
    },
    {
      "epoch": 1.9446121217065013,
      "grad_norm": 5.3125,
      "learning_rate": 3.982300167664788e-05,
      "loss": 2.2415,
      "num_input_tokens_seen": 24902048,
      "step": 11920
    },
    {
      "epoch": 1.9454278489273187,
      "grad_norm": 8.8125,
      "learning_rate": 3.981509489937868e-05,
      "loss": 2.2327,
      "num_input_tokens_seen": 24912096,
      "step": 11925
    },
    {
      "epoch": 1.946243576148136,
      "grad_norm": 9.8125,
      "learning_rate": 3.9807185837435643e-05,
      "loss": 2.5083,
      "num_input_tokens_seen": 24923632,
      "step": 11930
    },
    {
      "epoch": 1.9470593033689534,
      "grad_norm": 8.125,
      "learning_rate": 3.9799274492038484e-05,
      "loss": 1.4533,
      "num_input_tokens_seen": 24933408,
      "step": 11935
    },
    {
      "epoch": 1.9478750305897707,
      "grad_norm": 5.21875,
      "learning_rate": 3.979136086440722e-05,
      "loss": 2.7647,
      "num_input_tokens_seen": 24944400,
      "step": 11940
    },
    {
      "epoch": 1.948690757810588,
      "grad_norm": 2.890625,
      "learning_rate": 3.9783444955762226e-05,
      "loss": 1.7154,
      "num_input_tokens_seen": 24955440,
      "step": 11945
    },
    {
      "epoch": 1.9495064850314054,
      "grad_norm": 11.5,
      "learning_rate": 3.977552676732424e-05,
      "loss": 1.8168,
      "num_input_tokens_seen": 24966176,
      "step": 11950
    },
    {
      "epoch": 1.9503222122522228,
      "grad_norm": 3.78125,
      "learning_rate": 3.976760630031435e-05,
      "loss": 1.6403,
      "num_input_tokens_seen": 24976736,
      "step": 11955
    },
    {
      "epoch": 1.9511379394730404,
      "grad_norm": 6.875,
      "learning_rate": 3.975968355595398e-05,
      "loss": 2.311,
      "num_input_tokens_seen": 24985568,
      "step": 11960
    },
    {
      "epoch": 1.9519536666938575,
      "grad_norm": 6.53125,
      "learning_rate": 3.9751758535464935e-05,
      "loss": 1.1853,
      "num_input_tokens_seen": 24996992,
      "step": 11965
    },
    {
      "epoch": 1.952769393914675,
      "grad_norm": 6.3125,
      "learning_rate": 3.9743831240069326e-05,
      "loss": 2.3758,
      "num_input_tokens_seen": 25008896,
      "step": 11970
    },
    {
      "epoch": 1.9535851211354922,
      "grad_norm": 10.4375,
      "learning_rate": 3.9735901670989675e-05,
      "loss": 2.9113,
      "num_input_tokens_seen": 25020304,
      "step": 11975
    },
    {
      "epoch": 1.9544008483563098,
      "grad_norm": 8.25,
      "learning_rate": 3.97279698294488e-05,
      "loss": 2.9867,
      "num_input_tokens_seen": 25030384,
      "step": 11980
    },
    {
      "epoch": 1.955216575577127,
      "grad_norm": 6.34375,
      "learning_rate": 3.9720035716669876e-05,
      "loss": 1.8336,
      "num_input_tokens_seen": 25040608,
      "step": 11985
    },
    {
      "epoch": 1.9560323027979445,
      "grad_norm": 7.375,
      "learning_rate": 3.9712099333876474e-05,
      "loss": 1.9539,
      "num_input_tokens_seen": 25051680,
      "step": 11990
    },
    {
      "epoch": 1.9568480300187616,
      "grad_norm": 11.875,
      "learning_rate": 3.9704160682292475e-05,
      "loss": 2.7914,
      "num_input_tokens_seen": 25062624,
      "step": 11995
    },
    {
      "epoch": 1.9576637572395792,
      "grad_norm": 5.59375,
      "learning_rate": 3.9696219763142106e-05,
      "loss": 2.8321,
      "num_input_tokens_seen": 25074096,
      "step": 12000
    },
    {
      "epoch": 1.9576637572395792,
      "eval_loss": 2.5416414737701416,
      "eval_runtime": 134.7761,
      "eval_samples_per_second": 20.219,
      "eval_steps_per_second": 10.113,
      "num_input_tokens_seen": 25074096,
      "step": 12000
    },
    {
      "epoch": 1.9584794844603963,
      "grad_norm": 10.125,
      "learning_rate": 3.968827657764997e-05,
      "loss": 3.2021,
      "num_input_tokens_seen": 25084000,
      "step": 12005
    },
    {
      "epoch": 1.959295211681214,
      "grad_norm": 11.4375,
      "learning_rate": 3.9680331127041e-05,
      "loss": 2.1219,
      "num_input_tokens_seen": 25093232,
      "step": 12010
    },
    {
      "epoch": 1.960110938902031,
      "grad_norm": 1.8828125,
      "learning_rate": 3.9672383412540495e-05,
      "loss": 2.0452,
      "num_input_tokens_seen": 25103312,
      "step": 12015
    },
    {
      "epoch": 1.9609266661228486,
      "grad_norm": 6.125,
      "learning_rate": 3.966443343537407e-05,
      "loss": 2.8641,
      "num_input_tokens_seen": 25113136,
      "step": 12020
    },
    {
      "epoch": 1.9617423933436657,
      "grad_norm": 8.1875,
      "learning_rate": 3.965648119676772e-05,
      "loss": 2.1686,
      "num_input_tokens_seen": 25124128,
      "step": 12025
    },
    {
      "epoch": 1.9625581205644833,
      "grad_norm": 6.5625,
      "learning_rate": 3.96485266979478e-05,
      "loss": 1.6585,
      "num_input_tokens_seen": 25134960,
      "step": 12030
    },
    {
      "epoch": 1.9633738477853004,
      "grad_norm": 10.375,
      "learning_rate": 3.9640569940140974e-05,
      "loss": 2.6192,
      "num_input_tokens_seen": 25146896,
      "step": 12035
    },
    {
      "epoch": 1.964189575006118,
      "grad_norm": 4.65625,
      "learning_rate": 3.963261092457428e-05,
      "loss": 1.6255,
      "num_input_tokens_seen": 25157648,
      "step": 12040
    },
    {
      "epoch": 1.9650053022269351,
      "grad_norm": 4.625,
      "learning_rate": 3.962464965247509e-05,
      "loss": 2.2282,
      "num_input_tokens_seen": 25169040,
      "step": 12045
    },
    {
      "epoch": 1.9658210294477527,
      "grad_norm": 10.8125,
      "learning_rate": 3.9616686125071135e-05,
      "loss": 3.2049,
      "num_input_tokens_seen": 25178944,
      "step": 12050
    },
    {
      "epoch": 1.96663675666857,
      "grad_norm": 9.5625,
      "learning_rate": 3.9608720343590506e-05,
      "loss": 2.4419,
      "num_input_tokens_seen": 25187888,
      "step": 12055
    },
    {
      "epoch": 1.9674524838893874,
      "grad_norm": 11.5625,
      "learning_rate": 3.960075230926161e-05,
      "loss": 3.2146,
      "num_input_tokens_seen": 25199440,
      "step": 12060
    },
    {
      "epoch": 1.9682682111102048,
      "grad_norm": 4.125,
      "learning_rate": 3.959278202331322e-05,
      "loss": 2.3154,
      "num_input_tokens_seen": 25209616,
      "step": 12065
    },
    {
      "epoch": 1.9690839383310221,
      "grad_norm": 8.5625,
      "learning_rate": 3.958480948697446e-05,
      "loss": 3.2495,
      "num_input_tokens_seen": 25220544,
      "step": 12070
    },
    {
      "epoch": 1.9698996655518395,
      "grad_norm": 7.28125,
      "learning_rate": 3.95768347014748e-05,
      "loss": 1.944,
      "num_input_tokens_seen": 25231424,
      "step": 12075
    },
    {
      "epoch": 1.9707153927726568,
      "grad_norm": 6.1875,
      "learning_rate": 3.956885766804404e-05,
      "loss": 2.4192,
      "num_input_tokens_seen": 25241712,
      "step": 12080
    },
    {
      "epoch": 1.9715311199934742,
      "grad_norm": 6.6875,
      "learning_rate": 3.956087838791235e-05,
      "loss": 2.8942,
      "num_input_tokens_seen": 25252192,
      "step": 12085
    },
    {
      "epoch": 1.9723468472142915,
      "grad_norm": 11.6875,
      "learning_rate": 3.955289686231022e-05,
      "loss": 2.9327,
      "num_input_tokens_seen": 25262416,
      "step": 12090
    },
    {
      "epoch": 1.973162574435109,
      "grad_norm": 6.1875,
      "learning_rate": 3.9544913092468504e-05,
      "loss": 3.3473,
      "num_input_tokens_seen": 25272464,
      "step": 12095
    },
    {
      "epoch": 1.9739783016559262,
      "grad_norm": 6.5625,
      "learning_rate": 3.9536927079618425e-05,
      "loss": 1.9518,
      "num_input_tokens_seen": 25283248,
      "step": 12100
    },
    {
      "epoch": 1.9747940288767436,
      "grad_norm": 0.16015625,
      "learning_rate": 3.9528938824991494e-05,
      "loss": 2.5185,
      "num_input_tokens_seen": 25292512,
      "step": 12105
    },
    {
      "epoch": 1.975609756097561,
      "grad_norm": 10.1875,
      "learning_rate": 3.952094832981962e-05,
      "loss": 2.4884,
      "num_input_tokens_seen": 25303888,
      "step": 12110
    },
    {
      "epoch": 1.9764254833183783,
      "grad_norm": 3.734375,
      "learning_rate": 3.951295559533503e-05,
      "loss": 3.4112,
      "num_input_tokens_seen": 25314464,
      "step": 12115
    },
    {
      "epoch": 1.9772412105391957,
      "grad_norm": 3.46875,
      "learning_rate": 3.95049606227703e-05,
      "loss": 3.3551,
      "num_input_tokens_seen": 25324080,
      "step": 12120
    },
    {
      "epoch": 1.978056937760013,
      "grad_norm": 7.875,
      "learning_rate": 3.949696341335838e-05,
      "loss": 2.7883,
      "num_input_tokens_seen": 25334816,
      "step": 12125
    },
    {
      "epoch": 1.9788726649808304,
      "grad_norm": 7.90625,
      "learning_rate": 3.9488963968332503e-05,
      "loss": 3.4786,
      "num_input_tokens_seen": 25345280,
      "step": 12130
    },
    {
      "epoch": 1.9796883922016477,
      "grad_norm": 5.90625,
      "learning_rate": 3.948096228892631e-05,
      "loss": 2.2978,
      "num_input_tokens_seen": 25356624,
      "step": 12135
    },
    {
      "epoch": 1.980504119422465,
      "grad_norm": 10.125,
      "learning_rate": 3.947295837637375e-05,
      "loss": 3.4422,
      "num_input_tokens_seen": 25367376,
      "step": 12140
    },
    {
      "epoch": 1.9813198466432826,
      "grad_norm": 5.90625,
      "learning_rate": 3.9464952231909135e-05,
      "loss": 2.3039,
      "num_input_tokens_seen": 25377584,
      "step": 12145
    },
    {
      "epoch": 1.9821355738640998,
      "grad_norm": 7.375,
      "learning_rate": 3.945694385676711e-05,
      "loss": 1.616,
      "num_input_tokens_seen": 25386672,
      "step": 12150
    },
    {
      "epoch": 1.9829513010849174,
      "grad_norm": 16.375,
      "learning_rate": 3.944893325218265e-05,
      "loss": 2.4857,
      "num_input_tokens_seen": 25396320,
      "step": 12155
    },
    {
      "epoch": 1.9837670283057345,
      "grad_norm": 5.53125,
      "learning_rate": 3.944092041939112e-05,
      "loss": 1.5507,
      "num_input_tokens_seen": 25406208,
      "step": 12160
    },
    {
      "epoch": 1.984582755526552,
      "grad_norm": 8.0,
      "learning_rate": 3.943290535962818e-05,
      "loss": 2.3132,
      "num_input_tokens_seen": 25417344,
      "step": 12165
    },
    {
      "epoch": 1.9853984827473692,
      "grad_norm": 4.21875,
      "learning_rate": 3.942488807412985e-05,
      "loss": 2.2424,
      "num_input_tokens_seen": 25427200,
      "step": 12170
    },
    {
      "epoch": 1.9862142099681868,
      "grad_norm": 10.5625,
      "learning_rate": 3.941686856413251e-05,
      "loss": 3.1834,
      "num_input_tokens_seen": 25437872,
      "step": 12175
    },
    {
      "epoch": 1.987029937189004,
      "grad_norm": 6.90625,
      "learning_rate": 3.9408846830872874e-05,
      "loss": 1.4161,
      "num_input_tokens_seen": 25447520,
      "step": 12180
    },
    {
      "epoch": 1.9878456644098215,
      "grad_norm": 3.765625,
      "learning_rate": 3.940082287558798e-05,
      "loss": 2.4365,
      "num_input_tokens_seen": 25458304,
      "step": 12185
    },
    {
      "epoch": 1.9886613916306386,
      "grad_norm": 7.125,
      "learning_rate": 3.939279669951522e-05,
      "loss": 2.1,
      "num_input_tokens_seen": 25467952,
      "step": 12190
    },
    {
      "epoch": 1.9894771188514562,
      "grad_norm": 17.5,
      "learning_rate": 3.938476830389234e-05,
      "loss": 1.7848,
      "num_input_tokens_seen": 25478480,
      "step": 12195
    },
    {
      "epoch": 1.9902928460722733,
      "grad_norm": 6.40625,
      "learning_rate": 3.937673768995742e-05,
      "loss": 1.0373,
      "num_input_tokens_seen": 25489056,
      "step": 12200
    },
    {
      "epoch": 1.9902928460722733,
      "eval_loss": 2.535943031311035,
      "eval_runtime": 134.7977,
      "eval_samples_per_second": 20.215,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 25489056,
      "step": 12200
    },
    {
      "epoch": 1.9911085732930909,
      "grad_norm": 8.1875,
      "learning_rate": 3.936870485894888e-05,
      "loss": 2.4247,
      "num_input_tokens_seen": 25498816,
      "step": 12205
    },
    {
      "epoch": 1.991924300513908,
      "grad_norm": 3.328125,
      "learning_rate": 3.9360669812105475e-05,
      "loss": 1.4053,
      "num_input_tokens_seen": 25509696,
      "step": 12210
    },
    {
      "epoch": 1.9927400277347256,
      "grad_norm": 5.40625,
      "learning_rate": 3.9352632550666325e-05,
      "loss": 1.7624,
      "num_input_tokens_seen": 25519360,
      "step": 12215
    },
    {
      "epoch": 1.9935557549555427,
      "grad_norm": 5.4375,
      "learning_rate": 3.9344593075870866e-05,
      "loss": 1.7833,
      "num_input_tokens_seen": 25530288,
      "step": 12220
    },
    {
      "epoch": 1.9943714821763603,
      "grad_norm": 4.9375,
      "learning_rate": 3.933655138895889e-05,
      "loss": 1.2549,
      "num_input_tokens_seen": 25539632,
      "step": 12225
    },
    {
      "epoch": 1.9951872093971774,
      "grad_norm": 5.34375,
      "learning_rate": 3.932850749117053e-05,
      "loss": 3.268,
      "num_input_tokens_seen": 25549744,
      "step": 12230
    },
    {
      "epoch": 1.996002936617995,
      "grad_norm": 9.9375,
      "learning_rate": 3.932046138374624e-05,
      "loss": 1.9572,
      "num_input_tokens_seen": 25558864,
      "step": 12235
    },
    {
      "epoch": 1.9968186638388123,
      "grad_norm": 9.9375,
      "learning_rate": 3.9312413067926854e-05,
      "loss": 2.6345,
      "num_input_tokens_seen": 25568784,
      "step": 12240
    },
    {
      "epoch": 1.9976343910596297,
      "grad_norm": 7.875,
      "learning_rate": 3.9304362544953506e-05,
      "loss": 3.4014,
      "num_input_tokens_seen": 25578816,
      "step": 12245
    },
    {
      "epoch": 1.998450118280447,
      "grad_norm": 2.328125,
      "learning_rate": 3.929630981606769e-05,
      "loss": 2.5976,
      "num_input_tokens_seen": 25589696,
      "step": 12250
    },
    {
      "epoch": 1.9992658455012644,
      "grad_norm": 9.875,
      "learning_rate": 3.928825488251124e-05,
      "loss": 2.1884,
      "num_input_tokens_seen": 25601184,
      "step": 12255
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.06689453125,
      "learning_rate": 3.9280197745526344e-05,
      "loss": 1.8509,
      "num_input_tokens_seen": 25610688,
      "step": 12260
    },
    {
      "epoch": 2.0008157272208176,
      "grad_norm": 5.53125,
      "learning_rate": 3.9272138406355495e-05,
      "loss": 2.2133,
      "num_input_tokens_seen": 25620368,
      "step": 12265
    },
    {
      "epoch": 2.0016314544416347,
      "grad_norm": 4.5625,
      "learning_rate": 3.926407686624154e-05,
      "loss": 3.1083,
      "num_input_tokens_seen": 25629664,
      "step": 12270
    },
    {
      "epoch": 2.0024471816624523,
      "grad_norm": 2.875,
      "learning_rate": 3.9256013126427684e-05,
      "loss": 1.4889,
      "num_input_tokens_seen": 25640880,
      "step": 12275
    },
    {
      "epoch": 2.0032629088832694,
      "grad_norm": 5.03125,
      "learning_rate": 3.9247947188157455e-05,
      "loss": 2.3403,
      "num_input_tokens_seen": 25650480,
      "step": 12280
    },
    {
      "epoch": 2.004078636104087,
      "grad_norm": 5.65625,
      "learning_rate": 3.9239879052674715e-05,
      "loss": 1.2824,
      "num_input_tokens_seen": 25659424,
      "step": 12285
    },
    {
      "epoch": 2.004894363324904,
      "grad_norm": 11.4375,
      "learning_rate": 3.9231808721223673e-05,
      "loss": 2.865,
      "num_input_tokens_seen": 25669408,
      "step": 12290
    },
    {
      "epoch": 2.0057100905457217,
      "grad_norm": 1.5625,
      "learning_rate": 3.9223736195048886e-05,
      "loss": 2.9898,
      "num_input_tokens_seen": 25678880,
      "step": 12295
    },
    {
      "epoch": 2.006525817766539,
      "grad_norm": 7.53125,
      "learning_rate": 3.921566147539523e-05,
      "loss": 1.8816,
      "num_input_tokens_seen": 25689760,
      "step": 12300
    },
    {
      "epoch": 2.0073415449873564,
      "grad_norm": 7.625,
      "learning_rate": 3.920758456350792e-05,
      "loss": 2.2706,
      "num_input_tokens_seen": 25700656,
      "step": 12305
    },
    {
      "epoch": 2.0081572722081735,
      "grad_norm": 4.90625,
      "learning_rate": 3.919950546063253e-05,
      "loss": 1.9157,
      "num_input_tokens_seen": 25711920,
      "step": 12310
    },
    {
      "epoch": 2.008972999428991,
      "grad_norm": 5.28125,
      "learning_rate": 3.919142416801496e-05,
      "loss": 2.9542,
      "num_input_tokens_seen": 25721008,
      "step": 12315
    },
    {
      "epoch": 2.0097887266498082,
      "grad_norm": 6.8125,
      "learning_rate": 3.918334068690144e-05,
      "loss": 2.6724,
      "num_input_tokens_seen": 25732000,
      "step": 12320
    },
    {
      "epoch": 2.010604453870626,
      "grad_norm": 4.53125,
      "learning_rate": 3.917525501853855e-05,
      "loss": 1.734,
      "num_input_tokens_seen": 25742368,
      "step": 12325
    },
    {
      "epoch": 2.011420181091443,
      "grad_norm": 5.53125,
      "learning_rate": 3.916716716417319e-05,
      "loss": 1.4206,
      "num_input_tokens_seen": 25752880,
      "step": 12330
    },
    {
      "epoch": 2.0122359083122605,
      "grad_norm": 6.8125,
      "learning_rate": 3.915907712505263e-05,
      "loss": 2.1621,
      "num_input_tokens_seen": 25761968,
      "step": 12335
    },
    {
      "epoch": 2.0130516355330776,
      "grad_norm": 7.78125,
      "learning_rate": 3.915098490242444e-05,
      "loss": 2.5838,
      "num_input_tokens_seen": 25772688,
      "step": 12340
    },
    {
      "epoch": 2.013867362753895,
      "grad_norm": 7.53125,
      "learning_rate": 3.914289049753654e-05,
      "loss": 2.812,
      "num_input_tokens_seen": 25783792,
      "step": 12345
    },
    {
      "epoch": 2.0146830899747123,
      "grad_norm": 10.125,
      "learning_rate": 3.913479391163719e-05,
      "loss": 3.0933,
      "num_input_tokens_seen": 25794128,
      "step": 12350
    },
    {
      "epoch": 2.01549881719553,
      "grad_norm": 3.4375,
      "learning_rate": 3.9126695145975e-05,
      "loss": 0.9704,
      "num_input_tokens_seen": 25805056,
      "step": 12355
    },
    {
      "epoch": 2.016314544416347,
      "grad_norm": 2.8125,
      "learning_rate": 3.911859420179889e-05,
      "loss": 2.623,
      "num_input_tokens_seen": 25814320,
      "step": 12360
    },
    {
      "epoch": 2.0171302716371646,
      "grad_norm": 4.0625,
      "learning_rate": 3.911049108035813e-05,
      "loss": 2.7762,
      "num_input_tokens_seen": 25825520,
      "step": 12365
    },
    {
      "epoch": 2.0179459988579818,
      "grad_norm": 2.609375,
      "learning_rate": 3.910238578290232e-05,
      "loss": 1.5594,
      "num_input_tokens_seen": 25836208,
      "step": 12370
    },
    {
      "epoch": 2.0187617260787993,
      "grad_norm": 6.21875,
      "learning_rate": 3.90942783106814e-05,
      "loss": 2.3027,
      "num_input_tokens_seen": 25844720,
      "step": 12375
    },
    {
      "epoch": 2.0195774532996165,
      "grad_norm": 2.46875,
      "learning_rate": 3.908616866494564e-05,
      "loss": 1.9183,
      "num_input_tokens_seen": 25854864,
      "step": 12380
    },
    {
      "epoch": 2.020393180520434,
      "grad_norm": 5.375,
      "learning_rate": 3.907805684694566e-05,
      "loss": 2.7487,
      "num_input_tokens_seen": 25865968,
      "step": 12385
    },
    {
      "epoch": 2.021208907741251,
      "grad_norm": 3.265625,
      "learning_rate": 3.90699428579324e-05,
      "loss": 0.5792,
      "num_input_tokens_seen": 25876128,
      "step": 12390
    },
    {
      "epoch": 2.0220246349620687,
      "grad_norm": 9.375,
      "learning_rate": 3.906182669915713e-05,
      "loss": 3.8842,
      "num_input_tokens_seen": 25886736,
      "step": 12395
    },
    {
      "epoch": 2.022840362182886,
      "grad_norm": 3.140625,
      "learning_rate": 3.9053708371871476e-05,
      "loss": 1.368,
      "num_input_tokens_seen": 25898496,
      "step": 12400
    },
    {
      "epoch": 2.022840362182886,
      "eval_loss": 2.5435256958007812,
      "eval_runtime": 135.048,
      "eval_samples_per_second": 20.178,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 25898496,
      "step": 12400
    },
    {
      "epoch": 2.0236560894037035,
      "grad_norm": 9.25,
      "learning_rate": 3.904558787732738e-05,
      "loss": 2.179,
      "num_input_tokens_seen": 25908560,
      "step": 12405
    },
    {
      "epoch": 2.0244718166245206,
      "grad_norm": 3.203125,
      "learning_rate": 3.9037465216777135e-05,
      "loss": 1.9731,
      "num_input_tokens_seen": 25919824,
      "step": 12410
    },
    {
      "epoch": 2.025287543845338,
      "grad_norm": 9.5,
      "learning_rate": 3.902934039147334e-05,
      "loss": 2.8509,
      "num_input_tokens_seen": 25929968,
      "step": 12415
    },
    {
      "epoch": 2.0261032710661553,
      "grad_norm": 5.125,
      "learning_rate": 3.902121340266894e-05,
      "loss": 2.4208,
      "num_input_tokens_seen": 25941216,
      "step": 12420
    },
    {
      "epoch": 2.026918998286973,
      "grad_norm": 9.8125,
      "learning_rate": 3.9013084251617246e-05,
      "loss": 3.5501,
      "num_input_tokens_seen": 25950816,
      "step": 12425
    },
    {
      "epoch": 2.02773472550779,
      "grad_norm": 8.4375,
      "learning_rate": 3.9004952939571865e-05,
      "loss": 1.962,
      "num_input_tokens_seen": 25962336,
      "step": 12430
    },
    {
      "epoch": 2.0285504527286076,
      "grad_norm": 10.1875,
      "learning_rate": 3.899681946778673e-05,
      "loss": 3.5948,
      "num_input_tokens_seen": 25971072,
      "step": 12435
    },
    {
      "epoch": 2.0293661799494247,
      "grad_norm": 5.15625,
      "learning_rate": 3.898868383751615e-05,
      "loss": 2.3573,
      "num_input_tokens_seen": 25982448,
      "step": 12440
    },
    {
      "epoch": 2.0301819071702423,
      "grad_norm": 10.4375,
      "learning_rate": 3.8980546050014724e-05,
      "loss": 2.226,
      "num_input_tokens_seen": 25993344,
      "step": 12445
    },
    {
      "epoch": 2.03099763439106,
      "grad_norm": 4.90625,
      "learning_rate": 3.897240610653741e-05,
      "loss": 1.801,
      "num_input_tokens_seen": 26004288,
      "step": 12450
    },
    {
      "epoch": 2.031813361611877,
      "grad_norm": 13.5625,
      "learning_rate": 3.896426400833948e-05,
      "loss": 2.6098,
      "num_input_tokens_seen": 26014928,
      "step": 12455
    },
    {
      "epoch": 2.0326290888326946,
      "grad_norm": 3.4375,
      "learning_rate": 3.895611975667656e-05,
      "loss": 3.965,
      "num_input_tokens_seen": 26025088,
      "step": 12460
    },
    {
      "epoch": 2.0334448160535117,
      "grad_norm": 4.71875,
      "learning_rate": 3.8947973352804584e-05,
      "loss": 2.8017,
      "num_input_tokens_seen": 26036816,
      "step": 12465
    },
    {
      "epoch": 2.0342605432743293,
      "grad_norm": 10.125,
      "learning_rate": 3.893982479797984e-05,
      "loss": 4.9591,
      "num_input_tokens_seen": 26047488,
      "step": 12470
    },
    {
      "epoch": 2.0350762704951464,
      "grad_norm": 8.375,
      "learning_rate": 3.8931674093458926e-05,
      "loss": 3.0106,
      "num_input_tokens_seen": 26058448,
      "step": 12475
    },
    {
      "epoch": 2.035891997715964,
      "grad_norm": 6.46875,
      "learning_rate": 3.89235212404988e-05,
      "loss": 2.3259,
      "num_input_tokens_seen": 26069760,
      "step": 12480
    },
    {
      "epoch": 2.036707724936781,
      "grad_norm": 4.5,
      "learning_rate": 3.891536624035672e-05,
      "loss": 1.1691,
      "num_input_tokens_seen": 26079680,
      "step": 12485
    },
    {
      "epoch": 2.0375234521575987,
      "grad_norm": 4.28125,
      "learning_rate": 3.8907209094290295e-05,
      "loss": 2.4451,
      "num_input_tokens_seen": 26090800,
      "step": 12490
    },
    {
      "epoch": 2.038339179378416,
      "grad_norm": 1.8828125,
      "learning_rate": 3.8899049803557466e-05,
      "loss": 1.0834,
      "num_input_tokens_seen": 26100384,
      "step": 12495
    },
    {
      "epoch": 2.0391549065992334,
      "grad_norm": 6.15625,
      "learning_rate": 3.889088836941648e-05,
      "loss": 3.3338,
      "num_input_tokens_seen": 26109712,
      "step": 12500
    },
    {
      "epoch": 2.0399706338200505,
      "grad_norm": 8.125,
      "learning_rate": 3.8882724793125946e-05,
      "loss": 2.2804,
      "num_input_tokens_seen": 26121536,
      "step": 12505
    },
    {
      "epoch": 2.040786361040868,
      "grad_norm": 12.9375,
      "learning_rate": 3.8874559075944794e-05,
      "loss": 1.2882,
      "num_input_tokens_seen": 26133024,
      "step": 12510
    },
    {
      "epoch": 2.041602088261685,
      "grad_norm": 9.125,
      "learning_rate": 3.886639121913227e-05,
      "loss": 1.6742,
      "num_input_tokens_seen": 26143376,
      "step": 12515
    },
    {
      "epoch": 2.042417815482503,
      "grad_norm": 10.375,
      "learning_rate": 3.885822122394797e-05,
      "loss": 2.8512,
      "num_input_tokens_seen": 26154608,
      "step": 12520
    },
    {
      "epoch": 2.04323354270332,
      "grad_norm": 10.25,
      "learning_rate": 3.8850049091651794e-05,
      "loss": 2.3516,
      "num_input_tokens_seen": 26165248,
      "step": 12525
    },
    {
      "epoch": 2.0440492699241375,
      "grad_norm": 4.625,
      "learning_rate": 3.8841874823504e-05,
      "loss": 1.7105,
      "num_input_tokens_seen": 26175952,
      "step": 12530
    },
    {
      "epoch": 2.0448649971449546,
      "grad_norm": 9.375,
      "learning_rate": 3.8833698420765157e-05,
      "loss": 2.4398,
      "num_input_tokens_seen": 26185984,
      "step": 12535
    },
    {
      "epoch": 2.045680724365772,
      "grad_norm": 8.0625,
      "learning_rate": 3.882551988469618e-05,
      "loss": 1.8864,
      "num_input_tokens_seen": 26196656,
      "step": 12540
    },
    {
      "epoch": 2.0464964515865893,
      "grad_norm": 7.59375,
      "learning_rate": 3.881733921655829e-05,
      "loss": 2.3541,
      "num_input_tokens_seen": 26206848,
      "step": 12545
    },
    {
      "epoch": 2.047312178807407,
      "grad_norm": 9.125,
      "learning_rate": 3.8809156417613054e-05,
      "loss": 2.796,
      "num_input_tokens_seen": 26216768,
      "step": 12550
    },
    {
      "epoch": 2.048127906028224,
      "grad_norm": 3.515625,
      "learning_rate": 3.8800971489122364e-05,
      "loss": 1.947,
      "num_input_tokens_seen": 26226320,
      "step": 12555
    },
    {
      "epoch": 2.0489436332490416,
      "grad_norm": 10.3125,
      "learning_rate": 3.8792784432348434e-05,
      "loss": 2.4148,
      "num_input_tokens_seen": 26236448,
      "step": 12560
    },
    {
      "epoch": 2.0497593604698587,
      "grad_norm": 7.9375,
      "learning_rate": 3.878459524855381e-05,
      "loss": 2.6365,
      "num_input_tokens_seen": 26247392,
      "step": 12565
    },
    {
      "epoch": 2.0505750876906763,
      "grad_norm": 6.34375,
      "learning_rate": 3.8776403939001384e-05,
      "loss": 2.5982,
      "num_input_tokens_seen": 26257872,
      "step": 12570
    },
    {
      "epoch": 2.0513908149114934,
      "grad_norm": 0.06640625,
      "learning_rate": 3.876821050495433e-05,
      "loss": 2.456,
      "num_input_tokens_seen": 26267872,
      "step": 12575
    },
    {
      "epoch": 2.052206542132311,
      "grad_norm": 12.25,
      "learning_rate": 3.87600149476762e-05,
      "loss": 4.0875,
      "num_input_tokens_seen": 26279088,
      "step": 12580
    },
    {
      "epoch": 2.053022269353128,
      "grad_norm": 4.53125,
      "learning_rate": 3.8751817268430843e-05,
      "loss": 1.6644,
      "num_input_tokens_seen": 26289664,
      "step": 12585
    },
    {
      "epoch": 2.0538379965739457,
      "grad_norm": 0.0966796875,
      "learning_rate": 3.8743617468482464e-05,
      "loss": 1.612,
      "num_input_tokens_seen": 26298768,
      "step": 12590
    },
    {
      "epoch": 2.054653723794763,
      "grad_norm": 7.9375,
      "learning_rate": 3.8735415549095535e-05,
      "loss": 2.4845,
      "num_input_tokens_seen": 26309216,
      "step": 12595
    },
    {
      "epoch": 2.0554694510155804,
      "grad_norm": 4.4375,
      "learning_rate": 3.8727211511534934e-05,
      "loss": 2.2438,
      "num_input_tokens_seen": 26319696,
      "step": 12600
    },
    {
      "epoch": 2.0554694510155804,
      "eval_loss": 2.54951810836792,
      "eval_runtime": 134.8545,
      "eval_samples_per_second": 20.207,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 26319696,
      "step": 12600
    },
    {
      "epoch": 2.0562851782363976,
      "grad_norm": 8.6875,
      "learning_rate": 3.8719005357065804e-05,
      "loss": 2.1388,
      "num_input_tokens_seen": 26329824,
      "step": 12605
    },
    {
      "epoch": 2.057100905457215,
      "grad_norm": 7.78125,
      "learning_rate": 3.8710797086953645e-05,
      "loss": 2.6026,
      "num_input_tokens_seen": 26340432,
      "step": 12610
    },
    {
      "epoch": 2.0579166326780323,
      "grad_norm": 11.0,
      "learning_rate": 3.870258670246427e-05,
      "loss": 3.012,
      "num_input_tokens_seen": 26351760,
      "step": 12615
    },
    {
      "epoch": 2.05873235989885,
      "grad_norm": 5.46875,
      "learning_rate": 3.869437420486384e-05,
      "loss": 1.6749,
      "num_input_tokens_seen": 26361696,
      "step": 12620
    },
    {
      "epoch": 2.059548087119667,
      "grad_norm": 10.0625,
      "learning_rate": 3.8686159595418805e-05,
      "loss": 2.9085,
      "num_input_tokens_seen": 26370592,
      "step": 12625
    },
    {
      "epoch": 2.0603638143404845,
      "grad_norm": 3.5625,
      "learning_rate": 3.867794287539597e-05,
      "loss": 2.6304,
      "num_input_tokens_seen": 26382208,
      "step": 12630
    },
    {
      "epoch": 2.0611795415613017,
      "grad_norm": 18.125,
      "learning_rate": 3.866972404606245e-05,
      "loss": 1.9655,
      "num_input_tokens_seen": 26392192,
      "step": 12635
    },
    {
      "epoch": 2.0619952687821193,
      "grad_norm": 9.1875,
      "learning_rate": 3.866150310868571e-05,
      "loss": 2.4638,
      "num_input_tokens_seen": 26403872,
      "step": 12640
    },
    {
      "epoch": 2.062810996002937,
      "grad_norm": 8.1875,
      "learning_rate": 3.8653280064533506e-05,
      "loss": 1.692,
      "num_input_tokens_seen": 26415312,
      "step": 12645
    },
    {
      "epoch": 2.063626723223754,
      "grad_norm": 7.21875,
      "learning_rate": 3.864505491487394e-05,
      "loss": 3.9869,
      "num_input_tokens_seen": 26426144,
      "step": 12650
    },
    {
      "epoch": 2.0644424504445715,
      "grad_norm": 2.40625,
      "learning_rate": 3.8636827660975414e-05,
      "loss": 2.282,
      "num_input_tokens_seen": 26436112,
      "step": 12655
    },
    {
      "epoch": 2.0652581776653887,
      "grad_norm": 15.0625,
      "learning_rate": 3.862859830410671e-05,
      "loss": 2.1649,
      "num_input_tokens_seen": 26447408,
      "step": 12660
    },
    {
      "epoch": 2.0660739048862062,
      "grad_norm": 5.5625,
      "learning_rate": 3.862036684553688e-05,
      "loss": 2.5149,
      "num_input_tokens_seen": 26459504,
      "step": 12665
    },
    {
      "epoch": 2.0668896321070234,
      "grad_norm": 8.0,
      "learning_rate": 3.8612133286535314e-05,
      "loss": 2.5804,
      "num_input_tokens_seen": 26469904,
      "step": 12670
    },
    {
      "epoch": 2.067705359327841,
      "grad_norm": 4.375,
      "learning_rate": 3.860389762837173e-05,
      "loss": 1.7207,
      "num_input_tokens_seen": 26479040,
      "step": 12675
    },
    {
      "epoch": 2.068521086548658,
      "grad_norm": 3.28125,
      "learning_rate": 3.859565987231618e-05,
      "loss": 1.9384,
      "num_input_tokens_seen": 26488256,
      "step": 12680
    },
    {
      "epoch": 2.0693368137694756,
      "grad_norm": 5.53125,
      "learning_rate": 3.858742001963902e-05,
      "loss": 2.1978,
      "num_input_tokens_seen": 26500016,
      "step": 12685
    },
    {
      "epoch": 2.0701525409902928,
      "grad_norm": 7.3125,
      "learning_rate": 3.857917807161094e-05,
      "loss": 3.1785,
      "num_input_tokens_seen": 26510464,
      "step": 12690
    },
    {
      "epoch": 2.0709682682111104,
      "grad_norm": 8.3125,
      "learning_rate": 3.857093402950296e-05,
      "loss": 1.6891,
      "num_input_tokens_seen": 26521344,
      "step": 12695
    },
    {
      "epoch": 2.0717839954319275,
      "grad_norm": 14.0,
      "learning_rate": 3.8562687894586414e-05,
      "loss": 2.6856,
      "num_input_tokens_seen": 26532416,
      "step": 12700
    },
    {
      "epoch": 2.072599722652745,
      "grad_norm": 8.4375,
      "learning_rate": 3.8554439668132946e-05,
      "loss": 2.0034,
      "num_input_tokens_seen": 26543520,
      "step": 12705
    },
    {
      "epoch": 2.073415449873562,
      "grad_norm": 7.0,
      "learning_rate": 3.854618935141455e-05,
      "loss": 2.4925,
      "num_input_tokens_seen": 26554352,
      "step": 12710
    },
    {
      "epoch": 2.0742311770943798,
      "grad_norm": 4.96875,
      "learning_rate": 3.8537936945703525e-05,
      "loss": 2.1544,
      "num_input_tokens_seen": 26565488,
      "step": 12715
    },
    {
      "epoch": 2.075046904315197,
      "grad_norm": 11.25,
      "learning_rate": 3.852968245227249e-05,
      "loss": 3.0383,
      "num_input_tokens_seen": 26576080,
      "step": 12720
    },
    {
      "epoch": 2.0758626315360145,
      "grad_norm": 8.75,
      "learning_rate": 3.85214258723944e-05,
      "loss": 2.0196,
      "num_input_tokens_seen": 26585968,
      "step": 12725
    },
    {
      "epoch": 2.0766783587568316,
      "grad_norm": 8.4375,
      "learning_rate": 3.8513167207342524e-05,
      "loss": 2.1878,
      "num_input_tokens_seen": 26597760,
      "step": 12730
    },
    {
      "epoch": 2.077494085977649,
      "grad_norm": 5.1875,
      "learning_rate": 3.850490645839044e-05,
      "loss": 2.8964,
      "num_input_tokens_seen": 26607568,
      "step": 12735
    },
    {
      "epoch": 2.0783098131984663,
      "grad_norm": 11.3125,
      "learning_rate": 3.849664362681207e-05,
      "loss": 2.9705,
      "num_input_tokens_seen": 26616912,
      "step": 12740
    },
    {
      "epoch": 2.079125540419284,
      "grad_norm": 10.0625,
      "learning_rate": 3.848837871388165e-05,
      "loss": 2.3568,
      "num_input_tokens_seen": 26628064,
      "step": 12745
    },
    {
      "epoch": 2.079941267640101,
      "grad_norm": 7.75,
      "learning_rate": 3.848011172087371e-05,
      "loss": 2.1284,
      "num_input_tokens_seen": 26640128,
      "step": 12750
    },
    {
      "epoch": 2.0807569948609186,
      "grad_norm": 3.46875,
      "learning_rate": 3.847184264906315e-05,
      "loss": 1.8572,
      "num_input_tokens_seen": 26650512,
      "step": 12755
    },
    {
      "epoch": 2.0815727220817357,
      "grad_norm": 5.21875,
      "learning_rate": 3.846357149972516e-05,
      "loss": 2.0123,
      "num_input_tokens_seen": 26661488,
      "step": 12760
    },
    {
      "epoch": 2.0823884493025533,
      "grad_norm": 2.703125,
      "learning_rate": 3.8455298274135246e-05,
      "loss": 2.0347,
      "num_input_tokens_seen": 26671744,
      "step": 12765
    },
    {
      "epoch": 2.0832041765233704,
      "grad_norm": 4.125,
      "learning_rate": 3.8447022973569254e-05,
      "loss": 1.7539,
      "num_input_tokens_seen": 26681824,
      "step": 12770
    },
    {
      "epoch": 2.084019903744188,
      "grad_norm": 5.34375,
      "learning_rate": 3.843874559930332e-05,
      "loss": 2.0657,
      "num_input_tokens_seen": 26692400,
      "step": 12775
    },
    {
      "epoch": 2.084835630965005,
      "grad_norm": 6.34375,
      "learning_rate": 3.843046615261394e-05,
      "loss": 1.2089,
      "num_input_tokens_seen": 26703200,
      "step": 12780
    },
    {
      "epoch": 2.0856513581858227,
      "grad_norm": 5.375,
      "learning_rate": 3.842218463477791e-05,
      "loss": 1.5922,
      "num_input_tokens_seen": 26713824,
      "step": 12785
    },
    {
      "epoch": 2.08646708540664,
      "grad_norm": 13.4375,
      "learning_rate": 3.841390104707233e-05,
      "loss": 2.7453,
      "num_input_tokens_seen": 26724832,
      "step": 12790
    },
    {
      "epoch": 2.0872828126274574,
      "grad_norm": 4.9375,
      "learning_rate": 3.8405615390774643e-05,
      "loss": 2.648,
      "num_input_tokens_seen": 26735552,
      "step": 12795
    },
    {
      "epoch": 2.0880985398482745,
      "grad_norm": 5.40625,
      "learning_rate": 3.839732766716259e-05,
      "loss": 2.1512,
      "num_input_tokens_seen": 26744896,
      "step": 12800
    },
    {
      "epoch": 2.0880985398482745,
      "eval_loss": 2.5462207794189453,
      "eval_runtime": 135.1489,
      "eval_samples_per_second": 20.163,
      "eval_steps_per_second": 10.085,
      "num_input_tokens_seen": 26744896,
      "step": 12800
    },
    {
      "epoch": 2.088914267069092,
      "grad_norm": 10.0,
      "learning_rate": 3.838903787751425e-05,
      "loss": 2.6636,
      "num_input_tokens_seen": 26754400,
      "step": 12805
    },
    {
      "epoch": 2.0897299942899092,
      "grad_norm": 5.84375,
      "learning_rate": 3.838074602310802e-05,
      "loss": 1.5369,
      "num_input_tokens_seen": 26765168,
      "step": 12810
    },
    {
      "epoch": 2.090545721510727,
      "grad_norm": 4.03125,
      "learning_rate": 3.837245210522258e-05,
      "loss": 1.29,
      "num_input_tokens_seen": 26775520,
      "step": 12815
    },
    {
      "epoch": 2.0913614487315444,
      "grad_norm": 5.84375,
      "learning_rate": 3.8364156125136996e-05,
      "loss": 2.7491,
      "num_input_tokens_seen": 26784816,
      "step": 12820
    },
    {
      "epoch": 2.0921771759523615,
      "grad_norm": 13.625,
      "learning_rate": 3.835585808413059e-05,
      "loss": 2.2804,
      "num_input_tokens_seen": 26795056,
      "step": 12825
    },
    {
      "epoch": 2.092992903173179,
      "grad_norm": 5.15625,
      "learning_rate": 3.8347557983483024e-05,
      "loss": 1.6828,
      "num_input_tokens_seen": 26806992,
      "step": 12830
    },
    {
      "epoch": 2.0938086303939962,
      "grad_norm": 7.4375,
      "learning_rate": 3.833925582447428e-05,
      "loss": 3.3858,
      "num_input_tokens_seen": 26817088,
      "step": 12835
    },
    {
      "epoch": 2.094624357614814,
      "grad_norm": 7.09375,
      "learning_rate": 3.8330951608384656e-05,
      "loss": 2.9799,
      "num_input_tokens_seen": 26827392,
      "step": 12840
    },
    {
      "epoch": 2.095440084835631,
      "grad_norm": 5.03125,
      "learning_rate": 3.832264533649477e-05,
      "loss": 2.9696,
      "num_input_tokens_seen": 26836096,
      "step": 12845
    },
    {
      "epoch": 2.0962558120564485,
      "grad_norm": 1.78125,
      "learning_rate": 3.8314337010085555e-05,
      "loss": 1.6584,
      "num_input_tokens_seen": 26846768,
      "step": 12850
    },
    {
      "epoch": 2.0970715392772656,
      "grad_norm": 5.8125,
      "learning_rate": 3.830602663043824e-05,
      "loss": 2.2998,
      "num_input_tokens_seen": 26857536,
      "step": 12855
    },
    {
      "epoch": 2.097887266498083,
      "grad_norm": 8.5625,
      "learning_rate": 3.8297714198834414e-05,
      "loss": 2.8623,
      "num_input_tokens_seen": 26868368,
      "step": 12860
    },
    {
      "epoch": 2.0987029937189003,
      "grad_norm": 2.734375,
      "learning_rate": 3.828939971655595e-05,
      "loss": 2.1364,
      "num_input_tokens_seen": 26877584,
      "step": 12865
    },
    {
      "epoch": 2.099518720939718,
      "grad_norm": 3.765625,
      "learning_rate": 3.828108318488505e-05,
      "loss": 1.6517,
      "num_input_tokens_seen": 26889136,
      "step": 12870
    },
    {
      "epoch": 2.100334448160535,
      "grad_norm": 8.1875,
      "learning_rate": 3.8272764605104216e-05,
      "loss": 2.905,
      "num_input_tokens_seen": 26899952,
      "step": 12875
    },
    {
      "epoch": 2.1011501753813526,
      "grad_norm": 8.0625,
      "learning_rate": 3.826444397849628e-05,
      "loss": 2.2225,
      "num_input_tokens_seen": 26910688,
      "step": 12880
    },
    {
      "epoch": 2.1019659026021698,
      "grad_norm": 9.3125,
      "learning_rate": 3.825612130634439e-05,
      "loss": 2.1163,
      "num_input_tokens_seen": 26922000,
      "step": 12885
    },
    {
      "epoch": 2.1027816298229873,
      "grad_norm": 3.703125,
      "learning_rate": 3.824779658993202e-05,
      "loss": 1.7419,
      "num_input_tokens_seen": 26932608,
      "step": 12890
    },
    {
      "epoch": 2.1035973570438045,
      "grad_norm": 8.375,
      "learning_rate": 3.823946983054292e-05,
      "loss": 2.8928,
      "num_input_tokens_seen": 26943520,
      "step": 12895
    },
    {
      "epoch": 2.104413084264622,
      "grad_norm": 15.8125,
      "learning_rate": 3.82311410294612e-05,
      "loss": 3.0741,
      "num_input_tokens_seen": 26954048,
      "step": 12900
    },
    {
      "epoch": 2.105228811485439,
      "grad_norm": 7.8125,
      "learning_rate": 3.822281018797127e-05,
      "loss": 1.3691,
      "num_input_tokens_seen": 26965152,
      "step": 12905
    },
    {
      "epoch": 2.1060445387062567,
      "grad_norm": 13.4375,
      "learning_rate": 3.821447730735783e-05,
      "loss": 3.287,
      "num_input_tokens_seen": 26975120,
      "step": 12910
    },
    {
      "epoch": 2.106860265927074,
      "grad_norm": 13.375,
      "learning_rate": 3.820614238890592e-05,
      "loss": 3.7302,
      "num_input_tokens_seen": 26986816,
      "step": 12915
    },
    {
      "epoch": 2.1076759931478914,
      "grad_norm": 5.375,
      "learning_rate": 3.819780543390091e-05,
      "loss": 0.7774,
      "num_input_tokens_seen": 26998400,
      "step": 12920
    },
    {
      "epoch": 2.1084917203687086,
      "grad_norm": 3.609375,
      "learning_rate": 3.818946644362844e-05,
      "loss": 1.7324,
      "num_input_tokens_seen": 27007616,
      "step": 12925
    },
    {
      "epoch": 2.109307447589526,
      "grad_norm": 10.25,
      "learning_rate": 3.81811254193745e-05,
      "loss": 2.3223,
      "num_input_tokens_seen": 27019168,
      "step": 12930
    },
    {
      "epoch": 2.1101231748103433,
      "grad_norm": 4.0,
      "learning_rate": 3.8172782362425366e-05,
      "loss": 2.6146,
      "num_input_tokens_seen": 27030048,
      "step": 12935
    },
    {
      "epoch": 2.110938902031161,
      "grad_norm": 5.96875,
      "learning_rate": 3.816443727406765e-05,
      "loss": 1.7589,
      "num_input_tokens_seen": 27040080,
      "step": 12940
    },
    {
      "epoch": 2.111754629251978,
      "grad_norm": 7.375,
      "learning_rate": 3.815609015558829e-05,
      "loss": 1.9522,
      "num_input_tokens_seen": 27050496,
      "step": 12945
    },
    {
      "epoch": 2.1125703564727956,
      "grad_norm": 9.6875,
      "learning_rate": 3.814774100827448e-05,
      "loss": 3.4615,
      "num_input_tokens_seen": 27060784,
      "step": 12950
    },
    {
      "epoch": 2.1133860836936127,
      "grad_norm": 14.5625,
      "learning_rate": 3.813938983341379e-05,
      "loss": 2.9793,
      "num_input_tokens_seen": 27071056,
      "step": 12955
    },
    {
      "epoch": 2.1142018109144303,
      "grad_norm": 8.0625,
      "learning_rate": 3.813103663229407e-05,
      "loss": 2.3659,
      "num_input_tokens_seen": 27080096,
      "step": 12960
    },
    {
      "epoch": 2.1150175381352474,
      "grad_norm": 9.25,
      "learning_rate": 3.812268140620349e-05,
      "loss": 2.5049,
      "num_input_tokens_seen": 27091264,
      "step": 12965
    },
    {
      "epoch": 2.115833265356065,
      "grad_norm": 6.09375,
      "learning_rate": 3.811432415643051e-05,
      "loss": 2.0872,
      "num_input_tokens_seen": 27101040,
      "step": 12970
    },
    {
      "epoch": 2.116648992576882,
      "grad_norm": 4.84375,
      "learning_rate": 3.8105964884263954e-05,
      "loss": 2.8819,
      "num_input_tokens_seen": 27113424,
      "step": 12975
    },
    {
      "epoch": 2.1174647197976997,
      "grad_norm": 4.03125,
      "learning_rate": 3.809760359099291e-05,
      "loss": 2.963,
      "num_input_tokens_seen": 27123488,
      "step": 12980
    },
    {
      "epoch": 2.118280447018517,
      "grad_norm": 4.65625,
      "learning_rate": 3.8089240277906804e-05,
      "loss": 2.4296,
      "num_input_tokens_seen": 27134432,
      "step": 12985
    },
    {
      "epoch": 2.1190961742393344,
      "grad_norm": 5.59375,
      "learning_rate": 3.808087494629535e-05,
      "loss": 2.4118,
      "num_input_tokens_seen": 27144976,
      "step": 12990
    },
    {
      "epoch": 2.1199119014601515,
      "grad_norm": 7.46875,
      "learning_rate": 3.8072507597448595e-05,
      "loss": 2.7485,
      "num_input_tokens_seen": 27155152,
      "step": 12995
    },
    {
      "epoch": 2.120727628680969,
      "grad_norm": 10.3125,
      "learning_rate": 3.806413823265689e-05,
      "loss": 1.7513,
      "num_input_tokens_seen": 27166064,
      "step": 13000
    },
    {
      "epoch": 2.120727628680969,
      "eval_loss": 2.5287065505981445,
      "eval_runtime": 134.8393,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 27166064,
      "step": 13000
    },
    {
      "epoch": 2.121543355901786,
      "grad_norm": 3.203125,
      "learning_rate": 3.805576685321089e-05,
      "loss": 2.6723,
      "num_input_tokens_seen": 27177456,
      "step": 13005
    },
    {
      "epoch": 2.122359083122604,
      "grad_norm": 3.96875,
      "learning_rate": 3.804739346040158e-05,
      "loss": 1.2507,
      "num_input_tokens_seen": 27186464,
      "step": 13010
    },
    {
      "epoch": 2.1231748103434214,
      "grad_norm": 4.9375,
      "learning_rate": 3.8039018055520234e-05,
      "loss": 2.8962,
      "num_input_tokens_seen": 27196320,
      "step": 13015
    },
    {
      "epoch": 2.1239905375642385,
      "grad_norm": 7.90625,
      "learning_rate": 3.803064063985844e-05,
      "loss": 2.2396,
      "num_input_tokens_seen": 27206992,
      "step": 13020
    },
    {
      "epoch": 2.124806264785056,
      "grad_norm": 2.234375,
      "learning_rate": 3.802226121470811e-05,
      "loss": 1.8954,
      "num_input_tokens_seen": 27218304,
      "step": 13025
    },
    {
      "epoch": 2.125621992005873,
      "grad_norm": 3.578125,
      "learning_rate": 3.801387978136145e-05,
      "loss": 3.4184,
      "num_input_tokens_seen": 27229040,
      "step": 13030
    },
    {
      "epoch": 2.126437719226691,
      "grad_norm": 7.625,
      "learning_rate": 3.800549634111099e-05,
      "loss": 2.4736,
      "num_input_tokens_seen": 27240064,
      "step": 13035
    },
    {
      "epoch": 2.127253446447508,
      "grad_norm": 9.9375,
      "learning_rate": 3.799711089524955e-05,
      "loss": 3.6355,
      "num_input_tokens_seen": 27249600,
      "step": 13040
    },
    {
      "epoch": 2.1280691736683255,
      "grad_norm": 2.265625,
      "learning_rate": 3.7988723445070285e-05,
      "loss": 1.1616,
      "num_input_tokens_seen": 27259264,
      "step": 13045
    },
    {
      "epoch": 2.1288849008891426,
      "grad_norm": 7.0625,
      "learning_rate": 3.798033399186663e-05,
      "loss": 1.3842,
      "num_input_tokens_seen": 27270784,
      "step": 13050
    },
    {
      "epoch": 2.12970062810996,
      "grad_norm": 4.34375,
      "learning_rate": 3.797194253693237e-05,
      "loss": 1.4918,
      "num_input_tokens_seen": 27281168,
      "step": 13055
    },
    {
      "epoch": 2.1305163553307773,
      "grad_norm": 5.46875,
      "learning_rate": 3.796354908156153e-05,
      "loss": 2.7857,
      "num_input_tokens_seen": 27290976,
      "step": 13060
    },
    {
      "epoch": 2.131332082551595,
      "grad_norm": 7.21875,
      "learning_rate": 3.795515362704853e-05,
      "loss": 3.5724,
      "num_input_tokens_seen": 27301088,
      "step": 13065
    },
    {
      "epoch": 2.132147809772412,
      "grad_norm": 8.4375,
      "learning_rate": 3.794675617468803e-05,
      "loss": 1.7854,
      "num_input_tokens_seen": 27311856,
      "step": 13070
    },
    {
      "epoch": 2.1329635369932296,
      "grad_norm": 8.5,
      "learning_rate": 3.793835672577503e-05,
      "loss": 2.7045,
      "num_input_tokens_seen": 27323280,
      "step": 13075
    },
    {
      "epoch": 2.1337792642140467,
      "grad_norm": 5.125,
      "learning_rate": 3.7929955281604826e-05,
      "loss": 2.5457,
      "num_input_tokens_seen": 27332192,
      "step": 13080
    },
    {
      "epoch": 2.1345949914348643,
      "grad_norm": 10.25,
      "learning_rate": 3.7921551843473036e-05,
      "loss": 1.4884,
      "num_input_tokens_seen": 27341712,
      "step": 13085
    },
    {
      "epoch": 2.1354107186556814,
      "grad_norm": 6.125,
      "learning_rate": 3.791314641267557e-05,
      "loss": 2.5575,
      "num_input_tokens_seen": 27352880,
      "step": 13090
    },
    {
      "epoch": 2.136226445876499,
      "grad_norm": 6.0,
      "learning_rate": 3.790473899050864e-05,
      "loss": 2.2465,
      "num_input_tokens_seen": 27362432,
      "step": 13095
    },
    {
      "epoch": 2.137042173097316,
      "grad_norm": 12.0625,
      "learning_rate": 3.7896329578268794e-05,
      "loss": 2.9499,
      "num_input_tokens_seen": 27373728,
      "step": 13100
    },
    {
      "epoch": 2.1378579003181337,
      "grad_norm": 6.8125,
      "learning_rate": 3.7887918177252855e-05,
      "loss": 1.6174,
      "num_input_tokens_seen": 27383936,
      "step": 13105
    },
    {
      "epoch": 2.138673627538951,
      "grad_norm": 6.28125,
      "learning_rate": 3.787950478875798e-05,
      "loss": 2.5031,
      "num_input_tokens_seen": 27396240,
      "step": 13110
    },
    {
      "epoch": 2.1394893547597684,
      "grad_norm": 9.5,
      "learning_rate": 3.787108941408162e-05,
      "loss": 2.6774,
      "num_input_tokens_seen": 27406352,
      "step": 13115
    },
    {
      "epoch": 2.1403050819805856,
      "grad_norm": 5.78125,
      "learning_rate": 3.786267205452151e-05,
      "loss": 1.6694,
      "num_input_tokens_seen": 27415696,
      "step": 13120
    },
    {
      "epoch": 2.141120809201403,
      "grad_norm": 3.53125,
      "learning_rate": 3.785425271137573e-05,
      "loss": 2.0385,
      "num_input_tokens_seen": 27425504,
      "step": 13125
    },
    {
      "epoch": 2.1419365364222203,
      "grad_norm": 8.75,
      "learning_rate": 3.7845831385942655e-05,
      "loss": 3.7738,
      "num_input_tokens_seen": 27437184,
      "step": 13130
    },
    {
      "epoch": 2.142752263643038,
      "grad_norm": 0.1962890625,
      "learning_rate": 3.7837408079520944e-05,
      "loss": 2.089,
      "num_input_tokens_seen": 27446608,
      "step": 13135
    },
    {
      "epoch": 2.143567990863855,
      "grad_norm": 4.65625,
      "learning_rate": 3.782898279340957e-05,
      "loss": 1.5335,
      "num_input_tokens_seen": 27456128,
      "step": 13140
    },
    {
      "epoch": 2.1443837180846725,
      "grad_norm": 13.625,
      "learning_rate": 3.782055552890784e-05,
      "loss": 2.4087,
      "num_input_tokens_seen": 27466640,
      "step": 13145
    },
    {
      "epoch": 2.1451994453054897,
      "grad_norm": 15.3125,
      "learning_rate": 3.781212628731534e-05,
      "loss": 2.5935,
      "num_input_tokens_seen": 27476384,
      "step": 13150
    },
    {
      "epoch": 2.1460151725263072,
      "grad_norm": 15.0,
      "learning_rate": 3.7803695069931946e-05,
      "loss": 2.7839,
      "num_input_tokens_seen": 27487184,
      "step": 13155
    },
    {
      "epoch": 2.1468308997471244,
      "grad_norm": 6.0625,
      "learning_rate": 3.779526187805789e-05,
      "loss": 3.7321,
      "num_input_tokens_seen": 27497152,
      "step": 13160
    },
    {
      "epoch": 2.147646626967942,
      "grad_norm": 6.625,
      "learning_rate": 3.778682671299364e-05,
      "loss": 2.1731,
      "num_input_tokens_seen": 27507504,
      "step": 13165
    },
    {
      "epoch": 2.148462354188759,
      "grad_norm": 9.9375,
      "learning_rate": 3.777838957604003e-05,
      "loss": 2.1296,
      "num_input_tokens_seen": 27517008,
      "step": 13170
    },
    {
      "epoch": 2.1492780814095767,
      "grad_norm": 13.4375,
      "learning_rate": 3.776995046849816e-05,
      "loss": 2.4311,
      "num_input_tokens_seen": 27527184,
      "step": 13175
    },
    {
      "epoch": 2.150093808630394,
      "grad_norm": 12.375,
      "learning_rate": 3.776150939166945e-05,
      "loss": 1.663,
      "num_input_tokens_seen": 27537632,
      "step": 13180
    },
    {
      "epoch": 2.1509095358512114,
      "grad_norm": 7.40625,
      "learning_rate": 3.775306634685562e-05,
      "loss": 3.7676,
      "num_input_tokens_seen": 27550032,
      "step": 13185
    },
    {
      "epoch": 2.151725263072029,
      "grad_norm": 10.6875,
      "learning_rate": 3.7744621335358696e-05,
      "loss": 2.4983,
      "num_input_tokens_seen": 27560336,
      "step": 13190
    },
    {
      "epoch": 2.152540990292846,
      "grad_norm": 6.96875,
      "learning_rate": 3.7736174358481e-05,
      "loss": 2.2404,
      "num_input_tokens_seen": 27571200,
      "step": 13195
    },
    {
      "epoch": 2.153356717513663,
      "grad_norm": 3.09375,
      "learning_rate": 3.7727725417525175e-05,
      "loss": 0.9243,
      "num_input_tokens_seen": 27581264,
      "step": 13200
    },
    {
      "epoch": 2.153356717513663,
      "eval_loss": 2.5427470207214355,
      "eval_runtime": 134.7467,
      "eval_samples_per_second": 20.223,
      "eval_steps_per_second": 10.115,
      "num_input_tokens_seen": 27581264,
      "step": 13200
    },
    {
      "epoch": 2.1541724447344808,
      "grad_norm": 6.6875,
      "learning_rate": 3.771927451379414e-05,
      "loss": 2.3377,
      "num_input_tokens_seen": 27593168,
      "step": 13205
    },
    {
      "epoch": 2.1549881719552983,
      "grad_norm": 5.875,
      "learning_rate": 3.7710821648591135e-05,
      "loss": 2.5799,
      "num_input_tokens_seen": 27602352,
      "step": 13210
    },
    {
      "epoch": 2.1558038991761155,
      "grad_norm": 3.953125,
      "learning_rate": 3.7702366823219694e-05,
      "loss": 1.9942,
      "num_input_tokens_seen": 27612032,
      "step": 13215
    },
    {
      "epoch": 2.156619626396933,
      "grad_norm": 8.4375,
      "learning_rate": 3.769391003898366e-05,
      "loss": 1.3627,
      "num_input_tokens_seen": 27621792,
      "step": 13220
    },
    {
      "epoch": 2.15743535361775,
      "grad_norm": 6.6875,
      "learning_rate": 3.768545129718718e-05,
      "loss": 2.4639,
      "num_input_tokens_seen": 27633856,
      "step": 13225
    },
    {
      "epoch": 2.1582510808385678,
      "grad_norm": 6.6875,
      "learning_rate": 3.7676990599134686e-05,
      "loss": 2.2643,
      "num_input_tokens_seen": 27644656,
      "step": 13230
    },
    {
      "epoch": 2.159066808059385,
      "grad_norm": 9.4375,
      "learning_rate": 3.766852794613095e-05,
      "loss": 3.2836,
      "num_input_tokens_seen": 27654736,
      "step": 13235
    },
    {
      "epoch": 2.1598825352802025,
      "grad_norm": 4.09375,
      "learning_rate": 3.766006333948099e-05,
      "loss": 2.2872,
      "num_input_tokens_seen": 27664336,
      "step": 13240
    },
    {
      "epoch": 2.1606982625010196,
      "grad_norm": 10.875,
      "learning_rate": 3.765159678049017e-05,
      "loss": 2.0381,
      "num_input_tokens_seen": 27675728,
      "step": 13245
    },
    {
      "epoch": 2.161513989721837,
      "grad_norm": 9.125,
      "learning_rate": 3.7643128270464134e-05,
      "loss": 3.165,
      "num_input_tokens_seen": 27687088,
      "step": 13250
    },
    {
      "epoch": 2.1623297169426543,
      "grad_norm": 5.9375,
      "learning_rate": 3.763465781070884e-05,
      "loss": 1.7316,
      "num_input_tokens_seen": 27697088,
      "step": 13255
    },
    {
      "epoch": 2.163145444163472,
      "grad_norm": 9.25,
      "learning_rate": 3.762618540253052e-05,
      "loss": 2.5775,
      "num_input_tokens_seen": 27707600,
      "step": 13260
    },
    {
      "epoch": 2.163961171384289,
      "grad_norm": 7.21875,
      "learning_rate": 3.761771104723576e-05,
      "loss": 1.5108,
      "num_input_tokens_seen": 27717360,
      "step": 13265
    },
    {
      "epoch": 2.1647768986051066,
      "grad_norm": 3.46875,
      "learning_rate": 3.7609234746131386e-05,
      "loss": 2.9984,
      "num_input_tokens_seen": 27727952,
      "step": 13270
    },
    {
      "epoch": 2.1655926258259237,
      "grad_norm": 6.8125,
      "learning_rate": 3.7600756500524556e-05,
      "loss": 2.8128,
      "num_input_tokens_seen": 27737792,
      "step": 13275
    },
    {
      "epoch": 2.1664083530467413,
      "grad_norm": 8.1875,
      "learning_rate": 3.759227631172271e-05,
      "loss": 2.0977,
      "num_input_tokens_seen": 27747280,
      "step": 13280
    },
    {
      "epoch": 2.1672240802675584,
      "grad_norm": 7.28125,
      "learning_rate": 3.758379418103363e-05,
      "loss": 3.0331,
      "num_input_tokens_seen": 27757504,
      "step": 13285
    },
    {
      "epoch": 2.168039807488376,
      "grad_norm": 16.0,
      "learning_rate": 3.757531010976534e-05,
      "loss": 3.1703,
      "num_input_tokens_seen": 27767552,
      "step": 13290
    },
    {
      "epoch": 2.168855534709193,
      "grad_norm": 11.9375,
      "learning_rate": 3.75668240992262e-05,
      "loss": 3.727,
      "num_input_tokens_seen": 27778560,
      "step": 13295
    },
    {
      "epoch": 2.1696712619300107,
      "grad_norm": 3.453125,
      "learning_rate": 3.7558336150724865e-05,
      "loss": 1.8895,
      "num_input_tokens_seen": 27789344,
      "step": 13300
    },
    {
      "epoch": 2.170486989150828,
      "grad_norm": 9.5625,
      "learning_rate": 3.754984626557028e-05,
      "loss": 3.3177,
      "num_input_tokens_seen": 27799824,
      "step": 13305
    },
    {
      "epoch": 2.1713027163716454,
      "grad_norm": 2.703125,
      "learning_rate": 3.754135444507168e-05,
      "loss": 1.1676,
      "num_input_tokens_seen": 27810992,
      "step": 13310
    },
    {
      "epoch": 2.1721184435924625,
      "grad_norm": 7.3125,
      "learning_rate": 3.753286069053863e-05,
      "loss": 1.347,
      "num_input_tokens_seen": 27821392,
      "step": 13315
    },
    {
      "epoch": 2.17293417081328,
      "grad_norm": 6.5,
      "learning_rate": 3.7524365003280945e-05,
      "loss": 2.4593,
      "num_input_tokens_seen": 27831264,
      "step": 13320
    },
    {
      "epoch": 2.1737498980340972,
      "grad_norm": 10.0,
      "learning_rate": 3.75158673846088e-05,
      "loss": 2.4446,
      "num_input_tokens_seen": 27839360,
      "step": 13325
    },
    {
      "epoch": 2.174565625254915,
      "grad_norm": 3.453125,
      "learning_rate": 3.750736783583262e-05,
      "loss": 1.6798,
      "num_input_tokens_seen": 27848448,
      "step": 13330
    },
    {
      "epoch": 2.175381352475732,
      "grad_norm": 7.5,
      "learning_rate": 3.7498866358263144e-05,
      "loss": 2.3483,
      "num_input_tokens_seen": 27859440,
      "step": 13335
    },
    {
      "epoch": 2.1761970796965495,
      "grad_norm": 10.125,
      "learning_rate": 3.74903629532114e-05,
      "loss": 2.3944,
      "num_input_tokens_seen": 27869952,
      "step": 13340
    },
    {
      "epoch": 2.1770128069173666,
      "grad_norm": 4.0,
      "learning_rate": 3.748185762198873e-05,
      "loss": 1.3601,
      "num_input_tokens_seen": 27879376,
      "step": 13345
    },
    {
      "epoch": 2.1778285341381842,
      "grad_norm": 10.875,
      "learning_rate": 3.747335036590676e-05,
      "loss": 2.4343,
      "num_input_tokens_seen": 27888912,
      "step": 13350
    },
    {
      "epoch": 2.1786442613590014,
      "grad_norm": 6.71875,
      "learning_rate": 3.7464841186277405e-05,
      "loss": 1.9506,
      "num_input_tokens_seen": 27898512,
      "step": 13355
    },
    {
      "epoch": 2.179459988579819,
      "grad_norm": 9.625,
      "learning_rate": 3.7456330084412896e-05,
      "loss": 3.0834,
      "num_input_tokens_seen": 27908352,
      "step": 13360
    },
    {
      "epoch": 2.180275715800636,
      "grad_norm": 4.25,
      "learning_rate": 3.744781706162576e-05,
      "loss": 3.2379,
      "num_input_tokens_seen": 27917456,
      "step": 13365
    },
    {
      "epoch": 2.1810914430214536,
      "grad_norm": 9.375,
      "learning_rate": 3.743930211922879e-05,
      "loss": 2.627,
      "num_input_tokens_seen": 27929024,
      "step": 13370
    },
    {
      "epoch": 2.1819071702422708,
      "grad_norm": 8.75,
      "learning_rate": 3.743078525853513e-05,
      "loss": 2.2167,
      "num_input_tokens_seen": 27940192,
      "step": 13375
    },
    {
      "epoch": 2.1827228974630883,
      "grad_norm": 4.90625,
      "learning_rate": 3.7422266480858154e-05,
      "loss": 2.6997,
      "num_input_tokens_seen": 27949648,
      "step": 13380
    },
    {
      "epoch": 2.183538624683906,
      "grad_norm": 6.59375,
      "learning_rate": 3.741374578751158e-05,
      "loss": 1.8599,
      "num_input_tokens_seen": 27959968,
      "step": 13385
    },
    {
      "epoch": 2.184354351904723,
      "grad_norm": 4.40625,
      "learning_rate": 3.740522317980941e-05,
      "loss": 1.7879,
      "num_input_tokens_seen": 27968144,
      "step": 13390
    },
    {
      "epoch": 2.1851700791255406,
      "grad_norm": 12.4375,
      "learning_rate": 3.739669865906593e-05,
      "loss": 3.2361,
      "num_input_tokens_seen": 27978992,
      "step": 13395
    },
    {
      "epoch": 2.1859858063463578,
      "grad_norm": 8.25,
      "learning_rate": 3.738817222659573e-05,
      "loss": 1.8392,
      "num_input_tokens_seen": 27988880,
      "step": 13400
    },
    {
      "epoch": 2.1859858063463578,
      "eval_loss": 2.5632946491241455,
      "eval_runtime": 134.7942,
      "eval_samples_per_second": 20.216,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 27988880,
      "step": 13400
    },
    {
      "epoch": 2.1868015335671753,
      "grad_norm": 11.5,
      "learning_rate": 3.73796438837137e-05,
      "loss": 2.3795,
      "num_input_tokens_seen": 27999520,
      "step": 13405
    },
    {
      "epoch": 2.1876172607879925,
      "grad_norm": 2.390625,
      "learning_rate": 3.7371113631735e-05,
      "loss": 1.8136,
      "num_input_tokens_seen": 28010640,
      "step": 13410
    },
    {
      "epoch": 2.18843298800881,
      "grad_norm": 4.65625,
      "learning_rate": 3.736258147197512e-05,
      "loss": 2.3365,
      "num_input_tokens_seen": 28021504,
      "step": 13415
    },
    {
      "epoch": 2.189248715229627,
      "grad_norm": 2.109375,
      "learning_rate": 3.735404740574981e-05,
      "loss": 2.0843,
      "num_input_tokens_seen": 28030432,
      "step": 13420
    },
    {
      "epoch": 2.1900644424504447,
      "grad_norm": 4.0,
      "learning_rate": 3.7345511434375145e-05,
      "loss": 1.6312,
      "num_input_tokens_seen": 28040752,
      "step": 13425
    },
    {
      "epoch": 2.190880169671262,
      "grad_norm": 7.53125,
      "learning_rate": 3.733697355916748e-05,
      "loss": 2.3655,
      "num_input_tokens_seen": 28050464,
      "step": 13430
    },
    {
      "epoch": 2.1916958968920794,
      "grad_norm": 10.625,
      "learning_rate": 3.732843378144345e-05,
      "loss": 2.2273,
      "num_input_tokens_seen": 28060048,
      "step": 13435
    },
    {
      "epoch": 2.1925116241128966,
      "grad_norm": 6.8125,
      "learning_rate": 3.7319892102519995e-05,
      "loss": 3.9828,
      "num_input_tokens_seen": 28069696,
      "step": 13440
    },
    {
      "epoch": 2.193327351333714,
      "grad_norm": 4.75,
      "learning_rate": 3.731134852371436e-05,
      "loss": 2.2042,
      "num_input_tokens_seen": 28080464,
      "step": 13445
    },
    {
      "epoch": 2.1941430785545313,
      "grad_norm": 11.8125,
      "learning_rate": 3.730280304634408e-05,
      "loss": 2.9954,
      "num_input_tokens_seen": 28091296,
      "step": 13450
    },
    {
      "epoch": 2.194958805775349,
      "grad_norm": 5.8125,
      "learning_rate": 3.729425567172696e-05,
      "loss": 2.8598,
      "num_input_tokens_seen": 28103232,
      "step": 13455
    },
    {
      "epoch": 2.195774532996166,
      "grad_norm": 9.0,
      "learning_rate": 3.728570640118111e-05,
      "loss": 2.5514,
      "num_input_tokens_seen": 28114000,
      "step": 13460
    },
    {
      "epoch": 2.1965902602169836,
      "grad_norm": 6.0,
      "learning_rate": 3.727715523602494e-05,
      "loss": 1.6747,
      "num_input_tokens_seen": 28123856,
      "step": 13465
    },
    {
      "epoch": 2.1974059874378007,
      "grad_norm": 5.9375,
      "learning_rate": 3.726860217757715e-05,
      "loss": 1.7441,
      "num_input_tokens_seen": 28135040,
      "step": 13470
    },
    {
      "epoch": 2.1982217146586183,
      "grad_norm": 6.84375,
      "learning_rate": 3.726004722715673e-05,
      "loss": 3.6209,
      "num_input_tokens_seen": 28143808,
      "step": 13475
    },
    {
      "epoch": 2.1990374418794354,
      "grad_norm": 9.125,
      "learning_rate": 3.725149038608296e-05,
      "loss": 2.721,
      "num_input_tokens_seen": 28154528,
      "step": 13480
    },
    {
      "epoch": 2.199853169100253,
      "grad_norm": 11.0,
      "learning_rate": 3.7242931655675404e-05,
      "loss": 3.1101,
      "num_input_tokens_seen": 28164800,
      "step": 13485
    },
    {
      "epoch": 2.20066889632107,
      "grad_norm": 7.6875,
      "learning_rate": 3.7234371037253937e-05,
      "loss": 2.8425,
      "num_input_tokens_seen": 28174208,
      "step": 13490
    },
    {
      "epoch": 2.2014846235418877,
      "grad_norm": 17.375,
      "learning_rate": 3.7225808532138705e-05,
      "loss": 3.5908,
      "num_input_tokens_seen": 28183648,
      "step": 13495
    },
    {
      "epoch": 2.202300350762705,
      "grad_norm": 6.5625,
      "learning_rate": 3.721724414165016e-05,
      "loss": 2.6048,
      "num_input_tokens_seen": 28194656,
      "step": 13500
    },
    {
      "epoch": 2.2031160779835224,
      "grad_norm": 9.25,
      "learning_rate": 3.720867786710904e-05,
      "loss": 2.39,
      "num_input_tokens_seen": 28204624,
      "step": 13505
    },
    {
      "epoch": 2.2039318052043395,
      "grad_norm": 4.46875,
      "learning_rate": 3.7200109709836366e-05,
      "loss": 2.1693,
      "num_input_tokens_seen": 28214304,
      "step": 13510
    },
    {
      "epoch": 2.204747532425157,
      "grad_norm": 2.09375,
      "learning_rate": 3.7191539671153465e-05,
      "loss": 2.0736,
      "num_input_tokens_seen": 28225728,
      "step": 13515
    },
    {
      "epoch": 2.205563259645974,
      "grad_norm": 9.625,
      "learning_rate": 3.718296775238193e-05,
      "loss": 3.4614,
      "num_input_tokens_seen": 28235520,
      "step": 13520
    },
    {
      "epoch": 2.206378986866792,
      "grad_norm": 7.0625,
      "learning_rate": 3.7174393954843675e-05,
      "loss": 2.2725,
      "num_input_tokens_seen": 28245232,
      "step": 13525
    },
    {
      "epoch": 2.207194714087609,
      "grad_norm": 7.78125,
      "learning_rate": 3.716581827986087e-05,
      "loss": 1.7475,
      "num_input_tokens_seen": 28255552,
      "step": 13530
    },
    {
      "epoch": 2.2080104413084265,
      "grad_norm": 2.5625,
      "learning_rate": 3.7157240728756004e-05,
      "loss": 2.3929,
      "num_input_tokens_seen": 28265104,
      "step": 13535
    },
    {
      "epoch": 2.2088261685292436,
      "grad_norm": 10.6875,
      "learning_rate": 3.714866130285184e-05,
      "loss": 3.1104,
      "num_input_tokens_seen": 28275552,
      "step": 13540
    },
    {
      "epoch": 2.209641895750061,
      "grad_norm": 2.046875,
      "learning_rate": 3.714008000347143e-05,
      "loss": 1.5486,
      "num_input_tokens_seen": 28285296,
      "step": 13545
    },
    {
      "epoch": 2.2104576229708783,
      "grad_norm": 10.0625,
      "learning_rate": 3.7131496831938126e-05,
      "loss": 3.3411,
      "num_input_tokens_seen": 28295872,
      "step": 13550
    },
    {
      "epoch": 2.211273350191696,
      "grad_norm": 2.984375,
      "learning_rate": 3.7122911789575565e-05,
      "loss": 1.4784,
      "num_input_tokens_seen": 28304816,
      "step": 13555
    },
    {
      "epoch": 2.2120890774125135,
      "grad_norm": 4.4375,
      "learning_rate": 3.711432487770765e-05,
      "loss": 2.0286,
      "num_input_tokens_seen": 28315664,
      "step": 13560
    },
    {
      "epoch": 2.2129048046333306,
      "grad_norm": 6.34375,
      "learning_rate": 3.710573609765861e-05,
      "loss": 2.3015,
      "num_input_tokens_seen": 28326160,
      "step": 13565
    },
    {
      "epoch": 2.2137205318541477,
      "grad_norm": 7.15625,
      "learning_rate": 3.709714545075292e-05,
      "loss": 3.069,
      "num_input_tokens_seen": 28335872,
      "step": 13570
    },
    {
      "epoch": 2.2145362590749653,
      "grad_norm": 8.375,
      "learning_rate": 3.708855293831538e-05,
      "loss": 1.668,
      "num_input_tokens_seen": 28346192,
      "step": 13575
    },
    {
      "epoch": 2.215351986295783,
      "grad_norm": 2.59375,
      "learning_rate": 3.707995856167107e-05,
      "loss": 2.2027,
      "num_input_tokens_seen": 28355680,
      "step": 13580
    },
    {
      "epoch": 2.2161677135166,
      "grad_norm": 6.21875,
      "learning_rate": 3.707136232214534e-05,
      "loss": 2.1956,
      "num_input_tokens_seen": 28366080,
      "step": 13585
    },
    {
      "epoch": 2.2169834407374176,
      "grad_norm": 6.6875,
      "learning_rate": 3.7062764221063844e-05,
      "loss": 1.955,
      "num_input_tokens_seen": 28376176,
      "step": 13590
    },
    {
      "epoch": 2.2177991679582347,
      "grad_norm": 4.5625,
      "learning_rate": 3.705416425975252e-05,
      "loss": 2.4168,
      "num_input_tokens_seen": 28386832,
      "step": 13595
    },
    {
      "epoch": 2.2186148951790523,
      "grad_norm": 2.078125,
      "learning_rate": 3.704556243953758e-05,
      "loss": 2.2886,
      "num_input_tokens_seen": 28397472,
      "step": 13600
    },
    {
      "epoch": 2.2186148951790523,
      "eval_loss": 2.5419933795928955,
      "eval_runtime": 134.9612,
      "eval_samples_per_second": 20.191,
      "eval_steps_per_second": 10.099,
      "num_input_tokens_seen": 28397472,
      "step": 13600
    },
    {
      "epoch": 2.2194306223998694,
      "grad_norm": 6.625,
      "learning_rate": 3.7036958761745535e-05,
      "loss": 1.9117,
      "num_input_tokens_seen": 28406608,
      "step": 13605
    },
    {
      "epoch": 2.220246349620687,
      "grad_norm": 3.46875,
      "learning_rate": 3.702835322770318e-05,
      "loss": 1.927,
      "num_input_tokens_seen": 28416480,
      "step": 13610
    },
    {
      "epoch": 2.221062076841504,
      "grad_norm": 11.5625,
      "learning_rate": 3.701974583873761e-05,
      "loss": 3.7652,
      "num_input_tokens_seen": 28426144,
      "step": 13615
    },
    {
      "epoch": 2.2218778040623217,
      "grad_norm": 10.5625,
      "learning_rate": 3.701113659617618e-05,
      "loss": 2.1548,
      "num_input_tokens_seen": 28438080,
      "step": 13620
    },
    {
      "epoch": 2.222693531283139,
      "grad_norm": 11.6875,
      "learning_rate": 3.7002525501346535e-05,
      "loss": 2.6727,
      "num_input_tokens_seen": 28447968,
      "step": 13625
    },
    {
      "epoch": 2.2235092585039564,
      "grad_norm": 5.9375,
      "learning_rate": 3.699391255557664e-05,
      "loss": 2.3659,
      "num_input_tokens_seen": 28458752,
      "step": 13630
    },
    {
      "epoch": 2.2243249857247736,
      "grad_norm": 3.484375,
      "learning_rate": 3.69852977601947e-05,
      "loss": 1.9306,
      "num_input_tokens_seen": 28468912,
      "step": 13635
    },
    {
      "epoch": 2.225140712945591,
      "grad_norm": 3.4375,
      "learning_rate": 3.697668111652922e-05,
      "loss": 2.6419,
      "num_input_tokens_seen": 28478864,
      "step": 13640
    },
    {
      "epoch": 2.2259564401664083,
      "grad_norm": 8.375,
      "learning_rate": 3.6968062625909005e-05,
      "loss": 1.6479,
      "num_input_tokens_seen": 28489504,
      "step": 13645
    },
    {
      "epoch": 2.226772167387226,
      "grad_norm": 5.96875,
      "learning_rate": 3.6959442289663135e-05,
      "loss": 2.1652,
      "num_input_tokens_seen": 28500112,
      "step": 13650
    },
    {
      "epoch": 2.227587894608043,
      "grad_norm": 3.65625,
      "learning_rate": 3.695082010912098e-05,
      "loss": 1.927,
      "num_input_tokens_seen": 28511216,
      "step": 13655
    },
    {
      "epoch": 2.2284036218288605,
      "grad_norm": 8.5,
      "learning_rate": 3.694219608561217e-05,
      "loss": 3.0202,
      "num_input_tokens_seen": 28521776,
      "step": 13660
    },
    {
      "epoch": 2.2292193490496777,
      "grad_norm": 5.25,
      "learning_rate": 3.693357022046665e-05,
      "loss": 2.5021,
      "num_input_tokens_seen": 28531920,
      "step": 13665
    },
    {
      "epoch": 2.2300350762704952,
      "grad_norm": 11.625,
      "learning_rate": 3.6924942515014644e-05,
      "loss": 3.6784,
      "num_input_tokens_seen": 28543856,
      "step": 13670
    },
    {
      "epoch": 2.2308508034913124,
      "grad_norm": 5.09375,
      "learning_rate": 3.691631297058664e-05,
      "loss": 2.7551,
      "num_input_tokens_seen": 28554288,
      "step": 13675
    },
    {
      "epoch": 2.23166653071213,
      "grad_norm": 8.375,
      "learning_rate": 3.6907681588513424e-05,
      "loss": 2.895,
      "num_input_tokens_seen": 28564352,
      "step": 13680
    },
    {
      "epoch": 2.232482257932947,
      "grad_norm": 6.78125,
      "learning_rate": 3.689904837012606e-05,
      "loss": 2.7676,
      "num_input_tokens_seen": 28575712,
      "step": 13685
    },
    {
      "epoch": 2.2332979851537647,
      "grad_norm": 8.625,
      "learning_rate": 3.689041331675591e-05,
      "loss": 2.0033,
      "num_input_tokens_seen": 28587104,
      "step": 13690
    },
    {
      "epoch": 2.234113712374582,
      "grad_norm": 7.4375,
      "learning_rate": 3.688177642973461e-05,
      "loss": 2.4829,
      "num_input_tokens_seen": 28597840,
      "step": 13695
    },
    {
      "epoch": 2.2349294395953994,
      "grad_norm": 8.875,
      "learning_rate": 3.687313771039406e-05,
      "loss": 2.3574,
      "num_input_tokens_seen": 28608656,
      "step": 13700
    },
    {
      "epoch": 2.2357451668162165,
      "grad_norm": 7.28125,
      "learning_rate": 3.686449716006647e-05,
      "loss": 2.884,
      "num_input_tokens_seen": 28618576,
      "step": 13705
    },
    {
      "epoch": 2.236560894037034,
      "grad_norm": 7.875,
      "learning_rate": 3.685585478008432e-05,
      "loss": 2.5909,
      "num_input_tokens_seen": 28629008,
      "step": 13710
    },
    {
      "epoch": 2.237376621257851,
      "grad_norm": 17.375,
      "learning_rate": 3.6847210571780364e-05,
      "loss": 2.6273,
      "num_input_tokens_seen": 28639072,
      "step": 13715
    },
    {
      "epoch": 2.2381923484786688,
      "grad_norm": 6.9375,
      "learning_rate": 3.683856453648767e-05,
      "loss": 2.9438,
      "num_input_tokens_seen": 28648304,
      "step": 13720
    },
    {
      "epoch": 2.239008075699486,
      "grad_norm": 3.796875,
      "learning_rate": 3.682991667553954e-05,
      "loss": 2.1269,
      "num_input_tokens_seen": 28659168,
      "step": 13725
    },
    {
      "epoch": 2.2398238029203035,
      "grad_norm": 9.8125,
      "learning_rate": 3.6821266990269606e-05,
      "loss": 3.2685,
      "num_input_tokens_seen": 28668560,
      "step": 13730
    },
    {
      "epoch": 2.2406395301411206,
      "grad_norm": 10.3125,
      "learning_rate": 3.681261548201174e-05,
      "loss": 3.126,
      "num_input_tokens_seen": 28679888,
      "step": 13735
    },
    {
      "epoch": 2.241455257361938,
      "grad_norm": 10.0625,
      "learning_rate": 3.6803962152100125e-05,
      "loss": 2.3718,
      "num_input_tokens_seen": 28688992,
      "step": 13740
    },
    {
      "epoch": 2.2422709845827553,
      "grad_norm": 9.375,
      "learning_rate": 3.67953070018692e-05,
      "loss": 2.9371,
      "num_input_tokens_seen": 28700704,
      "step": 13745
    },
    {
      "epoch": 2.243086711803573,
      "grad_norm": 7.46875,
      "learning_rate": 3.678665003265371e-05,
      "loss": 1.72,
      "num_input_tokens_seen": 28711392,
      "step": 13750
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 10.6875,
      "learning_rate": 3.677799124578867e-05,
      "loss": 2.8431,
      "num_input_tokens_seen": 28721920,
      "step": 13755
    },
    {
      "epoch": 2.2447181662452076,
      "grad_norm": 10.25,
      "learning_rate": 3.676933064260937e-05,
      "loss": 1.663,
      "num_input_tokens_seen": 28733392,
      "step": 13760
    },
    {
      "epoch": 2.245533893466025,
      "grad_norm": 0.21484375,
      "learning_rate": 3.6760668224451365e-05,
      "loss": 1.6227,
      "num_input_tokens_seen": 28743088,
      "step": 13765
    },
    {
      "epoch": 2.2463496206868423,
      "grad_norm": 10.5,
      "learning_rate": 3.675200399265054e-05,
      "loss": 2.1069,
      "num_input_tokens_seen": 28752656,
      "step": 13770
    },
    {
      "epoch": 2.24716534790766,
      "grad_norm": 9.0,
      "learning_rate": 3.6743337948543014e-05,
      "loss": 2.1198,
      "num_input_tokens_seen": 28763440,
      "step": 13775
    },
    {
      "epoch": 2.247981075128477,
      "grad_norm": 2.328125,
      "learning_rate": 3.6734670093465204e-05,
      "loss": 1.2295,
      "num_input_tokens_seen": 28774128,
      "step": 13780
    },
    {
      "epoch": 2.2487968023492946,
      "grad_norm": 5.5,
      "learning_rate": 3.672600042875379e-05,
      "loss": 2.0569,
      "num_input_tokens_seen": 28782768,
      "step": 13785
    },
    {
      "epoch": 2.2496125295701117,
      "grad_norm": 6.84375,
      "learning_rate": 3.671732895574575e-05,
      "loss": 2.03,
      "num_input_tokens_seen": 28792224,
      "step": 13790
    },
    {
      "epoch": 2.2504282567909293,
      "grad_norm": 7.03125,
      "learning_rate": 3.670865567577834e-05,
      "loss": 2.5284,
      "num_input_tokens_seen": 28801664,
      "step": 13795
    },
    {
      "epoch": 2.2512439840117464,
      "grad_norm": 5.21875,
      "learning_rate": 3.669998059018909e-05,
      "loss": 3.2633,
      "num_input_tokens_seen": 28812800,
      "step": 13800
    },
    {
      "epoch": 2.2512439840117464,
      "eval_loss": 2.545050859451294,
      "eval_runtime": 134.9252,
      "eval_samples_per_second": 20.196,
      "eval_steps_per_second": 10.102,
      "num_input_tokens_seen": 28812800,
      "step": 13800
    },
    {
      "epoch": 2.252059711232564,
      "grad_norm": 1.6953125,
      "learning_rate": 3.6691303700315796e-05,
      "loss": 1.9925,
      "num_input_tokens_seen": 28822224,
      "step": 13805
    },
    {
      "epoch": 2.252875438453381,
      "grad_norm": 0.11962890625,
      "learning_rate": 3.668262500749655e-05,
      "loss": 2.4546,
      "num_input_tokens_seen": 28831536,
      "step": 13810
    },
    {
      "epoch": 2.2536911656741987,
      "grad_norm": 0.640625,
      "learning_rate": 3.667394451306971e-05,
      "loss": 2.2364,
      "num_input_tokens_seen": 28841728,
      "step": 13815
    },
    {
      "epoch": 2.254506892895016,
      "grad_norm": 11.375,
      "learning_rate": 3.666526221837393e-05,
      "loss": 3.1414,
      "num_input_tokens_seen": 28852208,
      "step": 13820
    },
    {
      "epoch": 2.2553226201158334,
      "grad_norm": 11.4375,
      "learning_rate": 3.665657812474812e-05,
      "loss": 2.4102,
      "num_input_tokens_seen": 28862912,
      "step": 13825
    },
    {
      "epoch": 2.2561383473366505,
      "grad_norm": 8.6875,
      "learning_rate": 3.664789223353147e-05,
      "loss": 4.0906,
      "num_input_tokens_seen": 28872560,
      "step": 13830
    },
    {
      "epoch": 2.256954074557468,
      "grad_norm": 7.6875,
      "learning_rate": 3.663920454606347e-05,
      "loss": 4.0156,
      "num_input_tokens_seen": 28884400,
      "step": 13835
    },
    {
      "epoch": 2.2577698017782852,
      "grad_norm": 4.1875,
      "learning_rate": 3.6630515063683856e-05,
      "loss": 3.4722,
      "num_input_tokens_seen": 28892976,
      "step": 13840
    },
    {
      "epoch": 2.258585528999103,
      "grad_norm": 4.59375,
      "learning_rate": 3.662182378773267e-05,
      "loss": 2.3658,
      "num_input_tokens_seen": 28904016,
      "step": 13845
    },
    {
      "epoch": 2.25940125621992,
      "grad_norm": 0.2158203125,
      "learning_rate": 3.66131307195502e-05,
      "loss": 2.4775,
      "num_input_tokens_seen": 28914288,
      "step": 13850
    },
    {
      "epoch": 2.2602169834407375,
      "grad_norm": 4.8125,
      "learning_rate": 3.6604435860477034e-05,
      "loss": 1.985,
      "num_input_tokens_seen": 28923968,
      "step": 13855
    },
    {
      "epoch": 2.2610327106615546,
      "grad_norm": 4.75,
      "learning_rate": 3.6595739211854025e-05,
      "loss": 2.2976,
      "num_input_tokens_seen": 28936352,
      "step": 13860
    },
    {
      "epoch": 2.261848437882372,
      "grad_norm": 6.5,
      "learning_rate": 3.658704077502231e-05,
      "loss": 1.6449,
      "num_input_tokens_seen": 28945232,
      "step": 13865
    },
    {
      "epoch": 2.2626641651031894,
      "grad_norm": 7.09375,
      "learning_rate": 3.65783405513233e-05,
      "loss": 1.9578,
      "num_input_tokens_seen": 28954992,
      "step": 13870
    },
    {
      "epoch": 2.263479892324007,
      "grad_norm": 10.625,
      "learning_rate": 3.656963854209867e-05,
      "loss": 3.1532,
      "num_input_tokens_seen": 28965136,
      "step": 13875
    },
    {
      "epoch": 2.264295619544824,
      "grad_norm": 7.3125,
      "learning_rate": 3.656093474869038e-05,
      "loss": 2.488,
      "num_input_tokens_seen": 28975520,
      "step": 13880
    },
    {
      "epoch": 2.2651113467656416,
      "grad_norm": 1.8359375,
      "learning_rate": 3.655222917244068e-05,
      "loss": 1.8568,
      "num_input_tokens_seen": 28985424,
      "step": 13885
    },
    {
      "epoch": 2.2659270739864588,
      "grad_norm": 7.0625,
      "learning_rate": 3.6543521814692054e-05,
      "loss": 1.8749,
      "num_input_tokens_seen": 28995040,
      "step": 13890
    },
    {
      "epoch": 2.2667428012072763,
      "grad_norm": 4.0625,
      "learning_rate": 3.653481267678731e-05,
      "loss": 1.5914,
      "num_input_tokens_seen": 29005008,
      "step": 13895
    },
    {
      "epoch": 2.2675585284280935,
      "grad_norm": 7.3125,
      "learning_rate": 3.652610176006949e-05,
      "loss": 1.7404,
      "num_input_tokens_seen": 29014944,
      "step": 13900
    },
    {
      "epoch": 2.268374255648911,
      "grad_norm": 7.90625,
      "learning_rate": 3.6517389065881925e-05,
      "loss": 2.9863,
      "num_input_tokens_seen": 29024880,
      "step": 13905
    },
    {
      "epoch": 2.269189982869728,
      "grad_norm": 8.4375,
      "learning_rate": 3.650867459556824e-05,
      "loss": 1.5652,
      "num_input_tokens_seen": 29034784,
      "step": 13910
    },
    {
      "epoch": 2.2700057100905457,
      "grad_norm": 5.0,
      "learning_rate": 3.64999583504723e-05,
      "loss": 2.2973,
      "num_input_tokens_seen": 29043744,
      "step": 13915
    },
    {
      "epoch": 2.270821437311363,
      "grad_norm": 16.125,
      "learning_rate": 3.649124033193827e-05,
      "loss": 1.6257,
      "num_input_tokens_seen": 29054128,
      "step": 13920
    },
    {
      "epoch": 2.2716371645321805,
      "grad_norm": 0.095703125,
      "learning_rate": 3.648252054131057e-05,
      "loss": 2.1929,
      "num_input_tokens_seen": 29063200,
      "step": 13925
    },
    {
      "epoch": 2.272452891752998,
      "grad_norm": 4.34375,
      "learning_rate": 3.647379897993391e-05,
      "loss": 0.9812,
      "num_input_tokens_seen": 29073488,
      "step": 13930
    },
    {
      "epoch": 2.273268618973815,
      "grad_norm": 2.09375,
      "learning_rate": 3.646507564915325e-05,
      "loss": 1.989,
      "num_input_tokens_seen": 29082896,
      "step": 13935
    },
    {
      "epoch": 2.2740843461946323,
      "grad_norm": 6.96875,
      "learning_rate": 3.645635055031385e-05,
      "loss": 2.7019,
      "num_input_tokens_seen": 29093968,
      "step": 13940
    },
    {
      "epoch": 2.27490007341545,
      "grad_norm": 10.1875,
      "learning_rate": 3.6447623684761224e-05,
      "loss": 3.7933,
      "num_input_tokens_seen": 29103984,
      "step": 13945
    },
    {
      "epoch": 2.2757158006362674,
      "grad_norm": 8.125,
      "learning_rate": 3.643889505384117e-05,
      "loss": 2.7107,
      "num_input_tokens_seen": 29115280,
      "step": 13950
    },
    {
      "epoch": 2.2765315278570846,
      "grad_norm": 7.875,
      "learning_rate": 3.6430164658899744e-05,
      "loss": 2.807,
      "num_input_tokens_seen": 29126464,
      "step": 13955
    },
    {
      "epoch": 2.2773472550779017,
      "grad_norm": 7.5,
      "learning_rate": 3.642143250128329e-05,
      "loss": 2.0492,
      "num_input_tokens_seen": 29137424,
      "step": 13960
    },
    {
      "epoch": 2.2781629822987193,
      "grad_norm": 2.28125,
      "learning_rate": 3.641269858233841e-05,
      "loss": 1.1254,
      "num_input_tokens_seen": 29148448,
      "step": 13965
    },
    {
      "epoch": 2.278978709519537,
      "grad_norm": 9.5,
      "learning_rate": 3.640396290341199e-05,
      "loss": 2.2028,
      "num_input_tokens_seen": 29159104,
      "step": 13970
    },
    {
      "epoch": 2.279794436740354,
      "grad_norm": 9.3125,
      "learning_rate": 3.639522546585118e-05,
      "loss": 2.7009,
      "num_input_tokens_seen": 29169232,
      "step": 13975
    },
    {
      "epoch": 2.2806101639611716,
      "grad_norm": 9.5,
      "learning_rate": 3.6386486271003404e-05,
      "loss": 2.7676,
      "num_input_tokens_seen": 29179232,
      "step": 13980
    },
    {
      "epoch": 2.2814258911819887,
      "grad_norm": 11.0,
      "learning_rate": 3.6377745320216346e-05,
      "loss": 3.3726,
      "num_input_tokens_seen": 29189408,
      "step": 13985
    },
    {
      "epoch": 2.2822416184028063,
      "grad_norm": 7.75,
      "learning_rate": 3.636900261483798e-05,
      "loss": 1.8711,
      "num_input_tokens_seen": 29200160,
      "step": 13990
    },
    {
      "epoch": 2.2830573456236234,
      "grad_norm": 7.5,
      "learning_rate": 3.636025815621654e-05,
      "loss": 2.4261,
      "num_input_tokens_seen": 29211872,
      "step": 13995
    },
    {
      "epoch": 2.283873072844441,
      "grad_norm": 8.9375,
      "learning_rate": 3.635151194570054e-05,
      "loss": 2.5096,
      "num_input_tokens_seen": 29222656,
      "step": 14000
    },
    {
      "epoch": 2.283873072844441,
      "eval_loss": 2.546114921569824,
      "eval_runtime": 135.0346,
      "eval_samples_per_second": 20.18,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 29222656,
      "step": 14000
    },
    {
      "epoch": 2.284688800065258,
      "grad_norm": 8.625,
      "learning_rate": 3.634276398463873e-05,
      "loss": 2.4827,
      "num_input_tokens_seen": 29233840,
      "step": 14005
    },
    {
      "epoch": 2.2855045272860757,
      "grad_norm": 6.21875,
      "learning_rate": 3.633401427438018e-05,
      "loss": 2.0139,
      "num_input_tokens_seen": 29244288,
      "step": 14010
    },
    {
      "epoch": 2.286320254506893,
      "grad_norm": 7.40625,
      "learning_rate": 3.63252628162742e-05,
      "loss": 2.8611,
      "num_input_tokens_seen": 29253968,
      "step": 14015
    },
    {
      "epoch": 2.2871359817277104,
      "grad_norm": 3.265625,
      "learning_rate": 3.6316509611670364e-05,
      "loss": 2.3223,
      "num_input_tokens_seen": 29265168,
      "step": 14020
    },
    {
      "epoch": 2.2879517089485275,
      "grad_norm": 5.96875,
      "learning_rate": 3.630775466191854e-05,
      "loss": 3.226,
      "num_input_tokens_seen": 29274496,
      "step": 14025
    },
    {
      "epoch": 2.288767436169345,
      "grad_norm": 8.0625,
      "learning_rate": 3.629899796836884e-05,
      "loss": 3.2147,
      "num_input_tokens_seen": 29284496,
      "step": 14030
    },
    {
      "epoch": 2.289583163390162,
      "grad_norm": 7.96875,
      "learning_rate": 3.6290239532371666e-05,
      "loss": 2.1345,
      "num_input_tokens_seen": 29295264,
      "step": 14035
    },
    {
      "epoch": 2.29039889061098,
      "grad_norm": 4.5625,
      "learning_rate": 3.628147935527767e-05,
      "loss": 1.7201,
      "num_input_tokens_seen": 29305280,
      "step": 14040
    },
    {
      "epoch": 2.291214617831797,
      "grad_norm": 7.375,
      "learning_rate": 3.627271743843779e-05,
      "loss": 1.9968,
      "num_input_tokens_seen": 29314640,
      "step": 14045
    },
    {
      "epoch": 2.2920303450526145,
      "grad_norm": 13.0625,
      "learning_rate": 3.626395378320321e-05,
      "loss": 3.2169,
      "num_input_tokens_seen": 29325504,
      "step": 14050
    },
    {
      "epoch": 2.2928460722734316,
      "grad_norm": 3.671875,
      "learning_rate": 3.625518839092541e-05,
      "loss": 1.8564,
      "num_input_tokens_seen": 29335632,
      "step": 14055
    },
    {
      "epoch": 2.293661799494249,
      "grad_norm": 7.6875,
      "learning_rate": 3.624642126295612e-05,
      "loss": 3.6122,
      "num_input_tokens_seen": 29346896,
      "step": 14060
    },
    {
      "epoch": 2.2944775267150663,
      "grad_norm": 3.609375,
      "learning_rate": 3.6237652400647345e-05,
      "loss": 1.6163,
      "num_input_tokens_seen": 29357008,
      "step": 14065
    },
    {
      "epoch": 2.295293253935884,
      "grad_norm": 8.1875,
      "learning_rate": 3.622888180535134e-05,
      "loss": 1.5268,
      "num_input_tokens_seen": 29366960,
      "step": 14070
    },
    {
      "epoch": 2.296108981156701,
      "grad_norm": 5.8125,
      "learning_rate": 3.6220109478420655e-05,
      "loss": 2.4028,
      "num_input_tokens_seen": 29377664,
      "step": 14075
    },
    {
      "epoch": 2.2969247083775186,
      "grad_norm": 4.0625,
      "learning_rate": 3.6211335421208084e-05,
      "loss": 2.1865,
      "num_input_tokens_seen": 29388528,
      "step": 14080
    },
    {
      "epoch": 2.2977404355983357,
      "grad_norm": 5.5,
      "learning_rate": 3.62025596350667e-05,
      "loss": 1.6842,
      "num_input_tokens_seen": 29398960,
      "step": 14085
    },
    {
      "epoch": 2.2985561628191533,
      "grad_norm": 10.375,
      "learning_rate": 3.619378212134984e-05,
      "loss": 3.0079,
      "num_input_tokens_seen": 29409936,
      "step": 14090
    },
    {
      "epoch": 2.2993718900399704,
      "grad_norm": 16.375,
      "learning_rate": 3.618500288141111e-05,
      "loss": 2.3269,
      "num_input_tokens_seen": 29421136,
      "step": 14095
    },
    {
      "epoch": 2.300187617260788,
      "grad_norm": 7.15625,
      "learning_rate": 3.617622191660438e-05,
      "loss": 3.0474,
      "num_input_tokens_seen": 29431168,
      "step": 14100
    },
    {
      "epoch": 2.3010033444816056,
      "grad_norm": 5.625,
      "learning_rate": 3.616743922828377e-05,
      "loss": 2.4032,
      "num_input_tokens_seen": 29441936,
      "step": 14105
    },
    {
      "epoch": 2.3018190717024227,
      "grad_norm": 4.125,
      "learning_rate": 3.615865481780371e-05,
      "loss": 2.403,
      "num_input_tokens_seen": 29451808,
      "step": 14110
    },
    {
      "epoch": 2.30263479892324,
      "grad_norm": 9.1875,
      "learning_rate": 3.614986868651883e-05,
      "loss": 2.0912,
      "num_input_tokens_seen": 29462464,
      "step": 14115
    },
    {
      "epoch": 2.3034505261440574,
      "grad_norm": 7.125,
      "learning_rate": 3.614108083578409e-05,
      "loss": 1.6459,
      "num_input_tokens_seen": 29473552,
      "step": 14120
    },
    {
      "epoch": 2.304266253364875,
      "grad_norm": 2.21875,
      "learning_rate": 3.613229126695467e-05,
      "loss": 2.6505,
      "num_input_tokens_seen": 29482896,
      "step": 14125
    },
    {
      "epoch": 2.305081980585692,
      "grad_norm": 3.734375,
      "learning_rate": 3.612349998138605e-05,
      "loss": 2.4459,
      "num_input_tokens_seen": 29492912,
      "step": 14130
    },
    {
      "epoch": 2.3058977078065093,
      "grad_norm": 7.15625,
      "learning_rate": 3.6114706980433946e-05,
      "loss": 3.7244,
      "num_input_tokens_seen": 29503536,
      "step": 14135
    },
    {
      "epoch": 2.306713435027327,
      "grad_norm": 4.40625,
      "learning_rate": 3.610591226545435e-05,
      "loss": 2.1316,
      "num_input_tokens_seen": 29513328,
      "step": 14140
    },
    {
      "epoch": 2.3075291622481444,
      "grad_norm": 6.125,
      "learning_rate": 3.6097115837803505e-05,
      "loss": 1.7471,
      "num_input_tokens_seen": 29523856,
      "step": 14145
    },
    {
      "epoch": 2.3083448894689615,
      "grad_norm": 12.5625,
      "learning_rate": 3.608831769883795e-05,
      "loss": 3.4881,
      "num_input_tokens_seen": 29535504,
      "step": 14150
    },
    {
      "epoch": 2.309160616689779,
      "grad_norm": 5.15625,
      "learning_rate": 3.607951784991446e-05,
      "loss": 2.2863,
      "num_input_tokens_seen": 29545376,
      "step": 14155
    },
    {
      "epoch": 2.3099763439105963,
      "grad_norm": 6.84375,
      "learning_rate": 3.6070716292390085e-05,
      "loss": 1.7744,
      "num_input_tokens_seen": 29554848,
      "step": 14160
    },
    {
      "epoch": 2.310792071131414,
      "grad_norm": 4.875,
      "learning_rate": 3.606191302762213e-05,
      "loss": 1.903,
      "num_input_tokens_seen": 29566080,
      "step": 14165
    },
    {
      "epoch": 2.311607798352231,
      "grad_norm": 4.84375,
      "learning_rate": 3.605310805696818e-05,
      "loss": 2.5293,
      "num_input_tokens_seen": 29575968,
      "step": 14170
    },
    {
      "epoch": 2.3124235255730485,
      "grad_norm": 12.375,
      "learning_rate": 3.6044301381786067e-05,
      "loss": 2.1236,
      "num_input_tokens_seen": 29586512,
      "step": 14175
    },
    {
      "epoch": 2.3132392527938657,
      "grad_norm": 6.8125,
      "learning_rate": 3.6035493003433883e-05,
      "loss": 3.2138,
      "num_input_tokens_seen": 29598816,
      "step": 14180
    },
    {
      "epoch": 2.3140549800146832,
      "grad_norm": 14.25,
      "learning_rate": 3.6026682923269994e-05,
      "loss": 3.3737,
      "num_input_tokens_seen": 29608768,
      "step": 14185
    },
    {
      "epoch": 2.3148707072355004,
      "grad_norm": 9.6875,
      "learning_rate": 3.6017871142653034e-05,
      "loss": 2.3755,
      "num_input_tokens_seen": 29620032,
      "step": 14190
    },
    {
      "epoch": 2.315686434456318,
      "grad_norm": 5.34375,
      "learning_rate": 3.600905766294189e-05,
      "loss": 1.867,
      "num_input_tokens_seen": 29631728,
      "step": 14195
    },
    {
      "epoch": 2.316502161677135,
      "grad_norm": 4.25,
      "learning_rate": 3.60002424854957e-05,
      "loss": 2.9558,
      "num_input_tokens_seen": 29642224,
      "step": 14200
    },
    {
      "epoch": 2.316502161677135,
      "eval_loss": 2.5656850337982178,
      "eval_runtime": 135.0129,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 29642224,
      "step": 14200
    },
    {
      "epoch": 2.3173178888979526,
      "grad_norm": 3.8125,
      "learning_rate": 3.5991425611673876e-05,
      "loss": 2.5722,
      "num_input_tokens_seen": 29653872,
      "step": 14205
    },
    {
      "epoch": 2.31813361611877,
      "grad_norm": 13.75,
      "learning_rate": 3.5982607042836105e-05,
      "loss": 3.2859,
      "num_input_tokens_seen": 29664560,
      "step": 14210
    },
    {
      "epoch": 2.3189493433395874,
      "grad_norm": 7.0,
      "learning_rate": 3.597378678034231e-05,
      "loss": 1.4872,
      "num_input_tokens_seen": 29676496,
      "step": 14215
    },
    {
      "epoch": 2.3197650705604045,
      "grad_norm": 10.4375,
      "learning_rate": 3.596496482555269e-05,
      "loss": 2.7543,
      "num_input_tokens_seen": 29687824,
      "step": 14220
    },
    {
      "epoch": 2.320580797781222,
      "grad_norm": 3.28125,
      "learning_rate": 3.595614117982769e-05,
      "loss": 1.1508,
      "num_input_tokens_seen": 29698720,
      "step": 14225
    },
    {
      "epoch": 2.321396525002039,
      "grad_norm": 9.3125,
      "learning_rate": 3.594731584452805e-05,
      "loss": 2.5484,
      "num_input_tokens_seen": 29708240,
      "step": 14230
    },
    {
      "epoch": 2.3222122522228568,
      "grad_norm": 6.15625,
      "learning_rate": 3.593848882101472e-05,
      "loss": 1.8081,
      "num_input_tokens_seen": 29718976,
      "step": 14235
    },
    {
      "epoch": 2.323027979443674,
      "grad_norm": 6.46875,
      "learning_rate": 3.592966011064896e-05,
      "loss": 2.3083,
      "num_input_tokens_seen": 29728784,
      "step": 14240
    },
    {
      "epoch": 2.3238437066644915,
      "grad_norm": 13.25,
      "learning_rate": 3.592082971479226e-05,
      "loss": 2.5457,
      "num_input_tokens_seen": 29738880,
      "step": 14245
    },
    {
      "epoch": 2.3246594338853086,
      "grad_norm": 5.0625,
      "learning_rate": 3.5911997634806385e-05,
      "loss": 4.0108,
      "num_input_tokens_seen": 29750480,
      "step": 14250
    },
    {
      "epoch": 2.325475161106126,
      "grad_norm": 11.625,
      "learning_rate": 3.5903163872053336e-05,
      "loss": 2.8508,
      "num_input_tokens_seen": 29761520,
      "step": 14255
    },
    {
      "epoch": 2.3262908883269433,
      "grad_norm": 1.90625,
      "learning_rate": 3.58943284278954e-05,
      "loss": 2.3385,
      "num_input_tokens_seen": 29771872,
      "step": 14260
    },
    {
      "epoch": 2.327106615547761,
      "grad_norm": 2.75,
      "learning_rate": 3.588549130369512e-05,
      "loss": 0.9379,
      "num_input_tokens_seen": 29779600,
      "step": 14265
    },
    {
      "epoch": 2.327922342768578,
      "grad_norm": 5.21875,
      "learning_rate": 3.5876652500815274e-05,
      "loss": 1.7726,
      "num_input_tokens_seen": 29790640,
      "step": 14270
    },
    {
      "epoch": 2.3287380699893956,
      "grad_norm": 4.5,
      "learning_rate": 3.586781202061894e-05,
      "loss": 3.997,
      "num_input_tokens_seen": 29800640,
      "step": 14275
    },
    {
      "epoch": 2.3295537972102127,
      "grad_norm": 5.53125,
      "learning_rate": 3.585896986446942e-05,
      "loss": 2.3434,
      "num_input_tokens_seen": 29811728,
      "step": 14280
    },
    {
      "epoch": 2.3303695244310303,
      "grad_norm": 7.46875,
      "learning_rate": 3.585012603373028e-05,
      "loss": 2.1559,
      "num_input_tokens_seen": 29822288,
      "step": 14285
    },
    {
      "epoch": 2.3311852516518474,
      "grad_norm": 8.0625,
      "learning_rate": 3.584128052976535e-05,
      "loss": 2.9035,
      "num_input_tokens_seen": 29833840,
      "step": 14290
    },
    {
      "epoch": 2.332000978872665,
      "grad_norm": 10.625,
      "learning_rate": 3.5832433353938724e-05,
      "loss": 1.7853,
      "num_input_tokens_seen": 29845760,
      "step": 14295
    },
    {
      "epoch": 2.3328167060934826,
      "grad_norm": 6.21875,
      "learning_rate": 3.5823584507614746e-05,
      "loss": 1.1862,
      "num_input_tokens_seen": 29854816,
      "step": 14300
    },
    {
      "epoch": 2.3336324333142997,
      "grad_norm": 6.65625,
      "learning_rate": 3.581473399215802e-05,
      "loss": 1.4872,
      "num_input_tokens_seen": 29864512,
      "step": 14305
    },
    {
      "epoch": 2.334448160535117,
      "grad_norm": 7.0,
      "learning_rate": 3.580588180893341e-05,
      "loss": 2.6081,
      "num_input_tokens_seen": 29874544,
      "step": 14310
    },
    {
      "epoch": 2.3352638877559344,
      "grad_norm": 9.5625,
      "learning_rate": 3.579702795930602e-05,
      "loss": 2.7013,
      "num_input_tokens_seen": 29883760,
      "step": 14315
    },
    {
      "epoch": 2.336079614976752,
      "grad_norm": 5.0625,
      "learning_rate": 3.578817244464125e-05,
      "loss": 2.9244,
      "num_input_tokens_seen": 29894480,
      "step": 14320
    },
    {
      "epoch": 2.336895342197569,
      "grad_norm": 2.171875,
      "learning_rate": 3.577931526630471e-05,
      "loss": 1.306,
      "num_input_tokens_seen": 29905280,
      "step": 14325
    },
    {
      "epoch": 2.3377110694183862,
      "grad_norm": 1.71875,
      "learning_rate": 3.577045642566229e-05,
      "loss": 1.1952,
      "num_input_tokens_seen": 29914832,
      "step": 14330
    },
    {
      "epoch": 2.338526796639204,
      "grad_norm": 4.0,
      "learning_rate": 3.576159592408014e-05,
      "loss": 2.855,
      "num_input_tokens_seen": 29925888,
      "step": 14335
    },
    {
      "epoch": 2.3393425238600214,
      "grad_norm": 0.2060546875,
      "learning_rate": 3.575273376292466e-05,
      "loss": 2.5254,
      "num_input_tokens_seen": 29936368,
      "step": 14340
    },
    {
      "epoch": 2.3401582510808385,
      "grad_norm": 14.0,
      "learning_rate": 3.574386994356251e-05,
      "loss": 3.0436,
      "num_input_tokens_seen": 29946672,
      "step": 14345
    },
    {
      "epoch": 2.340973978301656,
      "grad_norm": 9.25,
      "learning_rate": 3.573500446736059e-05,
      "loss": 1.878,
      "num_input_tokens_seen": 29956544,
      "step": 14350
    },
    {
      "epoch": 2.3417897055224732,
      "grad_norm": 9.0,
      "learning_rate": 3.5726137335686094e-05,
      "loss": 2.3938,
      "num_input_tokens_seen": 29966224,
      "step": 14355
    },
    {
      "epoch": 2.342605432743291,
      "grad_norm": 9.5625,
      "learning_rate": 3.571726854990642e-05,
      "loss": 2.4237,
      "num_input_tokens_seen": 29977936,
      "step": 14360
    },
    {
      "epoch": 2.343421159964108,
      "grad_norm": 4.5,
      "learning_rate": 3.570839811138925e-05,
      "loss": 1.9157,
      "num_input_tokens_seen": 29990224,
      "step": 14365
    },
    {
      "epoch": 2.3442368871849255,
      "grad_norm": 7.4375,
      "learning_rate": 3.569952602150252e-05,
      "loss": 1.1063,
      "num_input_tokens_seen": 30000672,
      "step": 14370
    },
    {
      "epoch": 2.3450526144057426,
      "grad_norm": 3.1875,
      "learning_rate": 3.569065228161442e-05,
      "loss": 1.9741,
      "num_input_tokens_seen": 30011664,
      "step": 14375
    },
    {
      "epoch": 2.34586834162656,
      "grad_norm": 2.40625,
      "learning_rate": 3.5681776893093395e-05,
      "loss": 2.5175,
      "num_input_tokens_seen": 30023536,
      "step": 14380
    },
    {
      "epoch": 2.3466840688473773,
      "grad_norm": 8.625,
      "learning_rate": 3.5672899857308134e-05,
      "loss": 2.5688,
      "num_input_tokens_seen": 30034016,
      "step": 14385
    },
    {
      "epoch": 2.347499796068195,
      "grad_norm": 6.5,
      "learning_rate": 3.566402117562759e-05,
      "loss": 3.0788,
      "num_input_tokens_seen": 30045520,
      "step": 14390
    },
    {
      "epoch": 2.348315523289012,
      "grad_norm": 8.6875,
      "learning_rate": 3.565514084942097e-05,
      "loss": 1.8852,
      "num_input_tokens_seen": 30055440,
      "step": 14395
    },
    {
      "epoch": 2.3491312505098296,
      "grad_norm": 8.25,
      "learning_rate": 3.564625888005773e-05,
      "loss": 3.4931,
      "num_input_tokens_seen": 30064704,
      "step": 14400
    },
    {
      "epoch": 2.3491312505098296,
      "eval_loss": 2.555907726287842,
      "eval_runtime": 134.9015,
      "eval_samples_per_second": 20.2,
      "eval_steps_per_second": 10.104,
      "num_input_tokens_seen": 30064704,
      "step": 14400
    },
    {
      "epoch": 2.3499469777306468,
      "grad_norm": 3.609375,
      "learning_rate": 3.563737526890759e-05,
      "loss": 1.3909,
      "num_input_tokens_seen": 30074832,
      "step": 14405
    },
    {
      "epoch": 2.3507627049514643,
      "grad_norm": 1.8125,
      "learning_rate": 3.562849001734049e-05,
      "loss": 1.7463,
      "num_input_tokens_seen": 30084592,
      "step": 14410
    },
    {
      "epoch": 2.3515784321722815,
      "grad_norm": 9.9375,
      "learning_rate": 3.561960312672667e-05,
      "loss": 1.4989,
      "num_input_tokens_seen": 30095216,
      "step": 14415
    },
    {
      "epoch": 2.352394159393099,
      "grad_norm": 6.84375,
      "learning_rate": 3.5610714598436596e-05,
      "loss": 1.7822,
      "num_input_tokens_seen": 30106320,
      "step": 14420
    },
    {
      "epoch": 2.353209886613916,
      "grad_norm": 7.84375,
      "learning_rate": 3.5601824433840986e-05,
      "loss": 3.1477,
      "num_input_tokens_seen": 30116320,
      "step": 14425
    },
    {
      "epoch": 2.3540256138347337,
      "grad_norm": 9.4375,
      "learning_rate": 3.559293263431082e-05,
      "loss": 2.4141,
      "num_input_tokens_seen": 30127712,
      "step": 14430
    },
    {
      "epoch": 2.354841341055551,
      "grad_norm": 7.28125,
      "learning_rate": 3.558403920121732e-05,
      "loss": 1.5888,
      "num_input_tokens_seen": 30137312,
      "step": 14435
    },
    {
      "epoch": 2.3556570682763684,
      "grad_norm": 6.46875,
      "learning_rate": 3.557514413593197e-05,
      "loss": 2.4824,
      "num_input_tokens_seen": 30148848,
      "step": 14440
    },
    {
      "epoch": 2.3564727954971856,
      "grad_norm": 3.234375,
      "learning_rate": 3.55662474398265e-05,
      "loss": 1.4398,
      "num_input_tokens_seen": 30158944,
      "step": 14445
    },
    {
      "epoch": 2.357288522718003,
      "grad_norm": 8.0,
      "learning_rate": 3.555734911427288e-05,
      "loss": 2.9362,
      "num_input_tokens_seen": 30167920,
      "step": 14450
    },
    {
      "epoch": 2.3581042499388203,
      "grad_norm": 6.78125,
      "learning_rate": 3.5548449160643363e-05,
      "loss": 1.6401,
      "num_input_tokens_seen": 30179664,
      "step": 14455
    },
    {
      "epoch": 2.358919977159638,
      "grad_norm": 3.515625,
      "learning_rate": 3.553954758031043e-05,
      "loss": 2.514,
      "num_input_tokens_seen": 30191488,
      "step": 14460
    },
    {
      "epoch": 2.359735704380455,
      "grad_norm": 6.625,
      "learning_rate": 3.5530644374646815e-05,
      "loss": 2.1916,
      "num_input_tokens_seen": 30202240,
      "step": 14465
    },
    {
      "epoch": 2.3605514316012726,
      "grad_norm": 14.25,
      "learning_rate": 3.552173954502549e-05,
      "loss": 2.2845,
      "num_input_tokens_seen": 30212496,
      "step": 14470
    },
    {
      "epoch": 2.36136715882209,
      "grad_norm": 5.15625,
      "learning_rate": 3.55128330928197e-05,
      "loss": 1.33,
      "num_input_tokens_seen": 30221872,
      "step": 14475
    },
    {
      "epoch": 2.3621828860429073,
      "grad_norm": 8.375,
      "learning_rate": 3.550392501940294e-05,
      "loss": 3.4351,
      "num_input_tokens_seen": 30232960,
      "step": 14480
    },
    {
      "epoch": 2.3629986132637244,
      "grad_norm": 3.046875,
      "learning_rate": 3.5495015326148945e-05,
      "loss": 2.874,
      "num_input_tokens_seen": 30244256,
      "step": 14485
    },
    {
      "epoch": 2.363814340484542,
      "grad_norm": 6.0625,
      "learning_rate": 3.548610401443169e-05,
      "loss": 2.5433,
      "num_input_tokens_seen": 30253104,
      "step": 14490
    },
    {
      "epoch": 2.3646300677053596,
      "grad_norm": 4.1875,
      "learning_rate": 3.547719108562543e-05,
      "loss": 2.1104,
      "num_input_tokens_seen": 30264096,
      "step": 14495
    },
    {
      "epoch": 2.3654457949261767,
      "grad_norm": 4.34375,
      "learning_rate": 3.546827654110464e-05,
      "loss": 1.8633,
      "num_input_tokens_seen": 30273744,
      "step": 14500
    },
    {
      "epoch": 2.366261522146994,
      "grad_norm": 12.5,
      "learning_rate": 3.545936038224405e-05,
      "loss": 2.1202,
      "num_input_tokens_seen": 30283232,
      "step": 14505
    },
    {
      "epoch": 2.3670772493678114,
      "grad_norm": 6.09375,
      "learning_rate": 3.545044261041864e-05,
      "loss": 1.4335,
      "num_input_tokens_seen": 30293040,
      "step": 14510
    },
    {
      "epoch": 2.367892976588629,
      "grad_norm": 6.78125,
      "learning_rate": 3.5441523227003657e-05,
      "loss": 2.5401,
      "num_input_tokens_seen": 30303968,
      "step": 14515
    },
    {
      "epoch": 2.368708703809446,
      "grad_norm": 9.25,
      "learning_rate": 3.543260223337459e-05,
      "loss": 2.9432,
      "num_input_tokens_seen": 30314576,
      "step": 14520
    },
    {
      "epoch": 2.3695244310302637,
      "grad_norm": 3.90625,
      "learning_rate": 3.542367963090714e-05,
      "loss": 3.546,
      "num_input_tokens_seen": 30324208,
      "step": 14525
    },
    {
      "epoch": 2.370340158251081,
      "grad_norm": 5.09375,
      "learning_rate": 3.5414755420977295e-05,
      "loss": 1.7623,
      "num_input_tokens_seen": 30333680,
      "step": 14530
    },
    {
      "epoch": 2.3711558854718984,
      "grad_norm": 11.8125,
      "learning_rate": 3.54058296049613e-05,
      "loss": 2.7064,
      "num_input_tokens_seen": 30342384,
      "step": 14535
    },
    {
      "epoch": 2.3719716126927155,
      "grad_norm": 7.09375,
      "learning_rate": 3.53969021842356e-05,
      "loss": 3.1097,
      "num_input_tokens_seen": 30352784,
      "step": 14540
    },
    {
      "epoch": 2.372787339913533,
      "grad_norm": 10.625,
      "learning_rate": 3.5387973160176926e-05,
      "loss": 2.4309,
      "num_input_tokens_seen": 30362976,
      "step": 14545
    },
    {
      "epoch": 2.37360306713435,
      "grad_norm": 5.4375,
      "learning_rate": 3.537904253416224e-05,
      "loss": 2.2821,
      "num_input_tokens_seen": 30373696,
      "step": 14550
    },
    {
      "epoch": 2.374418794355168,
      "grad_norm": 8.1875,
      "learning_rate": 3.537011030756878e-05,
      "loss": 2.0931,
      "num_input_tokens_seen": 30384000,
      "step": 14555
    },
    {
      "epoch": 2.375234521575985,
      "grad_norm": 5.21875,
      "learning_rate": 3.536117648177399e-05,
      "loss": 2.6989,
      "num_input_tokens_seen": 30394352,
      "step": 14560
    },
    {
      "epoch": 2.3760502487968025,
      "grad_norm": 1.5859375,
      "learning_rate": 3.535224105815558e-05,
      "loss": 1.5693,
      "num_input_tokens_seen": 30405232,
      "step": 14565
    },
    {
      "epoch": 2.3768659760176196,
      "grad_norm": 2.75,
      "learning_rate": 3.5343304038091494e-05,
      "loss": 1.9615,
      "num_input_tokens_seen": 30415088,
      "step": 14570
    },
    {
      "epoch": 2.377681703238437,
      "grad_norm": 7.21875,
      "learning_rate": 3.5334365422959955e-05,
      "loss": 3.2998,
      "num_input_tokens_seen": 30426272,
      "step": 14575
    },
    {
      "epoch": 2.3784974304592543,
      "grad_norm": 14.4375,
      "learning_rate": 3.5325425214139396e-05,
      "loss": 2.9735,
      "num_input_tokens_seen": 30436848,
      "step": 14580
    },
    {
      "epoch": 2.379313157680072,
      "grad_norm": 1.921875,
      "learning_rate": 3.531648341300851e-05,
      "loss": 2.2361,
      "num_input_tokens_seen": 30448144,
      "step": 14585
    },
    {
      "epoch": 2.380128884900889,
      "grad_norm": 3.6875,
      "learning_rate": 3.530754002094623e-05,
      "loss": 2.6064,
      "num_input_tokens_seen": 30459360,
      "step": 14590
    },
    {
      "epoch": 2.3809446121217066,
      "grad_norm": 3.171875,
      "learning_rate": 3.529859503933175e-05,
      "loss": 0.8127,
      "num_input_tokens_seen": 30470560,
      "step": 14595
    },
    {
      "epoch": 2.3817603393425237,
      "grad_norm": 4.0625,
      "learning_rate": 3.52896484695445e-05,
      "loss": 1.6007,
      "num_input_tokens_seen": 30481488,
      "step": 14600
    },
    {
      "epoch": 2.3817603393425237,
      "eval_loss": 2.5469157695770264,
      "eval_runtime": 134.8148,
      "eval_samples_per_second": 20.213,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 30481488,
      "step": 14600
    },
    {
      "epoch": 2.3825760665633413,
      "grad_norm": 9.875,
      "learning_rate": 3.528070031296414e-05,
      "loss": 2.3966,
      "num_input_tokens_seen": 30491696,
      "step": 14605
    },
    {
      "epoch": 2.3833917937841584,
      "grad_norm": 2.5,
      "learning_rate": 3.5271750570970605e-05,
      "loss": 2.1099,
      "num_input_tokens_seen": 30501200,
      "step": 14610
    },
    {
      "epoch": 2.384207521004976,
      "grad_norm": 11.0,
      "learning_rate": 3.526279924494405e-05,
      "loss": 1.9646,
      "num_input_tokens_seen": 30510432,
      "step": 14615
    },
    {
      "epoch": 2.385023248225793,
      "grad_norm": 8.5,
      "learning_rate": 3.5253846336264874e-05,
      "loss": 1.996,
      "num_input_tokens_seen": 30520592,
      "step": 14620
    },
    {
      "epoch": 2.3858389754466107,
      "grad_norm": 3.796875,
      "learning_rate": 3.5244891846313736e-05,
      "loss": 2.7506,
      "num_input_tokens_seen": 30531296,
      "step": 14625
    },
    {
      "epoch": 2.386654702667428,
      "grad_norm": 6.5625,
      "learning_rate": 3.5235935776471527e-05,
      "loss": 2.2318,
      "num_input_tokens_seen": 30543184,
      "step": 14630
    },
    {
      "epoch": 2.3874704298882454,
      "grad_norm": 6.375,
      "learning_rate": 3.522697812811939e-05,
      "loss": 2.2106,
      "num_input_tokens_seen": 30553072,
      "step": 14635
    },
    {
      "epoch": 2.3882861571090626,
      "grad_norm": 7.75,
      "learning_rate": 3.521801890263871e-05,
      "loss": 2.7049,
      "num_input_tokens_seen": 30563600,
      "step": 14640
    },
    {
      "epoch": 2.38910188432988,
      "grad_norm": 7.53125,
      "learning_rate": 3.5209058101411114e-05,
      "loss": 2.2348,
      "num_input_tokens_seen": 30574320,
      "step": 14645
    },
    {
      "epoch": 2.3899176115506973,
      "grad_norm": 5.09375,
      "learning_rate": 3.520009572581845e-05,
      "loss": 2.4331,
      "num_input_tokens_seen": 30585232,
      "step": 14650
    },
    {
      "epoch": 2.390733338771515,
      "grad_norm": 7.09375,
      "learning_rate": 3.519113177724285e-05,
      "loss": 2.1535,
      "num_input_tokens_seen": 30596096,
      "step": 14655
    },
    {
      "epoch": 2.391549065992332,
      "grad_norm": 7.0,
      "learning_rate": 3.5182166257066656e-05,
      "loss": 2.0447,
      "num_input_tokens_seen": 30607264,
      "step": 14660
    },
    {
      "epoch": 2.3923647932131495,
      "grad_norm": 6.65625,
      "learning_rate": 3.517319916667247e-05,
      "loss": 1.7979,
      "num_input_tokens_seen": 30617344,
      "step": 14665
    },
    {
      "epoch": 2.393180520433967,
      "grad_norm": 4.375,
      "learning_rate": 3.516423050744313e-05,
      "loss": 1.6519,
      "num_input_tokens_seen": 30629600,
      "step": 14670
    },
    {
      "epoch": 2.3939962476547842,
      "grad_norm": 4.8125,
      "learning_rate": 3.5155260280761704e-05,
      "loss": 3.1939,
      "num_input_tokens_seen": 30640864,
      "step": 14675
    },
    {
      "epoch": 2.3948119748756014,
      "grad_norm": 5.28125,
      "learning_rate": 3.514628848801154e-05,
      "loss": 1.392,
      "num_input_tokens_seen": 30650272,
      "step": 14680
    },
    {
      "epoch": 2.395627702096419,
      "grad_norm": 7.28125,
      "learning_rate": 3.5137315130576174e-05,
      "loss": 3.4961,
      "num_input_tokens_seen": 30662080,
      "step": 14685
    },
    {
      "epoch": 2.3964434293172365,
      "grad_norm": 7.84375,
      "learning_rate": 3.512834020983942e-05,
      "loss": 2.0489,
      "num_input_tokens_seen": 30672416,
      "step": 14690
    },
    {
      "epoch": 2.3972591565380537,
      "grad_norm": 4.0625,
      "learning_rate": 3.5119363727185334e-05,
      "loss": 3.0647,
      "num_input_tokens_seen": 30683440,
      "step": 14695
    },
    {
      "epoch": 2.398074883758871,
      "grad_norm": 10.0625,
      "learning_rate": 3.511038568399819e-05,
      "loss": 2.162,
      "num_input_tokens_seen": 30693136,
      "step": 14700
    },
    {
      "epoch": 2.3988906109796884,
      "grad_norm": 4.59375,
      "learning_rate": 3.510140608166251e-05,
      "loss": 1.4769,
      "num_input_tokens_seen": 30702672,
      "step": 14705
    },
    {
      "epoch": 2.399706338200506,
      "grad_norm": 7.46875,
      "learning_rate": 3.509242492156308e-05,
      "loss": 2.4022,
      "num_input_tokens_seen": 30713680,
      "step": 14710
    },
    {
      "epoch": 2.400522065421323,
      "grad_norm": 6.125,
      "learning_rate": 3.5083442205084896e-05,
      "loss": 2.8953,
      "num_input_tokens_seen": 30724016,
      "step": 14715
    },
    {
      "epoch": 2.4013377926421406,
      "grad_norm": 8.9375,
      "learning_rate": 3.507445793361321e-05,
      "loss": 2.3662,
      "num_input_tokens_seen": 30734096,
      "step": 14720
    },
    {
      "epoch": 2.4021535198629578,
      "grad_norm": 7.5,
      "learning_rate": 3.5065472108533505e-05,
      "loss": 1.5555,
      "num_input_tokens_seen": 30745136,
      "step": 14725
    },
    {
      "epoch": 2.4029692470837754,
      "grad_norm": 7.0,
      "learning_rate": 3.5056484731231504e-05,
      "loss": 1.8827,
      "num_input_tokens_seen": 30756032,
      "step": 14730
    },
    {
      "epoch": 2.4037849743045925,
      "grad_norm": 14.5,
      "learning_rate": 3.504749580309319e-05,
      "loss": 2.306,
      "num_input_tokens_seen": 30766208,
      "step": 14735
    },
    {
      "epoch": 2.40460070152541,
      "grad_norm": 9.1875,
      "learning_rate": 3.5038505325504753e-05,
      "loss": 1.7345,
      "num_input_tokens_seen": 30776864,
      "step": 14740
    },
    {
      "epoch": 2.405416428746227,
      "grad_norm": 3.234375,
      "learning_rate": 3.502951329985264e-05,
      "loss": 2.6317,
      "num_input_tokens_seen": 30786800,
      "step": 14745
    },
    {
      "epoch": 2.4062321559670448,
      "grad_norm": 3.984375,
      "learning_rate": 3.502051972752354e-05,
      "loss": 2.0613,
      "num_input_tokens_seen": 30797072,
      "step": 14750
    },
    {
      "epoch": 2.407047883187862,
      "grad_norm": 3.484375,
      "learning_rate": 3.5011524609904374e-05,
      "loss": 2.445,
      "num_input_tokens_seen": 30808688,
      "step": 14755
    },
    {
      "epoch": 2.4078636104086795,
      "grad_norm": 8.125,
      "learning_rate": 3.50025279483823e-05,
      "loss": 4.2698,
      "num_input_tokens_seen": 30818016,
      "step": 14760
    },
    {
      "epoch": 2.4086793376294966,
      "grad_norm": 8.3125,
      "learning_rate": 3.499352974434472e-05,
      "loss": 3.152,
      "num_input_tokens_seen": 30828160,
      "step": 14765
    },
    {
      "epoch": 2.409495064850314,
      "grad_norm": 7.875,
      "learning_rate": 3.498452999917926e-05,
      "loss": 1.0195,
      "num_input_tokens_seen": 30837824,
      "step": 14770
    },
    {
      "epoch": 2.4103107920711313,
      "grad_norm": 4.1875,
      "learning_rate": 3.4975528714273795e-05,
      "loss": 2.1821,
      "num_input_tokens_seen": 30848544,
      "step": 14775
    },
    {
      "epoch": 2.411126519291949,
      "grad_norm": 5.3125,
      "learning_rate": 3.4966525891016454e-05,
      "loss": 1.8296,
      "num_input_tokens_seen": 30860304,
      "step": 14780
    },
    {
      "epoch": 2.411942246512766,
      "grad_norm": 7.34375,
      "learning_rate": 3.495752153079557e-05,
      "loss": 3.2565,
      "num_input_tokens_seen": 30871360,
      "step": 14785
    },
    {
      "epoch": 2.4127579737335836,
      "grad_norm": 4.03125,
      "learning_rate": 3.494851563499974e-05,
      "loss": 1.044,
      "num_input_tokens_seen": 30880800,
      "step": 14790
    },
    {
      "epoch": 2.4135737009544007,
      "grad_norm": 19.625,
      "learning_rate": 3.493950820501777e-05,
      "loss": 2.8518,
      "num_input_tokens_seen": 30890192,
      "step": 14795
    },
    {
      "epoch": 2.4143894281752183,
      "grad_norm": 3.375,
      "learning_rate": 3.493049924223872e-05,
      "loss": 2.3649,
      "num_input_tokens_seen": 30900976,
      "step": 14800
    },
    {
      "epoch": 2.4143894281752183,
      "eval_loss": 2.544508457183838,
      "eval_runtime": 135.118,
      "eval_samples_per_second": 20.168,
      "eval_steps_per_second": 10.087,
      "num_input_tokens_seen": 30900976,
      "step": 14800
    },
    {
      "epoch": 2.4152051553960354,
      "grad_norm": 8.0,
      "learning_rate": 3.49214887480519e-05,
      "loss": 3.0263,
      "num_input_tokens_seen": 30911120,
      "step": 14805
    },
    {
      "epoch": 2.416020882616853,
      "grad_norm": 0.68359375,
      "learning_rate": 3.4912476723846834e-05,
      "loss": 1.6809,
      "num_input_tokens_seen": 30921008,
      "step": 14810
    },
    {
      "epoch": 2.41683660983767,
      "grad_norm": 3.65625,
      "learning_rate": 3.490346317101328e-05,
      "loss": 2.7736,
      "num_input_tokens_seen": 30931664,
      "step": 14815
    },
    {
      "epoch": 2.4176523370584877,
      "grad_norm": 4.5,
      "learning_rate": 3.4894448090941266e-05,
      "loss": 1.6923,
      "num_input_tokens_seen": 30941824,
      "step": 14820
    },
    {
      "epoch": 2.418468064279305,
      "grad_norm": 5.375,
      "learning_rate": 3.488543148502101e-05,
      "loss": 1.6397,
      "num_input_tokens_seen": 30952608,
      "step": 14825
    },
    {
      "epoch": 2.4192837915001224,
      "grad_norm": 14.75,
      "learning_rate": 3.487641335464299e-05,
      "loss": 2.9483,
      "num_input_tokens_seen": 30960944,
      "step": 14830
    },
    {
      "epoch": 2.4200995187209395,
      "grad_norm": 5.25,
      "learning_rate": 3.4867393701197914e-05,
      "loss": 2.0904,
      "num_input_tokens_seen": 30972512,
      "step": 14835
    },
    {
      "epoch": 2.420915245941757,
      "grad_norm": 7.5,
      "learning_rate": 3.485837252607673e-05,
      "loss": 3.2152,
      "num_input_tokens_seen": 30983520,
      "step": 14840
    },
    {
      "epoch": 2.4217309731625742,
      "grad_norm": 6.15625,
      "learning_rate": 3.4849349830670615e-05,
      "loss": 2.866,
      "num_input_tokens_seen": 30993792,
      "step": 14845
    },
    {
      "epoch": 2.422546700383392,
      "grad_norm": 3.828125,
      "learning_rate": 3.4840325616370976e-05,
      "loss": 2.7403,
      "num_input_tokens_seen": 31003552,
      "step": 14850
    },
    {
      "epoch": 2.423362427604209,
      "grad_norm": 9.625,
      "learning_rate": 3.483129988456947e-05,
      "loss": 2.6481,
      "num_input_tokens_seen": 31013232,
      "step": 14855
    },
    {
      "epoch": 2.4241781548250265,
      "grad_norm": 12.75,
      "learning_rate": 3.482227263665797e-05,
      "loss": 2.1846,
      "num_input_tokens_seen": 31022432,
      "step": 14860
    },
    {
      "epoch": 2.424993882045844,
      "grad_norm": 12.8125,
      "learning_rate": 3.48132438740286e-05,
      "loss": 2.2602,
      "num_input_tokens_seen": 31034512,
      "step": 14865
    },
    {
      "epoch": 2.4258096092666612,
      "grad_norm": 10.125,
      "learning_rate": 3.48042135980737e-05,
      "loss": 2.8863,
      "num_input_tokens_seen": 31044704,
      "step": 14870
    },
    {
      "epoch": 2.4266253364874784,
      "grad_norm": 7.03125,
      "learning_rate": 3.479518181018586e-05,
      "loss": 2.6415,
      "num_input_tokens_seen": 31055280,
      "step": 14875
    },
    {
      "epoch": 2.427441063708296,
      "grad_norm": 7.65625,
      "learning_rate": 3.4786148511757886e-05,
      "loss": 1.4341,
      "num_input_tokens_seen": 31066656,
      "step": 14880
    },
    {
      "epoch": 2.4282567909291135,
      "grad_norm": 5.1875,
      "learning_rate": 3.477711370418284e-05,
      "loss": 2.6218,
      "num_input_tokens_seen": 31079120,
      "step": 14885
    },
    {
      "epoch": 2.4290725181499306,
      "grad_norm": 8.6875,
      "learning_rate": 3.476807738885399e-05,
      "loss": 1.5148,
      "num_input_tokens_seen": 31089872,
      "step": 14890
    },
    {
      "epoch": 2.429888245370748,
      "grad_norm": 7.5625,
      "learning_rate": 3.475903956716485e-05,
      "loss": 2.7455,
      "num_input_tokens_seen": 31099744,
      "step": 14895
    },
    {
      "epoch": 2.4307039725915653,
      "grad_norm": 11.5,
      "learning_rate": 3.475000024050917e-05,
      "loss": 2.1032,
      "num_input_tokens_seen": 31109760,
      "step": 14900
    },
    {
      "epoch": 2.431519699812383,
      "grad_norm": 8.9375,
      "learning_rate": 3.4740959410280926e-05,
      "loss": 3.1514,
      "num_input_tokens_seen": 31119664,
      "step": 14905
    },
    {
      "epoch": 2.4323354270332,
      "grad_norm": 4.25,
      "learning_rate": 3.4731917077874324e-05,
      "loss": 2.6253,
      "num_input_tokens_seen": 31132128,
      "step": 14910
    },
    {
      "epoch": 2.4331511542540176,
      "grad_norm": 12.5,
      "learning_rate": 3.4722873244683816e-05,
      "loss": 3.7684,
      "num_input_tokens_seen": 31143344,
      "step": 14915
    },
    {
      "epoch": 2.4339668814748348,
      "grad_norm": 5.28125,
      "learning_rate": 3.4713827912104065e-05,
      "loss": 1.3152,
      "num_input_tokens_seen": 31153952,
      "step": 14920
    },
    {
      "epoch": 2.4347826086956523,
      "grad_norm": 8.5,
      "learning_rate": 3.470478108152998e-05,
      "loss": 1.5731,
      "num_input_tokens_seen": 31163680,
      "step": 14925
    },
    {
      "epoch": 2.4355983359164695,
      "grad_norm": 5.75,
      "learning_rate": 3.4695732754356695e-05,
      "loss": 2.6869,
      "num_input_tokens_seen": 31174816,
      "step": 14930
    },
    {
      "epoch": 2.436414063137287,
      "grad_norm": 7.59375,
      "learning_rate": 3.4686682931979576e-05,
      "loss": 3.86,
      "num_input_tokens_seen": 31186400,
      "step": 14935
    },
    {
      "epoch": 2.437229790358104,
      "grad_norm": 16.625,
      "learning_rate": 3.467763161579422e-05,
      "loss": 2.1023,
      "num_input_tokens_seen": 31195712,
      "step": 14940
    },
    {
      "epoch": 2.4380455175789217,
      "grad_norm": 4.90625,
      "learning_rate": 3.466857880719645e-05,
      "loss": 2.6019,
      "num_input_tokens_seen": 31205184,
      "step": 14945
    },
    {
      "epoch": 2.438861244799739,
      "grad_norm": 3.4375,
      "learning_rate": 3.465952450758233e-05,
      "loss": 1.8624,
      "num_input_tokens_seen": 31215504,
      "step": 14950
    },
    {
      "epoch": 2.4396769720205564,
      "grad_norm": 9.25,
      "learning_rate": 3.4650468718348126e-05,
      "loss": 3.2255,
      "num_input_tokens_seen": 31226688,
      "step": 14955
    },
    {
      "epoch": 2.4404926992413736,
      "grad_norm": 7.4375,
      "learning_rate": 3.464141144089038e-05,
      "loss": 2.5472,
      "num_input_tokens_seen": 31236048,
      "step": 14960
    },
    {
      "epoch": 2.441308426462191,
      "grad_norm": 8.3125,
      "learning_rate": 3.463235267660583e-05,
      "loss": 1.7156,
      "num_input_tokens_seen": 31246864,
      "step": 14965
    },
    {
      "epoch": 2.4421241536830083,
      "grad_norm": 9.375,
      "learning_rate": 3.462329242689145e-05,
      "loss": 2.487,
      "num_input_tokens_seen": 31257808,
      "step": 14970
    },
    {
      "epoch": 2.442939880903826,
      "grad_norm": 8.6875,
      "learning_rate": 3.461423069314444e-05,
      "loss": 2.8036,
      "num_input_tokens_seen": 31268976,
      "step": 14975
    },
    {
      "epoch": 2.443755608124643,
      "grad_norm": 6.625,
      "learning_rate": 3.460516747676224e-05,
      "loss": 2.4585,
      "num_input_tokens_seen": 31279488,
      "step": 14980
    },
    {
      "epoch": 2.4445713353454606,
      "grad_norm": 4.5625,
      "learning_rate": 3.459610277914251e-05,
      "loss": 2.3534,
      "num_input_tokens_seen": 31290400,
      "step": 14985
    },
    {
      "epoch": 2.4453870625662777,
      "grad_norm": 5.0625,
      "learning_rate": 3.458703660168314e-05,
      "loss": 2.583,
      "num_input_tokens_seen": 31301008,
      "step": 14990
    },
    {
      "epoch": 2.4462027897870953,
      "grad_norm": 9.375,
      "learning_rate": 3.457796894578224e-05,
      "loss": 2.8282,
      "num_input_tokens_seen": 31311360,
      "step": 14995
    },
    {
      "epoch": 2.4470185170079124,
      "grad_norm": 6.25,
      "learning_rate": 3.456889981283817e-05,
      "loss": 2.3624,
      "num_input_tokens_seen": 31321184,
      "step": 15000
    },
    {
      "epoch": 2.4470185170079124,
      "eval_loss": 2.5304582118988037,
      "eval_runtime": 135.1172,
      "eval_samples_per_second": 20.168,
      "eval_steps_per_second": 10.088,
      "num_input_tokens_seen": 31321184,
      "step": 15000
    },
    {
      "epoch": 2.44783424422873,
      "grad_norm": 11.6875,
      "learning_rate": 3.45598292042495e-05,
      "loss": 4.5703,
      "num_input_tokens_seen": 31330304,
      "step": 15005
    },
    {
      "epoch": 2.448649971449547,
      "grad_norm": 7.71875,
      "learning_rate": 3.4550757121415035e-05,
      "loss": 3.3948,
      "num_input_tokens_seen": 31339040,
      "step": 15010
    },
    {
      "epoch": 2.4494656986703647,
      "grad_norm": 4.8125,
      "learning_rate": 3.454168356573378e-05,
      "loss": 1.4794,
      "num_input_tokens_seen": 31349296,
      "step": 15015
    },
    {
      "epoch": 2.450281425891182,
      "grad_norm": 4.03125,
      "learning_rate": 3.453260853860503e-05,
      "loss": 1.7381,
      "num_input_tokens_seen": 31358976,
      "step": 15020
    },
    {
      "epoch": 2.4510971531119994,
      "grad_norm": 4.71875,
      "learning_rate": 3.452353204142824e-05,
      "loss": 2.2608,
      "num_input_tokens_seen": 31368864,
      "step": 15025
    },
    {
      "epoch": 2.4519128803328165,
      "grad_norm": 6.09375,
      "learning_rate": 3.4514454075603136e-05,
      "loss": 2.0529,
      "num_input_tokens_seen": 31378800,
      "step": 15030
    },
    {
      "epoch": 2.452728607553634,
      "grad_norm": 2.203125,
      "learning_rate": 3.450537464252964e-05,
      "loss": 2.0673,
      "num_input_tokens_seen": 31389072,
      "step": 15035
    },
    {
      "epoch": 2.4535443347744517,
      "grad_norm": 6.53125,
      "learning_rate": 3.4496293743607925e-05,
      "loss": 2.5837,
      "num_input_tokens_seen": 31399312,
      "step": 15040
    },
    {
      "epoch": 2.454360061995269,
      "grad_norm": 7.75,
      "learning_rate": 3.448721138023838e-05,
      "loss": 1.8439,
      "num_input_tokens_seen": 31408864,
      "step": 15045
    },
    {
      "epoch": 2.455175789216086,
      "grad_norm": 11.0625,
      "learning_rate": 3.447812755382162e-05,
      "loss": 2.3647,
      "num_input_tokens_seen": 31420176,
      "step": 15050
    },
    {
      "epoch": 2.4559915164369035,
      "grad_norm": 2.640625,
      "learning_rate": 3.446904226575847e-05,
      "loss": 2.6357,
      "num_input_tokens_seen": 31431296,
      "step": 15055
    },
    {
      "epoch": 2.456807243657721,
      "grad_norm": 1.21875,
      "learning_rate": 3.445995551745002e-05,
      "loss": 2.0267,
      "num_input_tokens_seen": 31441184,
      "step": 15060
    },
    {
      "epoch": 2.457622970878538,
      "grad_norm": 11.4375,
      "learning_rate": 3.445086731029753e-05,
      "loss": 2.3031,
      "num_input_tokens_seen": 31451168,
      "step": 15065
    },
    {
      "epoch": 2.4584386980993553,
      "grad_norm": 2.8125,
      "learning_rate": 3.444177764570255e-05,
      "loss": 2.7129,
      "num_input_tokens_seen": 31462624,
      "step": 15070
    },
    {
      "epoch": 2.459254425320173,
      "grad_norm": 9.3125,
      "learning_rate": 3.44326865250668e-05,
      "loss": 2.7599,
      "num_input_tokens_seen": 31473744,
      "step": 15075
    },
    {
      "epoch": 2.4600701525409905,
      "grad_norm": 5.4375,
      "learning_rate": 3.442359394979225e-05,
      "loss": 2.737,
      "num_input_tokens_seen": 31484640,
      "step": 15080
    },
    {
      "epoch": 2.4608858797618076,
      "grad_norm": 1.6875,
      "learning_rate": 3.441449992128108e-05,
      "loss": 1.5051,
      "num_input_tokens_seen": 31494944,
      "step": 15085
    },
    {
      "epoch": 2.461701606982625,
      "grad_norm": 6.1875,
      "learning_rate": 3.440540444093573e-05,
      "loss": 1.5543,
      "num_input_tokens_seen": 31504032,
      "step": 15090
    },
    {
      "epoch": 2.4625173342034423,
      "grad_norm": 3.828125,
      "learning_rate": 3.43963075101588e-05,
      "loss": 1.5137,
      "num_input_tokens_seen": 31514128,
      "step": 15095
    },
    {
      "epoch": 2.46333306142426,
      "grad_norm": 8.4375,
      "learning_rate": 3.438720913035318e-05,
      "loss": 3.3419,
      "num_input_tokens_seen": 31525008,
      "step": 15100
    },
    {
      "epoch": 2.464148788645077,
      "grad_norm": 2.90625,
      "learning_rate": 3.437810930292195e-05,
      "loss": 1.7543,
      "num_input_tokens_seen": 31535568,
      "step": 15105
    },
    {
      "epoch": 2.4649645158658946,
      "grad_norm": 7.03125,
      "learning_rate": 3.43690080292684e-05,
      "loss": 2.6299,
      "num_input_tokens_seen": 31544336,
      "step": 15110
    },
    {
      "epoch": 2.4657802430867117,
      "grad_norm": 3.015625,
      "learning_rate": 3.435990531079608e-05,
      "loss": 1.2907,
      "num_input_tokens_seen": 31553456,
      "step": 15115
    },
    {
      "epoch": 2.4665959703075293,
      "grad_norm": 3.0625,
      "learning_rate": 3.435080114890874e-05,
      "loss": 1.7295,
      "num_input_tokens_seen": 31564816,
      "step": 15120
    },
    {
      "epoch": 2.4674116975283464,
      "grad_norm": 0.173828125,
      "learning_rate": 3.434169554501035e-05,
      "loss": 2.0831,
      "num_input_tokens_seen": 31574768,
      "step": 15125
    },
    {
      "epoch": 2.468227424749164,
      "grad_norm": 9.375,
      "learning_rate": 3.433258850050511e-05,
      "loss": 1.5758,
      "num_input_tokens_seen": 31585808,
      "step": 15130
    },
    {
      "epoch": 2.469043151969981,
      "grad_norm": 9.5625,
      "learning_rate": 3.4323480016797446e-05,
      "loss": 2.5295,
      "num_input_tokens_seen": 31595216,
      "step": 15135
    },
    {
      "epoch": 2.4698588791907987,
      "grad_norm": 8.25,
      "learning_rate": 3.4314370095291995e-05,
      "loss": 2.0662,
      "num_input_tokens_seen": 31606384,
      "step": 15140
    },
    {
      "epoch": 2.470674606411616,
      "grad_norm": 9.875,
      "learning_rate": 3.430525873739363e-05,
      "loss": 1.9718,
      "num_input_tokens_seen": 31616320,
      "step": 15145
    },
    {
      "epoch": 2.4714903336324334,
      "grad_norm": 6.59375,
      "learning_rate": 3.429614594450743e-05,
      "loss": 1.8538,
      "num_input_tokens_seen": 31625984,
      "step": 15150
    },
    {
      "epoch": 2.4723060608532506,
      "grad_norm": 7.65625,
      "learning_rate": 3.428703171803869e-05,
      "loss": 2.9443,
      "num_input_tokens_seen": 31634848,
      "step": 15155
    },
    {
      "epoch": 2.473121788074068,
      "grad_norm": 5.78125,
      "learning_rate": 3.4277916059392964e-05,
      "loss": 1.8884,
      "num_input_tokens_seen": 31646256,
      "step": 15160
    },
    {
      "epoch": 2.4739375152948853,
      "grad_norm": 10.8125,
      "learning_rate": 3.426879896997598e-05,
      "loss": 3.4463,
      "num_input_tokens_seen": 31656448,
      "step": 15165
    },
    {
      "epoch": 2.474753242515703,
      "grad_norm": 3.390625,
      "learning_rate": 3.425968045119372e-05,
      "loss": 3.6269,
      "num_input_tokens_seen": 31666800,
      "step": 15170
    },
    {
      "epoch": 2.47556896973652,
      "grad_norm": 4.1875,
      "learning_rate": 3.425056050445237e-05,
      "loss": 1.1527,
      "num_input_tokens_seen": 31677056,
      "step": 15175
    },
    {
      "epoch": 2.4763846969573375,
      "grad_norm": 9.625,
      "learning_rate": 3.4241439131158336e-05,
      "loss": 1.7436,
      "num_input_tokens_seen": 31688336,
      "step": 15180
    },
    {
      "epoch": 2.4772004241781547,
      "grad_norm": 8.625,
      "learning_rate": 3.423231633271825e-05,
      "loss": 3.1376,
      "num_input_tokens_seen": 31699296,
      "step": 15185
    },
    {
      "epoch": 2.4780161513989722,
      "grad_norm": 8.375,
      "learning_rate": 3.4223192110538985e-05,
      "loss": 2.9628,
      "num_input_tokens_seen": 31710784,
      "step": 15190
    },
    {
      "epoch": 2.4788318786197894,
      "grad_norm": 9.5625,
      "learning_rate": 3.4214066466027575e-05,
      "loss": 2.9427,
      "num_input_tokens_seen": 31721616,
      "step": 15195
    },
    {
      "epoch": 2.479647605840607,
      "grad_norm": 4.28125,
      "learning_rate": 3.4204939400591325e-05,
      "loss": 1.8445,
      "num_input_tokens_seen": 31730928,
      "step": 15200
    },
    {
      "epoch": 2.479647605840607,
      "eval_loss": 2.553328037261963,
      "eval_runtime": 134.795,
      "eval_samples_per_second": 20.216,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 31730928,
      "step": 15200
    },
    {
      "epoch": 2.480463333061424,
      "grad_norm": 7.9375,
      "learning_rate": 3.419581091563775e-05,
      "loss": 3.1038,
      "num_input_tokens_seen": 31741152,
      "step": 15205
    },
    {
      "epoch": 2.4812790602822417,
      "grad_norm": 8.125,
      "learning_rate": 3.418668101257456e-05,
      "loss": 0.9535,
      "num_input_tokens_seen": 31750960,
      "step": 15210
    },
    {
      "epoch": 2.482094787503059,
      "grad_norm": 2.65625,
      "learning_rate": 3.417754969280971e-05,
      "loss": 2.2999,
      "num_input_tokens_seen": 31761584,
      "step": 15215
    },
    {
      "epoch": 2.4829105147238764,
      "grad_norm": 6.25,
      "learning_rate": 3.416841695775137e-05,
      "loss": 2.7431,
      "num_input_tokens_seen": 31772016,
      "step": 15220
    },
    {
      "epoch": 2.4837262419446935,
      "grad_norm": 1.53125,
      "learning_rate": 3.415928280880792e-05,
      "loss": 1.7103,
      "num_input_tokens_seen": 31782560,
      "step": 15225
    },
    {
      "epoch": 2.484541969165511,
      "grad_norm": 5.71875,
      "learning_rate": 3.4150147247387965e-05,
      "loss": 1.5234,
      "num_input_tokens_seen": 31792752,
      "step": 15230
    },
    {
      "epoch": 2.4853576963863286,
      "grad_norm": 4.40625,
      "learning_rate": 3.4141010274900306e-05,
      "loss": 2.962,
      "num_input_tokens_seen": 31801280,
      "step": 15235
    },
    {
      "epoch": 2.4861734236071458,
      "grad_norm": 5.8125,
      "learning_rate": 3.413187189275399e-05,
      "loss": 1.8907,
      "num_input_tokens_seen": 31812144,
      "step": 15240
    },
    {
      "epoch": 2.486989150827963,
      "grad_norm": 9.3125,
      "learning_rate": 3.4122732102358265e-05,
      "loss": 3.0662,
      "num_input_tokens_seen": 31822048,
      "step": 15245
    },
    {
      "epoch": 2.4878048780487805,
      "grad_norm": 8.25,
      "learning_rate": 3.411359090512261e-05,
      "loss": 4.1113,
      "num_input_tokens_seen": 31832688,
      "step": 15250
    },
    {
      "epoch": 2.488620605269598,
      "grad_norm": 6.1875,
      "learning_rate": 3.410444830245672e-05,
      "loss": 1.5104,
      "num_input_tokens_seen": 31842896,
      "step": 15255
    },
    {
      "epoch": 2.489436332490415,
      "grad_norm": 7.5,
      "learning_rate": 3.409530429577048e-05,
      "loss": 2.677,
      "num_input_tokens_seen": 31852768,
      "step": 15260
    },
    {
      "epoch": 2.4902520597112328,
      "grad_norm": 6.1875,
      "learning_rate": 3.408615888647402e-05,
      "loss": 3.061,
      "num_input_tokens_seen": 31862848,
      "step": 15265
    },
    {
      "epoch": 2.49106778693205,
      "grad_norm": 6.25,
      "learning_rate": 3.4077012075977675e-05,
      "loss": 3.2132,
      "num_input_tokens_seen": 31872800,
      "step": 15270
    },
    {
      "epoch": 2.4918835141528675,
      "grad_norm": 12.5625,
      "learning_rate": 3.4067863865692e-05,
      "loss": 2.2271,
      "num_input_tokens_seen": 31883360,
      "step": 15275
    },
    {
      "epoch": 2.4926992413736846,
      "grad_norm": 3.328125,
      "learning_rate": 3.4058714257027755e-05,
      "loss": 3.0222,
      "num_input_tokens_seen": 31893792,
      "step": 15280
    },
    {
      "epoch": 2.493514968594502,
      "grad_norm": 10.4375,
      "learning_rate": 3.404956325139594e-05,
      "loss": 2.8273,
      "num_input_tokens_seen": 31905584,
      "step": 15285
    },
    {
      "epoch": 2.4943306958153193,
      "grad_norm": 6.21875,
      "learning_rate": 3.404041085020775e-05,
      "loss": 2.6197,
      "num_input_tokens_seen": 31916208,
      "step": 15290
    },
    {
      "epoch": 2.495146423036137,
      "grad_norm": 13.5625,
      "learning_rate": 3.403125705487459e-05,
      "loss": 2.3621,
      "num_input_tokens_seen": 31926080,
      "step": 15295
    },
    {
      "epoch": 2.495962150256954,
      "grad_norm": 3.484375,
      "learning_rate": 3.402210186680811e-05,
      "loss": 2.1563,
      "num_input_tokens_seen": 31936576,
      "step": 15300
    },
    {
      "epoch": 2.4967778774777716,
      "grad_norm": 8.5,
      "learning_rate": 3.4012945287420137e-05,
      "loss": 1.9736,
      "num_input_tokens_seen": 31946544,
      "step": 15305
    },
    {
      "epoch": 2.4975936046985887,
      "grad_norm": 13.3125,
      "learning_rate": 3.400378731812274e-05,
      "loss": 3.9374,
      "num_input_tokens_seen": 31956288,
      "step": 15310
    },
    {
      "epoch": 2.4984093319194063,
      "grad_norm": 12.4375,
      "learning_rate": 3.399462796032817e-05,
      "loss": 2.2517,
      "num_input_tokens_seen": 31967632,
      "step": 15315
    },
    {
      "epoch": 2.4992250591402234,
      "grad_norm": 6.0,
      "learning_rate": 3.3985467215448954e-05,
      "loss": 1.6535,
      "num_input_tokens_seen": 31977712,
      "step": 15320
    },
    {
      "epoch": 2.500040786361041,
      "grad_norm": 14.25,
      "learning_rate": 3.3976305084897776e-05,
      "loss": 3.4787,
      "num_input_tokens_seen": 31988000,
      "step": 15325
    },
    {
      "epoch": 2.500856513581858,
      "grad_norm": 6.0625,
      "learning_rate": 3.3967141570087544e-05,
      "loss": 2.4876,
      "num_input_tokens_seen": 31998304,
      "step": 15330
    },
    {
      "epoch": 2.5016722408026757,
      "grad_norm": 8.4375,
      "learning_rate": 3.39579766724314e-05,
      "loss": 2.6039,
      "num_input_tokens_seen": 32009984,
      "step": 15335
    },
    {
      "epoch": 2.502487968023493,
      "grad_norm": 11.8125,
      "learning_rate": 3.3948810393342677e-05,
      "loss": 3.0548,
      "num_input_tokens_seen": 32020720,
      "step": 15340
    },
    {
      "epoch": 2.5033036952443104,
      "grad_norm": 10.0,
      "learning_rate": 3.3939642734234936e-05,
      "loss": 1.6656,
      "num_input_tokens_seen": 32031824,
      "step": 15345
    },
    {
      "epoch": 2.5041194224651275,
      "grad_norm": 7.46875,
      "learning_rate": 3.393047369652194e-05,
      "loss": 1.1684,
      "num_input_tokens_seen": 32042816,
      "step": 15350
    },
    {
      "epoch": 2.504935149685945,
      "grad_norm": 5.3125,
      "learning_rate": 3.3921303281617664e-05,
      "loss": 1.8827,
      "num_input_tokens_seen": 32052992,
      "step": 15355
    },
    {
      "epoch": 2.5057508769067622,
      "grad_norm": 4.1875,
      "learning_rate": 3.391213149093632e-05,
      "loss": 1.221,
      "num_input_tokens_seen": 32063216,
      "step": 15360
    },
    {
      "epoch": 2.50656660412758,
      "grad_norm": 11.1875,
      "learning_rate": 3.3902958325892303e-05,
      "loss": 3.292,
      "num_input_tokens_seen": 32075360,
      "step": 15365
    },
    {
      "epoch": 2.507382331348397,
      "grad_norm": 10.1875,
      "learning_rate": 3.389378378790023e-05,
      "loss": 2.2244,
      "num_input_tokens_seen": 32085072,
      "step": 15370
    },
    {
      "epoch": 2.5081980585692145,
      "grad_norm": 7.25,
      "learning_rate": 3.388460787837493e-05,
      "loss": 2.5915,
      "num_input_tokens_seen": 32095792,
      "step": 15375
    },
    {
      "epoch": 2.5090137857900316,
      "grad_norm": 5.15625,
      "learning_rate": 3.387543059873145e-05,
      "loss": 2.0859,
      "num_input_tokens_seen": 32106880,
      "step": 15380
    },
    {
      "epoch": 2.5098295130108492,
      "grad_norm": 4.90625,
      "learning_rate": 3.386625195038503e-05,
      "loss": 2.4389,
      "num_input_tokens_seen": 32117200,
      "step": 15385
    },
    {
      "epoch": 2.510645240231667,
      "grad_norm": 5.96875,
      "learning_rate": 3.3857071934751136e-05,
      "loss": 3.0424,
      "num_input_tokens_seen": 32127072,
      "step": 15390
    },
    {
      "epoch": 2.511460967452484,
      "grad_norm": 4.53125,
      "learning_rate": 3.384789055324544e-05,
      "loss": 1.921,
      "num_input_tokens_seen": 32136368,
      "step": 15395
    },
    {
      "epoch": 2.512276694673301,
      "grad_norm": 4.375,
      "learning_rate": 3.3838707807283843e-05,
      "loss": 2.4441,
      "num_input_tokens_seen": 32146304,
      "step": 15400
    },
    {
      "epoch": 2.512276694673301,
      "eval_loss": 2.5513195991516113,
      "eval_runtime": 134.7621,
      "eval_samples_per_second": 20.221,
      "eval_steps_per_second": 10.114,
      "num_input_tokens_seen": 32146304,
      "step": 15400
    },
    {
      "epoch": 2.5130924218941186,
      "grad_norm": 9.75,
      "learning_rate": 3.382952369828243e-05,
      "loss": 2.8509,
      "num_input_tokens_seen": 32154896,
      "step": 15405
    },
    {
      "epoch": 2.513908149114936,
      "grad_norm": 4.96875,
      "learning_rate": 3.38203382276575e-05,
      "loss": 1.9193,
      "num_input_tokens_seen": 32166304,
      "step": 15410
    },
    {
      "epoch": 2.5147238763357533,
      "grad_norm": 4.65625,
      "learning_rate": 3.381115139682557e-05,
      "loss": 2.4256,
      "num_input_tokens_seen": 32176656,
      "step": 15415
    },
    {
      "epoch": 2.5155396035565705,
      "grad_norm": 4.28125,
      "learning_rate": 3.3801963207203366e-05,
      "loss": 1.2782,
      "num_input_tokens_seen": 32188800,
      "step": 15420
    },
    {
      "epoch": 2.516355330777388,
      "grad_norm": 6.28125,
      "learning_rate": 3.379277366020782e-05,
      "loss": 2.4299,
      "num_input_tokens_seen": 32200256,
      "step": 15425
    },
    {
      "epoch": 2.5171710579982056,
      "grad_norm": 9.6875,
      "learning_rate": 3.3783582757256085e-05,
      "loss": 2.9336,
      "num_input_tokens_seen": 32211104,
      "step": 15430
    },
    {
      "epoch": 2.5179867852190227,
      "grad_norm": 7.1875,
      "learning_rate": 3.3774390499765504e-05,
      "loss": 2.0857,
      "num_input_tokens_seen": 32221104,
      "step": 15435
    },
    {
      "epoch": 2.51880251243984,
      "grad_norm": 11.375,
      "learning_rate": 3.376519688915364e-05,
      "loss": 3.0225,
      "num_input_tokens_seen": 32231184,
      "step": 15440
    },
    {
      "epoch": 2.5196182396606575,
      "grad_norm": 11.375,
      "learning_rate": 3.3756001926838273e-05,
      "loss": 2.5835,
      "num_input_tokens_seen": 32241536,
      "step": 15445
    },
    {
      "epoch": 2.520433966881475,
      "grad_norm": 1.4375,
      "learning_rate": 3.374680561423737e-05,
      "loss": 2.6224,
      "num_input_tokens_seen": 32252896,
      "step": 15450
    },
    {
      "epoch": 2.521249694102292,
      "grad_norm": 6.1875,
      "learning_rate": 3.373760795276912e-05,
      "loss": 3.0368,
      "num_input_tokens_seen": 32263472,
      "step": 15455
    },
    {
      "epoch": 2.5220654213231093,
      "grad_norm": 4.125,
      "learning_rate": 3.372840894385192e-05,
      "loss": 2.6807,
      "num_input_tokens_seen": 32273920,
      "step": 15460
    },
    {
      "epoch": 2.522881148543927,
      "grad_norm": 6.96875,
      "learning_rate": 3.3719208588904375e-05,
      "loss": 1.9394,
      "num_input_tokens_seen": 32282784,
      "step": 15465
    },
    {
      "epoch": 2.5236968757647444,
      "grad_norm": 6.71875,
      "learning_rate": 3.371000688934529e-05,
      "loss": 2.7863,
      "num_input_tokens_seen": 32293136,
      "step": 15470
    },
    {
      "epoch": 2.5245126029855616,
      "grad_norm": 7.15625,
      "learning_rate": 3.370080384659369e-05,
      "loss": 2.8565,
      "num_input_tokens_seen": 32303824,
      "step": 15475
    },
    {
      "epoch": 2.525328330206379,
      "grad_norm": 12.3125,
      "learning_rate": 3.36915994620688e-05,
      "loss": 3.9078,
      "num_input_tokens_seen": 32314352,
      "step": 15480
    },
    {
      "epoch": 2.5261440574271963,
      "grad_norm": 6.46875,
      "learning_rate": 3.3682393737190035e-05,
      "loss": 1.7043,
      "num_input_tokens_seen": 32324912,
      "step": 15485
    },
    {
      "epoch": 2.526959784648014,
      "grad_norm": 6.3125,
      "learning_rate": 3.3673186673377054e-05,
      "loss": 2.1298,
      "num_input_tokens_seen": 32335200,
      "step": 15490
    },
    {
      "epoch": 2.527775511868831,
      "grad_norm": 10.0,
      "learning_rate": 3.366397827204969e-05,
      "loss": 2.9514,
      "num_input_tokens_seen": 32346208,
      "step": 15495
    },
    {
      "epoch": 2.5285912390896486,
      "grad_norm": 3.875,
      "learning_rate": 3.3654768534628e-05,
      "loss": 0.9498,
      "num_input_tokens_seen": 32356368,
      "step": 15500
    },
    {
      "epoch": 2.5294069663104657,
      "grad_norm": 18.875,
      "learning_rate": 3.3645557462532245e-05,
      "loss": 1.9841,
      "num_input_tokens_seen": 32367376,
      "step": 15505
    },
    {
      "epoch": 2.5302226935312833,
      "grad_norm": 7.21875,
      "learning_rate": 3.363634505718288e-05,
      "loss": 2.0988,
      "num_input_tokens_seen": 32378048,
      "step": 15510
    },
    {
      "epoch": 2.5310384207521004,
      "grad_norm": 2.21875,
      "learning_rate": 3.362713132000057e-05,
      "loss": 2.065,
      "num_input_tokens_seen": 32388208,
      "step": 15515
    },
    {
      "epoch": 2.531854147972918,
      "grad_norm": 1.7109375,
      "learning_rate": 3.36179162524062e-05,
      "loss": 1.6412,
      "num_input_tokens_seen": 32398720,
      "step": 15520
    },
    {
      "epoch": 2.532669875193735,
      "grad_norm": 9.75,
      "learning_rate": 3.3608699855820846e-05,
      "loss": 2.5209,
      "num_input_tokens_seen": 32409328,
      "step": 15525
    },
    {
      "epoch": 2.5334856024145527,
      "grad_norm": 12.1875,
      "learning_rate": 3.359948213166578e-05,
      "loss": 2.6615,
      "num_input_tokens_seen": 32420800,
      "step": 15530
    },
    {
      "epoch": 2.53430132963537,
      "grad_norm": 0.1474609375,
      "learning_rate": 3.359026308136252e-05,
      "loss": 2.5416,
      "num_input_tokens_seen": 32430784,
      "step": 15535
    },
    {
      "epoch": 2.5351170568561874,
      "grad_norm": 10.0,
      "learning_rate": 3.358104270633272e-05,
      "loss": 3.4055,
      "num_input_tokens_seen": 32441296,
      "step": 15540
    },
    {
      "epoch": 2.5359327840770045,
      "grad_norm": 12.125,
      "learning_rate": 3.357182100799831e-05,
      "loss": 1.6179,
      "num_input_tokens_seen": 32451552,
      "step": 15545
    },
    {
      "epoch": 2.536748511297822,
      "grad_norm": 6.125,
      "learning_rate": 3.3562597987781384e-05,
      "loss": 2.2592,
      "num_input_tokens_seen": 32463136,
      "step": 15550
    },
    {
      "epoch": 2.537564238518639,
      "grad_norm": 4.03125,
      "learning_rate": 3.355337364710424e-05,
      "loss": 2.8074,
      "num_input_tokens_seen": 32472416,
      "step": 15555
    },
    {
      "epoch": 2.538379965739457,
      "grad_norm": 9.0625,
      "learning_rate": 3.354414798738939e-05,
      "loss": 2.3342,
      "num_input_tokens_seen": 32481344,
      "step": 15560
    },
    {
      "epoch": 2.539195692960274,
      "grad_norm": 4.03125,
      "learning_rate": 3.353492101005955e-05,
      "loss": 1.4057,
      "num_input_tokens_seen": 32491648,
      "step": 15565
    },
    {
      "epoch": 2.5400114201810915,
      "grad_norm": 3.265625,
      "learning_rate": 3.352569271653763e-05,
      "loss": 2.158,
      "num_input_tokens_seen": 32503072,
      "step": 15570
    },
    {
      "epoch": 2.5408271474019086,
      "grad_norm": 5.84375,
      "learning_rate": 3.351646310824675e-05,
      "loss": 2.0664,
      "num_input_tokens_seen": 32513504,
      "step": 15575
    },
    {
      "epoch": 2.541642874622726,
      "grad_norm": 5.75,
      "learning_rate": 3.350723218661023e-05,
      "loss": 2.6191,
      "num_input_tokens_seen": 32525040,
      "step": 15580
    },
    {
      "epoch": 2.5424586018435438,
      "grad_norm": 6.125,
      "learning_rate": 3.349799995305162e-05,
      "loss": 2.4312,
      "num_input_tokens_seen": 32536208,
      "step": 15585
    },
    {
      "epoch": 2.543274329064361,
      "grad_norm": 7.6875,
      "learning_rate": 3.348876640899461e-05,
      "loss": 1.7887,
      "num_input_tokens_seen": 32546080,
      "step": 15590
    },
    {
      "epoch": 2.544090056285178,
      "grad_norm": 7.28125,
      "learning_rate": 3.3479531555863144e-05,
      "loss": 3.3634,
      "num_input_tokens_seen": 32555104,
      "step": 15595
    },
    {
      "epoch": 2.5449057835059956,
      "grad_norm": 5.875,
      "learning_rate": 3.3470295395081344e-05,
      "loss": 1.2673,
      "num_input_tokens_seen": 32566096,
      "step": 15600
    },
    {
      "epoch": 2.5449057835059956,
      "eval_loss": 2.535526990890503,
      "eval_runtime": 135.1285,
      "eval_samples_per_second": 20.166,
      "eval_steps_per_second": 10.087,
      "num_input_tokens_seen": 32566096,
      "step": 15600
    },
    {
      "epoch": 2.545721510726813,
      "grad_norm": 3.1875,
      "learning_rate": 3.3461057928073556e-05,
      "loss": 3.0945,
      "num_input_tokens_seen": 32575488,
      "step": 15605
    },
    {
      "epoch": 2.5465372379476303,
      "grad_norm": 15.1875,
      "learning_rate": 3.345181915626431e-05,
      "loss": 2.1694,
      "num_input_tokens_seen": 32586816,
      "step": 15610
    },
    {
      "epoch": 2.5473529651684474,
      "grad_norm": 4.25,
      "learning_rate": 3.344257908107834e-05,
      "loss": 2.726,
      "num_input_tokens_seen": 32597904,
      "step": 15615
    },
    {
      "epoch": 2.548168692389265,
      "grad_norm": 7.71875,
      "learning_rate": 3.343333770394058e-05,
      "loss": 1.7838,
      "num_input_tokens_seen": 32607696,
      "step": 15620
    },
    {
      "epoch": 2.5489844196100826,
      "grad_norm": 2.0,
      "learning_rate": 3.342409502627616e-05,
      "loss": 2.8575,
      "num_input_tokens_seen": 32617664,
      "step": 15625
    },
    {
      "epoch": 2.5498001468308997,
      "grad_norm": 5.71875,
      "learning_rate": 3.341485104951043e-05,
      "loss": 1.4401,
      "num_input_tokens_seen": 32626624,
      "step": 15630
    },
    {
      "epoch": 2.550615874051717,
      "grad_norm": 13.375,
      "learning_rate": 3.340560577506892e-05,
      "loss": 3.2228,
      "num_input_tokens_seen": 32635344,
      "step": 15635
    },
    {
      "epoch": 2.5514316012725344,
      "grad_norm": 4.8125,
      "learning_rate": 3.339635920437735e-05,
      "loss": 1.6663,
      "num_input_tokens_seen": 32646112,
      "step": 15640
    },
    {
      "epoch": 2.552247328493352,
      "grad_norm": 7.84375,
      "learning_rate": 3.338711133886169e-05,
      "loss": 0.8318,
      "num_input_tokens_seen": 32656528,
      "step": 15645
    },
    {
      "epoch": 2.553063055714169,
      "grad_norm": 2.15625,
      "learning_rate": 3.3377862179948064e-05,
      "loss": 2.9447,
      "num_input_tokens_seen": 32666688,
      "step": 15650
    },
    {
      "epoch": 2.5538787829349863,
      "grad_norm": 6.15625,
      "learning_rate": 3.336861172906281e-05,
      "loss": 1.6852,
      "num_input_tokens_seen": 32677632,
      "step": 15655
    },
    {
      "epoch": 2.554694510155804,
      "grad_norm": 5.8125,
      "learning_rate": 3.335935998763245e-05,
      "loss": 2.5543,
      "num_input_tokens_seen": 32688016,
      "step": 15660
    },
    {
      "epoch": 2.5555102373766214,
      "grad_norm": 4.25,
      "learning_rate": 3.3350106957083744e-05,
      "loss": 2.012,
      "num_input_tokens_seen": 32699632,
      "step": 15665
    },
    {
      "epoch": 2.5563259645974385,
      "grad_norm": 5.40625,
      "learning_rate": 3.33408526388436e-05,
      "loss": 2.4287,
      "num_input_tokens_seen": 32709728,
      "step": 15670
    },
    {
      "epoch": 2.557141691818256,
      "grad_norm": 6.40625,
      "learning_rate": 3.3331597034339166e-05,
      "loss": 2.2699,
      "num_input_tokens_seen": 32719408,
      "step": 15675
    },
    {
      "epoch": 2.5579574190390733,
      "grad_norm": 6.34375,
      "learning_rate": 3.3322340144997764e-05,
      "loss": 4.4912,
      "num_input_tokens_seen": 32729776,
      "step": 15680
    },
    {
      "epoch": 2.558773146259891,
      "grad_norm": 5.96875,
      "learning_rate": 3.331308197224693e-05,
      "loss": 1.2452,
      "num_input_tokens_seen": 32739136,
      "step": 15685
    },
    {
      "epoch": 2.559588873480708,
      "grad_norm": 5.625,
      "learning_rate": 3.330382251751438e-05,
      "loss": 1.7019,
      "num_input_tokens_seen": 32750304,
      "step": 15690
    },
    {
      "epoch": 2.5604046007015255,
      "grad_norm": 10.8125,
      "learning_rate": 3.3294561782228054e-05,
      "loss": 2.8795,
      "num_input_tokens_seen": 32760592,
      "step": 15695
    },
    {
      "epoch": 2.5612203279223427,
      "grad_norm": 12.1875,
      "learning_rate": 3.328529976781607e-05,
      "loss": 3.287,
      "num_input_tokens_seen": 32770800,
      "step": 15700
    },
    {
      "epoch": 2.5620360551431602,
      "grad_norm": 8.1875,
      "learning_rate": 3.327603647570673e-05,
      "loss": 3.8624,
      "num_input_tokens_seen": 32781408,
      "step": 15705
    },
    {
      "epoch": 2.5628517823639774,
      "grad_norm": 10.5,
      "learning_rate": 3.326677190732857e-05,
      "loss": 1.5547,
      "num_input_tokens_seen": 32793712,
      "step": 15710
    },
    {
      "epoch": 2.563667509584795,
      "grad_norm": 8.3125,
      "learning_rate": 3.325750606411029e-05,
      "loss": 3.686,
      "num_input_tokens_seen": 32804608,
      "step": 15715
    },
    {
      "epoch": 2.564483236805612,
      "grad_norm": 3.484375,
      "learning_rate": 3.3248238947480804e-05,
      "loss": 2.6634,
      "num_input_tokens_seen": 32814416,
      "step": 15720
    },
    {
      "epoch": 2.5652989640264297,
      "grad_norm": 1.4296875,
      "learning_rate": 3.323897055886922e-05,
      "loss": 1.2281,
      "num_input_tokens_seen": 32824672,
      "step": 15725
    },
    {
      "epoch": 2.566114691247247,
      "grad_norm": 3.53125,
      "learning_rate": 3.322970089970484e-05,
      "loss": 2.3854,
      "num_input_tokens_seen": 32834528,
      "step": 15730
    },
    {
      "epoch": 2.5669304184680644,
      "grad_norm": 8.6875,
      "learning_rate": 3.3220429971417165e-05,
      "loss": 1.2117,
      "num_input_tokens_seen": 32844496,
      "step": 15735
    },
    {
      "epoch": 2.5677461456888815,
      "grad_norm": 8.5625,
      "learning_rate": 3.321115777543588e-05,
      "loss": 3.9821,
      "num_input_tokens_seen": 32852992,
      "step": 15740
    },
    {
      "epoch": 2.568561872909699,
      "grad_norm": 4.5625,
      "learning_rate": 3.320188431319088e-05,
      "loss": 2.4435,
      "num_input_tokens_seen": 32862832,
      "step": 15745
    },
    {
      "epoch": 2.569377600130516,
      "grad_norm": 7.28125,
      "learning_rate": 3.319260958611224e-05,
      "loss": 2.8928,
      "num_input_tokens_seen": 32873168,
      "step": 15750
    },
    {
      "epoch": 2.5701933273513338,
      "grad_norm": 15.25,
      "learning_rate": 3.3183333595630256e-05,
      "loss": 2.9744,
      "num_input_tokens_seen": 32882624,
      "step": 15755
    },
    {
      "epoch": 2.5710090545721513,
      "grad_norm": 9.5625,
      "learning_rate": 3.317405634317538e-05,
      "loss": 3.4959,
      "num_input_tokens_seen": 32893328,
      "step": 15760
    },
    {
      "epoch": 2.5718247817929685,
      "grad_norm": 12.0,
      "learning_rate": 3.3164777830178315e-05,
      "loss": 2.9176,
      "num_input_tokens_seen": 32904288,
      "step": 15765
    },
    {
      "epoch": 2.5726405090137856,
      "grad_norm": 6.875,
      "learning_rate": 3.315549805806989e-05,
      "loss": 2.0804,
      "num_input_tokens_seen": 32914384,
      "step": 15770
    },
    {
      "epoch": 2.573456236234603,
      "grad_norm": 7.1875,
      "learning_rate": 3.314621702828118e-05,
      "loss": 2.2735,
      "num_input_tokens_seen": 32924400,
      "step": 15775
    },
    {
      "epoch": 2.5742719634554208,
      "grad_norm": 4.71875,
      "learning_rate": 3.313693474224342e-05,
      "loss": 2.2886,
      "num_input_tokens_seen": 32935856,
      "step": 15780
    },
    {
      "epoch": 2.575087690676238,
      "grad_norm": 4.15625,
      "learning_rate": 3.312765120138809e-05,
      "loss": 1.0562,
      "num_input_tokens_seen": 32948160,
      "step": 15785
    },
    {
      "epoch": 2.575903417897055,
      "grad_norm": 5.125,
      "learning_rate": 3.311836640714679e-05,
      "loss": 1.2024,
      "num_input_tokens_seen": 32959840,
      "step": 15790
    },
    {
      "epoch": 2.5767191451178726,
      "grad_norm": 9.375,
      "learning_rate": 3.310908036095137e-05,
      "loss": 2.3837,
      "num_input_tokens_seen": 32969856,
      "step": 15795
    },
    {
      "epoch": 2.57753487233869,
      "grad_norm": 9.8125,
      "learning_rate": 3.309979306423386e-05,
      "loss": 2.9281,
      "num_input_tokens_seen": 32981664,
      "step": 15800
    },
    {
      "epoch": 2.57753487233869,
      "eval_loss": 2.5533969402313232,
      "eval_runtime": 134.8638,
      "eval_samples_per_second": 20.206,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 32981664,
      "step": 15800
    },
    {
      "epoch": 2.5783505995595073,
      "grad_norm": 2.078125,
      "learning_rate": 3.309050451842647e-05,
      "loss": 3.3276,
      "num_input_tokens_seen": 32991168,
      "step": 15805
    },
    {
      "epoch": 2.5791663267803244,
      "grad_norm": 10.0625,
      "learning_rate": 3.3081214724961604e-05,
      "loss": 2.1978,
      "num_input_tokens_seen": 33001728,
      "step": 15810
    },
    {
      "epoch": 2.579982054001142,
      "grad_norm": 9.125,
      "learning_rate": 3.307192368527188e-05,
      "loss": 2.7747,
      "num_input_tokens_seen": 33011744,
      "step": 15815
    },
    {
      "epoch": 2.5807977812219596,
      "grad_norm": 3.609375,
      "learning_rate": 3.306263140079008e-05,
      "loss": 3.1538,
      "num_input_tokens_seen": 33022016,
      "step": 15820
    },
    {
      "epoch": 2.5816135084427767,
      "grad_norm": 19.75,
      "learning_rate": 3.30533378729492e-05,
      "loss": 1.7303,
      "num_input_tokens_seen": 33032480,
      "step": 15825
    },
    {
      "epoch": 2.582429235663594,
      "grad_norm": 3.625,
      "learning_rate": 3.304404310318242e-05,
      "loss": 1.2063,
      "num_input_tokens_seen": 33042336,
      "step": 15830
    },
    {
      "epoch": 2.5832449628844114,
      "grad_norm": 7.28125,
      "learning_rate": 3.3034747092923105e-05,
      "loss": 2.3761,
      "num_input_tokens_seen": 33053264,
      "step": 15835
    },
    {
      "epoch": 2.584060690105229,
      "grad_norm": 11.5,
      "learning_rate": 3.3025449843604806e-05,
      "loss": 2.8879,
      "num_input_tokens_seen": 33064032,
      "step": 15840
    },
    {
      "epoch": 2.584876417326046,
      "grad_norm": 7.25,
      "learning_rate": 3.30161513566613e-05,
      "loss": 1.9474,
      "num_input_tokens_seen": 33074544,
      "step": 15845
    },
    {
      "epoch": 2.5856921445468637,
      "grad_norm": 6.5625,
      "learning_rate": 3.3006851633526506e-05,
      "loss": 2.4332,
      "num_input_tokens_seen": 33085008,
      "step": 15850
    },
    {
      "epoch": 2.586507871767681,
      "grad_norm": 8.25,
      "learning_rate": 3.2997550675634584e-05,
      "loss": 1.8788,
      "num_input_tokens_seen": 33095200,
      "step": 15855
    },
    {
      "epoch": 2.5873235989884984,
      "grad_norm": 8.875,
      "learning_rate": 3.2988248484419825e-05,
      "loss": 3.2416,
      "num_input_tokens_seen": 33105120,
      "step": 15860
    },
    {
      "epoch": 2.5881393262093155,
      "grad_norm": 6.625,
      "learning_rate": 3.2978945061316776e-05,
      "loss": 1.7235,
      "num_input_tokens_seen": 33115440,
      "step": 15865
    },
    {
      "epoch": 2.588955053430133,
      "grad_norm": 8.25,
      "learning_rate": 3.296964040776013e-05,
      "loss": 2.5833,
      "num_input_tokens_seen": 33125728,
      "step": 15870
    },
    {
      "epoch": 2.5897707806509502,
      "grad_norm": 8.375,
      "learning_rate": 3.296033452518478e-05,
      "loss": 2.4218,
      "num_input_tokens_seen": 33136224,
      "step": 15875
    },
    {
      "epoch": 2.590586507871768,
      "grad_norm": 2.515625,
      "learning_rate": 3.2951027415025806e-05,
      "loss": 1.9913,
      "num_input_tokens_seen": 33147344,
      "step": 15880
    },
    {
      "epoch": 2.591402235092585,
      "grad_norm": 5.125,
      "learning_rate": 3.294171907871849e-05,
      "loss": 3.3177,
      "num_input_tokens_seen": 33156096,
      "step": 15885
    },
    {
      "epoch": 2.5922179623134025,
      "grad_norm": 4.34375,
      "learning_rate": 3.293240951769828e-05,
      "loss": 2.6984,
      "num_input_tokens_seen": 33165744,
      "step": 15890
    },
    {
      "epoch": 2.5930336895342196,
      "grad_norm": 7.09375,
      "learning_rate": 3.2923098733400846e-05,
      "loss": 1.8384,
      "num_input_tokens_seen": 33176192,
      "step": 15895
    },
    {
      "epoch": 2.593849416755037,
      "grad_norm": 7.4375,
      "learning_rate": 3.291378672726202e-05,
      "loss": 2.2854,
      "num_input_tokens_seen": 33186432,
      "step": 15900
    },
    {
      "epoch": 2.5946651439758543,
      "grad_norm": 10.0625,
      "learning_rate": 3.2904473500717824e-05,
      "loss": 2.9531,
      "num_input_tokens_seen": 33197696,
      "step": 15905
    },
    {
      "epoch": 2.595480871196672,
      "grad_norm": 10.0,
      "learning_rate": 3.289515905520449e-05,
      "loss": 2.0581,
      "num_input_tokens_seen": 33206864,
      "step": 15910
    },
    {
      "epoch": 2.596296598417489,
      "grad_norm": 10.0625,
      "learning_rate": 3.288584339215841e-05,
      "loss": 2.2463,
      "num_input_tokens_seen": 33218080,
      "step": 15915
    },
    {
      "epoch": 2.5971123256383066,
      "grad_norm": 6.6875,
      "learning_rate": 3.287652651301617e-05,
      "loss": 2.7555,
      "num_input_tokens_seen": 33228848,
      "step": 15920
    },
    {
      "epoch": 2.5979280528591238,
      "grad_norm": 9.1875,
      "learning_rate": 3.286720841921457e-05,
      "loss": 3.1506,
      "num_input_tokens_seen": 33239872,
      "step": 15925
    },
    {
      "epoch": 2.5987437800799413,
      "grad_norm": 5.46875,
      "learning_rate": 3.285788911219056e-05,
      "loss": 2.1683,
      "num_input_tokens_seen": 33250592,
      "step": 15930
    },
    {
      "epoch": 2.5995595073007585,
      "grad_norm": 12.0625,
      "learning_rate": 3.284856859338131e-05,
      "loss": 2.8576,
      "num_input_tokens_seen": 33261488,
      "step": 15935
    },
    {
      "epoch": 2.600375234521576,
      "grad_norm": 6.90625,
      "learning_rate": 3.283924686422414e-05,
      "loss": 2.319,
      "num_input_tokens_seen": 33273872,
      "step": 15940
    },
    {
      "epoch": 2.601190961742393,
      "grad_norm": 9.1875,
      "learning_rate": 3.282992392615659e-05,
      "loss": 3.1293,
      "num_input_tokens_seen": 33284208,
      "step": 15945
    },
    {
      "epoch": 2.6020066889632107,
      "grad_norm": 11.25,
      "learning_rate": 3.282059978061638e-05,
      "loss": 2.5991,
      "num_input_tokens_seen": 33295584,
      "step": 15950
    },
    {
      "epoch": 2.6028224161840283,
      "grad_norm": 5.5625,
      "learning_rate": 3.28112744290414e-05,
      "loss": 2.2707,
      "num_input_tokens_seen": 33305872,
      "step": 15955
    },
    {
      "epoch": 2.6036381434048455,
      "grad_norm": 8.4375,
      "learning_rate": 3.280194787286974e-05,
      "loss": 3.0486,
      "num_input_tokens_seen": 33317712,
      "step": 15960
    },
    {
      "epoch": 2.6044538706256626,
      "grad_norm": 4.59375,
      "learning_rate": 3.2792620113539674e-05,
      "loss": 1.3371,
      "num_input_tokens_seen": 33328944,
      "step": 15965
    },
    {
      "epoch": 2.60526959784648,
      "grad_norm": 2.9375,
      "learning_rate": 3.278329115248966e-05,
      "loss": 2.552,
      "num_input_tokens_seen": 33340816,
      "step": 15970
    },
    {
      "epoch": 2.6060853250672977,
      "grad_norm": 3.65625,
      "learning_rate": 3.277396099115834e-05,
      "loss": 2.1152,
      "num_input_tokens_seen": 33351152,
      "step": 15975
    },
    {
      "epoch": 2.606901052288115,
      "grad_norm": 3.21875,
      "learning_rate": 3.276462963098454e-05,
      "loss": 2.3813,
      "num_input_tokens_seen": 33361680,
      "step": 15980
    },
    {
      "epoch": 2.607716779508932,
      "grad_norm": 4.34375,
      "learning_rate": 3.275529707340728e-05,
      "loss": 1.8114,
      "num_input_tokens_seen": 33372992,
      "step": 15985
    },
    {
      "epoch": 2.6085325067297496,
      "grad_norm": 10.1875,
      "learning_rate": 3.274596331986574e-05,
      "loss": 2.5709,
      "num_input_tokens_seen": 33382656,
      "step": 15990
    },
    {
      "epoch": 2.609348233950567,
      "grad_norm": 8.8125,
      "learning_rate": 3.273662837179932e-05,
      "loss": 2.1711,
      "num_input_tokens_seen": 33393824,
      "step": 15995
    },
    {
      "epoch": 2.6101639611713843,
      "grad_norm": 15.6875,
      "learning_rate": 3.272729223064758e-05,
      "loss": 3.8721,
      "num_input_tokens_seen": 33403328,
      "step": 16000
    },
    {
      "epoch": 2.6101639611713843,
      "eval_loss": 2.542147636413574,
      "eval_runtime": 134.8552,
      "eval_samples_per_second": 20.207,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 33403328,
      "step": 16000
    },
    {
      "epoch": 2.6109796883922014,
      "grad_norm": 8.0,
      "learning_rate": 3.2717954897850264e-05,
      "loss": 2.5919,
      "num_input_tokens_seen": 33413568,
      "step": 16005
    },
    {
      "epoch": 2.611795415613019,
      "grad_norm": 3.125,
      "learning_rate": 3.270861637484733e-05,
      "loss": 1.279,
      "num_input_tokens_seen": 33424032,
      "step": 16010
    },
    {
      "epoch": 2.6126111428338366,
      "grad_norm": 1.3671875,
      "learning_rate": 3.2699276663078867e-05,
      "loss": 1.663,
      "num_input_tokens_seen": 33435056,
      "step": 16015
    },
    {
      "epoch": 2.6134268700546537,
      "grad_norm": 4.96875,
      "learning_rate": 3.268993576398519e-05,
      "loss": 1.964,
      "num_input_tokens_seen": 33447312,
      "step": 16020
    },
    {
      "epoch": 2.614242597275471,
      "grad_norm": 6.96875,
      "learning_rate": 3.268059367900678e-05,
      "loss": 2.8165,
      "num_input_tokens_seen": 33458480,
      "step": 16025
    },
    {
      "epoch": 2.6150583244962884,
      "grad_norm": 5.75,
      "learning_rate": 3.26712504095843e-05,
      "loss": 1.5581,
      "num_input_tokens_seen": 33468512,
      "step": 16030
    },
    {
      "epoch": 2.615874051717106,
      "grad_norm": 4.53125,
      "learning_rate": 3.2661905957158615e-05,
      "loss": 1.5841,
      "num_input_tokens_seen": 33479328,
      "step": 16035
    },
    {
      "epoch": 2.616689778937923,
      "grad_norm": 9.6875,
      "learning_rate": 3.2652560323170734e-05,
      "loss": 2.9167,
      "num_input_tokens_seen": 33490416,
      "step": 16040
    },
    {
      "epoch": 2.6175055061587407,
      "grad_norm": 7.25,
      "learning_rate": 3.264321350906189e-05,
      "loss": 2.464,
      "num_input_tokens_seen": 33499504,
      "step": 16045
    },
    {
      "epoch": 2.618321233379558,
      "grad_norm": 10.625,
      "learning_rate": 3.263386551627346e-05,
      "loss": 3.0481,
      "num_input_tokens_seen": 33509056,
      "step": 16050
    },
    {
      "epoch": 2.6191369606003754,
      "grad_norm": 5.46875,
      "learning_rate": 3.2624516346247055e-05,
      "loss": 2.2434,
      "num_input_tokens_seen": 33519584,
      "step": 16055
    },
    {
      "epoch": 2.6199526878211925,
      "grad_norm": 7.09375,
      "learning_rate": 3.2615166000424404e-05,
      "loss": 1.7911,
      "num_input_tokens_seen": 33528752,
      "step": 16060
    },
    {
      "epoch": 2.62076841504201,
      "grad_norm": 6.125,
      "learning_rate": 3.260581448024745e-05,
      "loss": 1.6875,
      "num_input_tokens_seen": 33538528,
      "step": 16065
    },
    {
      "epoch": 2.621584142262827,
      "grad_norm": 9.3125,
      "learning_rate": 3.2596461787158335e-05,
      "loss": 3.255,
      "num_input_tokens_seen": 33549104,
      "step": 16070
    },
    {
      "epoch": 2.622399869483645,
      "grad_norm": 4.96875,
      "learning_rate": 3.258710792259934e-05,
      "loss": 3.9284,
      "num_input_tokens_seen": 33559680,
      "step": 16075
    },
    {
      "epoch": 2.623215596704462,
      "grad_norm": 3.90625,
      "learning_rate": 3.257775288801296e-05,
      "loss": 2.4367,
      "num_input_tokens_seen": 33570480,
      "step": 16080
    },
    {
      "epoch": 2.6240313239252795,
      "grad_norm": 10.625,
      "learning_rate": 3.256839668484186e-05,
      "loss": 1.7586,
      "num_input_tokens_seen": 33580912,
      "step": 16085
    },
    {
      "epoch": 2.6248470511460966,
      "grad_norm": 6.28125,
      "learning_rate": 3.255903931452888e-05,
      "loss": 1.8664,
      "num_input_tokens_seen": 33592112,
      "step": 16090
    },
    {
      "epoch": 2.625662778366914,
      "grad_norm": 6.5625,
      "learning_rate": 3.2549680778517045e-05,
      "loss": 2.1405,
      "num_input_tokens_seen": 33602320,
      "step": 16095
    },
    {
      "epoch": 2.6264785055877313,
      "grad_norm": 15.375,
      "learning_rate": 3.2540321078249556e-05,
      "loss": 3.7148,
      "num_input_tokens_seen": 33613648,
      "step": 16100
    },
    {
      "epoch": 2.627294232808549,
      "grad_norm": 6.1875,
      "learning_rate": 3.2530960215169795e-05,
      "loss": 2.2121,
      "num_input_tokens_seen": 33623328,
      "step": 16105
    },
    {
      "epoch": 2.628109960029366,
      "grad_norm": 2.046875,
      "learning_rate": 3.2521598190721345e-05,
      "loss": 1.9363,
      "num_input_tokens_seen": 33632848,
      "step": 16110
    },
    {
      "epoch": 2.6289256872501836,
      "grad_norm": 7.03125,
      "learning_rate": 3.251223500634792e-05,
      "loss": 1.5073,
      "num_input_tokens_seen": 33641904,
      "step": 16115
    },
    {
      "epoch": 2.6297414144710007,
      "grad_norm": 7.8125,
      "learning_rate": 3.2502870663493445e-05,
      "loss": 1.6046,
      "num_input_tokens_seen": 33653152,
      "step": 16120
    },
    {
      "epoch": 2.6305571416918183,
      "grad_norm": 3.84375,
      "learning_rate": 3.249350516360203e-05,
      "loss": 1.76,
      "num_input_tokens_seen": 33663984,
      "step": 16125
    },
    {
      "epoch": 2.631372868912636,
      "grad_norm": 5.8125,
      "learning_rate": 3.248413850811797e-05,
      "loss": 1.8946,
      "num_input_tokens_seen": 33675664,
      "step": 16130
    },
    {
      "epoch": 2.632188596133453,
      "grad_norm": 4.40625,
      "learning_rate": 3.2474770698485677e-05,
      "loss": 1.9752,
      "num_input_tokens_seen": 33686080,
      "step": 16135
    },
    {
      "epoch": 2.63300432335427,
      "grad_norm": 8.0625,
      "learning_rate": 3.246540173614983e-05,
      "loss": 2.8633,
      "num_input_tokens_seen": 33695568,
      "step": 16140
    },
    {
      "epoch": 2.6338200505750877,
      "grad_norm": 6.1875,
      "learning_rate": 3.2456031622555197e-05,
      "loss": 1.7717,
      "num_input_tokens_seen": 33708096,
      "step": 16145
    },
    {
      "epoch": 2.6346357777959053,
      "grad_norm": 7.875,
      "learning_rate": 3.2446660359146794e-05,
      "loss": 2.1496,
      "num_input_tokens_seen": 33718976,
      "step": 16150
    },
    {
      "epoch": 2.6354515050167224,
      "grad_norm": 5.0625,
      "learning_rate": 3.2437287947369786e-05,
      "loss": 3.0623,
      "num_input_tokens_seen": 33730048,
      "step": 16155
    },
    {
      "epoch": 2.6362672322375396,
      "grad_norm": 6.9375,
      "learning_rate": 3.2427914388669525e-05,
      "loss": 3.4004,
      "num_input_tokens_seen": 33740688,
      "step": 16160
    },
    {
      "epoch": 2.637082959458357,
      "grad_norm": 11.0,
      "learning_rate": 3.241853968449151e-05,
      "loss": 3.1018,
      "num_input_tokens_seen": 33750912,
      "step": 16165
    },
    {
      "epoch": 2.6378986866791747,
      "grad_norm": 11.3125,
      "learning_rate": 3.240916383628144e-05,
      "loss": 2.6743,
      "num_input_tokens_seen": 33761520,
      "step": 16170
    },
    {
      "epoch": 2.638714413899992,
      "grad_norm": 9.75,
      "learning_rate": 3.239978684548521e-05,
      "loss": 3.6831,
      "num_input_tokens_seen": 33772096,
      "step": 16175
    },
    {
      "epoch": 2.639530141120809,
      "grad_norm": 9.5,
      "learning_rate": 3.239040871354885e-05,
      "loss": 3.7123,
      "num_input_tokens_seen": 33783264,
      "step": 16180
    },
    {
      "epoch": 2.6403458683416265,
      "grad_norm": 3.4375,
      "learning_rate": 3.2381029441918596e-05,
      "loss": 1.7228,
      "num_input_tokens_seen": 33793840,
      "step": 16185
    },
    {
      "epoch": 2.641161595562444,
      "grad_norm": 5.8125,
      "learning_rate": 3.2371649032040845e-05,
      "loss": 2.4894,
      "num_input_tokens_seen": 33804768,
      "step": 16190
    },
    {
      "epoch": 2.6419773227832613,
      "grad_norm": 9.5625,
      "learning_rate": 3.2362267485362174e-05,
      "loss": 2.3563,
      "num_input_tokens_seen": 33815648,
      "step": 16195
    },
    {
      "epoch": 2.6427930500040784,
      "grad_norm": 6.375,
      "learning_rate": 3.235288480332934e-05,
      "loss": 1.8551,
      "num_input_tokens_seen": 33827808,
      "step": 16200
    },
    {
      "epoch": 2.6427930500040784,
      "eval_loss": 2.546783208847046,
      "eval_runtime": 134.8424,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 33827808,
      "step": 16200
    },
    {
      "epoch": 2.643608777224896,
      "grad_norm": 9.4375,
      "learning_rate": 3.234350098738927e-05,
      "loss": 2.7817,
      "num_input_tokens_seen": 33838544,
      "step": 16205
    },
    {
      "epoch": 2.6444245044457135,
      "grad_norm": 10.375,
      "learning_rate": 3.233411603898906e-05,
      "loss": 2.1702,
      "num_input_tokens_seen": 33849280,
      "step": 16210
    },
    {
      "epoch": 2.6452402316665307,
      "grad_norm": 7.96875,
      "learning_rate": 3.232472995957599e-05,
      "loss": 1.6365,
      "num_input_tokens_seen": 33859808,
      "step": 16215
    },
    {
      "epoch": 2.6460559588873482,
      "grad_norm": 10.0,
      "learning_rate": 3.231534275059751e-05,
      "loss": 3.3369,
      "num_input_tokens_seen": 33870112,
      "step": 16220
    },
    {
      "epoch": 2.6468716861081654,
      "grad_norm": 2.234375,
      "learning_rate": 3.230595441350125e-05,
      "loss": 2.2198,
      "num_input_tokens_seen": 33880832,
      "step": 16225
    },
    {
      "epoch": 2.647687413328983,
      "grad_norm": 8.25,
      "learning_rate": 3.2296564949735e-05,
      "loss": 4.5615,
      "num_input_tokens_seen": 33890960,
      "step": 16230
    },
    {
      "epoch": 2.6485031405498,
      "grad_norm": 8.3125,
      "learning_rate": 3.228717436074675e-05,
      "loss": 2.0312,
      "num_input_tokens_seen": 33900816,
      "step": 16235
    },
    {
      "epoch": 2.6493188677706176,
      "grad_norm": 9.0,
      "learning_rate": 3.227778264798463e-05,
      "loss": 2.5483,
      "num_input_tokens_seen": 33910464,
      "step": 16240
    },
    {
      "epoch": 2.6501345949914348,
      "grad_norm": 8.4375,
      "learning_rate": 3.226838981289698e-05,
      "loss": 2.0637,
      "num_input_tokens_seen": 33921520,
      "step": 16245
    },
    {
      "epoch": 2.6509503222122524,
      "grad_norm": 0.166015625,
      "learning_rate": 3.225899585693227e-05,
      "loss": 0.6785,
      "num_input_tokens_seen": 33931552,
      "step": 16250
    },
    {
      "epoch": 2.6517660494330695,
      "grad_norm": 8.375,
      "learning_rate": 3.224960078153918e-05,
      "loss": 2.1385,
      "num_input_tokens_seen": 33943072,
      "step": 16255
    },
    {
      "epoch": 2.652581776653887,
      "grad_norm": 7.0625,
      "learning_rate": 3.224020458816655e-05,
      "loss": 3.2142,
      "num_input_tokens_seen": 33952448,
      "step": 16260
    },
    {
      "epoch": 2.653397503874704,
      "grad_norm": 5.4375,
      "learning_rate": 3.223080727826337e-05,
      "loss": 3.8898,
      "num_input_tokens_seen": 33963264,
      "step": 16265
    },
    {
      "epoch": 2.6542132310955218,
      "grad_norm": 9.125,
      "learning_rate": 3.222140885327885e-05,
      "loss": 3.6363,
      "num_input_tokens_seen": 33974368,
      "step": 16270
    },
    {
      "epoch": 2.655028958316339,
      "grad_norm": 14.25,
      "learning_rate": 3.221200931466234e-05,
      "loss": 2.9565,
      "num_input_tokens_seen": 33985248,
      "step": 16275
    },
    {
      "epoch": 2.6558446855371565,
      "grad_norm": 3.703125,
      "learning_rate": 3.220260866386336e-05,
      "loss": 2.5862,
      "num_input_tokens_seen": 33995632,
      "step": 16280
    },
    {
      "epoch": 2.6566604127579736,
      "grad_norm": 9.625,
      "learning_rate": 3.21932069023316e-05,
      "loss": 3.0562,
      "num_input_tokens_seen": 34006160,
      "step": 16285
    },
    {
      "epoch": 2.657476139978791,
      "grad_norm": 5.875,
      "learning_rate": 3.218380403151695e-05,
      "loss": 2.0022,
      "num_input_tokens_seen": 34016576,
      "step": 16290
    },
    {
      "epoch": 2.6582918671996083,
      "grad_norm": 5.28125,
      "learning_rate": 3.217440005286943e-05,
      "loss": 1.3833,
      "num_input_tokens_seen": 34027776,
      "step": 16295
    },
    {
      "epoch": 2.659107594420426,
      "grad_norm": 4.78125,
      "learning_rate": 3.216499496783928e-05,
      "loss": 2.7437,
      "num_input_tokens_seen": 34038256,
      "step": 16300
    },
    {
      "epoch": 2.659923321641243,
      "grad_norm": 12.3125,
      "learning_rate": 3.2155588777876856e-05,
      "loss": 2.2299,
      "num_input_tokens_seen": 34046896,
      "step": 16305
    },
    {
      "epoch": 2.6607390488620606,
      "grad_norm": 8.0625,
      "learning_rate": 3.214618148443273e-05,
      "loss": 2.6666,
      "num_input_tokens_seen": 34056896,
      "step": 16310
    },
    {
      "epoch": 2.6615547760828777,
      "grad_norm": 4.875,
      "learning_rate": 3.2136773088957595e-05,
      "loss": 2.8164,
      "num_input_tokens_seen": 34066672,
      "step": 16315
    },
    {
      "epoch": 2.6623705033036953,
      "grad_norm": 7.9375,
      "learning_rate": 3.2127363592902374e-05,
      "loss": 3.217,
      "num_input_tokens_seen": 34078384,
      "step": 16320
    },
    {
      "epoch": 2.663186230524513,
      "grad_norm": 7.3125,
      "learning_rate": 3.211795299771812e-05,
      "loss": 3.2504,
      "num_input_tokens_seen": 34087568,
      "step": 16325
    },
    {
      "epoch": 2.66400195774533,
      "grad_norm": 5.21875,
      "learning_rate": 3.210854130485605e-05,
      "loss": 2.2324,
      "num_input_tokens_seen": 34098896,
      "step": 16330
    },
    {
      "epoch": 2.664817684966147,
      "grad_norm": 10.875,
      "learning_rate": 3.209912851576759e-05,
      "loss": 2.5861,
      "num_input_tokens_seen": 34108032,
      "step": 16335
    },
    {
      "epoch": 2.6656334121869647,
      "grad_norm": 2.484375,
      "learning_rate": 3.208971463190431e-05,
      "loss": 2.3717,
      "num_input_tokens_seen": 34118448,
      "step": 16340
    },
    {
      "epoch": 2.6664491394077823,
      "grad_norm": 10.3125,
      "learning_rate": 3.208029965471793e-05,
      "loss": 2.9142,
      "num_input_tokens_seen": 34129936,
      "step": 16345
    },
    {
      "epoch": 2.6672648666285994,
      "grad_norm": 4.84375,
      "learning_rate": 3.2070883585660364e-05,
      "loss": 1.8916,
      "num_input_tokens_seen": 34141632,
      "step": 16350
    },
    {
      "epoch": 2.6680805938494165,
      "grad_norm": 4.65625,
      "learning_rate": 3.20614664261837e-05,
      "loss": 2.1912,
      "num_input_tokens_seen": 34150912,
      "step": 16355
    },
    {
      "epoch": 2.668896321070234,
      "grad_norm": 2.390625,
      "learning_rate": 3.205204817774016e-05,
      "loss": 1.105,
      "num_input_tokens_seen": 34161312,
      "step": 16360
    },
    {
      "epoch": 2.6697120482910517,
      "grad_norm": 11.875,
      "learning_rate": 3.204262884178218e-05,
      "loss": 2.2465,
      "num_input_tokens_seen": 34173104,
      "step": 16365
    },
    {
      "epoch": 2.670527775511869,
      "grad_norm": 6.15625,
      "learning_rate": 3.2033208419762314e-05,
      "loss": 1.3373,
      "num_input_tokens_seen": 34183376,
      "step": 16370
    },
    {
      "epoch": 2.671343502732686,
      "grad_norm": 4.1875,
      "learning_rate": 3.2023786913133344e-05,
      "loss": 1.5983,
      "num_input_tokens_seen": 34193168,
      "step": 16375
    },
    {
      "epoch": 2.6721592299535035,
      "grad_norm": 6.28125,
      "learning_rate": 3.201436432334816e-05,
      "loss": 2.2728,
      "num_input_tokens_seen": 34203664,
      "step": 16380
    },
    {
      "epoch": 2.672974957174321,
      "grad_norm": 7.125,
      "learning_rate": 3.2004940651859844e-05,
      "loss": 3.2168,
      "num_input_tokens_seen": 34214544,
      "step": 16385
    },
    {
      "epoch": 2.6737906843951382,
      "grad_norm": 5.625,
      "learning_rate": 3.1995515900121655e-05,
      "loss": 2.6577,
      "num_input_tokens_seen": 34225424,
      "step": 16390
    },
    {
      "epoch": 2.6746064116159554,
      "grad_norm": 7.21875,
      "learning_rate": 3.1986090069587e-05,
      "loss": 2.5505,
      "num_input_tokens_seen": 34236496,
      "step": 16395
    },
    {
      "epoch": 2.675422138836773,
      "grad_norm": 3.8125,
      "learning_rate": 3.1976663161709466e-05,
      "loss": 2.6711,
      "num_input_tokens_seen": 34245456,
      "step": 16400
    },
    {
      "epoch": 2.675422138836773,
      "eval_loss": 2.5521464347839355,
      "eval_runtime": 135.1348,
      "eval_samples_per_second": 20.165,
      "eval_steps_per_second": 10.086,
      "num_input_tokens_seen": 34245456,
      "step": 16400
    },
    {
      "epoch": 2.6762378660575905,
      "grad_norm": 6.3125,
      "learning_rate": 3.196723517794279e-05,
      "loss": 1.8623,
      "num_input_tokens_seen": 34256464,
      "step": 16405
    },
    {
      "epoch": 2.6770535932784076,
      "grad_norm": 6.375,
      "learning_rate": 3.19578061197409e-05,
      "loss": 3.2433,
      "num_input_tokens_seen": 34267728,
      "step": 16410
    },
    {
      "epoch": 2.677869320499225,
      "grad_norm": 4.125,
      "learning_rate": 3.194837598855787e-05,
      "loss": 2.6012,
      "num_input_tokens_seen": 34278192,
      "step": 16415
    },
    {
      "epoch": 2.6786850477200423,
      "grad_norm": 8.25,
      "learning_rate": 3.193894478584794e-05,
      "loss": 3.3096,
      "num_input_tokens_seen": 34288624,
      "step": 16420
    },
    {
      "epoch": 2.67950077494086,
      "grad_norm": 11.3125,
      "learning_rate": 3.192951251306553e-05,
      "loss": 3.0019,
      "num_input_tokens_seen": 34299536,
      "step": 16425
    },
    {
      "epoch": 2.680316502161677,
      "grad_norm": 8.1875,
      "learning_rate": 3.192007917166521e-05,
      "loss": 1.7351,
      "num_input_tokens_seen": 34310640,
      "step": 16430
    },
    {
      "epoch": 2.6811322293824946,
      "grad_norm": 7.375,
      "learning_rate": 3.191064476310171e-05,
      "loss": 3.6402,
      "num_input_tokens_seen": 34321392,
      "step": 16435
    },
    {
      "epoch": 2.6819479566033118,
      "grad_norm": 6.9375,
      "learning_rate": 3.1901209288829944e-05,
      "loss": 4.4265,
      "num_input_tokens_seen": 34332544,
      "step": 16440
    },
    {
      "epoch": 2.6827636838241293,
      "grad_norm": 6.40625,
      "learning_rate": 3.1891772750304985e-05,
      "loss": 3.1058,
      "num_input_tokens_seen": 34342800,
      "step": 16445
    },
    {
      "epoch": 2.6835794110449465,
      "grad_norm": 8.75,
      "learning_rate": 3.188233514898206e-05,
      "loss": 2.7022,
      "num_input_tokens_seen": 34353200,
      "step": 16450
    },
    {
      "epoch": 2.684395138265764,
      "grad_norm": 7.84375,
      "learning_rate": 3.187289648631657e-05,
      "loss": 1.7889,
      "num_input_tokens_seen": 34364784,
      "step": 16455
    },
    {
      "epoch": 2.685210865486581,
      "grad_norm": 8.1875,
      "learning_rate": 3.186345676376406e-05,
      "loss": 2.9518,
      "num_input_tokens_seen": 34375824,
      "step": 16460
    },
    {
      "epoch": 2.6860265927073987,
      "grad_norm": 3.171875,
      "learning_rate": 3.1854015982780275e-05,
      "loss": 2.4943,
      "num_input_tokens_seen": 34387168,
      "step": 16465
    },
    {
      "epoch": 2.686842319928216,
      "grad_norm": 13.5,
      "learning_rate": 3.1844574144821084e-05,
      "loss": 2.1364,
      "num_input_tokens_seen": 34398128,
      "step": 16470
    },
    {
      "epoch": 2.6876580471490334,
      "grad_norm": 4.625,
      "learning_rate": 3.1835131251342554e-05,
      "loss": 2.8713,
      "num_input_tokens_seen": 34409296,
      "step": 16475
    },
    {
      "epoch": 2.6884737743698506,
      "grad_norm": 3.828125,
      "learning_rate": 3.182568730380089e-05,
      "loss": 1.813,
      "num_input_tokens_seen": 34418224,
      "step": 16480
    },
    {
      "epoch": 2.689289501590668,
      "grad_norm": 11.1875,
      "learning_rate": 3.181624230365245e-05,
      "loss": 3.4145,
      "num_input_tokens_seen": 34428720,
      "step": 16485
    },
    {
      "epoch": 2.6901052288114853,
      "grad_norm": 10.6875,
      "learning_rate": 3.180679625235381e-05,
      "loss": 2.5642,
      "num_input_tokens_seen": 34438752,
      "step": 16490
    },
    {
      "epoch": 2.690920956032303,
      "grad_norm": 9.8125,
      "learning_rate": 3.1797349151361646e-05,
      "loss": 2.0636,
      "num_input_tokens_seen": 34449840,
      "step": 16495
    },
    {
      "epoch": 2.6917366832531204,
      "grad_norm": 6.59375,
      "learning_rate": 3.178790100213281e-05,
      "loss": 2.8,
      "num_input_tokens_seen": 34460352,
      "step": 16500
    },
    {
      "epoch": 2.6925524104739376,
      "grad_norm": 10.625,
      "learning_rate": 3.1778451806124346e-05,
      "loss": 2.9267,
      "num_input_tokens_seen": 34472368,
      "step": 16505
    },
    {
      "epoch": 2.6933681376947547,
      "grad_norm": 12.0625,
      "learning_rate": 3.176900156479342e-05,
      "loss": 3.2725,
      "num_input_tokens_seen": 34484160,
      "step": 16510
    },
    {
      "epoch": 2.6941838649155723,
      "grad_norm": 7.78125,
      "learning_rate": 3.17595502795974e-05,
      "loss": 3.5583,
      "num_input_tokens_seen": 34494000,
      "step": 16515
    },
    {
      "epoch": 2.69499959213639,
      "grad_norm": 8.4375,
      "learning_rate": 3.175009795199377e-05,
      "loss": 2.0964,
      "num_input_tokens_seen": 34503968,
      "step": 16520
    },
    {
      "epoch": 2.695815319357207,
      "grad_norm": 12.5625,
      "learning_rate": 3.1740644583440224e-05,
      "loss": 3.9194,
      "num_input_tokens_seen": 34514832,
      "step": 16525
    },
    {
      "epoch": 2.696631046578024,
      "grad_norm": 13.25,
      "learning_rate": 3.173119017539457e-05,
      "loss": 3.6509,
      "num_input_tokens_seen": 34525824,
      "step": 16530
    },
    {
      "epoch": 2.6974467737988417,
      "grad_norm": 4.625,
      "learning_rate": 3.172173472931479e-05,
      "loss": 3.5547,
      "num_input_tokens_seen": 34536832,
      "step": 16535
    },
    {
      "epoch": 2.6982625010196593,
      "grad_norm": 7.59375,
      "learning_rate": 3.1712278246659055e-05,
      "loss": 3.1259,
      "num_input_tokens_seen": 34546720,
      "step": 16540
    },
    {
      "epoch": 2.6990782282404764,
      "grad_norm": 5.28125,
      "learning_rate": 3.170282072888566e-05,
      "loss": 2.0688,
      "num_input_tokens_seen": 34555904,
      "step": 16545
    },
    {
      "epoch": 2.6998939554612935,
      "grad_norm": 2.484375,
      "learning_rate": 3.169336217745307e-05,
      "loss": 2.6751,
      "num_input_tokens_seen": 34566944,
      "step": 16550
    },
    {
      "epoch": 2.700709682682111,
      "grad_norm": 8.75,
      "learning_rate": 3.1683902593819924e-05,
      "loss": 2.8128,
      "num_input_tokens_seen": 34578752,
      "step": 16555
    },
    {
      "epoch": 2.7015254099029287,
      "grad_norm": 4.6875,
      "learning_rate": 3.1674441979445e-05,
      "loss": 2.1039,
      "num_input_tokens_seen": 34589856,
      "step": 16560
    },
    {
      "epoch": 2.702341137123746,
      "grad_norm": 11.125,
      "learning_rate": 3.166498033578725e-05,
      "loss": 3.8616,
      "num_input_tokens_seen": 34600896,
      "step": 16565
    },
    {
      "epoch": 2.703156864344563,
      "grad_norm": 6.8125,
      "learning_rate": 3.165551766430578e-05,
      "loss": 2.9954,
      "num_input_tokens_seen": 34611376,
      "step": 16570
    },
    {
      "epoch": 2.7039725915653805,
      "grad_norm": 11.0,
      "learning_rate": 3.164605396645984e-05,
      "loss": 0.9768,
      "num_input_tokens_seen": 34622896,
      "step": 16575
    },
    {
      "epoch": 2.704788318786198,
      "grad_norm": 2.8125,
      "learning_rate": 3.163658924370886e-05,
      "loss": 3.186,
      "num_input_tokens_seen": 34634352,
      "step": 16580
    },
    {
      "epoch": 2.705604046007015,
      "grad_norm": 6.25,
      "learning_rate": 3.1627123497512415e-05,
      "loss": 2.7933,
      "num_input_tokens_seen": 34645040,
      "step": 16585
    },
    {
      "epoch": 2.7064197732278323,
      "grad_norm": 6.75,
      "learning_rate": 3.1617656729330245e-05,
      "loss": 2.1957,
      "num_input_tokens_seen": 34654080,
      "step": 16590
    },
    {
      "epoch": 2.70723550044865,
      "grad_norm": 8.125,
      "learning_rate": 3.1608188940622255e-05,
      "loss": 2.1157,
      "num_input_tokens_seen": 34663200,
      "step": 16595
    },
    {
      "epoch": 2.7080512276694675,
      "grad_norm": 1.9296875,
      "learning_rate": 3.159872013284847e-05,
      "loss": 1.1239,
      "num_input_tokens_seen": 34673616,
      "step": 16600
    },
    {
      "epoch": 2.7080512276694675,
      "eval_loss": 2.545654535293579,
      "eval_runtime": 134.8327,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.109,
      "num_input_tokens_seen": 34673616,
      "step": 16600
    },
    {
      "epoch": 2.7088669548902846,
      "grad_norm": 13.625,
      "learning_rate": 3.1589250307469134e-05,
      "loss": 2.0076,
      "num_input_tokens_seen": 34684144,
      "step": 16605
    },
    {
      "epoch": 2.709682682111102,
      "grad_norm": 5.15625,
      "learning_rate": 3.1579779465944586e-05,
      "loss": 1.5802,
      "num_input_tokens_seen": 34695328,
      "step": 16610
    },
    {
      "epoch": 2.7104984093319193,
      "grad_norm": 4.625,
      "learning_rate": 3.1570307609735363e-05,
      "loss": 2.5241,
      "num_input_tokens_seen": 34706288,
      "step": 16615
    },
    {
      "epoch": 2.711314136552737,
      "grad_norm": 10.5,
      "learning_rate": 3.156083474030213e-05,
      "loss": 2.6136,
      "num_input_tokens_seen": 34717520,
      "step": 16620
    },
    {
      "epoch": 2.712129863773554,
      "grad_norm": 11.125,
      "learning_rate": 3.155136085910573e-05,
      "loss": 3.1711,
      "num_input_tokens_seen": 34728480,
      "step": 16625
    },
    {
      "epoch": 2.7129455909943716,
      "grad_norm": 10.375,
      "learning_rate": 3.154188596760717e-05,
      "loss": 3.0228,
      "num_input_tokens_seen": 34738880,
      "step": 16630
    },
    {
      "epoch": 2.7137613182151887,
      "grad_norm": 5.125,
      "learning_rate": 3.153241006726757e-05,
      "loss": 1.6732,
      "num_input_tokens_seen": 34749200,
      "step": 16635
    },
    {
      "epoch": 2.7145770454360063,
      "grad_norm": 3.515625,
      "learning_rate": 3.152293315954825e-05,
      "loss": 2.7277,
      "num_input_tokens_seen": 34759968,
      "step": 16640
    },
    {
      "epoch": 2.7153927726568234,
      "grad_norm": 9.375,
      "learning_rate": 3.1513455245910666e-05,
      "loss": 2.0066,
      "num_input_tokens_seen": 34769664,
      "step": 16645
    },
    {
      "epoch": 2.716208499877641,
      "grad_norm": 8.0,
      "learning_rate": 3.150397632781643e-05,
      "loss": 2.9875,
      "num_input_tokens_seen": 34780544,
      "step": 16650
    },
    {
      "epoch": 2.717024227098458,
      "grad_norm": 8.0,
      "learning_rate": 3.149449640672731e-05,
      "loss": 0.8494,
      "num_input_tokens_seen": 34790144,
      "step": 16655
    },
    {
      "epoch": 2.7178399543192757,
      "grad_norm": 5.375,
      "learning_rate": 3.148501548410523e-05,
      "loss": 2.4132,
      "num_input_tokens_seen": 34800832,
      "step": 16660
    },
    {
      "epoch": 2.718655681540093,
      "grad_norm": 4.65625,
      "learning_rate": 3.1475533561412256e-05,
      "loss": 2.1299,
      "num_input_tokens_seen": 34810576,
      "step": 16665
    },
    {
      "epoch": 2.7194714087609104,
      "grad_norm": 8.125,
      "learning_rate": 3.146605064011065e-05,
      "loss": 2.204,
      "num_input_tokens_seen": 34820976,
      "step": 16670
    },
    {
      "epoch": 2.7202871359817276,
      "grad_norm": 7.0,
      "learning_rate": 3.145656672166277e-05,
      "loss": 2.6839,
      "num_input_tokens_seen": 34830256,
      "step": 16675
    },
    {
      "epoch": 2.721102863202545,
      "grad_norm": 2.578125,
      "learning_rate": 3.144708180753116e-05,
      "loss": 1.743,
      "num_input_tokens_seen": 34840848,
      "step": 16680
    },
    {
      "epoch": 2.7219185904233623,
      "grad_norm": 11.1875,
      "learning_rate": 3.143759589917851e-05,
      "loss": 1.5557,
      "num_input_tokens_seen": 34851184,
      "step": 16685
    },
    {
      "epoch": 2.72273431764418,
      "grad_norm": 6.625,
      "learning_rate": 3.142810899806768e-05,
      "loss": 1.928,
      "num_input_tokens_seen": 34861232,
      "step": 16690
    },
    {
      "epoch": 2.7235500448649974,
      "grad_norm": 5.125,
      "learning_rate": 3.141862110566166e-05,
      "loss": 2.1565,
      "num_input_tokens_seen": 34869968,
      "step": 16695
    },
    {
      "epoch": 2.7243657720858145,
      "grad_norm": 1.859375,
      "learning_rate": 3.1409132223423606e-05,
      "loss": 1.488,
      "num_input_tokens_seen": 34881120,
      "step": 16700
    },
    {
      "epoch": 2.7251814993066317,
      "grad_norm": 2.40625,
      "learning_rate": 3.139964235281682e-05,
      "loss": 2.4342,
      "num_input_tokens_seen": 34891696,
      "step": 16705
    },
    {
      "epoch": 2.7259972265274492,
      "grad_norm": 0.1298828125,
      "learning_rate": 3.139015149530476e-05,
      "loss": 2.2172,
      "num_input_tokens_seen": 34901808,
      "step": 16710
    },
    {
      "epoch": 2.726812953748267,
      "grad_norm": 6.6875,
      "learning_rate": 3.1380659652351034e-05,
      "loss": 1.335,
      "num_input_tokens_seen": 34911216,
      "step": 16715
    },
    {
      "epoch": 2.727628680969084,
      "grad_norm": 10.0,
      "learning_rate": 3.137116682541941e-05,
      "loss": 2.4858,
      "num_input_tokens_seen": 34922304,
      "step": 16720
    },
    {
      "epoch": 2.728444408189901,
      "grad_norm": 3.46875,
      "learning_rate": 3.136167301597379e-05,
      "loss": 1.628,
      "num_input_tokens_seen": 34932336,
      "step": 16725
    },
    {
      "epoch": 2.7292601354107187,
      "grad_norm": 5.71875,
      "learning_rate": 3.1352178225478254e-05,
      "loss": 1.7993,
      "num_input_tokens_seen": 34942880,
      "step": 16730
    },
    {
      "epoch": 2.7300758626315362,
      "grad_norm": 8.25,
      "learning_rate": 3.1342682455396996e-05,
      "loss": 1.4979,
      "num_input_tokens_seen": 34952592,
      "step": 16735
    },
    {
      "epoch": 2.7308915898523534,
      "grad_norm": 12.125,
      "learning_rate": 3.133318570719441e-05,
      "loss": 2.1121,
      "num_input_tokens_seen": 34964592,
      "step": 16740
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 7.0,
      "learning_rate": 3.132368798233499e-05,
      "loss": 1.6761,
      "num_input_tokens_seen": 34975328,
      "step": 16745
    },
    {
      "epoch": 2.732523044293988,
      "grad_norm": 9.8125,
      "learning_rate": 3.131418928228342e-05,
      "loss": 2.7917,
      "num_input_tokens_seen": 34985088,
      "step": 16750
    },
    {
      "epoch": 2.7333387715148056,
      "grad_norm": 5.15625,
      "learning_rate": 3.1304689608504514e-05,
      "loss": 3.6144,
      "num_input_tokens_seen": 34994960,
      "step": 16755
    },
    {
      "epoch": 2.7341544987356228,
      "grad_norm": 8.0,
      "learning_rate": 3.129518896246324e-05,
      "loss": 2.2451,
      "num_input_tokens_seen": 35005184,
      "step": 16760
    },
    {
      "epoch": 2.73497022595644,
      "grad_norm": 2.5,
      "learning_rate": 3.128568734562472e-05,
      "loss": 2.0785,
      "num_input_tokens_seen": 35018112,
      "step": 16765
    },
    {
      "epoch": 2.7357859531772575,
      "grad_norm": 8.1875,
      "learning_rate": 3.127618475945421e-05,
      "loss": 2.3159,
      "num_input_tokens_seen": 35029856,
      "step": 16770
    },
    {
      "epoch": 2.736601680398075,
      "grad_norm": 6.40625,
      "learning_rate": 3.126668120541715e-05,
      "loss": 3.1607,
      "num_input_tokens_seen": 35040416,
      "step": 16775
    },
    {
      "epoch": 2.737417407618892,
      "grad_norm": 15.0625,
      "learning_rate": 3.1257176684979096e-05,
      "loss": 3.0275,
      "num_input_tokens_seen": 35051920,
      "step": 16780
    },
    {
      "epoch": 2.7382331348397098,
      "grad_norm": 4.75,
      "learning_rate": 3.124767119960576e-05,
      "loss": 3.6674,
      "num_input_tokens_seen": 35060656,
      "step": 16785
    },
    {
      "epoch": 2.739048862060527,
      "grad_norm": 3.8125,
      "learning_rate": 3.123816475076301e-05,
      "loss": 1.5737,
      "num_input_tokens_seen": 35072064,
      "step": 16790
    },
    {
      "epoch": 2.7398645892813445,
      "grad_norm": 6.4375,
      "learning_rate": 3.122865733991687e-05,
      "loss": 1.9718,
      "num_input_tokens_seen": 35081360,
      "step": 16795
    },
    {
      "epoch": 2.7406803165021616,
      "grad_norm": 7.625,
      "learning_rate": 3.1219148968533486e-05,
      "loss": 1.4874,
      "num_input_tokens_seen": 35089872,
      "step": 16800
    },
    {
      "epoch": 2.7406803165021616,
      "eval_loss": 2.5342445373535156,
      "eval_runtime": 134.8029,
      "eval_samples_per_second": 20.215,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 35089872,
      "step": 16800
    },
    {
      "epoch": 2.741496043722979,
      "grad_norm": 4.84375,
      "learning_rate": 3.120963963807918e-05,
      "loss": 1.1901,
      "num_input_tokens_seen": 35100608,
      "step": 16805
    },
    {
      "epoch": 2.7423117709437963,
      "grad_norm": 12.5625,
      "learning_rate": 3.12001293500204e-05,
      "loss": 2.1622,
      "num_input_tokens_seen": 35111216,
      "step": 16810
    },
    {
      "epoch": 2.743127498164614,
      "grad_norm": 6.84375,
      "learning_rate": 3.1190618105823765e-05,
      "loss": 3.1671,
      "num_input_tokens_seen": 35120848,
      "step": 16815
    },
    {
      "epoch": 2.743943225385431,
      "grad_norm": 12.875,
      "learning_rate": 3.118110590695603e-05,
      "loss": 2.3103,
      "num_input_tokens_seen": 35130416,
      "step": 16820
    },
    {
      "epoch": 2.7447589526062486,
      "grad_norm": 1.7578125,
      "learning_rate": 3.117159275488407e-05,
      "loss": 1.6903,
      "num_input_tokens_seen": 35141152,
      "step": 16825
    },
    {
      "epoch": 2.7455746798270657,
      "grad_norm": 11.8125,
      "learning_rate": 3.1162078651074956e-05,
      "loss": 2.0126,
      "num_input_tokens_seen": 35151424,
      "step": 16830
    },
    {
      "epoch": 2.7463904070478833,
      "grad_norm": 6.3125,
      "learning_rate": 3.1152563596995885e-05,
      "loss": 2.7796,
      "num_input_tokens_seen": 35162128,
      "step": 16835
    },
    {
      "epoch": 2.7472061342687004,
      "grad_norm": 5.71875,
      "learning_rate": 3.1143047594114186e-05,
      "loss": 2.2402,
      "num_input_tokens_seen": 35172688,
      "step": 16840
    },
    {
      "epoch": 2.748021861489518,
      "grad_norm": 6.4375,
      "learning_rate": 3.113353064389734e-05,
      "loss": 2.0471,
      "num_input_tokens_seen": 35182672,
      "step": 16845
    },
    {
      "epoch": 2.748837588710335,
      "grad_norm": 5.8125,
      "learning_rate": 3.1124012747812993e-05,
      "loss": 1.8655,
      "num_input_tokens_seen": 35193920,
      "step": 16850
    },
    {
      "epoch": 2.7496533159311527,
      "grad_norm": 1.8828125,
      "learning_rate": 3.1114493907328936e-05,
      "loss": 2.5842,
      "num_input_tokens_seen": 35204576,
      "step": 16855
    },
    {
      "epoch": 2.75046904315197,
      "grad_norm": 3.828125,
      "learning_rate": 3.110497412391306e-05,
      "loss": 2.4617,
      "num_input_tokens_seen": 35215056,
      "step": 16860
    },
    {
      "epoch": 2.7512847703727874,
      "grad_norm": 9.4375,
      "learning_rate": 3.1095453399033466e-05,
      "loss": 2.5978,
      "num_input_tokens_seen": 35223456,
      "step": 16865
    },
    {
      "epoch": 2.7521004975936045,
      "grad_norm": 5.78125,
      "learning_rate": 3.108593173415835e-05,
      "loss": 1.5533,
      "num_input_tokens_seen": 35235344,
      "step": 16870
    },
    {
      "epoch": 2.752916224814422,
      "grad_norm": 5.1875,
      "learning_rate": 3.107640913075609e-05,
      "loss": 2.0874,
      "num_input_tokens_seen": 35245824,
      "step": 16875
    },
    {
      "epoch": 2.7537319520352392,
      "grad_norm": 7.40625,
      "learning_rate": 3.106688559029517e-05,
      "loss": 2.311,
      "num_input_tokens_seen": 35255808,
      "step": 16880
    },
    {
      "epoch": 2.754547679256057,
      "grad_norm": 6.65625,
      "learning_rate": 3.105736111424425e-05,
      "loss": 2.7496,
      "num_input_tokens_seen": 35265232,
      "step": 16885
    },
    {
      "epoch": 2.7553634064768744,
      "grad_norm": 8.6875,
      "learning_rate": 3.1047835704072136e-05,
      "loss": 4.0668,
      "num_input_tokens_seen": 35274480,
      "step": 16890
    },
    {
      "epoch": 2.7561791336976915,
      "grad_norm": 8.3125,
      "learning_rate": 3.103830936124775e-05,
      "loss": 4.0149,
      "num_input_tokens_seen": 35284816,
      "step": 16895
    },
    {
      "epoch": 2.7569948609185086,
      "grad_norm": 12.3125,
      "learning_rate": 3.102878208724018e-05,
      "loss": 3.4668,
      "num_input_tokens_seen": 35296592,
      "step": 16900
    },
    {
      "epoch": 2.7578105881393262,
      "grad_norm": 1.5,
      "learning_rate": 3.101925388351865e-05,
      "loss": 2.1166,
      "num_input_tokens_seen": 35307152,
      "step": 16905
    },
    {
      "epoch": 2.758626315360144,
      "grad_norm": 8.0,
      "learning_rate": 3.1009724751552515e-05,
      "loss": 2.6486,
      "num_input_tokens_seen": 35317264,
      "step": 16910
    },
    {
      "epoch": 2.759442042580961,
      "grad_norm": 6.40625,
      "learning_rate": 3.100019469281131e-05,
      "loss": 1.6806,
      "num_input_tokens_seen": 35328528,
      "step": 16915
    },
    {
      "epoch": 2.760257769801778,
      "grad_norm": 8.8125,
      "learning_rate": 3.0990663708764685e-05,
      "loss": 1.6484,
      "num_input_tokens_seen": 35339936,
      "step": 16920
    },
    {
      "epoch": 2.7610734970225956,
      "grad_norm": 7.03125,
      "learning_rate": 3.098113180088243e-05,
      "loss": 2.4061,
      "num_input_tokens_seen": 35350704,
      "step": 16925
    },
    {
      "epoch": 2.761889224243413,
      "grad_norm": 4.875,
      "learning_rate": 3.097159897063448e-05,
      "loss": 1.5002,
      "num_input_tokens_seen": 35361584,
      "step": 16930
    },
    {
      "epoch": 2.7627049514642303,
      "grad_norm": 5.5,
      "learning_rate": 3.096206521949094e-05,
      "loss": 2.7435,
      "num_input_tokens_seen": 35371440,
      "step": 16935
    },
    {
      "epoch": 2.7635206786850475,
      "grad_norm": 7.71875,
      "learning_rate": 3.0952530548922006e-05,
      "loss": 1.5029,
      "num_input_tokens_seen": 35382048,
      "step": 16940
    },
    {
      "epoch": 2.764336405905865,
      "grad_norm": 11.875,
      "learning_rate": 3.0942994960398064e-05,
      "loss": 2.207,
      "num_input_tokens_seen": 35392880,
      "step": 16945
    },
    {
      "epoch": 2.7651521331266826,
      "grad_norm": 5.75,
      "learning_rate": 3.093345845538961e-05,
      "loss": 3.9481,
      "num_input_tokens_seen": 35402944,
      "step": 16950
    },
    {
      "epoch": 2.7659678603474998,
      "grad_norm": 5.40625,
      "learning_rate": 3.09239210353673e-05,
      "loss": 1.6424,
      "num_input_tokens_seen": 35413168,
      "step": 16955
    },
    {
      "epoch": 2.766783587568317,
      "grad_norm": 3.375,
      "learning_rate": 3.0914382701801926e-05,
      "loss": 1.8041,
      "num_input_tokens_seen": 35423888,
      "step": 16960
    },
    {
      "epoch": 2.7675993147891345,
      "grad_norm": 9.3125,
      "learning_rate": 3.090484345616441e-05,
      "loss": 3.4789,
      "num_input_tokens_seen": 35434368,
      "step": 16965
    },
    {
      "epoch": 2.768415042009952,
      "grad_norm": 11.6875,
      "learning_rate": 3.0895303299925825e-05,
      "loss": 3.8258,
      "num_input_tokens_seen": 35444096,
      "step": 16970
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 3.359375,
      "learning_rate": 3.0885762234557393e-05,
      "loss": 2.3049,
      "num_input_tokens_seen": 35455280,
      "step": 16975
    },
    {
      "epoch": 2.7700464964515867,
      "grad_norm": 6.96875,
      "learning_rate": 3.087622026153045e-05,
      "loss": 1.9618,
      "num_input_tokens_seen": 35466160,
      "step": 16980
    },
    {
      "epoch": 2.770862223672404,
      "grad_norm": 9.5625,
      "learning_rate": 3.086667738231651e-05,
      "loss": 3.5833,
      "num_input_tokens_seen": 35475824,
      "step": 16985
    },
    {
      "epoch": 2.7716779508932214,
      "grad_norm": 19.25,
      "learning_rate": 3.085713359838718e-05,
      "loss": 3.2067,
      "num_input_tokens_seen": 35486272,
      "step": 16990
    },
    {
      "epoch": 2.7724936781140386,
      "grad_norm": 9.9375,
      "learning_rate": 3.084758891121425e-05,
      "loss": 2.129,
      "num_input_tokens_seen": 35497424,
      "step": 16995
    },
    {
      "epoch": 2.773309405334856,
      "grad_norm": 3.953125,
      "learning_rate": 3.083804332226963e-05,
      "loss": 1.152,
      "num_input_tokens_seen": 35508944,
      "step": 17000
    },
    {
      "epoch": 2.773309405334856,
      "eval_loss": 2.5512051582336426,
      "eval_runtime": 135.0119,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 35508944,
      "step": 17000
    },
    {
      "epoch": 2.7741251325556733,
      "grad_norm": 5.15625,
      "learning_rate": 3.082849683302536e-05,
      "loss": 3.032,
      "num_input_tokens_seen": 35519872,
      "step": 17005
    },
    {
      "epoch": 2.774940859776491,
      "grad_norm": 8.125,
      "learning_rate": 3.081894944495363e-05,
      "loss": 1.8765,
      "num_input_tokens_seen": 35530032,
      "step": 17010
    },
    {
      "epoch": 2.775756586997308,
      "grad_norm": 10.5625,
      "learning_rate": 3.080940115952677e-05,
      "loss": 4.8067,
      "num_input_tokens_seen": 35538944,
      "step": 17015
    },
    {
      "epoch": 2.7765723142181256,
      "grad_norm": 3.40625,
      "learning_rate": 3.0799851978217245e-05,
      "loss": 1.8712,
      "num_input_tokens_seen": 35549136,
      "step": 17020
    },
    {
      "epoch": 2.7773880414389427,
      "grad_norm": 3.96875,
      "learning_rate": 3.0790301902497666e-05,
      "loss": 3.0485,
      "num_input_tokens_seen": 35559376,
      "step": 17025
    },
    {
      "epoch": 2.7782037686597603,
      "grad_norm": 1.921875,
      "learning_rate": 3.078075093384076e-05,
      "loss": 1.2675,
      "num_input_tokens_seen": 35570352,
      "step": 17030
    },
    {
      "epoch": 2.7790194958805774,
      "grad_norm": 5.65625,
      "learning_rate": 3.077119907371942e-05,
      "loss": 2.1657,
      "num_input_tokens_seen": 35582144,
      "step": 17035
    },
    {
      "epoch": 2.779835223101395,
      "grad_norm": 5.3125,
      "learning_rate": 3.076164632360666e-05,
      "loss": 1.9921,
      "num_input_tokens_seen": 35592304,
      "step": 17040
    },
    {
      "epoch": 2.780650950322212,
      "grad_norm": 9.6875,
      "learning_rate": 3.075209268497563e-05,
      "loss": 2.7261,
      "num_input_tokens_seen": 35603264,
      "step": 17045
    },
    {
      "epoch": 2.7814666775430297,
      "grad_norm": 11.1875,
      "learning_rate": 3.074253815929961e-05,
      "loss": 1.976,
      "num_input_tokens_seen": 35613904,
      "step": 17050
    },
    {
      "epoch": 2.782282404763847,
      "grad_norm": 7.65625,
      "learning_rate": 3.0732982748052054e-05,
      "loss": 4.0768,
      "num_input_tokens_seen": 35625424,
      "step": 17055
    },
    {
      "epoch": 2.7830981319846644,
      "grad_norm": 17.5,
      "learning_rate": 3.072342645270651e-05,
      "loss": 2.4043,
      "num_input_tokens_seen": 35636720,
      "step": 17060
    },
    {
      "epoch": 2.783913859205482,
      "grad_norm": 4.0625,
      "learning_rate": 3.071386927473668e-05,
      "loss": 2.4074,
      "num_input_tokens_seen": 35645360,
      "step": 17065
    },
    {
      "epoch": 2.784729586426299,
      "grad_norm": 10.6875,
      "learning_rate": 3.0704311215616404e-05,
      "loss": 3.319,
      "num_input_tokens_seen": 35655600,
      "step": 17070
    },
    {
      "epoch": 2.785545313647116,
      "grad_norm": 0.1318359375,
      "learning_rate": 3.0694752276819656e-05,
      "loss": 1.2539,
      "num_input_tokens_seen": 35664736,
      "step": 17075
    },
    {
      "epoch": 2.786361040867934,
      "grad_norm": 2.421875,
      "learning_rate": 3.068519245982054e-05,
      "loss": 2.0095,
      "num_input_tokens_seen": 35675040,
      "step": 17080
    },
    {
      "epoch": 2.7871767680887514,
      "grad_norm": 10.9375,
      "learning_rate": 3.0675631766093304e-05,
      "loss": 2.9543,
      "num_input_tokens_seen": 35684384,
      "step": 17085
    },
    {
      "epoch": 2.7879924953095685,
      "grad_norm": 5.21875,
      "learning_rate": 3.066607019711232e-05,
      "loss": 1.7485,
      "num_input_tokens_seen": 35695216,
      "step": 17090
    },
    {
      "epoch": 2.7888082225303856,
      "grad_norm": 6.90625,
      "learning_rate": 3.065650775435211e-05,
      "loss": 2.0713,
      "num_input_tokens_seen": 35705920,
      "step": 17095
    },
    {
      "epoch": 2.789623949751203,
      "grad_norm": 8.1875,
      "learning_rate": 3.0646944439287326e-05,
      "loss": 2.7004,
      "num_input_tokens_seen": 35715104,
      "step": 17100
    },
    {
      "epoch": 2.7904396769720208,
      "grad_norm": 9.75,
      "learning_rate": 3.0637380253392736e-05,
      "loss": 2.2188,
      "num_input_tokens_seen": 35724960,
      "step": 17105
    },
    {
      "epoch": 2.791255404192838,
      "grad_norm": 3.5,
      "learning_rate": 3.062781519814327e-05,
      "loss": 2.2592,
      "num_input_tokens_seen": 35734432,
      "step": 17110
    },
    {
      "epoch": 2.792071131413655,
      "grad_norm": 5.4375,
      "learning_rate": 3.0618249275013985e-05,
      "loss": 1.9567,
      "num_input_tokens_seen": 35745392,
      "step": 17115
    },
    {
      "epoch": 2.7928868586344726,
      "grad_norm": 3.109375,
      "learning_rate": 3.060868248548005e-05,
      "loss": 1.3854,
      "num_input_tokens_seen": 35756096,
      "step": 17120
    },
    {
      "epoch": 2.79370258585529,
      "grad_norm": 13.8125,
      "learning_rate": 3.0599114831016796e-05,
      "loss": 2.242,
      "num_input_tokens_seen": 35767056,
      "step": 17125
    },
    {
      "epoch": 2.7945183130761073,
      "grad_norm": 8.75,
      "learning_rate": 3.0589546313099666e-05,
      "loss": 1.7485,
      "num_input_tokens_seen": 35775424,
      "step": 17130
    },
    {
      "epoch": 2.7953340402969244,
      "grad_norm": 2.953125,
      "learning_rate": 3.0579976933204255e-05,
      "loss": 1.3898,
      "num_input_tokens_seen": 35785376,
      "step": 17135
    },
    {
      "epoch": 2.796149767517742,
      "grad_norm": 4.71875,
      "learning_rate": 3.0570406692806284e-05,
      "loss": 2.9007,
      "num_input_tokens_seen": 35796928,
      "step": 17140
    },
    {
      "epoch": 2.7969654947385596,
      "grad_norm": 8.6875,
      "learning_rate": 3.05608355933816e-05,
      "loss": 2.9269,
      "num_input_tokens_seen": 35807216,
      "step": 17145
    },
    {
      "epoch": 2.7977812219593767,
      "grad_norm": 4.96875,
      "learning_rate": 3.055126363640618e-05,
      "loss": 2.1375,
      "num_input_tokens_seen": 35817088,
      "step": 17150
    },
    {
      "epoch": 2.7985969491801943,
      "grad_norm": 9.875,
      "learning_rate": 3.0541690823356146e-05,
      "loss": 2.3497,
      "num_input_tokens_seen": 35828160,
      "step": 17155
    },
    {
      "epoch": 2.7994126764010114,
      "grad_norm": 4.5625,
      "learning_rate": 3.053211715570775e-05,
      "loss": 2.8608,
      "num_input_tokens_seen": 35839072,
      "step": 17160
    },
    {
      "epoch": 2.800228403621829,
      "grad_norm": 2.546875,
      "learning_rate": 3.052254263493736e-05,
      "loss": 1.4995,
      "num_input_tokens_seen": 35849472,
      "step": 17165
    },
    {
      "epoch": 2.801044130842646,
      "grad_norm": 9.9375,
      "learning_rate": 3.0512967262521498e-05,
      "loss": 3.3004,
      "num_input_tokens_seen": 35858000,
      "step": 17170
    },
    {
      "epoch": 2.8018598580634637,
      "grad_norm": 10.375,
      "learning_rate": 3.0503391039936803e-05,
      "loss": 1.9098,
      "num_input_tokens_seen": 35867488,
      "step": 17175
    },
    {
      "epoch": 2.802675585284281,
      "grad_norm": 10.875,
      "learning_rate": 3.0493813968660056e-05,
      "loss": 2.473,
      "num_input_tokens_seen": 35878608,
      "step": 17180
    },
    {
      "epoch": 2.8034913125050984,
      "grad_norm": 8.9375,
      "learning_rate": 3.0484236050168153e-05,
      "loss": 2.3895,
      "num_input_tokens_seen": 35889840,
      "step": 17185
    },
    {
      "epoch": 2.8043070397259156,
      "grad_norm": 7.34375,
      "learning_rate": 3.0474657285938123e-05,
      "loss": 1.8708,
      "num_input_tokens_seen": 35901072,
      "step": 17190
    },
    {
      "epoch": 2.805122766946733,
      "grad_norm": 11.5,
      "learning_rate": 3.046507767744715e-05,
      "loss": 3.7846,
      "num_input_tokens_seen": 35911616,
      "step": 17195
    },
    {
      "epoch": 2.8059384941675503,
      "grad_norm": 2.171875,
      "learning_rate": 3.045549722617252e-05,
      "loss": 2.1899,
      "num_input_tokens_seen": 35922144,
      "step": 17200
    },
    {
      "epoch": 2.8059384941675503,
      "eval_loss": 2.556351900100708,
      "eval_runtime": 135.0275,
      "eval_samples_per_second": 20.181,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 35922144,
      "step": 17200
    },
    {
      "epoch": 2.806754221388368,
      "grad_norm": 4.875,
      "learning_rate": 3.0445915933591658e-05,
      "loss": 2.6447,
      "num_input_tokens_seen": 35933024,
      "step": 17205
    },
    {
      "epoch": 2.807569948609185,
      "grad_norm": 6.40625,
      "learning_rate": 3.0436333801182114e-05,
      "loss": 2.0107,
      "num_input_tokens_seen": 35944192,
      "step": 17210
    },
    {
      "epoch": 2.8083856758300025,
      "grad_norm": 5.15625,
      "learning_rate": 3.0426750830421596e-05,
      "loss": 2.6096,
      "num_input_tokens_seen": 35954208,
      "step": 17215
    },
    {
      "epoch": 2.8092014030508197,
      "grad_norm": 3.703125,
      "learning_rate": 3.0417167022787897e-05,
      "loss": 2.6197,
      "num_input_tokens_seen": 35963440,
      "step": 17220
    },
    {
      "epoch": 2.8100171302716372,
      "grad_norm": 9.5,
      "learning_rate": 3.0407582379758966e-05,
      "loss": 2.4956,
      "num_input_tokens_seen": 35974224,
      "step": 17225
    },
    {
      "epoch": 2.8108328574924544,
      "grad_norm": 1.1484375,
      "learning_rate": 3.039799690281287e-05,
      "loss": 2.0583,
      "num_input_tokens_seen": 35982752,
      "step": 17230
    },
    {
      "epoch": 2.811648584713272,
      "grad_norm": 5.0,
      "learning_rate": 3.0388410593427823e-05,
      "loss": 2.7041,
      "num_input_tokens_seen": 35991984,
      "step": 17235
    },
    {
      "epoch": 2.812464311934089,
      "grad_norm": 10.125,
      "learning_rate": 3.0378823453082146e-05,
      "loss": 2.176,
      "num_input_tokens_seen": 36002720,
      "step": 17240
    },
    {
      "epoch": 2.8132800391549067,
      "grad_norm": 7.625,
      "learning_rate": 3.03692354832543e-05,
      "loss": 3.2001,
      "num_input_tokens_seen": 36014592,
      "step": 17245
    },
    {
      "epoch": 2.814095766375724,
      "grad_norm": 3.265625,
      "learning_rate": 3.0359646685422865e-05,
      "loss": 3.0239,
      "num_input_tokens_seen": 36026720,
      "step": 17250
    },
    {
      "epoch": 2.8149114935965414,
      "grad_norm": 7.5625,
      "learning_rate": 3.035005706106656e-05,
      "loss": 2.1807,
      "num_input_tokens_seen": 36037200,
      "step": 17255
    },
    {
      "epoch": 2.815727220817359,
      "grad_norm": 4.15625,
      "learning_rate": 3.034046661166422e-05,
      "loss": 2.3683,
      "num_input_tokens_seen": 36048144,
      "step": 17260
    },
    {
      "epoch": 2.816542948038176,
      "grad_norm": 7.25,
      "learning_rate": 3.033087533869482e-05,
      "loss": 2.7169,
      "num_input_tokens_seen": 36059312,
      "step": 17265
    },
    {
      "epoch": 2.817358675258993,
      "grad_norm": 7.375,
      "learning_rate": 3.0321283243637444e-05,
      "loss": 3.0105,
      "num_input_tokens_seen": 36070640,
      "step": 17270
    },
    {
      "epoch": 2.8181744024798108,
      "grad_norm": 11.875,
      "learning_rate": 3.0311690327971326e-05,
      "loss": 3.4156,
      "num_input_tokens_seen": 36081104,
      "step": 17275
    },
    {
      "epoch": 2.8189901297006283,
      "grad_norm": 8.125,
      "learning_rate": 3.030209659317581e-05,
      "loss": 2.627,
      "num_input_tokens_seen": 36092160,
      "step": 17280
    },
    {
      "epoch": 2.8198058569214455,
      "grad_norm": 0.85546875,
      "learning_rate": 3.0292502040730362e-05,
      "loss": 3.1023,
      "num_input_tokens_seen": 36103200,
      "step": 17285
    },
    {
      "epoch": 2.8206215841422626,
      "grad_norm": 7.1875,
      "learning_rate": 3.0282906672114597e-05,
      "loss": 1.2399,
      "num_input_tokens_seen": 36114512,
      "step": 17290
    },
    {
      "epoch": 2.82143731136308,
      "grad_norm": 1.609375,
      "learning_rate": 3.027331048880823e-05,
      "loss": 1.3071,
      "num_input_tokens_seen": 36125808,
      "step": 17295
    },
    {
      "epoch": 2.8222530385838978,
      "grad_norm": 7.59375,
      "learning_rate": 3.0263713492291123e-05,
      "loss": 2.8007,
      "num_input_tokens_seen": 36136112,
      "step": 17300
    },
    {
      "epoch": 2.823068765804715,
      "grad_norm": 7.5,
      "learning_rate": 3.0254115684043242e-05,
      "loss": 2.1973,
      "num_input_tokens_seen": 36147872,
      "step": 17305
    },
    {
      "epoch": 2.823884493025532,
      "grad_norm": 4.0625,
      "learning_rate": 3.024451706554469e-05,
      "loss": 3.9647,
      "num_input_tokens_seen": 36157840,
      "step": 17310
    },
    {
      "epoch": 2.8247002202463496,
      "grad_norm": 7.0625,
      "learning_rate": 3.0234917638275705e-05,
      "loss": 3.0103,
      "num_input_tokens_seen": 36167696,
      "step": 17315
    },
    {
      "epoch": 2.825515947467167,
      "grad_norm": 6.75,
      "learning_rate": 3.0225317403716635e-05,
      "loss": 2.9,
      "num_input_tokens_seen": 36179024,
      "step": 17320
    },
    {
      "epoch": 2.8263316746879843,
      "grad_norm": 10.0,
      "learning_rate": 3.0215716363347956e-05,
      "loss": 2.3599,
      "num_input_tokens_seen": 36189584,
      "step": 17325
    },
    {
      "epoch": 2.8271474019088014,
      "grad_norm": 1.6953125,
      "learning_rate": 3.0206114518650275e-05,
      "loss": 1.9153,
      "num_input_tokens_seen": 36201184,
      "step": 17330
    },
    {
      "epoch": 2.827963129129619,
      "grad_norm": 5.5625,
      "learning_rate": 3.0196511871104304e-05,
      "loss": 3.1447,
      "num_input_tokens_seen": 36212912,
      "step": 17335
    },
    {
      "epoch": 2.8287788563504366,
      "grad_norm": 9.375,
      "learning_rate": 3.01869084221909e-05,
      "loss": 2.2019,
      "num_input_tokens_seen": 36222608,
      "step": 17340
    },
    {
      "epoch": 2.8295945835712537,
      "grad_norm": 10.3125,
      "learning_rate": 3.0177304173391037e-05,
      "loss": 3.1339,
      "num_input_tokens_seen": 36232848,
      "step": 17345
    },
    {
      "epoch": 2.8304103107920713,
      "grad_norm": 10.3125,
      "learning_rate": 3.01676991261858e-05,
      "loss": 2.4359,
      "num_input_tokens_seen": 36243776,
      "step": 17350
    },
    {
      "epoch": 2.8312260380128884,
      "grad_norm": 4.59375,
      "learning_rate": 3.015809328205642e-05,
      "loss": 3.0128,
      "num_input_tokens_seen": 36252960,
      "step": 17355
    },
    {
      "epoch": 2.832041765233706,
      "grad_norm": 11.9375,
      "learning_rate": 3.0148486642484248e-05,
      "loss": 3.9857,
      "num_input_tokens_seen": 36262832,
      "step": 17360
    },
    {
      "epoch": 2.832857492454523,
      "grad_norm": 9.4375,
      "learning_rate": 3.0138879208950722e-05,
      "loss": 2.0324,
      "num_input_tokens_seen": 36272544,
      "step": 17365
    },
    {
      "epoch": 2.8336732196753407,
      "grad_norm": 6.59375,
      "learning_rate": 3.012927098293744e-05,
      "loss": 2.5986,
      "num_input_tokens_seen": 36283392,
      "step": 17370
    },
    {
      "epoch": 2.834488946896158,
      "grad_norm": 8.125,
      "learning_rate": 3.0119661965926123e-05,
      "loss": 2.4543,
      "num_input_tokens_seen": 36292272,
      "step": 17375
    },
    {
      "epoch": 2.8353046741169754,
      "grad_norm": 10.8125,
      "learning_rate": 3.0110052159398587e-05,
      "loss": 3.0589,
      "num_input_tokens_seen": 36302400,
      "step": 17380
    },
    {
      "epoch": 2.8361204013377925,
      "grad_norm": 8.6875,
      "learning_rate": 3.0100441564836802e-05,
      "loss": 2.7999,
      "num_input_tokens_seen": 36313072,
      "step": 17385
    },
    {
      "epoch": 2.83693612855861,
      "grad_norm": 7.6875,
      "learning_rate": 3.0090830183722817e-05,
      "loss": 2.2582,
      "num_input_tokens_seen": 36324048,
      "step": 17390
    },
    {
      "epoch": 2.8377518557794272,
      "grad_norm": 8.8125,
      "learning_rate": 3.0081218017538852e-05,
      "loss": 2.415,
      "num_input_tokens_seen": 36335360,
      "step": 17395
    },
    {
      "epoch": 2.838567583000245,
      "grad_norm": 2.890625,
      "learning_rate": 3.0071605067767212e-05,
      "loss": 3.1566,
      "num_input_tokens_seen": 36345856,
      "step": 17400
    },
    {
      "epoch": 2.838567583000245,
      "eval_loss": 2.540825843811035,
      "eval_runtime": 135.0255,
      "eval_samples_per_second": 20.181,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 36345856,
      "step": 17400
    },
    {
      "epoch": 2.839383310221062,
      "grad_norm": 4.5,
      "learning_rate": 3.006199133589034e-05,
      "loss": 2.7889,
      "num_input_tokens_seen": 36357392,
      "step": 17405
    },
    {
      "epoch": 2.8401990374418795,
      "grad_norm": 2.203125,
      "learning_rate": 3.005237682339079e-05,
      "loss": 2.587,
      "num_input_tokens_seen": 36367664,
      "step": 17410
    },
    {
      "epoch": 2.8410147646626966,
      "grad_norm": 4.46875,
      "learning_rate": 3.0042761531751228e-05,
      "loss": 2.8813,
      "num_input_tokens_seen": 36377856,
      "step": 17415
    },
    {
      "epoch": 2.841830491883514,
      "grad_norm": 5.71875,
      "learning_rate": 3.0033145462454482e-05,
      "loss": 1.8902,
      "num_input_tokens_seen": 36387200,
      "step": 17420
    },
    {
      "epoch": 2.8426462191043314,
      "grad_norm": 10.8125,
      "learning_rate": 3.002352861698345e-05,
      "loss": 1.8144,
      "num_input_tokens_seen": 36397888,
      "step": 17425
    },
    {
      "epoch": 2.843461946325149,
      "grad_norm": 6.84375,
      "learning_rate": 3.0013910996821178e-05,
      "loss": 2.1199,
      "num_input_tokens_seen": 36408752,
      "step": 17430
    },
    {
      "epoch": 2.8442776735459665,
      "grad_norm": 10.3125,
      "learning_rate": 3.0004292603450817e-05,
      "loss": 1.1044,
      "num_input_tokens_seen": 36420272,
      "step": 17435
    },
    {
      "epoch": 2.8450934007667836,
      "grad_norm": 5.625,
      "learning_rate": 2.9994673438355653e-05,
      "loss": 1.8123,
      "num_input_tokens_seen": 36431040,
      "step": 17440
    },
    {
      "epoch": 2.8459091279876008,
      "grad_norm": 3.75,
      "learning_rate": 2.9985053503019078e-05,
      "loss": 2.1149,
      "num_input_tokens_seen": 36441520,
      "step": 17445
    },
    {
      "epoch": 2.8467248552084183,
      "grad_norm": 5.625,
      "learning_rate": 2.99754327989246e-05,
      "loss": 2.4335,
      "num_input_tokens_seen": 36452272,
      "step": 17450
    },
    {
      "epoch": 2.847540582429236,
      "grad_norm": 8.375,
      "learning_rate": 2.9965811327555864e-05,
      "loss": 1.3532,
      "num_input_tokens_seen": 36462832,
      "step": 17455
    },
    {
      "epoch": 2.848356309650053,
      "grad_norm": 0.16796875,
      "learning_rate": 2.995618909039662e-05,
      "loss": 2.0719,
      "num_input_tokens_seen": 36473952,
      "step": 17460
    },
    {
      "epoch": 2.84917203687087,
      "grad_norm": 7.15625,
      "learning_rate": 2.9946566088930727e-05,
      "loss": 2.202,
      "num_input_tokens_seen": 36484384,
      "step": 17465
    },
    {
      "epoch": 2.8499877640916877,
      "grad_norm": 6.65625,
      "learning_rate": 2.9936942324642192e-05,
      "loss": 1.8019,
      "num_input_tokens_seen": 36493968,
      "step": 17470
    },
    {
      "epoch": 2.8508034913125053,
      "grad_norm": 8.0625,
      "learning_rate": 2.9927317799015097e-05,
      "loss": 2.2342,
      "num_input_tokens_seen": 36504144,
      "step": 17475
    },
    {
      "epoch": 2.8516192185333225,
      "grad_norm": 5.34375,
      "learning_rate": 2.9917692513533685e-05,
      "loss": 2.025,
      "num_input_tokens_seen": 36514512,
      "step": 17480
    },
    {
      "epoch": 2.8524349457541396,
      "grad_norm": 7.84375,
      "learning_rate": 2.990806646968229e-05,
      "loss": 1.4646,
      "num_input_tokens_seen": 36525648,
      "step": 17485
    },
    {
      "epoch": 2.853250672974957,
      "grad_norm": 7.90625,
      "learning_rate": 2.989843966894536e-05,
      "loss": 4.3998,
      "num_input_tokens_seen": 36536144,
      "step": 17490
    },
    {
      "epoch": 2.8540664001957747,
      "grad_norm": 6.25,
      "learning_rate": 2.9888812112807472e-05,
      "loss": 3.5881,
      "num_input_tokens_seen": 36546000,
      "step": 17495
    },
    {
      "epoch": 2.854882127416592,
      "grad_norm": 7.8125,
      "learning_rate": 2.987918380275333e-05,
      "loss": 2.1008,
      "num_input_tokens_seen": 36556688,
      "step": 17500
    },
    {
      "epoch": 2.855697854637409,
      "grad_norm": 9.375,
      "learning_rate": 2.9869554740267724e-05,
      "loss": 3.3469,
      "num_input_tokens_seen": 36567232,
      "step": 17505
    },
    {
      "epoch": 2.8565135818582266,
      "grad_norm": 6.34375,
      "learning_rate": 2.9859924926835585e-05,
      "loss": 2.5967,
      "num_input_tokens_seen": 36576832,
      "step": 17510
    },
    {
      "epoch": 2.857329309079044,
      "grad_norm": 8.9375,
      "learning_rate": 2.9850294363941944e-05,
      "loss": 3.5461,
      "num_input_tokens_seen": 36586576,
      "step": 17515
    },
    {
      "epoch": 2.8581450362998613,
      "grad_norm": 4.125,
      "learning_rate": 2.9840663053071967e-05,
      "loss": 1.3531,
      "num_input_tokens_seen": 36598400,
      "step": 17520
    },
    {
      "epoch": 2.858960763520679,
      "grad_norm": 7.65625,
      "learning_rate": 2.983103099571091e-05,
      "loss": 2.4773,
      "num_input_tokens_seen": 36609696,
      "step": 17525
    },
    {
      "epoch": 2.859776490741496,
      "grad_norm": 11.375,
      "learning_rate": 2.9821398193344164e-05,
      "loss": 3.664,
      "num_input_tokens_seen": 36620944,
      "step": 17530
    },
    {
      "epoch": 2.8605922179623136,
      "grad_norm": 3.3125,
      "learning_rate": 2.9811764647457226e-05,
      "loss": 3.1181,
      "num_input_tokens_seen": 36630976,
      "step": 17535
    },
    {
      "epoch": 2.8614079451831307,
      "grad_norm": 9.5625,
      "learning_rate": 2.9802130359535714e-05,
      "loss": 2.042,
      "num_input_tokens_seen": 36642704,
      "step": 17540
    },
    {
      "epoch": 2.8622236724039483,
      "grad_norm": 9.0,
      "learning_rate": 2.979249533106535e-05,
      "loss": 3.0682,
      "num_input_tokens_seen": 36652464,
      "step": 17545
    },
    {
      "epoch": 2.8630393996247654,
      "grad_norm": 7.28125,
      "learning_rate": 2.9782859563531986e-05,
      "loss": 3.3983,
      "num_input_tokens_seen": 36663120,
      "step": 17550
    },
    {
      "epoch": 2.863855126845583,
      "grad_norm": 7.4375,
      "learning_rate": 2.977322305842156e-05,
      "loss": 2.2612,
      "num_input_tokens_seen": 36673872,
      "step": 17555
    },
    {
      "epoch": 2.8646708540664,
      "grad_norm": 9.0625,
      "learning_rate": 2.9763585817220162e-05,
      "loss": 1.81,
      "num_input_tokens_seen": 36683760,
      "step": 17560
    },
    {
      "epoch": 2.8654865812872177,
      "grad_norm": 3.828125,
      "learning_rate": 2.975394784141397e-05,
      "loss": 2.5872,
      "num_input_tokens_seen": 36694112,
      "step": 17565
    },
    {
      "epoch": 2.866302308508035,
      "grad_norm": 5.25,
      "learning_rate": 2.974430913248928e-05,
      "loss": 3.3368,
      "num_input_tokens_seen": 36705472,
      "step": 17570
    },
    {
      "epoch": 2.8671180357288524,
      "grad_norm": 10.1875,
      "learning_rate": 2.9734669691932497e-05,
      "loss": 2.6282,
      "num_input_tokens_seen": 36716288,
      "step": 17575
    },
    {
      "epoch": 2.8679337629496695,
      "grad_norm": 6.125,
      "learning_rate": 2.9725029521230147e-05,
      "loss": 2.7525,
      "num_input_tokens_seen": 36727216,
      "step": 17580
    },
    {
      "epoch": 2.868749490170487,
      "grad_norm": 9.4375,
      "learning_rate": 2.9715388621868873e-05,
      "loss": 1.658,
      "num_input_tokens_seen": 36737648,
      "step": 17585
    },
    {
      "epoch": 2.869565217391304,
      "grad_norm": 3.9375,
      "learning_rate": 2.970574699533541e-05,
      "loss": 3.0057,
      "num_input_tokens_seen": 36748784,
      "step": 17590
    },
    {
      "epoch": 2.870380944612122,
      "grad_norm": 3.0,
      "learning_rate": 2.969610464311662e-05,
      "loss": 1.2724,
      "num_input_tokens_seen": 36758976,
      "step": 17595
    },
    {
      "epoch": 2.871196671832939,
      "grad_norm": 8.125,
      "learning_rate": 2.9686461566699487e-05,
      "loss": 1.2781,
      "num_input_tokens_seen": 36770688,
      "step": 17600
    },
    {
      "epoch": 2.871196671832939,
      "eval_loss": 2.5483222007751465,
      "eval_runtime": 134.9733,
      "eval_samples_per_second": 20.189,
      "eval_steps_per_second": 10.098,
      "num_input_tokens_seen": 36770688,
      "step": 17600
    },
    {
      "epoch": 2.8720123990537565,
      "grad_norm": 9.3125,
      "learning_rate": 2.9676817767571086e-05,
      "loss": 2.8454,
      "num_input_tokens_seen": 36782448,
      "step": 17605
    },
    {
      "epoch": 2.8728281262745736,
      "grad_norm": 7.875,
      "learning_rate": 2.966717324721861e-05,
      "loss": 1.6154,
      "num_input_tokens_seen": 36793232,
      "step": 17610
    },
    {
      "epoch": 2.873643853495391,
      "grad_norm": 3.234375,
      "learning_rate": 2.9657528007129366e-05,
      "loss": 3.3533,
      "num_input_tokens_seen": 36803664,
      "step": 17615
    },
    {
      "epoch": 2.8744595807162083,
      "grad_norm": 8.625,
      "learning_rate": 2.9647882048790777e-05,
      "loss": 2.1044,
      "num_input_tokens_seen": 36815504,
      "step": 17620
    },
    {
      "epoch": 2.875275307937026,
      "grad_norm": 16.125,
      "learning_rate": 2.963823537369037e-05,
      "loss": 2.6641,
      "num_input_tokens_seen": 36826224,
      "step": 17625
    },
    {
      "epoch": 2.8760910351578435,
      "grad_norm": 7.375,
      "learning_rate": 2.9628587983315775e-05,
      "loss": 2.3371,
      "num_input_tokens_seen": 36835872,
      "step": 17630
    },
    {
      "epoch": 2.8769067623786606,
      "grad_norm": 5.125,
      "learning_rate": 2.9618939879154746e-05,
      "loss": 2.5781,
      "num_input_tokens_seen": 36846880,
      "step": 17635
    },
    {
      "epoch": 2.8777224895994777,
      "grad_norm": 9.6875,
      "learning_rate": 2.9609291062695143e-05,
      "loss": 2.0684,
      "num_input_tokens_seen": 36856528,
      "step": 17640
    },
    {
      "epoch": 2.8785382168202953,
      "grad_norm": 10.6875,
      "learning_rate": 2.9599641535424938e-05,
      "loss": 1.6072,
      "num_input_tokens_seen": 36866592,
      "step": 17645
    },
    {
      "epoch": 2.879353944041113,
      "grad_norm": 5.59375,
      "learning_rate": 2.9589991298832202e-05,
      "loss": 2.0815,
      "num_input_tokens_seen": 36875232,
      "step": 17650
    },
    {
      "epoch": 2.88016967126193,
      "grad_norm": 8.75,
      "learning_rate": 2.958034035440513e-05,
      "loss": 1.5411,
      "num_input_tokens_seen": 36885696,
      "step": 17655
    },
    {
      "epoch": 2.880985398482747,
      "grad_norm": 7.34375,
      "learning_rate": 2.957068870363201e-05,
      "loss": 1.6153,
      "num_input_tokens_seen": 36895456,
      "step": 17660
    },
    {
      "epoch": 2.8818011257035647,
      "grad_norm": 4.75,
      "learning_rate": 2.956103634800126e-05,
      "loss": 3.3167,
      "num_input_tokens_seen": 36905728,
      "step": 17665
    },
    {
      "epoch": 2.8826168529243823,
      "grad_norm": 15.75,
      "learning_rate": 2.9551383289001384e-05,
      "loss": 3.4195,
      "num_input_tokens_seen": 36916784,
      "step": 17670
    },
    {
      "epoch": 2.8834325801451994,
      "grad_norm": 8.875,
      "learning_rate": 2.9541729528121005e-05,
      "loss": 3.0218,
      "num_input_tokens_seen": 36927488,
      "step": 17675
    },
    {
      "epoch": 2.8842483073660166,
      "grad_norm": 6.15625,
      "learning_rate": 2.9532075066848856e-05,
      "loss": 1.8761,
      "num_input_tokens_seen": 36938704,
      "step": 17680
    },
    {
      "epoch": 2.885064034586834,
      "grad_norm": 13.5625,
      "learning_rate": 2.9522419906673786e-05,
      "loss": 2.72,
      "num_input_tokens_seen": 36947904,
      "step": 17685
    },
    {
      "epoch": 2.8858797618076517,
      "grad_norm": 6.84375,
      "learning_rate": 2.951276404908474e-05,
      "loss": 2.8413,
      "num_input_tokens_seen": 36958672,
      "step": 17690
    },
    {
      "epoch": 2.886695489028469,
      "grad_norm": 6.625,
      "learning_rate": 2.9503107495570752e-05,
      "loss": 3.1267,
      "num_input_tokens_seen": 36970336,
      "step": 17695
    },
    {
      "epoch": 2.887511216249286,
      "grad_norm": 5.40625,
      "learning_rate": 2.9493450247621003e-05,
      "loss": 0.9952,
      "num_input_tokens_seen": 36979680,
      "step": 17700
    },
    {
      "epoch": 2.8883269434701035,
      "grad_norm": 4.65625,
      "learning_rate": 2.948379230672476e-05,
      "loss": 1.0621,
      "num_input_tokens_seen": 36990912,
      "step": 17705
    },
    {
      "epoch": 2.889142670690921,
      "grad_norm": 6.5,
      "learning_rate": 2.9474133674371396e-05,
      "loss": 3.3345,
      "num_input_tokens_seen": 37001408,
      "step": 17710
    },
    {
      "epoch": 2.8899583979117383,
      "grad_norm": 14.1875,
      "learning_rate": 2.9464474352050387e-05,
      "loss": 2.4321,
      "num_input_tokens_seen": 37011600,
      "step": 17715
    },
    {
      "epoch": 2.890774125132556,
      "grad_norm": 7.28125,
      "learning_rate": 2.9454814341251336e-05,
      "loss": 1.6818,
      "num_input_tokens_seen": 37022160,
      "step": 17720
    },
    {
      "epoch": 2.891589852353373,
      "grad_norm": 6.40625,
      "learning_rate": 2.9445153643463942e-05,
      "loss": 1.7129,
      "num_input_tokens_seen": 37033504,
      "step": 17725
    },
    {
      "epoch": 2.8924055795741905,
      "grad_norm": 6.78125,
      "learning_rate": 2.943549226017798e-05,
      "loss": 1.9233,
      "num_input_tokens_seen": 37045696,
      "step": 17730
    },
    {
      "epoch": 2.8932213067950077,
      "grad_norm": 4.34375,
      "learning_rate": 2.942583019288337e-05,
      "loss": 2.4831,
      "num_input_tokens_seen": 37055840,
      "step": 17735
    },
    {
      "epoch": 2.8940370340158252,
      "grad_norm": 12.5,
      "learning_rate": 2.9416167443070132e-05,
      "loss": 2.6092,
      "num_input_tokens_seen": 37066144,
      "step": 17740
    },
    {
      "epoch": 2.8948527612366424,
      "grad_norm": 5.40625,
      "learning_rate": 2.9406504012228375e-05,
      "loss": 2.9348,
      "num_input_tokens_seen": 37076144,
      "step": 17745
    },
    {
      "epoch": 2.89566848845746,
      "grad_norm": 12.0625,
      "learning_rate": 2.939683990184832e-05,
      "loss": 1.5366,
      "num_input_tokens_seen": 37087440,
      "step": 17750
    },
    {
      "epoch": 2.896484215678277,
      "grad_norm": 6.21875,
      "learning_rate": 2.93871751134203e-05,
      "loss": 2.1692,
      "num_input_tokens_seen": 37097520,
      "step": 17755
    },
    {
      "epoch": 2.8972999428990946,
      "grad_norm": 6.1875,
      "learning_rate": 2.9377509648434752e-05,
      "loss": 1.9937,
      "num_input_tokens_seen": 37109024,
      "step": 17760
    },
    {
      "epoch": 2.898115670119912,
      "grad_norm": 4.1875,
      "learning_rate": 2.9367843508382203e-05,
      "loss": 3.1007,
      "num_input_tokens_seen": 37118336,
      "step": 17765
    },
    {
      "epoch": 2.8989313973407294,
      "grad_norm": 7.9375,
      "learning_rate": 2.9358176694753293e-05,
      "loss": 1.047,
      "num_input_tokens_seen": 37128624,
      "step": 17770
    },
    {
      "epoch": 2.8997471245615465,
      "grad_norm": 5.53125,
      "learning_rate": 2.9348509209038766e-05,
      "loss": 1.3671,
      "num_input_tokens_seen": 37139824,
      "step": 17775
    },
    {
      "epoch": 2.900562851782364,
      "grad_norm": 7.8125,
      "learning_rate": 2.933884105272947e-05,
      "loss": 2.157,
      "num_input_tokens_seen": 37150288,
      "step": 17780
    },
    {
      "epoch": 2.901378579003181,
      "grad_norm": 5.375,
      "learning_rate": 2.9329172227316366e-05,
      "loss": 0.8971,
      "num_input_tokens_seen": 37162576,
      "step": 17785
    },
    {
      "epoch": 2.9021943062239988,
      "grad_norm": 5.625,
      "learning_rate": 2.93195027342905e-05,
      "loss": 1.5663,
      "num_input_tokens_seen": 37173104,
      "step": 17790
    },
    {
      "epoch": 2.903010033444816,
      "grad_norm": 5.84375,
      "learning_rate": 2.9309832575143024e-05,
      "loss": 2.5671,
      "num_input_tokens_seen": 37184544,
      "step": 17795
    },
    {
      "epoch": 2.9038257606656335,
      "grad_norm": 5.40625,
      "learning_rate": 2.930016175136521e-05,
      "loss": 2.3983,
      "num_input_tokens_seen": 37194864,
      "step": 17800
    },
    {
      "epoch": 2.9038257606656335,
      "eval_loss": 2.533569574356079,
      "eval_runtime": 134.8985,
      "eval_samples_per_second": 20.2,
      "eval_steps_per_second": 10.104,
      "num_input_tokens_seen": 37194864,
      "step": 17800
    },
    {
      "epoch": 2.904641487886451,
      "grad_norm": 4.0,
      "learning_rate": 2.9290490264448412e-05,
      "loss": 2.2953,
      "num_input_tokens_seen": 37204528,
      "step": 17805
    },
    {
      "epoch": 2.905457215107268,
      "grad_norm": 8.8125,
      "learning_rate": 2.9280818115884094e-05,
      "loss": 2.2755,
      "num_input_tokens_seen": 37214704,
      "step": 17810
    },
    {
      "epoch": 2.9062729423280853,
      "grad_norm": 10.375,
      "learning_rate": 2.9271145307163828e-05,
      "loss": 2.8866,
      "num_input_tokens_seen": 37225456,
      "step": 17815
    },
    {
      "epoch": 2.907088669548903,
      "grad_norm": 10.875,
      "learning_rate": 2.9261471839779287e-05,
      "loss": 3.1576,
      "num_input_tokens_seen": 37235872,
      "step": 17820
    },
    {
      "epoch": 2.9079043967697205,
      "grad_norm": 4.59375,
      "learning_rate": 2.925179771522223e-05,
      "loss": 1.6748,
      "num_input_tokens_seen": 37246032,
      "step": 17825
    },
    {
      "epoch": 2.9087201239905376,
      "grad_norm": 6.65625,
      "learning_rate": 2.9242122934984535e-05,
      "loss": 2.7183,
      "num_input_tokens_seen": 37256912,
      "step": 17830
    },
    {
      "epoch": 2.9095358512113547,
      "grad_norm": 7.21875,
      "learning_rate": 2.9232447500558176e-05,
      "loss": 2.2345,
      "num_input_tokens_seen": 37268896,
      "step": 17835
    },
    {
      "epoch": 2.9103515784321723,
      "grad_norm": 3.953125,
      "learning_rate": 2.9222771413435225e-05,
      "loss": 2.6866,
      "num_input_tokens_seen": 37278672,
      "step": 17840
    },
    {
      "epoch": 2.91116730565299,
      "grad_norm": 0.93359375,
      "learning_rate": 2.9213094675107848e-05,
      "loss": 2.5689,
      "num_input_tokens_seen": 37288208,
      "step": 17845
    },
    {
      "epoch": 2.911983032873807,
      "grad_norm": 8.1875,
      "learning_rate": 2.9203417287068335e-05,
      "loss": 2.7114,
      "num_input_tokens_seen": 37298112,
      "step": 17850
    },
    {
      "epoch": 2.912798760094624,
      "grad_norm": 11.375,
      "learning_rate": 2.9193739250809042e-05,
      "loss": 2.8746,
      "num_input_tokens_seen": 37308032,
      "step": 17855
    },
    {
      "epoch": 2.9136144873154417,
      "grad_norm": 13.1875,
      "learning_rate": 2.9184060567822463e-05,
      "loss": 4.1829,
      "num_input_tokens_seen": 37320720,
      "step": 17860
    },
    {
      "epoch": 2.9144302145362593,
      "grad_norm": 4.875,
      "learning_rate": 2.9174381239601166e-05,
      "loss": 2.7385,
      "num_input_tokens_seen": 37329840,
      "step": 17865
    },
    {
      "epoch": 2.9152459417570764,
      "grad_norm": 6.34375,
      "learning_rate": 2.916470126763783e-05,
      "loss": 2.5973,
      "num_input_tokens_seen": 37339488,
      "step": 17870
    },
    {
      "epoch": 2.9160616689778935,
      "grad_norm": 8.8125,
      "learning_rate": 2.9155020653425203e-05,
      "loss": 3.386,
      "num_input_tokens_seen": 37348608,
      "step": 17875
    },
    {
      "epoch": 2.916877396198711,
      "grad_norm": 6.09375,
      "learning_rate": 2.9145339398456184e-05,
      "loss": 2.7162,
      "num_input_tokens_seen": 37358672,
      "step": 17880
    },
    {
      "epoch": 2.9176931234195287,
      "grad_norm": 12.375,
      "learning_rate": 2.913565750422374e-05,
      "loss": 4.2184,
      "num_input_tokens_seen": 37368752,
      "step": 17885
    },
    {
      "epoch": 2.918508850640346,
      "grad_norm": 9.375,
      "learning_rate": 2.9125974972220938e-05,
      "loss": 1.9727,
      "num_input_tokens_seen": 37379168,
      "step": 17890
    },
    {
      "epoch": 2.919324577861163,
      "grad_norm": 21.0,
      "learning_rate": 2.9116291803940932e-05,
      "loss": 2.4604,
      "num_input_tokens_seen": 37388944,
      "step": 17895
    },
    {
      "epoch": 2.9201403050819805,
      "grad_norm": 8.9375,
      "learning_rate": 2.910660800087701e-05,
      "loss": 2.3198,
      "num_input_tokens_seen": 37399216,
      "step": 17900
    },
    {
      "epoch": 2.920956032302798,
      "grad_norm": 11.875,
      "learning_rate": 2.909692356452254e-05,
      "loss": 4.4498,
      "num_input_tokens_seen": 37410144,
      "step": 17905
    },
    {
      "epoch": 2.9217717595236152,
      "grad_norm": 8.4375,
      "learning_rate": 2.9087238496370962e-05,
      "loss": 2.4111,
      "num_input_tokens_seen": 37421264,
      "step": 17910
    },
    {
      "epoch": 2.922587486744433,
      "grad_norm": 2.125,
      "learning_rate": 2.907755279791583e-05,
      "loss": 1.0407,
      "num_input_tokens_seen": 37431408,
      "step": 17915
    },
    {
      "epoch": 2.92340321396525,
      "grad_norm": 1.9296875,
      "learning_rate": 2.906786647065083e-05,
      "loss": 3.126,
      "num_input_tokens_seen": 37442976,
      "step": 17920
    },
    {
      "epoch": 2.9242189411860675,
      "grad_norm": 3.28125,
      "learning_rate": 2.9058179516069695e-05,
      "loss": 2.0501,
      "num_input_tokens_seen": 37452608,
      "step": 17925
    },
    {
      "epoch": 2.9250346684068846,
      "grad_norm": 4.8125,
      "learning_rate": 2.9048491935666282e-05,
      "loss": 3.5146,
      "num_input_tokens_seen": 37462960,
      "step": 17930
    },
    {
      "epoch": 2.925850395627702,
      "grad_norm": 5.0,
      "learning_rate": 2.9038803730934534e-05,
      "loss": 2.8024,
      "num_input_tokens_seen": 37473776,
      "step": 17935
    },
    {
      "epoch": 2.9266661228485193,
      "grad_norm": 5.6875,
      "learning_rate": 2.9029114903368503e-05,
      "loss": 1.4991,
      "num_input_tokens_seen": 37484896,
      "step": 17940
    },
    {
      "epoch": 2.927481850069337,
      "grad_norm": 6.53125,
      "learning_rate": 2.9019425454462318e-05,
      "loss": 2.23,
      "num_input_tokens_seen": 37496272,
      "step": 17945
    },
    {
      "epoch": 2.928297577290154,
      "grad_norm": 12.0,
      "learning_rate": 2.9009735385710212e-05,
      "loss": 2.9989,
      "num_input_tokens_seen": 37505872,
      "step": 17950
    },
    {
      "epoch": 2.9291133045109716,
      "grad_norm": 6.96875,
      "learning_rate": 2.900004469860652e-05,
      "loss": 3.0661,
      "num_input_tokens_seen": 37516864,
      "step": 17955
    },
    {
      "epoch": 2.9299290317317888,
      "grad_norm": 8.125,
      "learning_rate": 2.8990353394645668e-05,
      "loss": 1.5856,
      "num_input_tokens_seen": 37528080,
      "step": 17960
    },
    {
      "epoch": 2.9307447589526063,
      "grad_norm": 6.09375,
      "learning_rate": 2.8980661475322186e-05,
      "loss": 2.3062,
      "num_input_tokens_seen": 37539600,
      "step": 17965
    },
    {
      "epoch": 2.9315604861734235,
      "grad_norm": 5.90625,
      "learning_rate": 2.897096894213067e-05,
      "loss": 1.5724,
      "num_input_tokens_seen": 37550176,
      "step": 17970
    },
    {
      "epoch": 2.932376213394241,
      "grad_norm": 7.625,
      "learning_rate": 2.8961275796565845e-05,
      "loss": 3.0958,
      "num_input_tokens_seen": 37560304,
      "step": 17975
    },
    {
      "epoch": 2.933191940615058,
      "grad_norm": 6.15625,
      "learning_rate": 2.8951582040122517e-05,
      "loss": 2.1767,
      "num_input_tokens_seen": 37571488,
      "step": 17980
    },
    {
      "epoch": 2.9340076678358757,
      "grad_norm": 6.21875,
      "learning_rate": 2.894188767429557e-05,
      "loss": 1.8617,
      "num_input_tokens_seen": 37581136,
      "step": 17985
    },
    {
      "epoch": 2.934823395056693,
      "grad_norm": 8.0625,
      "learning_rate": 2.8932192700580014e-05,
      "loss": 3.1924,
      "num_input_tokens_seen": 37591968,
      "step": 17990
    },
    {
      "epoch": 2.9356391222775104,
      "grad_norm": 11.625,
      "learning_rate": 2.8922497120470916e-05,
      "loss": 3.033,
      "num_input_tokens_seen": 37603072,
      "step": 17995
    },
    {
      "epoch": 2.936454849498328,
      "grad_norm": 6.53125,
      "learning_rate": 2.891280093546348e-05,
      "loss": 1.9318,
      "num_input_tokens_seen": 37615344,
      "step": 18000
    },
    {
      "epoch": 2.936454849498328,
      "eval_loss": 2.53791880607605,
      "eval_runtime": 134.9454,
      "eval_samples_per_second": 20.193,
      "eval_steps_per_second": 10.1,
      "num_input_tokens_seen": 37615344,
      "step": 18000
    },
    {
      "epoch": 2.937270576719145,
      "grad_norm": 8.0625,
      "learning_rate": 2.890310414705297e-05,
      "loss": 1.4782,
      "num_input_tokens_seen": 37626608,
      "step": 18005
    },
    {
      "epoch": 2.9380863039399623,
      "grad_norm": 6.3125,
      "learning_rate": 2.8893406756734742e-05,
      "loss": 2.6036,
      "num_input_tokens_seen": 37637136,
      "step": 18010
    },
    {
      "epoch": 2.93890203116078,
      "grad_norm": 5.34375,
      "learning_rate": 2.888370876600427e-05,
      "loss": 1.9082,
      "num_input_tokens_seen": 37647664,
      "step": 18015
    },
    {
      "epoch": 2.9397177583815974,
      "grad_norm": 9.25,
      "learning_rate": 2.8874010176357104e-05,
      "loss": 1.8654,
      "num_input_tokens_seen": 37658352,
      "step": 18020
    },
    {
      "epoch": 2.9405334856024146,
      "grad_norm": 6.96875,
      "learning_rate": 2.886431098928888e-05,
      "loss": 3.079,
      "num_input_tokens_seen": 37669504,
      "step": 18025
    },
    {
      "epoch": 2.9413492128232317,
      "grad_norm": 9.375,
      "learning_rate": 2.885461120629534e-05,
      "loss": 2.9154,
      "num_input_tokens_seen": 37679520,
      "step": 18030
    },
    {
      "epoch": 2.9421649400440493,
      "grad_norm": 6.90625,
      "learning_rate": 2.8844910828872317e-05,
      "loss": 2.5521,
      "num_input_tokens_seen": 37688176,
      "step": 18035
    },
    {
      "epoch": 2.942980667264867,
      "grad_norm": 11.125,
      "learning_rate": 2.8835209858515715e-05,
      "loss": 2.7315,
      "num_input_tokens_seen": 37698912,
      "step": 18040
    },
    {
      "epoch": 2.943796394485684,
      "grad_norm": 8.1875,
      "learning_rate": 2.8825508296721566e-05,
      "loss": 2.1624,
      "num_input_tokens_seen": 37710480,
      "step": 18045
    },
    {
      "epoch": 2.944612121706501,
      "grad_norm": 5.5,
      "learning_rate": 2.881580614498596e-05,
      "loss": 2.2693,
      "num_input_tokens_seen": 37721216,
      "step": 18050
    },
    {
      "epoch": 2.9454278489273187,
      "grad_norm": 5.78125,
      "learning_rate": 2.8806103404805103e-05,
      "loss": 2.5076,
      "num_input_tokens_seen": 37732480,
      "step": 18055
    },
    {
      "epoch": 2.9462435761481363,
      "grad_norm": 10.375,
      "learning_rate": 2.8796400077675257e-05,
      "loss": 2.3798,
      "num_input_tokens_seen": 37743744,
      "step": 18060
    },
    {
      "epoch": 2.9470593033689534,
      "grad_norm": 12.4375,
      "learning_rate": 2.8786696165092812e-05,
      "loss": 2.1377,
      "num_input_tokens_seen": 37754848,
      "step": 18065
    },
    {
      "epoch": 2.9478750305897705,
      "grad_norm": 4.125,
      "learning_rate": 2.8776991668554236e-05,
      "loss": 1.4124,
      "num_input_tokens_seen": 37764960,
      "step": 18070
    },
    {
      "epoch": 2.948690757810588,
      "grad_norm": 3.546875,
      "learning_rate": 2.876728658955608e-05,
      "loss": 1.1722,
      "num_input_tokens_seen": 37775856,
      "step": 18075
    },
    {
      "epoch": 2.9495064850314057,
      "grad_norm": 4.625,
      "learning_rate": 2.8757580929594986e-05,
      "loss": 2.5072,
      "num_input_tokens_seen": 37786416,
      "step": 18080
    },
    {
      "epoch": 2.950322212252223,
      "grad_norm": 0.478515625,
      "learning_rate": 2.87478746901677e-05,
      "loss": 1.9796,
      "num_input_tokens_seen": 37795104,
      "step": 18085
    },
    {
      "epoch": 2.9511379394730404,
      "grad_norm": 7.28125,
      "learning_rate": 2.873816787277103e-05,
      "loss": 1.3089,
      "num_input_tokens_seen": 37805536,
      "step": 18090
    },
    {
      "epoch": 2.9519536666938575,
      "grad_norm": 6.5625,
      "learning_rate": 2.8728460478901903e-05,
      "loss": 3.0849,
      "num_input_tokens_seen": 37813488,
      "step": 18095
    },
    {
      "epoch": 2.952769393914675,
      "grad_norm": 8.0625,
      "learning_rate": 2.8718752510057307e-05,
      "loss": 2.0511,
      "num_input_tokens_seen": 37823648,
      "step": 18100
    },
    {
      "epoch": 2.953585121135492,
      "grad_norm": 12.125,
      "learning_rate": 2.870904396773435e-05,
      "loss": 2.1328,
      "num_input_tokens_seen": 37833744,
      "step": 18105
    },
    {
      "epoch": 2.95440084835631,
      "grad_norm": 5.0625,
      "learning_rate": 2.86993348534302e-05,
      "loss": 1.3966,
      "num_input_tokens_seen": 37844752,
      "step": 18110
    },
    {
      "epoch": 2.955216575577127,
      "grad_norm": 1.1328125,
      "learning_rate": 2.868962516864212e-05,
      "loss": 2.2523,
      "num_input_tokens_seen": 37854704,
      "step": 18115
    },
    {
      "epoch": 2.9560323027979445,
      "grad_norm": 12.0,
      "learning_rate": 2.8679914914867477e-05,
      "loss": 2.3021,
      "num_input_tokens_seen": 37864976,
      "step": 18120
    },
    {
      "epoch": 2.9568480300187616,
      "grad_norm": 13.0,
      "learning_rate": 2.8670204093603713e-05,
      "loss": 2.3064,
      "num_input_tokens_seen": 37876352,
      "step": 18125
    },
    {
      "epoch": 2.957663757239579,
      "grad_norm": 8.8125,
      "learning_rate": 2.8660492706348357e-05,
      "loss": 1.9308,
      "num_input_tokens_seen": 37886000,
      "step": 18130
    },
    {
      "epoch": 2.9584794844603963,
      "grad_norm": 9.3125,
      "learning_rate": 2.8650780754599022e-05,
      "loss": 2.4612,
      "num_input_tokens_seen": 37896960,
      "step": 18135
    },
    {
      "epoch": 2.959295211681214,
      "grad_norm": 6.875,
      "learning_rate": 2.8641068239853407e-05,
      "loss": 1.7649,
      "num_input_tokens_seen": 37906224,
      "step": 18140
    },
    {
      "epoch": 2.960110938902031,
      "grad_norm": 5.75,
      "learning_rate": 2.863135516360932e-05,
      "loss": 2.873,
      "num_input_tokens_seen": 37917632,
      "step": 18145
    },
    {
      "epoch": 2.9609266661228486,
      "grad_norm": 10.125,
      "learning_rate": 2.8621641527364633e-05,
      "loss": 2.4413,
      "num_input_tokens_seen": 37929440,
      "step": 18150
    },
    {
      "epoch": 2.9617423933436657,
      "grad_norm": 5.15625,
      "learning_rate": 2.8611927332617313e-05,
      "loss": 2.2025,
      "num_input_tokens_seen": 37939968,
      "step": 18155
    },
    {
      "epoch": 2.9625581205644833,
      "grad_norm": 14.8125,
      "learning_rate": 2.8602212580865405e-05,
      "loss": 2.6721,
      "num_input_tokens_seen": 37949728,
      "step": 18160
    },
    {
      "epoch": 2.9633738477853004,
      "grad_norm": 8.3125,
      "learning_rate": 2.859249727360705e-05,
      "loss": 2.6591,
      "num_input_tokens_seen": 37960832,
      "step": 18165
    },
    {
      "epoch": 2.964189575006118,
      "grad_norm": 10.0,
      "learning_rate": 2.8582781412340465e-05,
      "loss": 2.8765,
      "num_input_tokens_seen": 37972704,
      "step": 18170
    },
    {
      "epoch": 2.965005302226935,
      "grad_norm": 12.4375,
      "learning_rate": 2.857306499856397e-05,
      "loss": 2.853,
      "num_input_tokens_seen": 37983120,
      "step": 18175
    },
    {
      "epoch": 2.9658210294477527,
      "grad_norm": 10.5625,
      "learning_rate": 2.856334803377594e-05,
      "loss": 2.9062,
      "num_input_tokens_seen": 37992672,
      "step": 18180
    },
    {
      "epoch": 2.96663675666857,
      "grad_norm": 11.75,
      "learning_rate": 2.8553630519474867e-05,
      "loss": 2.4334,
      "num_input_tokens_seen": 38002944,
      "step": 18185
    },
    {
      "epoch": 2.9674524838893874,
      "grad_norm": 8.125,
      "learning_rate": 2.8543912457159317e-05,
      "loss": 2.8965,
      "num_input_tokens_seen": 38012096,
      "step": 18190
    },
    {
      "epoch": 2.968268211110205,
      "grad_norm": 7.3125,
      "learning_rate": 2.853419384832792e-05,
      "loss": 3.5969,
      "num_input_tokens_seen": 38020576,
      "step": 18195
    },
    {
      "epoch": 2.969083938331022,
      "grad_norm": 2.328125,
      "learning_rate": 2.8524474694479423e-05,
      "loss": 1.6621,
      "num_input_tokens_seen": 38030400,
      "step": 18200
    },
    {
      "epoch": 2.969083938331022,
      "eval_loss": 2.5444138050079346,
      "eval_runtime": 135.0548,
      "eval_samples_per_second": 20.177,
      "eval_steps_per_second": 10.092,
      "num_input_tokens_seen": 38030400,
      "step": 18200
    },
    {
      "epoch": 2.9698996655518393,
      "grad_norm": 9.9375,
      "learning_rate": 2.851475499711264e-05,
      "loss": 3.4106,
      "num_input_tokens_seen": 38039856,
      "step": 18205
    },
    {
      "epoch": 2.970715392772657,
      "grad_norm": 6.3125,
      "learning_rate": 2.8505034757726468e-05,
      "loss": 1.4542,
      "num_input_tokens_seen": 38049600,
      "step": 18210
    },
    {
      "epoch": 2.9715311199934744,
      "grad_norm": 8.5625,
      "learning_rate": 2.8495313977819886e-05,
      "loss": 3.4219,
      "num_input_tokens_seen": 38060512,
      "step": 18215
    },
    {
      "epoch": 2.9723468472142915,
      "grad_norm": 11.125,
      "learning_rate": 2.8485592658891956e-05,
      "loss": 3.0502,
      "num_input_tokens_seen": 38069968,
      "step": 18220
    },
    {
      "epoch": 2.9731625744351087,
      "grad_norm": 6.0,
      "learning_rate": 2.8475870802441844e-05,
      "loss": 3.5948,
      "num_input_tokens_seen": 38080560,
      "step": 18225
    },
    {
      "epoch": 2.9739783016559262,
      "grad_norm": 9.5,
      "learning_rate": 2.8466148409968774e-05,
      "loss": 2.6381,
      "num_input_tokens_seen": 38089328,
      "step": 18230
    },
    {
      "epoch": 2.974794028876744,
      "grad_norm": 4.53125,
      "learning_rate": 2.8456425482972067e-05,
      "loss": 1.267,
      "num_input_tokens_seen": 38098864,
      "step": 18235
    },
    {
      "epoch": 2.975609756097561,
      "grad_norm": 5.625,
      "learning_rate": 2.84467020229511e-05,
      "loss": 2.028,
      "num_input_tokens_seen": 38108464,
      "step": 18240
    },
    {
      "epoch": 2.976425483318378,
      "grad_norm": 9.25,
      "learning_rate": 2.8436978031405375e-05,
      "loss": 2.8224,
      "num_input_tokens_seen": 38118272,
      "step": 18245
    },
    {
      "epoch": 2.9772412105391957,
      "grad_norm": 8.625,
      "learning_rate": 2.842725350983445e-05,
      "loss": 4.1128,
      "num_input_tokens_seen": 38128688,
      "step": 18250
    },
    {
      "epoch": 2.9780569377600132,
      "grad_norm": 2.984375,
      "learning_rate": 2.8417528459737957e-05,
      "loss": 1.9387,
      "num_input_tokens_seen": 38137520,
      "step": 18255
    },
    {
      "epoch": 2.9788726649808304,
      "grad_norm": 16.625,
      "learning_rate": 2.8407802882615624e-05,
      "loss": 1.9687,
      "num_input_tokens_seen": 38148480,
      "step": 18260
    },
    {
      "epoch": 2.9796883922016475,
      "grad_norm": 11.1875,
      "learning_rate": 2.8398076779967277e-05,
      "loss": 2.9298,
      "num_input_tokens_seen": 38158160,
      "step": 18265
    },
    {
      "epoch": 2.980504119422465,
      "grad_norm": 2.59375,
      "learning_rate": 2.8388350153292774e-05,
      "loss": 1.3153,
      "num_input_tokens_seen": 38168544,
      "step": 18270
    },
    {
      "epoch": 2.9813198466432826,
      "grad_norm": 4.03125,
      "learning_rate": 2.8378623004092103e-05,
      "loss": 1.5028,
      "num_input_tokens_seen": 38178848,
      "step": 18275
    },
    {
      "epoch": 2.9821355738640998,
      "grad_norm": 5.25,
      "learning_rate": 2.8368895333865302e-05,
      "loss": 2.29,
      "num_input_tokens_seen": 38188288,
      "step": 18280
    },
    {
      "epoch": 2.9829513010849174,
      "grad_norm": 8.5,
      "learning_rate": 2.835916714411251e-05,
      "loss": 2.9891,
      "num_input_tokens_seen": 38197904,
      "step": 18285
    },
    {
      "epoch": 2.9837670283057345,
      "grad_norm": 10.5,
      "learning_rate": 2.8349438436333926e-05,
      "loss": 3.2737,
      "num_input_tokens_seen": 38208640,
      "step": 18290
    },
    {
      "epoch": 2.984582755526552,
      "grad_norm": 2.1875,
      "learning_rate": 2.833970921202984e-05,
      "loss": 1.5725,
      "num_input_tokens_seen": 38221136,
      "step": 18295
    },
    {
      "epoch": 2.985398482747369,
      "grad_norm": 7.75,
      "learning_rate": 2.8329979472700628e-05,
      "loss": 2.4457,
      "num_input_tokens_seen": 38230800,
      "step": 18300
    },
    {
      "epoch": 2.9862142099681868,
      "grad_norm": 6.6875,
      "learning_rate": 2.832024921984674e-05,
      "loss": 2.9475,
      "num_input_tokens_seen": 38240128,
      "step": 18305
    },
    {
      "epoch": 2.987029937189004,
      "grad_norm": 1.609375,
      "learning_rate": 2.8310518454968693e-05,
      "loss": 1.8237,
      "num_input_tokens_seen": 38251888,
      "step": 18310
    },
    {
      "epoch": 2.9878456644098215,
      "grad_norm": 7.84375,
      "learning_rate": 2.8300787179567095e-05,
      "loss": 1.9924,
      "num_input_tokens_seen": 38262384,
      "step": 18315
    },
    {
      "epoch": 2.9886613916306386,
      "grad_norm": 4.59375,
      "learning_rate": 2.8291055395142636e-05,
      "loss": 2.4782,
      "num_input_tokens_seen": 38270528,
      "step": 18320
    },
    {
      "epoch": 2.989477118851456,
      "grad_norm": 5.875,
      "learning_rate": 2.8281323103196073e-05,
      "loss": 1.7856,
      "num_input_tokens_seen": 38282320,
      "step": 18325
    },
    {
      "epoch": 2.9902928460722733,
      "grad_norm": 7.21875,
      "learning_rate": 2.8271590305228256e-05,
      "loss": 2.7209,
      "num_input_tokens_seen": 38293984,
      "step": 18330
    },
    {
      "epoch": 2.991108573293091,
      "grad_norm": 1.9765625,
      "learning_rate": 2.82618570027401e-05,
      "loss": 2.3179,
      "num_input_tokens_seen": 38304048,
      "step": 18335
    },
    {
      "epoch": 2.991924300513908,
      "grad_norm": 2.78125,
      "learning_rate": 2.8252123197232604e-05,
      "loss": 2.617,
      "num_input_tokens_seen": 38314816,
      "step": 18340
    },
    {
      "epoch": 2.9927400277347256,
      "grad_norm": 6.71875,
      "learning_rate": 2.8242388890206843e-05,
      "loss": 2.7697,
      "num_input_tokens_seen": 38325424,
      "step": 18345
    },
    {
      "epoch": 2.9935557549555427,
      "grad_norm": 8.75,
      "learning_rate": 2.8232654083163967e-05,
      "loss": 1.7537,
      "num_input_tokens_seen": 38335776,
      "step": 18350
    },
    {
      "epoch": 2.9943714821763603,
      "grad_norm": 6.96875,
      "learning_rate": 2.822291877760521e-05,
      "loss": 2.5797,
      "num_input_tokens_seen": 38346608,
      "step": 18355
    },
    {
      "epoch": 2.9951872093971774,
      "grad_norm": 4.375,
      "learning_rate": 2.8213182975031864e-05,
      "loss": 1.0546,
      "num_input_tokens_seen": 38357200,
      "step": 18360
    },
    {
      "epoch": 2.996002936617995,
      "grad_norm": 8.6875,
      "learning_rate": 2.8203446676945337e-05,
      "loss": 2.0592,
      "num_input_tokens_seen": 38367104,
      "step": 18365
    },
    {
      "epoch": 2.9968186638388126,
      "grad_norm": 1.1953125,
      "learning_rate": 2.8193709884847075e-05,
      "loss": 2.3507,
      "num_input_tokens_seen": 38377984,
      "step": 18370
    },
    {
      "epoch": 2.9976343910596297,
      "grad_norm": 0.365234375,
      "learning_rate": 2.8183972600238605e-05,
      "loss": 1.4517,
      "num_input_tokens_seen": 38387936,
      "step": 18375
    },
    {
      "epoch": 2.998450118280447,
      "grad_norm": 9.8125,
      "learning_rate": 2.817423482462156e-05,
      "loss": 4.0664,
      "num_input_tokens_seen": 38397024,
      "step": 18380
    },
    {
      "epoch": 2.9992658455012644,
      "grad_norm": 7.75,
      "learning_rate": 2.8164496559497605e-05,
      "loss": 1.9576,
      "num_input_tokens_seen": 38405936,
      "step": 18385
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.0,
      "learning_rate": 2.815475780636852e-05,
      "loss": 2.1702,
      "num_input_tokens_seen": 38413760,
      "step": 18390
    },
    {
      "epoch": 3.0008157272208176,
      "grad_norm": 0.76953125,
      "learning_rate": 2.814501856673613e-05,
      "loss": 1.7922,
      "num_input_tokens_seen": 38423248,
      "step": 18395
    },
    {
      "epoch": 3.0016314544416347,
      "grad_norm": 8.4375,
      "learning_rate": 2.8135278842102353e-05,
      "loss": 2.3376,
      "num_input_tokens_seen": 38435312,
      "step": 18400
    },
    {
      "epoch": 3.0016314544416347,
      "eval_loss": 2.535127878189087,
      "eval_runtime": 134.6644,
      "eval_samples_per_second": 20.235,
      "eval_steps_per_second": 10.121,
      "num_input_tokens_seen": 38435312,
      "step": 18400
    },
    {
      "epoch": 3.0024471816624523,
      "grad_norm": 2.578125,
      "learning_rate": 2.8125538633969183e-05,
      "loss": 2.2135,
      "num_input_tokens_seen": 38446352,
      "step": 18405
    },
    {
      "epoch": 3.0032629088832694,
      "grad_norm": 5.59375,
      "learning_rate": 2.8115797943838677e-05,
      "loss": 2.1805,
      "num_input_tokens_seen": 38455520,
      "step": 18410
    },
    {
      "epoch": 3.004078636104087,
      "grad_norm": 8.75,
      "learning_rate": 2.810605677321298e-05,
      "loss": 2.539,
      "num_input_tokens_seen": 38467056,
      "step": 18415
    },
    {
      "epoch": 3.004894363324904,
      "grad_norm": 3.359375,
      "learning_rate": 2.809631512359428e-05,
      "loss": 0.9996,
      "num_input_tokens_seen": 38476688,
      "step": 18420
    },
    {
      "epoch": 3.0057100905457217,
      "grad_norm": 8.75,
      "learning_rate": 2.8086572996484884e-05,
      "loss": 4.1607,
      "num_input_tokens_seen": 38487936,
      "step": 18425
    },
    {
      "epoch": 3.006525817766539,
      "grad_norm": 1.765625,
      "learning_rate": 2.8076830393387143e-05,
      "loss": 2.9166,
      "num_input_tokens_seen": 38497984,
      "step": 18430
    },
    {
      "epoch": 3.0073415449873564,
      "grad_norm": 3.390625,
      "learning_rate": 2.8067087315803497e-05,
      "loss": 1.1785,
      "num_input_tokens_seen": 38508944,
      "step": 18435
    },
    {
      "epoch": 3.0081572722081735,
      "grad_norm": 9.875,
      "learning_rate": 2.8057343765236433e-05,
      "loss": 3.0023,
      "num_input_tokens_seen": 38519808,
      "step": 18440
    },
    {
      "epoch": 3.008972999428991,
      "grad_norm": 12.4375,
      "learning_rate": 2.804759974318854e-05,
      "loss": 2.3482,
      "num_input_tokens_seen": 38530928,
      "step": 18445
    },
    {
      "epoch": 3.0097887266498082,
      "grad_norm": 3.875,
      "learning_rate": 2.8037855251162482e-05,
      "loss": 3.0993,
      "num_input_tokens_seen": 38542160,
      "step": 18450
    },
    {
      "epoch": 3.010604453870626,
      "grad_norm": 2.609375,
      "learning_rate": 2.802811029066096e-05,
      "loss": 3.5326,
      "num_input_tokens_seen": 38554224,
      "step": 18455
    },
    {
      "epoch": 3.011420181091443,
      "grad_norm": 7.3125,
      "learning_rate": 2.8018364863186764e-05,
      "loss": 2.1594,
      "num_input_tokens_seen": 38565952,
      "step": 18460
    },
    {
      "epoch": 3.0122359083122605,
      "grad_norm": 7.59375,
      "learning_rate": 2.800861897024279e-05,
      "loss": 2.3561,
      "num_input_tokens_seen": 38576800,
      "step": 18465
    },
    {
      "epoch": 3.0130516355330776,
      "grad_norm": 13.0,
      "learning_rate": 2.799887261333196e-05,
      "loss": 2.2839,
      "num_input_tokens_seen": 38587584,
      "step": 18470
    },
    {
      "epoch": 3.013867362753895,
      "grad_norm": 6.84375,
      "learning_rate": 2.798912579395728e-05,
      "loss": 1.8395,
      "num_input_tokens_seen": 38597520,
      "step": 18475
    },
    {
      "epoch": 3.0146830899747123,
      "grad_norm": 5.28125,
      "learning_rate": 2.797937851362185e-05,
      "loss": 1.2295,
      "num_input_tokens_seen": 38608288,
      "step": 18480
    },
    {
      "epoch": 3.01549881719553,
      "grad_norm": 7.5,
      "learning_rate": 2.7969630773828802e-05,
      "loss": 2.3853,
      "num_input_tokens_seen": 38617328,
      "step": 18485
    },
    {
      "epoch": 3.016314544416347,
      "grad_norm": 7.09375,
      "learning_rate": 2.7959882576081382e-05,
      "loss": 2.8429,
      "num_input_tokens_seen": 38627536,
      "step": 18490
    },
    {
      "epoch": 3.0171302716371646,
      "grad_norm": 7.125,
      "learning_rate": 2.795013392188286e-05,
      "loss": 2.3691,
      "num_input_tokens_seen": 38638480,
      "step": 18495
    },
    {
      "epoch": 3.0179459988579818,
      "grad_norm": 11.625,
      "learning_rate": 2.7940384812736614e-05,
      "loss": 1.9965,
      "num_input_tokens_seen": 38649872,
      "step": 18500
    },
    {
      "epoch": 3.0187617260787993,
      "grad_norm": 5.6875,
      "learning_rate": 2.7930635250146087e-05,
      "loss": 2.4363,
      "num_input_tokens_seen": 38660688,
      "step": 18505
    },
    {
      "epoch": 3.0195774532996165,
      "grad_norm": 4.75,
      "learning_rate": 2.792088523561477e-05,
      "loss": 1.6412,
      "num_input_tokens_seen": 38672800,
      "step": 18510
    },
    {
      "epoch": 3.020393180520434,
      "grad_norm": 2.9375,
      "learning_rate": 2.7911134770646246e-05,
      "loss": 2.1075,
      "num_input_tokens_seen": 38684128,
      "step": 18515
    },
    {
      "epoch": 3.021208907741251,
      "grad_norm": 6.34375,
      "learning_rate": 2.7901383856744157e-05,
      "loss": 2.8286,
      "num_input_tokens_seen": 38694736,
      "step": 18520
    },
    {
      "epoch": 3.0220246349620687,
      "grad_norm": 6.96875,
      "learning_rate": 2.7891632495412217e-05,
      "loss": 1.8855,
      "num_input_tokens_seen": 38706064,
      "step": 18525
    },
    {
      "epoch": 3.022840362182886,
      "grad_norm": 6.375,
      "learning_rate": 2.7881880688154205e-05,
      "loss": 2.2283,
      "num_input_tokens_seen": 38717632,
      "step": 18530
    },
    {
      "epoch": 3.0236560894037035,
      "grad_norm": 4.375,
      "learning_rate": 2.7872128436473977e-05,
      "loss": 1.8838,
      "num_input_tokens_seen": 38728528,
      "step": 18535
    },
    {
      "epoch": 3.0244718166245206,
      "grad_norm": 9.8125,
      "learning_rate": 2.7862375741875448e-05,
      "loss": 4.1259,
      "num_input_tokens_seen": 38739536,
      "step": 18540
    },
    {
      "epoch": 3.025287543845338,
      "grad_norm": 0.60546875,
      "learning_rate": 2.785262260586261e-05,
      "loss": 2.8911,
      "num_input_tokens_seen": 38750880,
      "step": 18545
    },
    {
      "epoch": 3.0261032710661553,
      "grad_norm": 7.5,
      "learning_rate": 2.7842869029939517e-05,
      "loss": 2.1592,
      "num_input_tokens_seen": 38760576,
      "step": 18550
    },
    {
      "epoch": 3.026918998286973,
      "grad_norm": 6.0,
      "learning_rate": 2.7833115015610296e-05,
      "loss": 2.9689,
      "num_input_tokens_seen": 38771152,
      "step": 18555
    },
    {
      "epoch": 3.02773472550779,
      "grad_norm": 7.8125,
      "learning_rate": 2.7823360564379136e-05,
      "loss": 1.89,
      "num_input_tokens_seen": 38780672,
      "step": 18560
    },
    {
      "epoch": 3.0285504527286076,
      "grad_norm": 3.859375,
      "learning_rate": 2.7813605677750297e-05,
      "loss": 1.6542,
      "num_input_tokens_seen": 38789696,
      "step": 18565
    },
    {
      "epoch": 3.0293661799494247,
      "grad_norm": 5.375,
      "learning_rate": 2.7803850357228102e-05,
      "loss": 3.9287,
      "num_input_tokens_seen": 38801872,
      "step": 18570
    },
    {
      "epoch": 3.0301819071702423,
      "grad_norm": 7.25,
      "learning_rate": 2.779409460431695e-05,
      "loss": 2.5079,
      "num_input_tokens_seen": 38813200,
      "step": 18575
    },
    {
      "epoch": 3.03099763439106,
      "grad_norm": 6.28125,
      "learning_rate": 2.778433842052129e-05,
      "loss": 2.2778,
      "num_input_tokens_seen": 38824160,
      "step": 18580
    },
    {
      "epoch": 3.031813361611877,
      "grad_norm": 4.78125,
      "learning_rate": 2.7774581807345664e-05,
      "loss": 1.9163,
      "num_input_tokens_seen": 38835856,
      "step": 18585
    },
    {
      "epoch": 3.0326290888326946,
      "grad_norm": 4.09375,
      "learning_rate": 2.776482476629465e-05,
      "loss": 1.7937,
      "num_input_tokens_seen": 38846672,
      "step": 18590
    },
    {
      "epoch": 3.0334448160535117,
      "grad_norm": 17.25,
      "learning_rate": 2.7755067298872924e-05,
      "loss": 2.5019,
      "num_input_tokens_seen": 38856576,
      "step": 18595
    },
    {
      "epoch": 3.0342605432743293,
      "grad_norm": 4.84375,
      "learning_rate": 2.774530940658518e-05,
      "loss": 3.3812,
      "num_input_tokens_seen": 38869040,
      "step": 18600
    },
    {
      "epoch": 3.0342605432743293,
      "eval_loss": 2.5402324199676514,
      "eval_runtime": 134.6647,
      "eval_samples_per_second": 20.235,
      "eval_steps_per_second": 10.121,
      "num_input_tokens_seen": 38869040,
      "step": 18600
    },
    {
      "epoch": 3.0350762704951464,
      "grad_norm": 6.0,
      "learning_rate": 2.7735551090936236e-05,
      "loss": 2.1983,
      "num_input_tokens_seen": 38879216,
      "step": 18605
    },
    {
      "epoch": 3.035891997715964,
      "grad_norm": 9.375,
      "learning_rate": 2.7725792353430934e-05,
      "loss": 2.0366,
      "num_input_tokens_seen": 38890208,
      "step": 18610
    },
    {
      "epoch": 3.036707724936781,
      "grad_norm": 5.84375,
      "learning_rate": 2.77160331955742e-05,
      "loss": 2.2987,
      "num_input_tokens_seen": 38900800,
      "step": 18615
    },
    {
      "epoch": 3.0375234521575987,
      "grad_norm": 6.03125,
      "learning_rate": 2.7706273618871008e-05,
      "loss": 1.7196,
      "num_input_tokens_seen": 38912096,
      "step": 18620
    },
    {
      "epoch": 3.038339179378416,
      "grad_norm": 4.5625,
      "learning_rate": 2.769651362482642e-05,
      "loss": 3.3059,
      "num_input_tokens_seen": 38922192,
      "step": 18625
    },
    {
      "epoch": 3.0391549065992334,
      "grad_norm": 5.71875,
      "learning_rate": 2.768675321494555e-05,
      "loss": 2.3266,
      "num_input_tokens_seen": 38932528,
      "step": 18630
    },
    {
      "epoch": 3.0399706338200505,
      "grad_norm": 5.875,
      "learning_rate": 2.7676992390733565e-05,
      "loss": 1.4866,
      "num_input_tokens_seen": 38943424,
      "step": 18635
    },
    {
      "epoch": 3.040786361040868,
      "grad_norm": 1.203125,
      "learning_rate": 2.766723115369571e-05,
      "loss": 1.9037,
      "num_input_tokens_seen": 38953312,
      "step": 18640
    },
    {
      "epoch": 3.041602088261685,
      "grad_norm": 11.3125,
      "learning_rate": 2.765746950533729e-05,
      "loss": 3.0202,
      "num_input_tokens_seen": 38963648,
      "step": 18645
    },
    {
      "epoch": 3.042417815482503,
      "grad_norm": 9.3125,
      "learning_rate": 2.7647707447163684e-05,
      "loss": 1.8829,
      "num_input_tokens_seen": 38975440,
      "step": 18650
    },
    {
      "epoch": 3.04323354270332,
      "grad_norm": 9.875,
      "learning_rate": 2.7637944980680315e-05,
      "loss": 1.6415,
      "num_input_tokens_seen": 38985712,
      "step": 18655
    },
    {
      "epoch": 3.0440492699241375,
      "grad_norm": 7.9375,
      "learning_rate": 2.762818210739268e-05,
      "loss": 3.2137,
      "num_input_tokens_seen": 38996432,
      "step": 18660
    },
    {
      "epoch": 3.0448649971449546,
      "grad_norm": 5.375,
      "learning_rate": 2.7618418828806332e-05,
      "loss": 2.4147,
      "num_input_tokens_seen": 39006176,
      "step": 18665
    },
    {
      "epoch": 3.045680724365772,
      "grad_norm": 6.625,
      "learning_rate": 2.76086551464269e-05,
      "loss": 2.1611,
      "num_input_tokens_seen": 39016032,
      "step": 18670
    },
    {
      "epoch": 3.0464964515865893,
      "grad_norm": 10.5625,
      "learning_rate": 2.759889106176006e-05,
      "loss": 2.0296,
      "num_input_tokens_seen": 39028112,
      "step": 18675
    },
    {
      "epoch": 3.047312178807407,
      "grad_norm": 7.34375,
      "learning_rate": 2.758912657631156e-05,
      "loss": 2.7356,
      "num_input_tokens_seen": 39039104,
      "step": 18680
    },
    {
      "epoch": 3.048127906028224,
      "grad_norm": 6.4375,
      "learning_rate": 2.7579361691587198e-05,
      "loss": 1.9054,
      "num_input_tokens_seen": 39050944,
      "step": 18685
    },
    {
      "epoch": 3.0489436332490416,
      "grad_norm": 4.78125,
      "learning_rate": 2.756959640909285e-05,
      "loss": 2.6923,
      "num_input_tokens_seen": 39062608,
      "step": 18690
    },
    {
      "epoch": 3.0497593604698587,
      "grad_norm": 5.0625,
      "learning_rate": 2.7559830730334452e-05,
      "loss": 1.8744,
      "num_input_tokens_seen": 39073088,
      "step": 18695
    },
    {
      "epoch": 3.0505750876906763,
      "grad_norm": 6.21875,
      "learning_rate": 2.7550064656817988e-05,
      "loss": 1.3693,
      "num_input_tokens_seen": 39082720,
      "step": 18700
    },
    {
      "epoch": 3.0513908149114934,
      "grad_norm": 18.625,
      "learning_rate": 2.7540298190049503e-05,
      "loss": 1.7674,
      "num_input_tokens_seen": 39094272,
      "step": 18705
    },
    {
      "epoch": 3.052206542132311,
      "grad_norm": 4.0,
      "learning_rate": 2.7530531331535107e-05,
      "loss": 2.6544,
      "num_input_tokens_seen": 39104816,
      "step": 18710
    },
    {
      "epoch": 3.053022269353128,
      "grad_norm": 7.0,
      "learning_rate": 2.752076408278099e-05,
      "loss": 2.6126,
      "num_input_tokens_seen": 39114880,
      "step": 18715
    },
    {
      "epoch": 3.0538379965739457,
      "grad_norm": 1.2578125,
      "learning_rate": 2.751099644529337e-05,
      "loss": 1.7647,
      "num_input_tokens_seen": 39126880,
      "step": 18720
    },
    {
      "epoch": 3.054653723794763,
      "grad_norm": 4.09375,
      "learning_rate": 2.7501228420578533e-05,
      "loss": 1.33,
      "num_input_tokens_seen": 39136016,
      "step": 18725
    },
    {
      "epoch": 3.0554694510155804,
      "grad_norm": 8.1875,
      "learning_rate": 2.7491460010142857e-05,
      "loss": 2.4768,
      "num_input_tokens_seen": 39146688,
      "step": 18730
    },
    {
      "epoch": 3.0562851782363976,
      "grad_norm": 1.8203125,
      "learning_rate": 2.7481691215492727e-05,
      "loss": 2.3832,
      "num_input_tokens_seen": 39156928,
      "step": 18735
    },
    {
      "epoch": 3.057100905457215,
      "grad_norm": 8.5625,
      "learning_rate": 2.747192203813463e-05,
      "loss": 2.1745,
      "num_input_tokens_seen": 39168096,
      "step": 18740
    },
    {
      "epoch": 3.0579166326780323,
      "grad_norm": 18.25,
      "learning_rate": 2.7462152479575087e-05,
      "loss": 3.1783,
      "num_input_tokens_seen": 39178272,
      "step": 18745
    },
    {
      "epoch": 3.05873235989885,
      "grad_norm": 9.0,
      "learning_rate": 2.7452382541320697e-05,
      "loss": 2.4667,
      "num_input_tokens_seen": 39188880,
      "step": 18750
    },
    {
      "epoch": 3.059548087119667,
      "grad_norm": 7.28125,
      "learning_rate": 2.7442612224878096e-05,
      "loss": 2.8692,
      "num_input_tokens_seen": 39199712,
      "step": 18755
    },
    {
      "epoch": 3.0603638143404845,
      "grad_norm": 8.375,
      "learning_rate": 2.7432841531753994e-05,
      "loss": 3.7592,
      "num_input_tokens_seen": 39210928,
      "step": 18760
    },
    {
      "epoch": 3.0611795415613017,
      "grad_norm": 7.0,
      "learning_rate": 2.7423070463455147e-05,
      "loss": 2.7395,
      "num_input_tokens_seen": 39221392,
      "step": 18765
    },
    {
      "epoch": 3.0619952687821193,
      "grad_norm": 8.4375,
      "learning_rate": 2.7413299021488397e-05,
      "loss": 3.383,
      "num_input_tokens_seen": 39230624,
      "step": 18770
    },
    {
      "epoch": 3.062810996002937,
      "grad_norm": 9.6875,
      "learning_rate": 2.7403527207360615e-05,
      "loss": 2.5962,
      "num_input_tokens_seen": 39241776,
      "step": 18775
    },
    {
      "epoch": 3.063626723223754,
      "grad_norm": 0.216796875,
      "learning_rate": 2.7393755022578722e-05,
      "loss": 1.927,
      "num_input_tokens_seen": 39251312,
      "step": 18780
    },
    {
      "epoch": 3.0644424504445715,
      "grad_norm": 6.5,
      "learning_rate": 2.7383982468649714e-05,
      "loss": 2.0895,
      "num_input_tokens_seen": 39262512,
      "step": 18785
    },
    {
      "epoch": 3.0652581776653887,
      "grad_norm": 4.375,
      "learning_rate": 2.7374209547080665e-05,
      "loss": 2.3992,
      "num_input_tokens_seen": 39272320,
      "step": 18790
    },
    {
      "epoch": 3.0660739048862062,
      "grad_norm": 8.1875,
      "learning_rate": 2.7364436259378663e-05,
      "loss": 1.4882,
      "num_input_tokens_seen": 39283888,
      "step": 18795
    },
    {
      "epoch": 3.0668896321070234,
      "grad_norm": 6.90625,
      "learning_rate": 2.735466260705088e-05,
      "loss": 2.2142,
      "num_input_tokens_seen": 39294832,
      "step": 18800
    },
    {
      "epoch": 3.0668896321070234,
      "eval_loss": 2.548177480697632,
      "eval_runtime": 135.1277,
      "eval_samples_per_second": 20.166,
      "eval_steps_per_second": 10.087,
      "num_input_tokens_seen": 39294832,
      "step": 18800
    },
    {
      "epoch": 3.067705359327841,
      "grad_norm": 10.4375,
      "learning_rate": 2.7344888591604524e-05,
      "loss": 1.2988,
      "num_input_tokens_seen": 39305520,
      "step": 18805
    },
    {
      "epoch": 3.068521086548658,
      "grad_norm": 3.796875,
      "learning_rate": 2.7335114214546893e-05,
      "loss": 2.2595,
      "num_input_tokens_seen": 39314640,
      "step": 18810
    },
    {
      "epoch": 3.0693368137694756,
      "grad_norm": 7.78125,
      "learning_rate": 2.7325339477385293e-05,
      "loss": 2.6693,
      "num_input_tokens_seen": 39324720,
      "step": 18815
    },
    {
      "epoch": 3.0701525409902928,
      "grad_norm": 7.0625,
      "learning_rate": 2.7315564381627128e-05,
      "loss": 2.535,
      "num_input_tokens_seen": 39335504,
      "step": 18820
    },
    {
      "epoch": 3.0709682682111104,
      "grad_norm": 3.65625,
      "learning_rate": 2.7305788928779835e-05,
      "loss": 2.8521,
      "num_input_tokens_seen": 39346160,
      "step": 18825
    },
    {
      "epoch": 3.0717839954319275,
      "grad_norm": 5.53125,
      "learning_rate": 2.729601312035091e-05,
      "loss": 1.8358,
      "num_input_tokens_seen": 39354576,
      "step": 18830
    },
    {
      "epoch": 3.072599722652745,
      "grad_norm": 10.0,
      "learning_rate": 2.7286236957847915e-05,
      "loss": 2.5139,
      "num_input_tokens_seen": 39363920,
      "step": 18835
    },
    {
      "epoch": 3.073415449873562,
      "grad_norm": 10.3125,
      "learning_rate": 2.7276460442778446e-05,
      "loss": 1.8985,
      "num_input_tokens_seen": 39374368,
      "step": 18840
    },
    {
      "epoch": 3.0742311770943798,
      "grad_norm": 7.25,
      "learning_rate": 2.726668357665017e-05,
      "loss": 2.1724,
      "num_input_tokens_seen": 39386048,
      "step": 18845
    },
    {
      "epoch": 3.075046904315197,
      "grad_norm": 7.84375,
      "learning_rate": 2.7256906360970808e-05,
      "loss": 1.0069,
      "num_input_tokens_seen": 39397056,
      "step": 18850
    },
    {
      "epoch": 3.0758626315360145,
      "grad_norm": 5.75,
      "learning_rate": 2.7247128797248117e-05,
      "loss": 3.2035,
      "num_input_tokens_seen": 39406592,
      "step": 18855
    },
    {
      "epoch": 3.0766783587568316,
      "grad_norm": 1.9296875,
      "learning_rate": 2.7237350886989925e-05,
      "loss": 2.4079,
      "num_input_tokens_seen": 39416016,
      "step": 18860
    },
    {
      "epoch": 3.077494085977649,
      "grad_norm": 12.5625,
      "learning_rate": 2.7227572631704107e-05,
      "loss": 2.8224,
      "num_input_tokens_seen": 39426144,
      "step": 18865
    },
    {
      "epoch": 3.0783098131984663,
      "grad_norm": 4.6875,
      "learning_rate": 2.7217794032898596e-05,
      "loss": 2.8631,
      "num_input_tokens_seen": 39436880,
      "step": 18870
    },
    {
      "epoch": 3.079125540419284,
      "grad_norm": 6.34375,
      "learning_rate": 2.7208015092081384e-05,
      "loss": 2.592,
      "num_input_tokens_seen": 39447856,
      "step": 18875
    },
    {
      "epoch": 3.079941267640101,
      "grad_norm": 9.8125,
      "learning_rate": 2.719823581076049e-05,
      "loss": 3.2701,
      "num_input_tokens_seen": 39456720,
      "step": 18880
    },
    {
      "epoch": 3.0807569948609186,
      "grad_norm": 7.6875,
      "learning_rate": 2.718845619044401e-05,
      "loss": 2.3436,
      "num_input_tokens_seen": 39466992,
      "step": 18885
    },
    {
      "epoch": 3.0815727220817357,
      "grad_norm": 11.3125,
      "learning_rate": 2.7178676232640088e-05,
      "loss": 2.0766,
      "num_input_tokens_seen": 39477488,
      "step": 18890
    },
    {
      "epoch": 3.0823884493025533,
      "grad_norm": 4.96875,
      "learning_rate": 2.716889593885691e-05,
      "loss": 2.3528,
      "num_input_tokens_seen": 39486432,
      "step": 18895
    },
    {
      "epoch": 3.0832041765233704,
      "grad_norm": 5.34375,
      "learning_rate": 2.7159115310602716e-05,
      "loss": 1.709,
      "num_input_tokens_seen": 39498208,
      "step": 18900
    },
    {
      "epoch": 3.084019903744188,
      "grad_norm": 5.03125,
      "learning_rate": 2.7149334349385814e-05,
      "loss": 1.2992,
      "num_input_tokens_seen": 39508832,
      "step": 18905
    },
    {
      "epoch": 3.084835630965005,
      "grad_norm": 5.8125,
      "learning_rate": 2.713955305671454e-05,
      "loss": 1.6824,
      "num_input_tokens_seen": 39519248,
      "step": 18910
    },
    {
      "epoch": 3.0856513581858227,
      "grad_norm": 11.25,
      "learning_rate": 2.71297714340973e-05,
      "loss": 2.8482,
      "num_input_tokens_seen": 39529936,
      "step": 18915
    },
    {
      "epoch": 3.08646708540664,
      "grad_norm": 6.0625,
      "learning_rate": 2.7119989483042545e-05,
      "loss": 1.9419,
      "num_input_tokens_seen": 39540752,
      "step": 18920
    },
    {
      "epoch": 3.0872828126274574,
      "grad_norm": 9.3125,
      "learning_rate": 2.7110207205058768e-05,
      "loss": 2.4596,
      "num_input_tokens_seen": 39550272,
      "step": 18925
    },
    {
      "epoch": 3.0880985398482745,
      "grad_norm": 3.015625,
      "learning_rate": 2.7100424601654517e-05,
      "loss": 2.2172,
      "num_input_tokens_seen": 39559920,
      "step": 18930
    },
    {
      "epoch": 3.088914267069092,
      "grad_norm": 1.6953125,
      "learning_rate": 2.7090641674338403e-05,
      "loss": 1.0733,
      "num_input_tokens_seen": 39569632,
      "step": 18935
    },
    {
      "epoch": 3.0897299942899092,
      "grad_norm": 11.625,
      "learning_rate": 2.7080858424619072e-05,
      "loss": 2.7145,
      "num_input_tokens_seen": 39580704,
      "step": 18940
    },
    {
      "epoch": 3.090545721510727,
      "grad_norm": 8.6875,
      "learning_rate": 2.707107485400521e-05,
      "loss": 2.8611,
      "num_input_tokens_seen": 39590160,
      "step": 18945
    },
    {
      "epoch": 3.0913614487315444,
      "grad_norm": 3.34375,
      "learning_rate": 2.7061290964005586e-05,
      "loss": 1.4562,
      "num_input_tokens_seen": 39600768,
      "step": 18950
    },
    {
      "epoch": 3.0921771759523615,
      "grad_norm": 8.5,
      "learning_rate": 2.7051506756129e-05,
      "loss": 2.1127,
      "num_input_tokens_seen": 39612256,
      "step": 18955
    },
    {
      "epoch": 3.092992903173179,
      "grad_norm": 10.9375,
      "learning_rate": 2.704172223188428e-05,
      "loss": 3.5645,
      "num_input_tokens_seen": 39621536,
      "step": 18960
    },
    {
      "epoch": 3.0938086303939962,
      "grad_norm": 3.953125,
      "learning_rate": 2.7031937392780334e-05,
      "loss": 3.572,
      "num_input_tokens_seen": 39631904,
      "step": 18965
    },
    {
      "epoch": 3.094624357614814,
      "grad_norm": 6.6875,
      "learning_rate": 2.702215224032611e-05,
      "loss": 2.5679,
      "num_input_tokens_seen": 39641872,
      "step": 18970
    },
    {
      "epoch": 3.095440084835631,
      "grad_norm": 5.9375,
      "learning_rate": 2.70123667760306e-05,
      "loss": 2.0799,
      "num_input_tokens_seen": 39652656,
      "step": 18975
    },
    {
      "epoch": 3.0962558120564485,
      "grad_norm": 7.78125,
      "learning_rate": 2.7002581001402845e-05,
      "loss": 1.7741,
      "num_input_tokens_seen": 39664400,
      "step": 18980
    },
    {
      "epoch": 3.0970715392772656,
      "grad_norm": 11.125,
      "learning_rate": 2.6992794917951923e-05,
      "loss": 2.7907,
      "num_input_tokens_seen": 39674944,
      "step": 18985
    },
    {
      "epoch": 3.097887266498083,
      "grad_norm": 4.5625,
      "learning_rate": 2.6983008527187e-05,
      "loss": 1.2365,
      "num_input_tokens_seen": 39685440,
      "step": 18990
    },
    {
      "epoch": 3.0987029937189003,
      "grad_norm": 9.1875,
      "learning_rate": 2.697322183061723e-05,
      "loss": 1.362,
      "num_input_tokens_seen": 39696608,
      "step": 18995
    },
    {
      "epoch": 3.099518720939718,
      "grad_norm": 5.96875,
      "learning_rate": 2.696343482975186e-05,
      "loss": 2.0647,
      "num_input_tokens_seen": 39706928,
      "step": 19000
    },
    {
      "epoch": 3.099518720939718,
      "eval_loss": 2.538421630859375,
      "eval_runtime": 134.7507,
      "eval_samples_per_second": 20.223,
      "eval_steps_per_second": 10.115,
      "num_input_tokens_seen": 39706928,
      "step": 19000
    },
    {
      "epoch": 3.100334448160535,
      "grad_norm": 4.875,
      "learning_rate": 2.695364752610016e-05,
      "loss": 1.3137,
      "num_input_tokens_seen": 39717904,
      "step": 19005
    },
    {
      "epoch": 3.1011501753813526,
      "grad_norm": 8.5625,
      "learning_rate": 2.6943859921171467e-05,
      "loss": 2.3466,
      "num_input_tokens_seen": 39728368,
      "step": 19010
    },
    {
      "epoch": 3.1019659026021698,
      "grad_norm": 5.59375,
      "learning_rate": 2.6934072016475143e-05,
      "loss": 2.2962,
      "num_input_tokens_seen": 39738544,
      "step": 19015
    },
    {
      "epoch": 3.1027816298229873,
      "grad_norm": 7.65625,
      "learning_rate": 2.6924283813520606e-05,
      "loss": 1.6056,
      "num_input_tokens_seen": 39748848,
      "step": 19020
    },
    {
      "epoch": 3.1035973570438045,
      "grad_norm": 9.4375,
      "learning_rate": 2.691449531381733e-05,
      "loss": 2.5877,
      "num_input_tokens_seen": 39759024,
      "step": 19025
    },
    {
      "epoch": 3.104413084264622,
      "grad_norm": 2.078125,
      "learning_rate": 2.6904706518874816e-05,
      "loss": 2.9983,
      "num_input_tokens_seen": 39770576,
      "step": 19030
    },
    {
      "epoch": 3.105228811485439,
      "grad_norm": 10.6875,
      "learning_rate": 2.6894917430202615e-05,
      "loss": 2.3436,
      "num_input_tokens_seen": 39780272,
      "step": 19035
    },
    {
      "epoch": 3.1060445387062567,
      "grad_norm": 3.9375,
      "learning_rate": 2.6885128049310343e-05,
      "loss": 2.1036,
      "num_input_tokens_seen": 39793072,
      "step": 19040
    },
    {
      "epoch": 3.106860265927074,
      "grad_norm": 11.4375,
      "learning_rate": 2.687533837770762e-05,
      "loss": 1.4843,
      "num_input_tokens_seen": 39803264,
      "step": 19045
    },
    {
      "epoch": 3.1076759931478914,
      "grad_norm": 6.53125,
      "learning_rate": 2.6865548416904162e-05,
      "loss": 0.7208,
      "num_input_tokens_seen": 39813824,
      "step": 19050
    },
    {
      "epoch": 3.1084917203687086,
      "grad_norm": 0.44921875,
      "learning_rate": 2.68557581684097e-05,
      "loss": 2.4978,
      "num_input_tokens_seen": 39822560,
      "step": 19055
    },
    {
      "epoch": 3.109307447589526,
      "grad_norm": 9.6875,
      "learning_rate": 2.6845967633733998e-05,
      "loss": 2.3114,
      "num_input_tokens_seen": 39832432,
      "step": 19060
    },
    {
      "epoch": 3.1101231748103433,
      "grad_norm": 9.125,
      "learning_rate": 2.683617681438689e-05,
      "loss": 4.1923,
      "num_input_tokens_seen": 39843424,
      "step": 19065
    },
    {
      "epoch": 3.110938902031161,
      "grad_norm": 8.75,
      "learning_rate": 2.682638571187825e-05,
      "loss": 2.5996,
      "num_input_tokens_seen": 39852128,
      "step": 19070
    },
    {
      "epoch": 3.111754629251978,
      "grad_norm": 7.0,
      "learning_rate": 2.6816594327717976e-05,
      "loss": 2.7419,
      "num_input_tokens_seen": 39863744,
      "step": 19075
    },
    {
      "epoch": 3.1125703564727956,
      "grad_norm": 5.09375,
      "learning_rate": 2.680680266341603e-05,
      "loss": 1.6439,
      "num_input_tokens_seen": 39873584,
      "step": 19080
    },
    {
      "epoch": 3.1133860836936127,
      "grad_norm": 6.96875,
      "learning_rate": 2.67970107204824e-05,
      "loss": 1.4829,
      "num_input_tokens_seen": 39882704,
      "step": 19085
    },
    {
      "epoch": 3.1142018109144303,
      "grad_norm": 9.0625,
      "learning_rate": 2.6787218500427142e-05,
      "loss": 3.2501,
      "num_input_tokens_seen": 39893568,
      "step": 19090
    },
    {
      "epoch": 3.1150175381352474,
      "grad_norm": 7.65625,
      "learning_rate": 2.6777426004760332e-05,
      "loss": 3.3429,
      "num_input_tokens_seen": 39905648,
      "step": 19095
    },
    {
      "epoch": 3.115833265356065,
      "grad_norm": 7.6875,
      "learning_rate": 2.6767633234992094e-05,
      "loss": 2.7122,
      "num_input_tokens_seen": 39914608,
      "step": 19100
    },
    {
      "epoch": 3.116648992576882,
      "grad_norm": 8.6875,
      "learning_rate": 2.6757840192632598e-05,
      "loss": 4.4158,
      "num_input_tokens_seen": 39925312,
      "step": 19105
    },
    {
      "epoch": 3.1174647197976997,
      "grad_norm": 6.5625,
      "learning_rate": 2.6748046879192052e-05,
      "loss": 1.5997,
      "num_input_tokens_seen": 39934672,
      "step": 19110
    },
    {
      "epoch": 3.118280447018517,
      "grad_norm": 13.1875,
      "learning_rate": 2.673825329618071e-05,
      "loss": 2.7824,
      "num_input_tokens_seen": 39945040,
      "step": 19115
    },
    {
      "epoch": 3.1190961742393344,
      "grad_norm": 6.625,
      "learning_rate": 2.6728459445108866e-05,
      "loss": 3.0114,
      "num_input_tokens_seen": 39955840,
      "step": 19120
    },
    {
      "epoch": 3.1199119014601515,
      "grad_norm": 6.5,
      "learning_rate": 2.6718665327486854e-05,
      "loss": 2.5034,
      "num_input_tokens_seen": 39966688,
      "step": 19125
    },
    {
      "epoch": 3.120727628680969,
      "grad_norm": 11.375,
      "learning_rate": 2.6708870944825048e-05,
      "loss": 3.1015,
      "num_input_tokens_seen": 39977392,
      "step": 19130
    },
    {
      "epoch": 3.121543355901786,
      "grad_norm": 3.3125,
      "learning_rate": 2.6699076298633874e-05,
      "loss": 2.2812,
      "num_input_tokens_seen": 39987968,
      "step": 19135
    },
    {
      "epoch": 3.122359083122604,
      "grad_norm": 6.15625,
      "learning_rate": 2.6689281390423788e-05,
      "loss": 2.9054,
      "num_input_tokens_seen": 39998272,
      "step": 19140
    },
    {
      "epoch": 3.1231748103434214,
      "grad_norm": 11.375,
      "learning_rate": 2.667948622170527e-05,
      "loss": 2.3876,
      "num_input_tokens_seen": 40009344,
      "step": 19145
    },
    {
      "epoch": 3.1239905375642385,
      "grad_norm": 2.546875,
      "learning_rate": 2.6669690793988873e-05,
      "loss": 2.3684,
      "num_input_tokens_seen": 40019200,
      "step": 19150
    },
    {
      "epoch": 3.124806264785056,
      "grad_norm": 6.40625,
      "learning_rate": 2.665989510878518e-05,
      "loss": 2.8145,
      "num_input_tokens_seen": 40028880,
      "step": 19155
    },
    {
      "epoch": 3.125621992005873,
      "grad_norm": 7.5,
      "learning_rate": 2.6650099167604793e-05,
      "loss": 2.9376,
      "num_input_tokens_seen": 40039824,
      "step": 19160
    },
    {
      "epoch": 3.126437719226691,
      "grad_norm": 2.1875,
      "learning_rate": 2.6640302971958376e-05,
      "loss": 2.1377,
      "num_input_tokens_seen": 40048944,
      "step": 19165
    },
    {
      "epoch": 3.127253446447508,
      "grad_norm": 7.71875,
      "learning_rate": 2.6630506523356635e-05,
      "loss": 3.9531,
      "num_input_tokens_seen": 40059312,
      "step": 19170
    },
    {
      "epoch": 3.1280691736683255,
      "grad_norm": 12.5,
      "learning_rate": 2.6620709823310297e-05,
      "loss": 3.2793,
      "num_input_tokens_seen": 40070928,
      "step": 19175
    },
    {
      "epoch": 3.1288849008891426,
      "grad_norm": 12.0,
      "learning_rate": 2.661091287333014e-05,
      "loss": 2.8367,
      "num_input_tokens_seen": 40080368,
      "step": 19180
    },
    {
      "epoch": 3.12970062810996,
      "grad_norm": 7.8125,
      "learning_rate": 2.660111567492696e-05,
      "loss": 2.9917,
      "num_input_tokens_seen": 40091376,
      "step": 19185
    },
    {
      "epoch": 3.1305163553307773,
      "grad_norm": 8.9375,
      "learning_rate": 2.6591318229611635e-05,
      "loss": 2.7834,
      "num_input_tokens_seen": 40101408,
      "step": 19190
    },
    {
      "epoch": 3.131332082551595,
      "grad_norm": 5.90625,
      "learning_rate": 2.6581520538895037e-05,
      "loss": 2.5444,
      "num_input_tokens_seen": 40112880,
      "step": 19195
    },
    {
      "epoch": 3.132147809772412,
      "grad_norm": 6.3125,
      "learning_rate": 2.6571722604288102e-05,
      "loss": 2.2582,
      "num_input_tokens_seen": 40121472,
      "step": 19200
    },
    {
      "epoch": 3.132147809772412,
      "eval_loss": 2.5614473819732666,
      "eval_runtime": 135.0803,
      "eval_samples_per_second": 20.173,
      "eval_steps_per_second": 10.09,
      "num_input_tokens_seen": 40121472,
      "step": 19200
    },
    {
      "epoch": 3.1329635369932296,
      "grad_norm": 11.4375,
      "learning_rate": 2.656192442730179e-05,
      "loss": 1.7784,
      "num_input_tokens_seen": 40131728,
      "step": 19205
    },
    {
      "epoch": 3.1337792642140467,
      "grad_norm": 10.5625,
      "learning_rate": 2.6552126009447098e-05,
      "loss": 2.3919,
      "num_input_tokens_seen": 40141408,
      "step": 19210
    },
    {
      "epoch": 3.1345949914348643,
      "grad_norm": 7.65625,
      "learning_rate": 2.654232735223507e-05,
      "loss": 2.6903,
      "num_input_tokens_seen": 40150880,
      "step": 19215
    },
    {
      "epoch": 3.1354107186556814,
      "grad_norm": 8.375,
      "learning_rate": 2.6532528457176787e-05,
      "loss": 2.7103,
      "num_input_tokens_seen": 40160608,
      "step": 19220
    },
    {
      "epoch": 3.136226445876499,
      "grad_norm": 6.25,
      "learning_rate": 2.6522729325783348e-05,
      "loss": 2.6182,
      "num_input_tokens_seen": 40170480,
      "step": 19225
    },
    {
      "epoch": 3.137042173097316,
      "grad_norm": 2.921875,
      "learning_rate": 2.6512929959565914e-05,
      "loss": 2.015,
      "num_input_tokens_seen": 40181536,
      "step": 19230
    },
    {
      "epoch": 3.1378579003181337,
      "grad_norm": 8.6875,
      "learning_rate": 2.6503130360035673e-05,
      "loss": 2.1905,
      "num_input_tokens_seen": 40191952,
      "step": 19235
    },
    {
      "epoch": 3.138673627538951,
      "grad_norm": 8.5625,
      "learning_rate": 2.6493330528703835e-05,
      "loss": 2.9658,
      "num_input_tokens_seen": 40201808,
      "step": 19240
    },
    {
      "epoch": 3.1394893547597684,
      "grad_norm": 9.25,
      "learning_rate": 2.648353046708167e-05,
      "loss": 0.716,
      "num_input_tokens_seen": 40214720,
      "step": 19245
    },
    {
      "epoch": 3.1403050819805856,
      "grad_norm": 8.75,
      "learning_rate": 2.647373017668046e-05,
      "loss": 3.0051,
      "num_input_tokens_seen": 40225920,
      "step": 19250
    },
    {
      "epoch": 3.141120809201403,
      "grad_norm": 11.75,
      "learning_rate": 2.6463929659011537e-05,
      "loss": 3.3569,
      "num_input_tokens_seen": 40236416,
      "step": 19255
    },
    {
      "epoch": 3.1419365364222203,
      "grad_norm": 10.0,
      "learning_rate": 2.6454128915586262e-05,
      "loss": 3.1072,
      "num_input_tokens_seen": 40245776,
      "step": 19260
    },
    {
      "epoch": 3.142752263643038,
      "grad_norm": 12.625,
      "learning_rate": 2.6444327947916036e-05,
      "loss": 2.3197,
      "num_input_tokens_seen": 40256160,
      "step": 19265
    },
    {
      "epoch": 3.143567990863855,
      "grad_norm": 10.75,
      "learning_rate": 2.6434526757512292e-05,
      "loss": 2.4474,
      "num_input_tokens_seen": 40267136,
      "step": 19270
    },
    {
      "epoch": 3.1443837180846725,
      "grad_norm": 0.12158203125,
      "learning_rate": 2.6424725345886486e-05,
      "loss": 2.1871,
      "num_input_tokens_seen": 40277360,
      "step": 19275
    },
    {
      "epoch": 3.1451994453054897,
      "grad_norm": 8.125,
      "learning_rate": 2.641492371455014e-05,
      "loss": 2.6652,
      "num_input_tokens_seen": 40287152,
      "step": 19280
    },
    {
      "epoch": 3.1460151725263072,
      "grad_norm": 7.4375,
      "learning_rate": 2.640512186501477e-05,
      "loss": 2.3262,
      "num_input_tokens_seen": 40297744,
      "step": 19285
    },
    {
      "epoch": 3.1468308997471244,
      "grad_norm": 6.53125,
      "learning_rate": 2.639531979879195e-05,
      "loss": 2.8425,
      "num_input_tokens_seen": 40307808,
      "step": 19290
    },
    {
      "epoch": 3.147646626967942,
      "grad_norm": 4.09375,
      "learning_rate": 2.638551751739328e-05,
      "loss": 2.9993,
      "num_input_tokens_seen": 40319152,
      "step": 19295
    },
    {
      "epoch": 3.148462354188759,
      "grad_norm": 8.0,
      "learning_rate": 2.6375715022330404e-05,
      "loss": 2.4484,
      "num_input_tokens_seen": 40330512,
      "step": 19300
    },
    {
      "epoch": 3.1492780814095767,
      "grad_norm": 5.875,
      "learning_rate": 2.6365912315114976e-05,
      "loss": 2.0357,
      "num_input_tokens_seen": 40341744,
      "step": 19305
    },
    {
      "epoch": 3.150093808630394,
      "grad_norm": 3.5625,
      "learning_rate": 2.6356109397258704e-05,
      "loss": 1.5845,
      "num_input_tokens_seen": 40353744,
      "step": 19310
    },
    {
      "epoch": 3.1509095358512114,
      "grad_norm": 4.40625,
      "learning_rate": 2.6346306270273325e-05,
      "loss": 2.025,
      "num_input_tokens_seen": 40364288,
      "step": 19315
    },
    {
      "epoch": 3.151725263072029,
      "grad_norm": 7.25,
      "learning_rate": 2.6336502935670608e-05,
      "loss": 2.7205,
      "num_input_tokens_seen": 40375152,
      "step": 19320
    },
    {
      "epoch": 3.152540990292846,
      "grad_norm": 8.25,
      "learning_rate": 2.6326699394962333e-05,
      "loss": 1.8616,
      "num_input_tokens_seen": 40384496,
      "step": 19325
    },
    {
      "epoch": 3.153356717513663,
      "grad_norm": 5.96875,
      "learning_rate": 2.6316895649660334e-05,
      "loss": 2.255,
      "num_input_tokens_seen": 40394080,
      "step": 19330
    },
    {
      "epoch": 3.1541724447344808,
      "grad_norm": 5.78125,
      "learning_rate": 2.6307091701276486e-05,
      "loss": 2.293,
      "num_input_tokens_seen": 40404240,
      "step": 19335
    },
    {
      "epoch": 3.1549881719552983,
      "grad_norm": 4.6875,
      "learning_rate": 2.629728755132267e-05,
      "loss": 1.6456,
      "num_input_tokens_seen": 40415696,
      "step": 19340
    },
    {
      "epoch": 3.1558038991761155,
      "grad_norm": 7.96875,
      "learning_rate": 2.628748320131081e-05,
      "loss": 2.3573,
      "num_input_tokens_seen": 40425376,
      "step": 19345
    },
    {
      "epoch": 3.156619626396933,
      "grad_norm": 4.0,
      "learning_rate": 2.6277678652752856e-05,
      "loss": 4.3231,
      "num_input_tokens_seen": 40435040,
      "step": 19350
    },
    {
      "epoch": 3.15743535361775,
      "grad_norm": 3.203125,
      "learning_rate": 2.6267873907160807e-05,
      "loss": 1.2957,
      "num_input_tokens_seen": 40445680,
      "step": 19355
    },
    {
      "epoch": 3.1582510808385678,
      "grad_norm": 8.6875,
      "learning_rate": 2.6258068966046668e-05,
      "loss": 2.794,
      "num_input_tokens_seen": 40457040,
      "step": 19360
    },
    {
      "epoch": 3.159066808059385,
      "grad_norm": 6.6875,
      "learning_rate": 2.6248263830922475e-05,
      "loss": 2.3179,
      "num_input_tokens_seen": 40467776,
      "step": 19365
    },
    {
      "epoch": 3.1598825352802025,
      "grad_norm": 7.0,
      "learning_rate": 2.6238458503300318e-05,
      "loss": 2.1722,
      "num_input_tokens_seen": 40478464,
      "step": 19370
    },
    {
      "epoch": 3.1606982625010196,
      "grad_norm": 9.125,
      "learning_rate": 2.6228652984692292e-05,
      "loss": 3.241,
      "num_input_tokens_seen": 40487760,
      "step": 19375
    },
    {
      "epoch": 3.161513989721837,
      "grad_norm": 9.875,
      "learning_rate": 2.621884727661054e-05,
      "loss": 1.3619,
      "num_input_tokens_seen": 40497584,
      "step": 19380
    },
    {
      "epoch": 3.1623297169426543,
      "grad_norm": 1.84375,
      "learning_rate": 2.6209041380567222e-05,
      "loss": 1.8896,
      "num_input_tokens_seen": 40507808,
      "step": 19385
    },
    {
      "epoch": 3.163145444163472,
      "grad_norm": 4.625,
      "learning_rate": 2.6199235298074527e-05,
      "loss": 2.5161,
      "num_input_tokens_seen": 40518256,
      "step": 19390
    },
    {
      "epoch": 3.163961171384289,
      "grad_norm": 0.126953125,
      "learning_rate": 2.618942903064468e-05,
      "loss": 2.0876,
      "num_input_tokens_seen": 40528144,
      "step": 19395
    },
    {
      "epoch": 3.1647768986051066,
      "grad_norm": 2.546875,
      "learning_rate": 2.6179622579789932e-05,
      "loss": 2.1523,
      "num_input_tokens_seen": 40537792,
      "step": 19400
    },
    {
      "epoch": 3.1647768986051066,
      "eval_loss": 2.54274582862854,
      "eval_runtime": 135.1021,
      "eval_samples_per_second": 20.17,
      "eval_steps_per_second": 10.089,
      "num_input_tokens_seen": 40537792,
      "step": 19400
    },
    {
      "epoch": 3.1655926258259237,
      "grad_norm": 18.5,
      "learning_rate": 2.6169815947022553e-05,
      "loss": 4.3618,
      "num_input_tokens_seen": 40548320,
      "step": 19405
    },
    {
      "epoch": 3.1664083530467413,
      "grad_norm": 3.546875,
      "learning_rate": 2.6160009133854853e-05,
      "loss": 1.5637,
      "num_input_tokens_seen": 40559536,
      "step": 19410
    },
    {
      "epoch": 3.1672240802675584,
      "grad_norm": 7.28125,
      "learning_rate": 2.6150202141799168e-05,
      "loss": 1.3557,
      "num_input_tokens_seen": 40571136,
      "step": 19415
    },
    {
      "epoch": 3.168039807488376,
      "grad_norm": 4.5,
      "learning_rate": 2.614039497236786e-05,
      "loss": 1.9108,
      "num_input_tokens_seen": 40581776,
      "step": 19420
    },
    {
      "epoch": 3.168855534709193,
      "grad_norm": 8.5,
      "learning_rate": 2.6130587627073315e-05,
      "loss": 2.6794,
      "num_input_tokens_seen": 40592656,
      "step": 19425
    },
    {
      "epoch": 3.1696712619300107,
      "grad_norm": 10.8125,
      "learning_rate": 2.6120780107427956e-05,
      "loss": 2.988,
      "num_input_tokens_seen": 40601984,
      "step": 19430
    },
    {
      "epoch": 3.170486989150828,
      "grad_norm": 8.25,
      "learning_rate": 2.6110972414944214e-05,
      "loss": 3.5976,
      "num_input_tokens_seen": 40612368,
      "step": 19435
    },
    {
      "epoch": 3.1713027163716454,
      "grad_norm": 9.75,
      "learning_rate": 2.6101164551134565e-05,
      "loss": 3.5858,
      "num_input_tokens_seen": 40623376,
      "step": 19440
    },
    {
      "epoch": 3.1721184435924625,
      "grad_norm": 4.0625,
      "learning_rate": 2.6091356517511505e-05,
      "loss": 1.5686,
      "num_input_tokens_seen": 40633616,
      "step": 19445
    },
    {
      "epoch": 3.17293417081328,
      "grad_norm": 9.75,
      "learning_rate": 2.608154831558755e-05,
      "loss": 2.89,
      "num_input_tokens_seen": 40644192,
      "step": 19450
    },
    {
      "epoch": 3.1737498980340972,
      "grad_norm": 4.71875,
      "learning_rate": 2.607173994687526e-05,
      "loss": 2.9278,
      "num_input_tokens_seen": 40655584,
      "step": 19455
    },
    {
      "epoch": 3.174565625254915,
      "grad_norm": 8.9375,
      "learning_rate": 2.6061931412887196e-05,
      "loss": 2.379,
      "num_input_tokens_seen": 40664544,
      "step": 19460
    },
    {
      "epoch": 3.175381352475732,
      "grad_norm": 5.21875,
      "learning_rate": 2.6052122715135973e-05,
      "loss": 2.7792,
      "num_input_tokens_seen": 40675968,
      "step": 19465
    },
    {
      "epoch": 3.1761970796965495,
      "grad_norm": 3.03125,
      "learning_rate": 2.60423138551342e-05,
      "loss": 1.3348,
      "num_input_tokens_seen": 40686736,
      "step": 19470
    },
    {
      "epoch": 3.1770128069173666,
      "grad_norm": 8.9375,
      "learning_rate": 2.6032504834394527e-05,
      "loss": 3.083,
      "num_input_tokens_seen": 40696656,
      "step": 19475
    },
    {
      "epoch": 3.1778285341381842,
      "grad_norm": 11.0,
      "learning_rate": 2.602269565442964e-05,
      "loss": 1.9149,
      "num_input_tokens_seen": 40707856,
      "step": 19480
    },
    {
      "epoch": 3.1786442613590014,
      "grad_norm": 10.5,
      "learning_rate": 2.6012886316752227e-05,
      "loss": 2.7468,
      "num_input_tokens_seen": 40718752,
      "step": 19485
    },
    {
      "epoch": 3.179459988579819,
      "grad_norm": 6.65625,
      "learning_rate": 2.6003076822875018e-05,
      "loss": 1.717,
      "num_input_tokens_seen": 40730240,
      "step": 19490
    },
    {
      "epoch": 3.180275715800636,
      "grad_norm": 5.625,
      "learning_rate": 2.5993267174310755e-05,
      "loss": 0.9695,
      "num_input_tokens_seen": 40740416,
      "step": 19495
    },
    {
      "epoch": 3.1810914430214536,
      "grad_norm": 4.96875,
      "learning_rate": 2.5983457372572218e-05,
      "loss": 2.1982,
      "num_input_tokens_seen": 40749968,
      "step": 19500
    },
    {
      "epoch": 3.1819071702422708,
      "grad_norm": 3.734375,
      "learning_rate": 2.597364741917219e-05,
      "loss": 3.3268,
      "num_input_tokens_seen": 40759088,
      "step": 19505
    },
    {
      "epoch": 3.1827228974630883,
      "grad_norm": 2.234375,
      "learning_rate": 2.5963837315623492e-05,
      "loss": 1.4316,
      "num_input_tokens_seen": 40769504,
      "step": 19510
    },
    {
      "epoch": 3.183538624683906,
      "grad_norm": 8.75,
      "learning_rate": 2.595402706343897e-05,
      "loss": 2.6635,
      "num_input_tokens_seen": 40780048,
      "step": 19515
    },
    {
      "epoch": 3.184354351904723,
      "grad_norm": 6.375,
      "learning_rate": 2.594421666413148e-05,
      "loss": 1.2298,
      "num_input_tokens_seen": 40790256,
      "step": 19520
    },
    {
      "epoch": 3.1851700791255406,
      "grad_norm": 5.6875,
      "learning_rate": 2.5934406119213928e-05,
      "loss": 1.7631,
      "num_input_tokens_seen": 40802192,
      "step": 19525
    },
    {
      "epoch": 3.1859858063463578,
      "grad_norm": 10.5,
      "learning_rate": 2.5924595430199193e-05,
      "loss": 3.1036,
      "num_input_tokens_seen": 40812880,
      "step": 19530
    },
    {
      "epoch": 3.1868015335671753,
      "grad_norm": 8.0625,
      "learning_rate": 2.5914784598600238e-05,
      "loss": 3.3569,
      "num_input_tokens_seen": 40824320,
      "step": 19535
    },
    {
      "epoch": 3.1876172607879925,
      "grad_norm": 8.375,
      "learning_rate": 2.5904973625930002e-05,
      "loss": 2.2934,
      "num_input_tokens_seen": 40835888,
      "step": 19540
    },
    {
      "epoch": 3.18843298800881,
      "grad_norm": 5.03125,
      "learning_rate": 2.5895162513701456e-05,
      "loss": 1.9237,
      "num_input_tokens_seen": 40847680,
      "step": 19545
    },
    {
      "epoch": 3.189248715229627,
      "grad_norm": 3.25,
      "learning_rate": 2.5885351263427593e-05,
      "loss": 1.9952,
      "num_input_tokens_seen": 40857424,
      "step": 19550
    },
    {
      "epoch": 3.1900644424504447,
      "grad_norm": 7.09375,
      "learning_rate": 2.5875539876621448e-05,
      "loss": 3.0426,
      "num_input_tokens_seen": 40868864,
      "step": 19555
    },
    {
      "epoch": 3.190880169671262,
      "grad_norm": 6.03125,
      "learning_rate": 2.586572835479605e-05,
      "loss": 2.5765,
      "num_input_tokens_seen": 40878672,
      "step": 19560
    },
    {
      "epoch": 3.1916958968920794,
      "grad_norm": 7.4375,
      "learning_rate": 2.585591669946446e-05,
      "loss": 3.1153,
      "num_input_tokens_seen": 40888256,
      "step": 19565
    },
    {
      "epoch": 3.1925116241128966,
      "grad_norm": 11.5,
      "learning_rate": 2.5846104912139756e-05,
      "loss": 2.7339,
      "num_input_tokens_seen": 40899104,
      "step": 19570
    },
    {
      "epoch": 3.193327351333714,
      "grad_norm": 10.25,
      "learning_rate": 2.583629299433505e-05,
      "loss": 1.8835,
      "num_input_tokens_seen": 40909856,
      "step": 19575
    },
    {
      "epoch": 3.1941430785545313,
      "grad_norm": 4.1875,
      "learning_rate": 2.582648094756345e-05,
      "loss": 2.3613,
      "num_input_tokens_seen": 40920000,
      "step": 19580
    },
    {
      "epoch": 3.194958805775349,
      "grad_norm": 4.71875,
      "learning_rate": 2.5816668773338098e-05,
      "loss": 1.7474,
      "num_input_tokens_seen": 40929168,
      "step": 19585
    },
    {
      "epoch": 3.195774532996166,
      "grad_norm": 6.9375,
      "learning_rate": 2.580685647317216e-05,
      "loss": 2.2862,
      "num_input_tokens_seen": 40939328,
      "step": 19590
    },
    {
      "epoch": 3.1965902602169836,
      "grad_norm": 9.0,
      "learning_rate": 2.5797044048578818e-05,
      "loss": 4.9899,
      "num_input_tokens_seen": 40949664,
      "step": 19595
    },
    {
      "epoch": 3.1974059874378007,
      "grad_norm": 8.5625,
      "learning_rate": 2.5787231501071262e-05,
      "loss": 3.2834,
      "num_input_tokens_seen": 40958928,
      "step": 19600
    },
    {
      "epoch": 3.1974059874378007,
      "eval_loss": 2.5320518016815186,
      "eval_runtime": 134.8504,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 40958928,
      "step": 19600
    },
    {
      "epoch": 3.1982217146586183,
      "grad_norm": 5.375,
      "learning_rate": 2.577741883216272e-05,
      "loss": 2.2181,
      "num_input_tokens_seen": 40969952,
      "step": 19605
    },
    {
      "epoch": 3.1990374418794354,
      "grad_norm": 4.03125,
      "learning_rate": 2.576760604336642e-05,
      "loss": 3.4751,
      "num_input_tokens_seen": 40978208,
      "step": 19610
    },
    {
      "epoch": 3.199853169100253,
      "grad_norm": 14.5,
      "learning_rate": 2.575779313619563e-05,
      "loss": 3.2397,
      "num_input_tokens_seen": 40990016,
      "step": 19615
    },
    {
      "epoch": 3.20066889632107,
      "grad_norm": 7.59375,
      "learning_rate": 2.5747980112163605e-05,
      "loss": 1.2285,
      "num_input_tokens_seen": 41000672,
      "step": 19620
    },
    {
      "epoch": 3.2014846235418877,
      "grad_norm": 7.375,
      "learning_rate": 2.5738166972783656e-05,
      "loss": 2.1506,
      "num_input_tokens_seen": 41011200,
      "step": 19625
    },
    {
      "epoch": 3.202300350762705,
      "grad_norm": 6.15625,
      "learning_rate": 2.5728353719569075e-05,
      "loss": 1.664,
      "num_input_tokens_seen": 41022320,
      "step": 19630
    },
    {
      "epoch": 3.2031160779835224,
      "grad_norm": 6.90625,
      "learning_rate": 2.57185403540332e-05,
      "loss": 1.8824,
      "num_input_tokens_seen": 41031760,
      "step": 19635
    },
    {
      "epoch": 3.2039318052043395,
      "grad_norm": 8.625,
      "learning_rate": 2.5708726877689375e-05,
      "loss": 2.1367,
      "num_input_tokens_seen": 41042448,
      "step": 19640
    },
    {
      "epoch": 3.204747532425157,
      "grad_norm": 4.0,
      "learning_rate": 2.5698913292050964e-05,
      "loss": 2.8153,
      "num_input_tokens_seen": 41051344,
      "step": 19645
    },
    {
      "epoch": 3.205563259645974,
      "grad_norm": 2.578125,
      "learning_rate": 2.568909959863133e-05,
      "loss": 1.9878,
      "num_input_tokens_seen": 41062000,
      "step": 19650
    },
    {
      "epoch": 3.206378986866792,
      "grad_norm": 6.53125,
      "learning_rate": 2.5679285798943887e-05,
      "loss": 2.8459,
      "num_input_tokens_seen": 41070912,
      "step": 19655
    },
    {
      "epoch": 3.207194714087609,
      "grad_norm": 7.0,
      "learning_rate": 2.5669471894502035e-05,
      "loss": 2.5157,
      "num_input_tokens_seen": 41082080,
      "step": 19660
    },
    {
      "epoch": 3.2080104413084265,
      "grad_norm": 3.3125,
      "learning_rate": 2.56596578868192e-05,
      "loss": 1.2955,
      "num_input_tokens_seen": 41092096,
      "step": 19665
    },
    {
      "epoch": 3.2088261685292436,
      "grad_norm": 3.734375,
      "learning_rate": 2.564984377740883e-05,
      "loss": 1.6861,
      "num_input_tokens_seen": 41102816,
      "step": 19670
    },
    {
      "epoch": 3.209641895750061,
      "grad_norm": 7.125,
      "learning_rate": 2.564002956778438e-05,
      "loss": 0.7973,
      "num_input_tokens_seen": 41112032,
      "step": 19675
    },
    {
      "epoch": 3.2104576229708783,
      "grad_norm": 7.53125,
      "learning_rate": 2.563021525945934e-05,
      "loss": 1.7472,
      "num_input_tokens_seen": 41123952,
      "step": 19680
    },
    {
      "epoch": 3.211273350191696,
      "grad_norm": 7.53125,
      "learning_rate": 2.562040085394718e-05,
      "loss": 2.7136,
      "num_input_tokens_seen": 41133648,
      "step": 19685
    },
    {
      "epoch": 3.2120890774125135,
      "grad_norm": 12.375,
      "learning_rate": 2.56105863527614e-05,
      "loss": 3.2519,
      "num_input_tokens_seen": 41146128,
      "step": 19690
    },
    {
      "epoch": 3.2129048046333306,
      "grad_norm": 9.125,
      "learning_rate": 2.5600771757415548e-05,
      "loss": 2.1781,
      "num_input_tokens_seen": 41156144,
      "step": 19695
    },
    {
      "epoch": 3.2137205318541477,
      "grad_norm": 7.25,
      "learning_rate": 2.5590957069423134e-05,
      "loss": 2.0801,
      "num_input_tokens_seen": 41165552,
      "step": 19700
    },
    {
      "epoch": 3.2145362590749653,
      "grad_norm": 9.5,
      "learning_rate": 2.5581142290297716e-05,
      "loss": 1.9684,
      "num_input_tokens_seen": 41175968,
      "step": 19705
    },
    {
      "epoch": 3.215351986295783,
      "grad_norm": 6.59375,
      "learning_rate": 2.557132742155285e-05,
      "loss": 1.7726,
      "num_input_tokens_seen": 41187264,
      "step": 19710
    },
    {
      "epoch": 3.2161677135166,
      "grad_norm": 11.4375,
      "learning_rate": 2.556151246470212e-05,
      "loss": 2.1975,
      "num_input_tokens_seen": 41197616,
      "step": 19715
    },
    {
      "epoch": 3.2169834407374176,
      "grad_norm": 11.375,
      "learning_rate": 2.5551697421259114e-05,
      "loss": 2.1005,
      "num_input_tokens_seen": 41208896,
      "step": 19720
    },
    {
      "epoch": 3.2177991679582347,
      "grad_norm": 0.458984375,
      "learning_rate": 2.554188229273743e-05,
      "loss": 2.2241,
      "num_input_tokens_seen": 41219552,
      "step": 19725
    },
    {
      "epoch": 3.2186148951790523,
      "grad_norm": 10.1875,
      "learning_rate": 2.5532067080650678e-05,
      "loss": 2.104,
      "num_input_tokens_seen": 41230448,
      "step": 19730
    },
    {
      "epoch": 3.2194306223998694,
      "grad_norm": 7.0625,
      "learning_rate": 2.55222517865125e-05,
      "loss": 1.8107,
      "num_input_tokens_seen": 41241104,
      "step": 19735
    },
    {
      "epoch": 3.220246349620687,
      "grad_norm": 2.84375,
      "learning_rate": 2.5512436411836538e-05,
      "loss": 3.1106,
      "num_input_tokens_seen": 41253232,
      "step": 19740
    },
    {
      "epoch": 3.221062076841504,
      "grad_norm": 3.5,
      "learning_rate": 2.5502620958136443e-05,
      "loss": 1.2471,
      "num_input_tokens_seen": 41264640,
      "step": 19745
    },
    {
      "epoch": 3.2218778040623217,
      "grad_norm": 11.125,
      "learning_rate": 2.5492805426925874e-05,
      "loss": 2.4015,
      "num_input_tokens_seen": 41275200,
      "step": 19750
    },
    {
      "epoch": 3.222693531283139,
      "grad_norm": 5.4375,
      "learning_rate": 2.5482989819718523e-05,
      "loss": 2.1809,
      "num_input_tokens_seen": 41285568,
      "step": 19755
    },
    {
      "epoch": 3.2235092585039564,
      "grad_norm": 1.6328125,
      "learning_rate": 2.5473174138028065e-05,
      "loss": 2.5299,
      "num_input_tokens_seen": 41295600,
      "step": 19760
    },
    {
      "epoch": 3.2243249857247736,
      "grad_norm": 5.03125,
      "learning_rate": 2.5463358383368212e-05,
      "loss": 1.6698,
      "num_input_tokens_seen": 41304880,
      "step": 19765
    },
    {
      "epoch": 3.225140712945591,
      "grad_norm": 2.5,
      "learning_rate": 2.545354255725267e-05,
      "loss": 2.4432,
      "num_input_tokens_seen": 41313616,
      "step": 19770
    },
    {
      "epoch": 3.2259564401664083,
      "grad_norm": 10.5625,
      "learning_rate": 2.5443726661195165e-05,
      "loss": 3.9232,
      "num_input_tokens_seen": 41324176,
      "step": 19775
    },
    {
      "epoch": 3.226772167387226,
      "grad_norm": 6.40625,
      "learning_rate": 2.543391069670944e-05,
      "loss": 4.0166,
      "num_input_tokens_seen": 41335824,
      "step": 19780
    },
    {
      "epoch": 3.227587894608043,
      "grad_norm": 4.75,
      "learning_rate": 2.5424094665309228e-05,
      "loss": 1.6907,
      "num_input_tokens_seen": 41346272,
      "step": 19785
    },
    {
      "epoch": 3.2284036218288605,
      "grad_norm": 3.78125,
      "learning_rate": 2.5414278568508292e-05,
      "loss": 2.3165,
      "num_input_tokens_seen": 41356880,
      "step": 19790
    },
    {
      "epoch": 3.2292193490496777,
      "grad_norm": 7.65625,
      "learning_rate": 2.540446240782039e-05,
      "loss": 1.7776,
      "num_input_tokens_seen": 41366608,
      "step": 19795
    },
    {
      "epoch": 3.2300350762704952,
      "grad_norm": 9.5,
      "learning_rate": 2.5394646184759307e-05,
      "loss": 2.2189,
      "num_input_tokens_seen": 41378128,
      "step": 19800
    },
    {
      "epoch": 3.2300350762704952,
      "eval_loss": 2.543567419052124,
      "eval_runtime": 134.8524,
      "eval_samples_per_second": 20.207,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 41378128,
      "step": 19800
    },
    {
      "epoch": 3.2308508034913124,
      "grad_norm": 6.75,
      "learning_rate": 2.538482990083882e-05,
      "loss": 1.7004,
      "num_input_tokens_seen": 41387392,
      "step": 19805
    },
    {
      "epoch": 3.23166653071213,
      "grad_norm": 7.4375,
      "learning_rate": 2.5375013557572725e-05,
      "loss": 2.3691,
      "num_input_tokens_seen": 41397776,
      "step": 19810
    },
    {
      "epoch": 3.232482257932947,
      "grad_norm": 5.09375,
      "learning_rate": 2.536519715647483e-05,
      "loss": 2.3283,
      "num_input_tokens_seen": 41408080,
      "step": 19815
    },
    {
      "epoch": 3.2332979851537647,
      "grad_norm": 6.15625,
      "learning_rate": 2.535538069905894e-05,
      "loss": 2.4024,
      "num_input_tokens_seen": 41418032,
      "step": 19820
    },
    {
      "epoch": 3.234113712374582,
      "grad_norm": 7.09375,
      "learning_rate": 2.534556418683888e-05,
      "loss": 1.686,
      "num_input_tokens_seen": 41428048,
      "step": 19825
    },
    {
      "epoch": 3.2349294395953994,
      "grad_norm": 4.6875,
      "learning_rate": 2.5335747621328486e-05,
      "loss": 1.852,
      "num_input_tokens_seen": 41438704,
      "step": 19830
    },
    {
      "epoch": 3.2357451668162165,
      "grad_norm": 7.09375,
      "learning_rate": 2.5325931004041586e-05,
      "loss": 2.7015,
      "num_input_tokens_seen": 41449008,
      "step": 19835
    },
    {
      "epoch": 3.236560894037034,
      "grad_norm": 9.3125,
      "learning_rate": 2.5316114336492032e-05,
      "loss": 2.7264,
      "num_input_tokens_seen": 41460176,
      "step": 19840
    },
    {
      "epoch": 3.237376621257851,
      "grad_norm": 6.5625,
      "learning_rate": 2.530629762019367e-05,
      "loss": 2.3789,
      "num_input_tokens_seen": 41471168,
      "step": 19845
    },
    {
      "epoch": 3.2381923484786688,
      "grad_norm": 7.4375,
      "learning_rate": 2.5296480856660364e-05,
      "loss": 2.2269,
      "num_input_tokens_seen": 41482192,
      "step": 19850
    },
    {
      "epoch": 3.239008075699486,
      "grad_norm": 6.40625,
      "learning_rate": 2.528666404740599e-05,
      "loss": 3.8866,
      "num_input_tokens_seen": 41492384,
      "step": 19855
    },
    {
      "epoch": 3.2398238029203035,
      "grad_norm": 0.197265625,
      "learning_rate": 2.527684719394442e-05,
      "loss": 1.754,
      "num_input_tokens_seen": 41500896,
      "step": 19860
    },
    {
      "epoch": 3.2406395301411206,
      "grad_norm": 7.21875,
      "learning_rate": 2.526703029778953e-05,
      "loss": 1.5633,
      "num_input_tokens_seen": 41511920,
      "step": 19865
    },
    {
      "epoch": 3.241455257361938,
      "grad_norm": 0.07470703125,
      "learning_rate": 2.5257213360455208e-05,
      "loss": 2.5769,
      "num_input_tokens_seen": 41522720,
      "step": 19870
    },
    {
      "epoch": 3.2422709845827553,
      "grad_norm": 13.8125,
      "learning_rate": 2.5247396383455353e-05,
      "loss": 4.4329,
      "num_input_tokens_seen": 41533840,
      "step": 19875
    },
    {
      "epoch": 3.243086711803573,
      "grad_norm": 6.34375,
      "learning_rate": 2.523757936830387e-05,
      "loss": 2.8715,
      "num_input_tokens_seen": 41544368,
      "step": 19880
    },
    {
      "epoch": 3.2439024390243905,
      "grad_norm": 13.5,
      "learning_rate": 2.5227762316514662e-05,
      "loss": 4.3807,
      "num_input_tokens_seen": 41555952,
      "step": 19885
    },
    {
      "epoch": 3.2447181662452076,
      "grad_norm": 9.6875,
      "learning_rate": 2.5217945229601648e-05,
      "loss": 2.6528,
      "num_input_tokens_seen": 41566096,
      "step": 19890
    },
    {
      "epoch": 3.245533893466025,
      "grad_norm": 7.3125,
      "learning_rate": 2.5208128109078738e-05,
      "loss": 2.2239,
      "num_input_tokens_seen": 41576928,
      "step": 19895
    },
    {
      "epoch": 3.2463496206868423,
      "grad_norm": 7.09375,
      "learning_rate": 2.5198310956459853e-05,
      "loss": 4.3729,
      "num_input_tokens_seen": 41588112,
      "step": 19900
    },
    {
      "epoch": 3.24716534790766,
      "grad_norm": 6.53125,
      "learning_rate": 2.518849377325893e-05,
      "loss": 2.3434,
      "num_input_tokens_seen": 41597744,
      "step": 19905
    },
    {
      "epoch": 3.247981075128477,
      "grad_norm": 10.625,
      "learning_rate": 2.51786765609899e-05,
      "loss": 1.9911,
      "num_input_tokens_seen": 41609232,
      "step": 19910
    },
    {
      "epoch": 3.2487968023492946,
      "grad_norm": 8.5,
      "learning_rate": 2.5168859321166694e-05,
      "loss": 3.9234,
      "num_input_tokens_seen": 41619408,
      "step": 19915
    },
    {
      "epoch": 3.2496125295701117,
      "grad_norm": 5.40625,
      "learning_rate": 2.515904205530326e-05,
      "loss": 3.744,
      "num_input_tokens_seen": 41630112,
      "step": 19920
    },
    {
      "epoch": 3.2504282567909293,
      "grad_norm": 3.484375,
      "learning_rate": 2.514922476491355e-05,
      "loss": 2.2678,
      "num_input_tokens_seen": 41639920,
      "step": 19925
    },
    {
      "epoch": 3.2512439840117464,
      "grad_norm": 6.25,
      "learning_rate": 2.51394074515115e-05,
      "loss": 2.1572,
      "num_input_tokens_seen": 41650608,
      "step": 19930
    },
    {
      "epoch": 3.252059711232564,
      "grad_norm": 5.3125,
      "learning_rate": 2.5129590116611067e-05,
      "loss": 3.117,
      "num_input_tokens_seen": 41661808,
      "step": 19935
    },
    {
      "epoch": 3.252875438453381,
      "grad_norm": 6.09375,
      "learning_rate": 2.5119772761726212e-05,
      "loss": 1.6687,
      "num_input_tokens_seen": 41672320,
      "step": 19940
    },
    {
      "epoch": 3.2536911656741987,
      "grad_norm": 6.65625,
      "learning_rate": 2.5109955388370893e-05,
      "loss": 3.0645,
      "num_input_tokens_seen": 41683392,
      "step": 19945
    },
    {
      "epoch": 3.254506892895016,
      "grad_norm": 1.9453125,
      "learning_rate": 2.510013799805907e-05,
      "loss": 3.329,
      "num_input_tokens_seen": 41692528,
      "step": 19950
    },
    {
      "epoch": 3.2553226201158334,
      "grad_norm": 8.1875,
      "learning_rate": 2.5090320592304706e-05,
      "loss": 1.9377,
      "num_input_tokens_seen": 41701984,
      "step": 19955
    },
    {
      "epoch": 3.2561383473366505,
      "grad_norm": 5.3125,
      "learning_rate": 2.5080503172621777e-05,
      "loss": 1.776,
      "num_input_tokens_seen": 41711760,
      "step": 19960
    },
    {
      "epoch": 3.256954074557468,
      "grad_norm": 8.5,
      "learning_rate": 2.5070685740524246e-05,
      "loss": 3.2439,
      "num_input_tokens_seen": 41723376,
      "step": 19965
    },
    {
      "epoch": 3.2577698017782852,
      "grad_norm": 15.0,
      "learning_rate": 2.5060868297526084e-05,
      "loss": 1.657,
      "num_input_tokens_seen": 41733392,
      "step": 19970
    },
    {
      "epoch": 3.258585528999103,
      "grad_norm": 10.4375,
      "learning_rate": 2.5051050845141267e-05,
      "loss": 3.1965,
      "num_input_tokens_seen": 41744704,
      "step": 19975
    },
    {
      "epoch": 3.25940125621992,
      "grad_norm": 8.5625,
      "learning_rate": 2.5041233384883765e-05,
      "loss": 2.1927,
      "num_input_tokens_seen": 41754176,
      "step": 19980
    },
    {
      "epoch": 3.2602169834407375,
      "grad_norm": 11.0625,
      "learning_rate": 2.5031415918267564e-05,
      "loss": 1.7513,
      "num_input_tokens_seen": 41763776,
      "step": 19985
    },
    {
      "epoch": 3.2610327106615546,
      "grad_norm": 7.90625,
      "learning_rate": 2.5021598446806626e-05,
      "loss": 2.6045,
      "num_input_tokens_seen": 41775120,
      "step": 19990
    },
    {
      "epoch": 3.261848437882372,
      "grad_norm": 15.1875,
      "learning_rate": 2.5011780972014937e-05,
      "loss": 1.1522,
      "num_input_tokens_seen": 41783792,
      "step": 19995
    },
    {
      "epoch": 3.2626641651031894,
      "grad_norm": 4.5,
      "learning_rate": 2.5001963495406478e-05,
      "loss": 3.4322,
      "num_input_tokens_seen": 41794480,
      "step": 20000
    },
    {
      "epoch": 3.2626641651031894,
      "eval_loss": 2.537410259246826,
      "eval_runtime": 135.0333,
      "eval_samples_per_second": 20.18,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 41794480,
      "step": 20000
    },
    {
      "epoch": 3.263479892324007,
      "grad_norm": 6.84375,
      "learning_rate": 2.499214601849522e-05,
      "loss": 2.1072,
      "num_input_tokens_seen": 41804528,
      "step": 20005
    },
    {
      "epoch": 3.264295619544824,
      "grad_norm": 6.34375,
      "learning_rate": 2.4982328542795148e-05,
      "loss": 3.2755,
      "num_input_tokens_seen": 41814624,
      "step": 20010
    },
    {
      "epoch": 3.2651113467656416,
      "grad_norm": 19.25,
      "learning_rate": 2.497251106982024e-05,
      "loss": 2.9295,
      "num_input_tokens_seen": 41825856,
      "step": 20015
    },
    {
      "epoch": 3.2659270739864588,
      "grad_norm": 9.3125,
      "learning_rate": 2.4962693601084458e-05,
      "loss": 2.649,
      "num_input_tokens_seen": 41834864,
      "step": 20020
    },
    {
      "epoch": 3.2667428012072763,
      "grad_norm": 5.8125,
      "learning_rate": 2.4952876138101794e-05,
      "loss": 3.0694,
      "num_input_tokens_seen": 41845344,
      "step": 20025
    },
    {
      "epoch": 3.2675585284280935,
      "grad_norm": 0.267578125,
      "learning_rate": 2.4943058682386233e-05,
      "loss": 1.7387,
      "num_input_tokens_seen": 41855808,
      "step": 20030
    },
    {
      "epoch": 3.268374255648911,
      "grad_norm": 3.5,
      "learning_rate": 2.493324123545173e-05,
      "loss": 1.1589,
      "num_input_tokens_seen": 41866176,
      "step": 20035
    },
    {
      "epoch": 3.269189982869728,
      "grad_norm": 11.125,
      "learning_rate": 2.4923423798812272e-05,
      "loss": 2.8565,
      "num_input_tokens_seen": 41878032,
      "step": 20040
    },
    {
      "epoch": 3.2700057100905457,
      "grad_norm": 6.84375,
      "learning_rate": 2.4913606373981825e-05,
      "loss": 2.7191,
      "num_input_tokens_seen": 41889280,
      "step": 20045
    },
    {
      "epoch": 3.270821437311363,
      "grad_norm": 4.21875,
      "learning_rate": 2.4903788962474357e-05,
      "loss": 2.2296,
      "num_input_tokens_seen": 41898304,
      "step": 20050
    },
    {
      "epoch": 3.2716371645321805,
      "grad_norm": 8.0,
      "learning_rate": 2.489397156580385e-05,
      "loss": 2.8063,
      "num_input_tokens_seen": 41907136,
      "step": 20055
    },
    {
      "epoch": 3.272452891752998,
      "grad_norm": 3.078125,
      "learning_rate": 2.4884154185484246e-05,
      "loss": 2.6455,
      "num_input_tokens_seen": 41917536,
      "step": 20060
    },
    {
      "epoch": 3.273268618973815,
      "grad_norm": 4.4375,
      "learning_rate": 2.4874336823029526e-05,
      "loss": 2.4361,
      "num_input_tokens_seen": 41927856,
      "step": 20065
    },
    {
      "epoch": 3.2740843461946323,
      "grad_norm": 6.90625,
      "learning_rate": 2.4864519479953656e-05,
      "loss": 2.8108,
      "num_input_tokens_seen": 41937584,
      "step": 20070
    },
    {
      "epoch": 3.27490007341545,
      "grad_norm": 5.8125,
      "learning_rate": 2.485470215777058e-05,
      "loss": 2.1643,
      "num_input_tokens_seen": 41946448,
      "step": 20075
    },
    {
      "epoch": 3.2757158006362674,
      "grad_norm": 3.90625,
      "learning_rate": 2.4844884857994258e-05,
      "loss": 3.3462,
      "num_input_tokens_seen": 41957904,
      "step": 20080
    },
    {
      "epoch": 3.2765315278570846,
      "grad_norm": 4.5,
      "learning_rate": 2.4835067582138638e-05,
      "loss": 1.7505,
      "num_input_tokens_seen": 41968816,
      "step": 20085
    },
    {
      "epoch": 3.2773472550779017,
      "grad_norm": 7.53125,
      "learning_rate": 2.4825250331717666e-05,
      "loss": 2.5283,
      "num_input_tokens_seen": 41979344,
      "step": 20090
    },
    {
      "epoch": 3.2781629822987193,
      "grad_norm": 10.1875,
      "learning_rate": 2.4815433108245298e-05,
      "loss": 2.3034,
      "num_input_tokens_seen": 41990896,
      "step": 20095
    },
    {
      "epoch": 3.278978709519537,
      "grad_norm": 6.03125,
      "learning_rate": 2.4805615913235456e-05,
      "loss": 2.6634,
      "num_input_tokens_seen": 42000880,
      "step": 20100
    },
    {
      "epoch": 3.279794436740354,
      "grad_norm": 1.890625,
      "learning_rate": 2.479579874820208e-05,
      "loss": 1.5421,
      "num_input_tokens_seen": 42012592,
      "step": 20105
    },
    {
      "epoch": 3.2806101639611716,
      "grad_norm": 9.9375,
      "learning_rate": 2.4785981614659115e-05,
      "loss": 1.7741,
      "num_input_tokens_seen": 42022608,
      "step": 20110
    },
    {
      "epoch": 3.2814258911819887,
      "grad_norm": 9.1875,
      "learning_rate": 2.477616451412047e-05,
      "loss": 3.941,
      "num_input_tokens_seen": 42034208,
      "step": 20115
    },
    {
      "epoch": 3.2822416184028063,
      "grad_norm": 7.0625,
      "learning_rate": 2.476634744810007e-05,
      "loss": 4.598,
      "num_input_tokens_seen": 42043744,
      "step": 20120
    },
    {
      "epoch": 3.2830573456236234,
      "grad_norm": 5.25,
      "learning_rate": 2.475653041811183e-05,
      "loss": 2.1986,
      "num_input_tokens_seen": 42054272,
      "step": 20125
    },
    {
      "epoch": 3.283873072844441,
      "grad_norm": 10.6875,
      "learning_rate": 2.4746713425669652e-05,
      "loss": 3.1769,
      "num_input_tokens_seen": 42065472,
      "step": 20130
    },
    {
      "epoch": 3.284688800065258,
      "grad_norm": 3.75,
      "learning_rate": 2.4736896472287458e-05,
      "loss": 3.018,
      "num_input_tokens_seen": 42075008,
      "step": 20135
    },
    {
      "epoch": 3.2855045272860757,
      "grad_norm": 11.5625,
      "learning_rate": 2.4727079559479124e-05,
      "loss": 3.1353,
      "num_input_tokens_seen": 42084736,
      "step": 20140
    },
    {
      "epoch": 3.286320254506893,
      "grad_norm": 8.9375,
      "learning_rate": 2.4717262688758557e-05,
      "loss": 3.8477,
      "num_input_tokens_seen": 42095104,
      "step": 20145
    },
    {
      "epoch": 3.2871359817277104,
      "grad_norm": 5.09375,
      "learning_rate": 2.4707445861639637e-05,
      "loss": 3.0957,
      "num_input_tokens_seen": 42104928,
      "step": 20150
    },
    {
      "epoch": 3.2879517089485275,
      "grad_norm": 1.7109375,
      "learning_rate": 2.4697629079636244e-05,
      "loss": 2.6156,
      "num_input_tokens_seen": 42114704,
      "step": 20155
    },
    {
      "epoch": 3.288767436169345,
      "grad_norm": 9.125,
      "learning_rate": 2.4687812344262244e-05,
      "loss": 4.717,
      "num_input_tokens_seen": 42124720,
      "step": 20160
    },
    {
      "epoch": 3.289583163390162,
      "grad_norm": 3.296875,
      "learning_rate": 2.46779956570315e-05,
      "loss": 2.7845,
      "num_input_tokens_seen": 42134256,
      "step": 20165
    },
    {
      "epoch": 3.29039889061098,
      "grad_norm": 11.0625,
      "learning_rate": 2.466817901945787e-05,
      "loss": 2.5101,
      "num_input_tokens_seen": 42144736,
      "step": 20170
    },
    {
      "epoch": 3.291214617831797,
      "grad_norm": 5.125,
      "learning_rate": 2.4658362433055217e-05,
      "loss": 2.7502,
      "num_input_tokens_seen": 42155008,
      "step": 20175
    },
    {
      "epoch": 3.2920303450526145,
      "grad_norm": 10.6875,
      "learning_rate": 2.4648545899337356e-05,
      "loss": 3.3808,
      "num_input_tokens_seen": 42164896,
      "step": 20180
    },
    {
      "epoch": 3.2928460722734316,
      "grad_norm": 5.125,
      "learning_rate": 2.4638729419818143e-05,
      "loss": 2.2589,
      "num_input_tokens_seen": 42175216,
      "step": 20185
    },
    {
      "epoch": 3.293661799494249,
      "grad_norm": 3.5,
      "learning_rate": 2.46289129960114e-05,
      "loss": 1.9207,
      "num_input_tokens_seen": 42186368,
      "step": 20190
    },
    {
      "epoch": 3.2944775267150663,
      "grad_norm": 4.4375,
      "learning_rate": 2.4619096629430924e-05,
      "loss": 1.8192,
      "num_input_tokens_seen": 42198400,
      "step": 20195
    },
    {
      "epoch": 3.295293253935884,
      "grad_norm": 3.828125,
      "learning_rate": 2.4609280321590543e-05,
      "loss": 3.3739,
      "num_input_tokens_seen": 42208144,
      "step": 20200
    },
    {
      "epoch": 3.295293253935884,
      "eval_loss": 2.5433285236358643,
      "eval_runtime": 134.8846,
      "eval_samples_per_second": 20.202,
      "eval_steps_per_second": 10.105,
      "num_input_tokens_seen": 42208144,
      "step": 20200
    },
    {
      "epoch": 3.296108981156701,
      "grad_norm": 1.6953125,
      "learning_rate": 2.4599464074004037e-05,
      "loss": 1.9925,
      "num_input_tokens_seen": 42219360,
      "step": 20205
    },
    {
      "epoch": 3.2969247083775186,
      "grad_norm": 3.859375,
      "learning_rate": 2.4589647888185204e-05,
      "loss": 1.2414,
      "num_input_tokens_seen": 42229456,
      "step": 20210
    },
    {
      "epoch": 3.2977404355983357,
      "grad_norm": 10.75,
      "learning_rate": 2.4579831765647836e-05,
      "loss": 2.4535,
      "num_input_tokens_seen": 42240688,
      "step": 20215
    },
    {
      "epoch": 3.2985561628191533,
      "grad_norm": 7.25,
      "learning_rate": 2.4570015707905676e-05,
      "loss": 1.4684,
      "num_input_tokens_seen": 42249680,
      "step": 20220
    },
    {
      "epoch": 3.2993718900399704,
      "grad_norm": 9.8125,
      "learning_rate": 2.4560199716472508e-05,
      "loss": 2.7598,
      "num_input_tokens_seen": 42259936,
      "step": 20225
    },
    {
      "epoch": 3.300187617260788,
      "grad_norm": 8.75,
      "learning_rate": 2.455038379286207e-05,
      "loss": 1.9112,
      "num_input_tokens_seen": 42271776,
      "step": 20230
    },
    {
      "epoch": 3.3010033444816056,
      "grad_norm": 11.5,
      "learning_rate": 2.4540567938588095e-05,
      "loss": 1.1607,
      "num_input_tokens_seen": 42283536,
      "step": 20235
    },
    {
      "epoch": 3.3018190717024227,
      "grad_norm": 6.0,
      "learning_rate": 2.4530752155164328e-05,
      "loss": 2.773,
      "num_input_tokens_seen": 42294464,
      "step": 20240
    },
    {
      "epoch": 3.30263479892324,
      "grad_norm": 9.375,
      "learning_rate": 2.4520936444104463e-05,
      "loss": 2.0663,
      "num_input_tokens_seen": 42304464,
      "step": 20245
    },
    {
      "epoch": 3.3034505261440574,
      "grad_norm": 18.375,
      "learning_rate": 2.4511120806922218e-05,
      "loss": 3.7125,
      "num_input_tokens_seen": 42316048,
      "step": 20250
    },
    {
      "epoch": 3.304266253364875,
      "grad_norm": 9.0625,
      "learning_rate": 2.45013052451313e-05,
      "loss": 3.4228,
      "num_input_tokens_seen": 42327616,
      "step": 20255
    },
    {
      "epoch": 3.305081980585692,
      "grad_norm": 6.5,
      "learning_rate": 2.4491489760245376e-05,
      "loss": 1.983,
      "num_input_tokens_seen": 42336704,
      "step": 20260
    },
    {
      "epoch": 3.3058977078065093,
      "grad_norm": 6.96875,
      "learning_rate": 2.4481674353778115e-05,
      "loss": 2.5131,
      "num_input_tokens_seen": 42346672,
      "step": 20265
    },
    {
      "epoch": 3.306713435027327,
      "grad_norm": 8.375,
      "learning_rate": 2.447185902724319e-05,
      "loss": 3.5484,
      "num_input_tokens_seen": 42356848,
      "step": 20270
    },
    {
      "epoch": 3.3075291622481444,
      "grad_norm": 0.87109375,
      "learning_rate": 2.4462043782154233e-05,
      "loss": 1.6148,
      "num_input_tokens_seen": 42367392,
      "step": 20275
    },
    {
      "epoch": 3.3083448894689615,
      "grad_norm": 4.78125,
      "learning_rate": 2.4452228620024895e-05,
      "loss": 2.4417,
      "num_input_tokens_seen": 42377856,
      "step": 20280
    },
    {
      "epoch": 3.309160616689779,
      "grad_norm": 5.8125,
      "learning_rate": 2.4442413542368776e-05,
      "loss": 2.4401,
      "num_input_tokens_seen": 42387808,
      "step": 20285
    },
    {
      "epoch": 3.3099763439105963,
      "grad_norm": 5.09375,
      "learning_rate": 2.4432598550699502e-05,
      "loss": 1.7985,
      "num_input_tokens_seen": 42399008,
      "step": 20290
    },
    {
      "epoch": 3.310792071131414,
      "grad_norm": 8.1875,
      "learning_rate": 2.4422783646530663e-05,
      "loss": 2.7012,
      "num_input_tokens_seen": 42409408,
      "step": 20295
    },
    {
      "epoch": 3.311607798352231,
      "grad_norm": 12.1875,
      "learning_rate": 2.441296883137584e-05,
      "loss": 4.0959,
      "num_input_tokens_seen": 42420736,
      "step": 20300
    },
    {
      "epoch": 3.3124235255730485,
      "grad_norm": 2.03125,
      "learning_rate": 2.4403154106748592e-05,
      "loss": 1.7947,
      "num_input_tokens_seen": 42430352,
      "step": 20305
    },
    {
      "epoch": 3.3132392527938657,
      "grad_norm": 6.15625,
      "learning_rate": 2.4393339474162494e-05,
      "loss": 1.0776,
      "num_input_tokens_seen": 42440464,
      "step": 20310
    },
    {
      "epoch": 3.3140549800146832,
      "grad_norm": 9.0,
      "learning_rate": 2.4383524935131062e-05,
      "loss": 2.5704,
      "num_input_tokens_seen": 42450896,
      "step": 20315
    },
    {
      "epoch": 3.3148707072355004,
      "grad_norm": 2.25,
      "learning_rate": 2.437371049116784e-05,
      "loss": 1.1739,
      "num_input_tokens_seen": 42461856,
      "step": 20320
    },
    {
      "epoch": 3.315686434456318,
      "grad_norm": 6.03125,
      "learning_rate": 2.436389614378632e-05,
      "loss": 1.6025,
      "num_input_tokens_seen": 42472768,
      "step": 20325
    },
    {
      "epoch": 3.316502161677135,
      "grad_norm": 3.359375,
      "learning_rate": 2.435408189450002e-05,
      "loss": 2.0135,
      "num_input_tokens_seen": 42482928,
      "step": 20330
    },
    {
      "epoch": 3.3173178888979526,
      "grad_norm": 1.8359375,
      "learning_rate": 2.4344267744822406e-05,
      "loss": 1.7818,
      "num_input_tokens_seen": 42493696,
      "step": 20335
    },
    {
      "epoch": 3.31813361611877,
      "grad_norm": 8.0,
      "learning_rate": 2.4334453696266944e-05,
      "loss": 2.3737,
      "num_input_tokens_seen": 42503920,
      "step": 20340
    },
    {
      "epoch": 3.3189493433395874,
      "grad_norm": 4.46875,
      "learning_rate": 2.432463975034708e-05,
      "loss": 1.9545,
      "num_input_tokens_seen": 42514560,
      "step": 20345
    },
    {
      "epoch": 3.3197650705604045,
      "grad_norm": 9.3125,
      "learning_rate": 2.4314825908576265e-05,
      "loss": 2.9223,
      "num_input_tokens_seen": 42524272,
      "step": 20350
    },
    {
      "epoch": 3.320580797781222,
      "grad_norm": 6.53125,
      "learning_rate": 2.4305012172467897e-05,
      "loss": 2.3135,
      "num_input_tokens_seen": 42534704,
      "step": 20355
    },
    {
      "epoch": 3.321396525002039,
      "grad_norm": 6.125,
      "learning_rate": 2.4295198543535393e-05,
      "loss": 1.9922,
      "num_input_tokens_seen": 42545616,
      "step": 20360
    },
    {
      "epoch": 3.3222122522228568,
      "grad_norm": 6.09375,
      "learning_rate": 2.4285385023292124e-05,
      "loss": 2.089,
      "num_input_tokens_seen": 42555136,
      "step": 20365
    },
    {
      "epoch": 3.323027979443674,
      "grad_norm": 4.125,
      "learning_rate": 2.427557161325147e-05,
      "loss": 2.8918,
      "num_input_tokens_seen": 42565456,
      "step": 20370
    },
    {
      "epoch": 3.3238437066644915,
      "grad_norm": 4.0625,
      "learning_rate": 2.4265758314926778e-05,
      "loss": 1.4327,
      "num_input_tokens_seen": 42577072,
      "step": 20375
    },
    {
      "epoch": 3.3246594338853086,
      "grad_norm": 6.40625,
      "learning_rate": 2.4255945129831373e-05,
      "loss": 1.5703,
      "num_input_tokens_seen": 42587424,
      "step": 20380
    },
    {
      "epoch": 3.325475161106126,
      "grad_norm": 6.9375,
      "learning_rate": 2.4246132059478578e-05,
      "loss": 2.4988,
      "num_input_tokens_seen": 42597888,
      "step": 20385
    },
    {
      "epoch": 3.3262908883269433,
      "grad_norm": 6.03125,
      "learning_rate": 2.4236319105381706e-05,
      "loss": 1.9234,
      "num_input_tokens_seen": 42606192,
      "step": 20390
    },
    {
      "epoch": 3.327106615547761,
      "grad_norm": 1.796875,
      "learning_rate": 2.422650626905401e-05,
      "loss": 2.1471,
      "num_input_tokens_seen": 42616096,
      "step": 20395
    },
    {
      "epoch": 3.327922342768578,
      "grad_norm": 5.59375,
      "learning_rate": 2.4216693552008785e-05,
      "loss": 2.6455,
      "num_input_tokens_seen": 42625520,
      "step": 20400
    },
    {
      "epoch": 3.327922342768578,
      "eval_loss": 2.5354511737823486,
      "eval_runtime": 135.0546,
      "eval_samples_per_second": 20.177,
      "eval_steps_per_second": 10.092,
      "num_input_tokens_seen": 42625520,
      "step": 20400
    },
    {
      "epoch": 3.3287380699893956,
      "grad_norm": 3.984375,
      "learning_rate": 2.4206880955759247e-05,
      "loss": 2.0619,
      "num_input_tokens_seen": 42636448,
      "step": 20405
    },
    {
      "epoch": 3.3295537972102127,
      "grad_norm": 13.25,
      "learning_rate": 2.419706848181863e-05,
      "loss": 2.969,
      "num_input_tokens_seen": 42647328,
      "step": 20410
    },
    {
      "epoch": 3.3303695244310303,
      "grad_norm": 9.3125,
      "learning_rate": 2.4187256131700153e-05,
      "loss": 3.5411,
      "num_input_tokens_seen": 42657552,
      "step": 20415
    },
    {
      "epoch": 3.3311852516518474,
      "grad_norm": 5.96875,
      "learning_rate": 2.4177443906916985e-05,
      "loss": 2.1037,
      "num_input_tokens_seen": 42668800,
      "step": 20420
    },
    {
      "epoch": 3.332000978872665,
      "grad_norm": 10.5625,
      "learning_rate": 2.4167631808982303e-05,
      "loss": 2.661,
      "num_input_tokens_seen": 42679600,
      "step": 20425
    },
    {
      "epoch": 3.3328167060934826,
      "grad_norm": 11.625,
      "learning_rate": 2.4157819839409264e-05,
      "loss": 2.026,
      "num_input_tokens_seen": 42690832,
      "step": 20430
    },
    {
      "epoch": 3.3336324333142997,
      "grad_norm": 5.75,
      "learning_rate": 2.414800799971098e-05,
      "loss": 2.5018,
      "num_input_tokens_seen": 42700912,
      "step": 20435
    },
    {
      "epoch": 3.334448160535117,
      "grad_norm": 10.875,
      "learning_rate": 2.4138196291400582e-05,
      "loss": 1.784,
      "num_input_tokens_seen": 42712016,
      "step": 20440
    },
    {
      "epoch": 3.3352638877559344,
      "grad_norm": 13.1875,
      "learning_rate": 2.412838471599114e-05,
      "loss": 3.6619,
      "num_input_tokens_seen": 42722336,
      "step": 20445
    },
    {
      "epoch": 3.336079614976752,
      "grad_norm": 12.3125,
      "learning_rate": 2.411857327499572e-05,
      "loss": 2.7701,
      "num_input_tokens_seen": 42731808,
      "step": 20450
    },
    {
      "epoch": 3.336895342197569,
      "grad_norm": 6.4375,
      "learning_rate": 2.410876196992739e-05,
      "loss": 2.526,
      "num_input_tokens_seen": 42742512,
      "step": 20455
    },
    {
      "epoch": 3.3377110694183862,
      "grad_norm": 2.953125,
      "learning_rate": 2.4098950802299156e-05,
      "loss": 1.7085,
      "num_input_tokens_seen": 42753680,
      "step": 20460
    },
    {
      "epoch": 3.338526796639204,
      "grad_norm": 3.09375,
      "learning_rate": 2.4089139773624027e-05,
      "loss": 1.8869,
      "num_input_tokens_seen": 42765376,
      "step": 20465
    },
    {
      "epoch": 3.3393425238600214,
      "grad_norm": 7.0625,
      "learning_rate": 2.4079328885415007e-05,
      "loss": 2.6333,
      "num_input_tokens_seen": 42775072,
      "step": 20470
    },
    {
      "epoch": 3.3401582510808385,
      "grad_norm": 2.359375,
      "learning_rate": 2.4069518139185036e-05,
      "loss": 3.2955,
      "num_input_tokens_seen": 42785872,
      "step": 20475
    },
    {
      "epoch": 3.340973978301656,
      "grad_norm": 4.96875,
      "learning_rate": 2.405970753644706e-05,
      "loss": 3.4296,
      "num_input_tokens_seen": 42796800,
      "step": 20480
    },
    {
      "epoch": 3.3417897055224732,
      "grad_norm": 6.125,
      "learning_rate": 2.4049897078714e-05,
      "loss": 2.823,
      "num_input_tokens_seen": 42807312,
      "step": 20485
    },
    {
      "epoch": 3.342605432743291,
      "grad_norm": 1.859375,
      "learning_rate": 2.404008676749874e-05,
      "loss": 3.3115,
      "num_input_tokens_seen": 42818704,
      "step": 20490
    },
    {
      "epoch": 3.343421159964108,
      "grad_norm": 7.0625,
      "learning_rate": 2.403027660431418e-05,
      "loss": 1.3676,
      "num_input_tokens_seen": 42829088,
      "step": 20495
    },
    {
      "epoch": 3.3442368871849255,
      "grad_norm": 4.96875,
      "learning_rate": 2.402046659067314e-05,
      "loss": 2.2335,
      "num_input_tokens_seen": 42840000,
      "step": 20500
    },
    {
      "epoch": 3.3450526144057426,
      "grad_norm": 2.25,
      "learning_rate": 2.401065672808847e-05,
      "loss": 1.8677,
      "num_input_tokens_seen": 42850240,
      "step": 20505
    },
    {
      "epoch": 3.34586834162656,
      "grad_norm": 8.3125,
      "learning_rate": 2.400084701807296e-05,
      "loss": 1.9798,
      "num_input_tokens_seen": 42862704,
      "step": 20510
    },
    {
      "epoch": 3.3466840688473773,
      "grad_norm": 12.625,
      "learning_rate": 2.39910374621394e-05,
      "loss": 3.1174,
      "num_input_tokens_seen": 42873392,
      "step": 20515
    },
    {
      "epoch": 3.347499796068195,
      "grad_norm": 7.25,
      "learning_rate": 2.3981228061800544e-05,
      "loss": 2.9435,
      "num_input_tokens_seen": 42883440,
      "step": 20520
    },
    {
      "epoch": 3.348315523289012,
      "grad_norm": 7.09375,
      "learning_rate": 2.3971418818569115e-05,
      "loss": 2.9212,
      "num_input_tokens_seen": 42893728,
      "step": 20525
    },
    {
      "epoch": 3.3491312505098296,
      "grad_norm": 7.40625,
      "learning_rate": 2.3961609733957832e-05,
      "loss": 3.6041,
      "num_input_tokens_seen": 42903968,
      "step": 20530
    },
    {
      "epoch": 3.3499469777306468,
      "grad_norm": 4.96875,
      "learning_rate": 2.395180080947939e-05,
      "loss": 2.7136,
      "num_input_tokens_seen": 42914464,
      "step": 20535
    },
    {
      "epoch": 3.3507627049514643,
      "grad_norm": 11.25,
      "learning_rate": 2.394199204664642e-05,
      "loss": 3.2381,
      "num_input_tokens_seen": 42925872,
      "step": 20540
    },
    {
      "epoch": 3.3515784321722815,
      "grad_norm": 13.5625,
      "learning_rate": 2.3932183446971583e-05,
      "loss": 2.6707,
      "num_input_tokens_seen": 42936208,
      "step": 20545
    },
    {
      "epoch": 3.352394159393099,
      "grad_norm": 3.328125,
      "learning_rate": 2.3922375011967473e-05,
      "loss": 3.7772,
      "num_input_tokens_seen": 42947456,
      "step": 20550
    },
    {
      "epoch": 3.353209886613916,
      "grad_norm": 4.3125,
      "learning_rate": 2.3912566743146676e-05,
      "loss": 2.3992,
      "num_input_tokens_seen": 42959088,
      "step": 20555
    },
    {
      "epoch": 3.3540256138347337,
      "grad_norm": 8.125,
      "learning_rate": 2.390275864202176e-05,
      "loss": 1.8851,
      "num_input_tokens_seen": 42969424,
      "step": 20560
    },
    {
      "epoch": 3.354841341055551,
      "grad_norm": 5.46875,
      "learning_rate": 2.3892950710105243e-05,
      "loss": 2.056,
      "num_input_tokens_seen": 42980192,
      "step": 20565
    },
    {
      "epoch": 3.3556570682763684,
      "grad_norm": 8.5625,
      "learning_rate": 2.3883142948909635e-05,
      "loss": 1.5327,
      "num_input_tokens_seen": 42991168,
      "step": 20570
    },
    {
      "epoch": 3.3564727954971856,
      "grad_norm": 7.375,
      "learning_rate": 2.3873335359947433e-05,
      "loss": 2.3814,
      "num_input_tokens_seen": 43000928,
      "step": 20575
    },
    {
      "epoch": 3.357288522718003,
      "grad_norm": 8.6875,
      "learning_rate": 2.3863527944731066e-05,
      "loss": 1.9113,
      "num_input_tokens_seen": 43011936,
      "step": 20580
    },
    {
      "epoch": 3.3581042499388203,
      "grad_norm": 6.65625,
      "learning_rate": 2.385372070477298e-05,
      "loss": 2.3879,
      "num_input_tokens_seen": 43023840,
      "step": 20585
    },
    {
      "epoch": 3.358919977159638,
      "grad_norm": 15.9375,
      "learning_rate": 2.384391364158556e-05,
      "loss": 2.972,
      "num_input_tokens_seen": 43034288,
      "step": 20590
    },
    {
      "epoch": 3.359735704380455,
      "grad_norm": 6.0,
      "learning_rate": 2.3834106756681185e-05,
      "loss": 2.527,
      "num_input_tokens_seen": 43044768,
      "step": 20595
    },
    {
      "epoch": 3.3605514316012726,
      "grad_norm": 11.875,
      "learning_rate": 2.3824300051572206e-05,
      "loss": 2.4638,
      "num_input_tokens_seen": 43054848,
      "step": 20600
    },
    {
      "epoch": 3.3605514316012726,
      "eval_loss": 2.5471649169921875,
      "eval_runtime": 135.0498,
      "eval_samples_per_second": 20.178,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 43054848,
      "step": 20600
    },
    {
      "epoch": 3.36136715882209,
      "grad_norm": 7.25,
      "learning_rate": 2.3814493527770923e-05,
      "loss": 2.5281,
      "num_input_tokens_seen": 43064880,
      "step": 20605
    },
    {
      "epoch": 3.3621828860429073,
      "grad_norm": 7.8125,
      "learning_rate": 2.3804687186789637e-05,
      "loss": 1.9726,
      "num_input_tokens_seen": 43076000,
      "step": 20610
    },
    {
      "epoch": 3.3629986132637244,
      "grad_norm": 9.9375,
      "learning_rate": 2.379488103014062e-05,
      "loss": 2.747,
      "num_input_tokens_seen": 43085296,
      "step": 20615
    },
    {
      "epoch": 3.363814340484542,
      "grad_norm": 8.1875,
      "learning_rate": 2.3785075059336086e-05,
      "loss": 2.0082,
      "num_input_tokens_seen": 43096432,
      "step": 20620
    },
    {
      "epoch": 3.3646300677053596,
      "grad_norm": 12.625,
      "learning_rate": 2.3775269275888248e-05,
      "loss": 3.2529,
      "num_input_tokens_seen": 43108512,
      "step": 20625
    },
    {
      "epoch": 3.3654457949261767,
      "grad_norm": 9.5,
      "learning_rate": 2.3765463681309274e-05,
      "loss": 2.348,
      "num_input_tokens_seen": 43118416,
      "step": 20630
    },
    {
      "epoch": 3.366261522146994,
      "grad_norm": 10.9375,
      "learning_rate": 2.3755658277111313e-05,
      "loss": 2.4894,
      "num_input_tokens_seen": 43129488,
      "step": 20635
    },
    {
      "epoch": 3.3670772493678114,
      "grad_norm": 11.3125,
      "learning_rate": 2.374585306480649e-05,
      "loss": 4.1974,
      "num_input_tokens_seen": 43140688,
      "step": 20640
    },
    {
      "epoch": 3.367892976588629,
      "grad_norm": 3.90625,
      "learning_rate": 2.3736048045906877e-05,
      "loss": 1.6057,
      "num_input_tokens_seen": 43149344,
      "step": 20645
    },
    {
      "epoch": 3.368708703809446,
      "grad_norm": 8.125,
      "learning_rate": 2.372624322192454e-05,
      "loss": 2.8391,
      "num_input_tokens_seen": 43160080,
      "step": 20650
    },
    {
      "epoch": 3.3695244310302637,
      "grad_norm": 1.6953125,
      "learning_rate": 2.3716438594371516e-05,
      "loss": 2.0155,
      "num_input_tokens_seen": 43171104,
      "step": 20655
    },
    {
      "epoch": 3.370340158251081,
      "grad_norm": 6.8125,
      "learning_rate": 2.3706634164759784e-05,
      "loss": 2.3163,
      "num_input_tokens_seen": 43181728,
      "step": 20660
    },
    {
      "epoch": 3.3711558854718984,
      "grad_norm": 5.8125,
      "learning_rate": 2.3696829934601323e-05,
      "loss": 2.3781,
      "num_input_tokens_seen": 43191984,
      "step": 20665
    },
    {
      "epoch": 3.3719716126927155,
      "grad_norm": 5.0,
      "learning_rate": 2.3687025905408053e-05,
      "loss": 2.1347,
      "num_input_tokens_seen": 43202240,
      "step": 20670
    },
    {
      "epoch": 3.372787339913533,
      "grad_norm": 8.125,
      "learning_rate": 2.3677222078691886e-05,
      "loss": 2.2269,
      "num_input_tokens_seen": 43213408,
      "step": 20675
    },
    {
      "epoch": 3.37360306713435,
      "grad_norm": 8.75,
      "learning_rate": 2.366741845596471e-05,
      "loss": 2.1443,
      "num_input_tokens_seen": 43223568,
      "step": 20680
    },
    {
      "epoch": 3.374418794355168,
      "grad_norm": 9.5625,
      "learning_rate": 2.3657615038738343e-05,
      "loss": 3.006,
      "num_input_tokens_seen": 43235024,
      "step": 20685
    },
    {
      "epoch": 3.375234521575985,
      "grad_norm": 4.8125,
      "learning_rate": 2.3647811828524614e-05,
      "loss": 2.3502,
      "num_input_tokens_seen": 43246640,
      "step": 20690
    },
    {
      "epoch": 3.3760502487968025,
      "grad_norm": 9.0,
      "learning_rate": 2.363800882683529e-05,
      "loss": 1.9453,
      "num_input_tokens_seen": 43258448,
      "step": 20695
    },
    {
      "epoch": 3.3768659760176196,
      "grad_norm": 10.8125,
      "learning_rate": 2.3628206035182125e-05,
      "loss": 1.7205,
      "num_input_tokens_seen": 43268800,
      "step": 20700
    },
    {
      "epoch": 3.377681703238437,
      "grad_norm": 9.4375,
      "learning_rate": 2.361840345507683e-05,
      "loss": 2.2791,
      "num_input_tokens_seen": 43279584,
      "step": 20705
    },
    {
      "epoch": 3.3784974304592543,
      "grad_norm": 10.25,
      "learning_rate": 2.3608601088031073e-05,
      "loss": 3.7969,
      "num_input_tokens_seen": 43290352,
      "step": 20710
    },
    {
      "epoch": 3.379313157680072,
      "grad_norm": 9.375,
      "learning_rate": 2.3598798935556516e-05,
      "loss": 2.0891,
      "num_input_tokens_seen": 43300704,
      "step": 20715
    },
    {
      "epoch": 3.380128884900889,
      "grad_norm": 3.078125,
      "learning_rate": 2.3588996999164784e-05,
      "loss": 1.9274,
      "num_input_tokens_seen": 43309936,
      "step": 20720
    },
    {
      "epoch": 3.3809446121217066,
      "grad_norm": 8.9375,
      "learning_rate": 2.3579195280367434e-05,
      "loss": 3.228,
      "num_input_tokens_seen": 43320720,
      "step": 20725
    },
    {
      "epoch": 3.3817603393425237,
      "grad_norm": 14.9375,
      "learning_rate": 2.356939378067603e-05,
      "loss": 2.214,
      "num_input_tokens_seen": 43332080,
      "step": 20730
    },
    {
      "epoch": 3.3825760665633413,
      "grad_norm": 1.046875,
      "learning_rate": 2.3559592501602092e-05,
      "loss": 2.0855,
      "num_input_tokens_seen": 43342944,
      "step": 20735
    },
    {
      "epoch": 3.3833917937841584,
      "grad_norm": 13.8125,
      "learning_rate": 2.3549791444657076e-05,
      "loss": 3.7724,
      "num_input_tokens_seen": 43352864,
      "step": 20740
    },
    {
      "epoch": 3.384207521004976,
      "grad_norm": 8.4375,
      "learning_rate": 2.353999061135246e-05,
      "loss": 2.5566,
      "num_input_tokens_seen": 43362480,
      "step": 20745
    },
    {
      "epoch": 3.385023248225793,
      "grad_norm": 6.15625,
      "learning_rate": 2.3530190003199626e-05,
      "loss": 2.5185,
      "num_input_tokens_seen": 43371840,
      "step": 20750
    },
    {
      "epoch": 3.3858389754466107,
      "grad_norm": 9.0625,
      "learning_rate": 2.3520389621709965e-05,
      "loss": 3.7439,
      "num_input_tokens_seen": 43383200,
      "step": 20755
    },
    {
      "epoch": 3.386654702667428,
      "grad_norm": 3.515625,
      "learning_rate": 2.351058946839483e-05,
      "loss": 2.2205,
      "num_input_tokens_seen": 43393792,
      "step": 20760
    },
    {
      "epoch": 3.3874704298882454,
      "grad_norm": 6.09375,
      "learning_rate": 2.350078954476551e-05,
      "loss": 1.9372,
      "num_input_tokens_seen": 43404704,
      "step": 20765
    },
    {
      "epoch": 3.3882861571090626,
      "grad_norm": 8.5,
      "learning_rate": 2.3490989852333272e-05,
      "loss": 2.715,
      "num_input_tokens_seen": 43413344,
      "step": 20770
    },
    {
      "epoch": 3.38910188432988,
      "grad_norm": 4.71875,
      "learning_rate": 2.3481190392609377e-05,
      "loss": 2.1407,
      "num_input_tokens_seen": 43423200,
      "step": 20775
    },
    {
      "epoch": 3.3899176115506973,
      "grad_norm": 6.0625,
      "learning_rate": 2.3471391167105e-05,
      "loss": 2.9771,
      "num_input_tokens_seen": 43432320,
      "step": 20780
    },
    {
      "epoch": 3.390733338771515,
      "grad_norm": 9.9375,
      "learning_rate": 2.3461592177331325e-05,
      "loss": 3.4726,
      "num_input_tokens_seen": 43441616,
      "step": 20785
    },
    {
      "epoch": 3.391549065992332,
      "grad_norm": 13.4375,
      "learning_rate": 2.345179342479946e-05,
      "loss": 2.6923,
      "num_input_tokens_seen": 43452752,
      "step": 20790
    },
    {
      "epoch": 3.3923647932131495,
      "grad_norm": 5.96875,
      "learning_rate": 2.3441994911020503e-05,
      "loss": 2.2664,
      "num_input_tokens_seen": 43463200,
      "step": 20795
    },
    {
      "epoch": 3.393180520433967,
      "grad_norm": 1.375,
      "learning_rate": 2.3432196637505522e-05,
      "loss": 0.7799,
      "num_input_tokens_seen": 43472928,
      "step": 20800
    },
    {
      "epoch": 3.393180520433967,
      "eval_loss": 2.5441832542419434,
      "eval_runtime": 135.0363,
      "eval_samples_per_second": 20.18,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 43472928,
      "step": 20800
    },
    {
      "epoch": 3.3939962476547842,
      "grad_norm": 4.96875,
      "learning_rate": 2.3422398605765515e-05,
      "loss": 2.1707,
      "num_input_tokens_seen": 43482976,
      "step": 20805
    },
    {
      "epoch": 3.3948119748756014,
      "grad_norm": 6.3125,
      "learning_rate": 2.3412600817311462e-05,
      "loss": 1.6711,
      "num_input_tokens_seen": 43493936,
      "step": 20810
    },
    {
      "epoch": 3.395627702096419,
      "grad_norm": 4.78125,
      "learning_rate": 2.3402803273654326e-05,
      "loss": 1.9102,
      "num_input_tokens_seen": 43503936,
      "step": 20815
    },
    {
      "epoch": 3.3964434293172365,
      "grad_norm": 10.6875,
      "learning_rate": 2.3393005976304983e-05,
      "loss": 3.3553,
      "num_input_tokens_seen": 43514416,
      "step": 20820
    },
    {
      "epoch": 3.3972591565380537,
      "grad_norm": 5.0,
      "learning_rate": 2.338320892677432e-05,
      "loss": 1.8994,
      "num_input_tokens_seen": 43525536,
      "step": 20825
    },
    {
      "epoch": 3.398074883758871,
      "grad_norm": 7.0625,
      "learning_rate": 2.3373412126573155e-05,
      "loss": 1.9248,
      "num_input_tokens_seen": 43537152,
      "step": 20830
    },
    {
      "epoch": 3.3988906109796884,
      "grad_norm": 4.03125,
      "learning_rate": 2.3363615577212285e-05,
      "loss": 2.3896,
      "num_input_tokens_seen": 43546944,
      "step": 20835
    },
    {
      "epoch": 3.399706338200506,
      "grad_norm": 3.984375,
      "learning_rate": 2.3353819280202455e-05,
      "loss": 1.9429,
      "num_input_tokens_seen": 43556096,
      "step": 20840
    },
    {
      "epoch": 3.400522065421323,
      "grad_norm": 8.625,
      "learning_rate": 2.334402323705438e-05,
      "loss": 2.7314,
      "num_input_tokens_seen": 43566256,
      "step": 20845
    },
    {
      "epoch": 3.4013377926421406,
      "grad_norm": 7.40625,
      "learning_rate": 2.3334227449278725e-05,
      "loss": 2.7681,
      "num_input_tokens_seen": 43576896,
      "step": 20850
    },
    {
      "epoch": 3.4021535198629578,
      "grad_norm": 9.3125,
      "learning_rate": 2.3324431918386143e-05,
      "loss": 2.326,
      "num_input_tokens_seen": 43587392,
      "step": 20855
    },
    {
      "epoch": 3.4029692470837754,
      "grad_norm": 9.0625,
      "learning_rate": 2.3314636645887207e-05,
      "loss": 3.8835,
      "num_input_tokens_seen": 43596608,
      "step": 20860
    },
    {
      "epoch": 3.4037849743045925,
      "grad_norm": 5.96875,
      "learning_rate": 2.3304841633292487e-05,
      "loss": 2.3571,
      "num_input_tokens_seen": 43607840,
      "step": 20865
    },
    {
      "epoch": 3.40460070152541,
      "grad_norm": 6.90625,
      "learning_rate": 2.329504688211248e-05,
      "loss": 1.7162,
      "num_input_tokens_seen": 43616752,
      "step": 20870
    },
    {
      "epoch": 3.405416428746227,
      "grad_norm": 7.5,
      "learning_rate": 2.3285252393857677e-05,
      "loss": 2.2454,
      "num_input_tokens_seen": 43628080,
      "step": 20875
    },
    {
      "epoch": 3.4062321559670448,
      "grad_norm": 7.0625,
      "learning_rate": 2.327545817003851e-05,
      "loss": 1.7572,
      "num_input_tokens_seen": 43638208,
      "step": 20880
    },
    {
      "epoch": 3.407047883187862,
      "grad_norm": 1.734375,
      "learning_rate": 2.326566421216535e-05,
      "loss": 2.4163,
      "num_input_tokens_seen": 43648688,
      "step": 20885
    },
    {
      "epoch": 3.4078636104086795,
      "grad_norm": 9.125,
      "learning_rate": 2.3255870521748565e-05,
      "loss": 3.6,
      "num_input_tokens_seen": 43659616,
      "step": 20890
    },
    {
      "epoch": 3.4086793376294966,
      "grad_norm": 8.625,
      "learning_rate": 2.3246077100298474e-05,
      "loss": 1.899,
      "num_input_tokens_seen": 43669744,
      "step": 20895
    },
    {
      "epoch": 3.409495064850314,
      "grad_norm": 7.09375,
      "learning_rate": 2.3236283949325328e-05,
      "loss": 1.8893,
      "num_input_tokens_seen": 43682080,
      "step": 20900
    },
    {
      "epoch": 3.4103107920711313,
      "grad_norm": 6.375,
      "learning_rate": 2.3226491070339368e-05,
      "loss": 2.3363,
      "num_input_tokens_seen": 43692640,
      "step": 20905
    },
    {
      "epoch": 3.411126519291949,
      "grad_norm": 8.1875,
      "learning_rate": 2.3216698464850762e-05,
      "loss": 1.8996,
      "num_input_tokens_seen": 43702496,
      "step": 20910
    },
    {
      "epoch": 3.411942246512766,
      "grad_norm": 5.25,
      "learning_rate": 2.320690613436967e-05,
      "loss": 2.5471,
      "num_input_tokens_seen": 43713552,
      "step": 20915
    },
    {
      "epoch": 3.4127579737335836,
      "grad_norm": 11.25,
      "learning_rate": 2.3197114080406192e-05,
      "loss": 2.8929,
      "num_input_tokens_seen": 43725264,
      "step": 20920
    },
    {
      "epoch": 3.4135737009544007,
      "grad_norm": 7.15625,
      "learning_rate": 2.3187322304470365e-05,
      "loss": 2.3528,
      "num_input_tokens_seen": 43734864,
      "step": 20925
    },
    {
      "epoch": 3.4143894281752183,
      "grad_norm": 13.125,
      "learning_rate": 2.3177530808072222e-05,
      "loss": 3.1845,
      "num_input_tokens_seen": 43746240,
      "step": 20930
    },
    {
      "epoch": 3.4152051553960354,
      "grad_norm": 6.0,
      "learning_rate": 2.316773959272174e-05,
      "loss": 1.975,
      "num_input_tokens_seen": 43757072,
      "step": 20935
    },
    {
      "epoch": 3.416020882616853,
      "grad_norm": 4.96875,
      "learning_rate": 2.3157948659928823e-05,
      "loss": 2.5901,
      "num_input_tokens_seen": 43766576,
      "step": 20940
    },
    {
      "epoch": 3.41683660983767,
      "grad_norm": 3.4375,
      "learning_rate": 2.3148158011203388e-05,
      "loss": 2.8722,
      "num_input_tokens_seen": 43777056,
      "step": 20945
    },
    {
      "epoch": 3.4176523370584877,
      "grad_norm": 7.8125,
      "learning_rate": 2.3138367648055253e-05,
      "loss": 2.8573,
      "num_input_tokens_seen": 43786704,
      "step": 20950
    },
    {
      "epoch": 3.418468064279305,
      "grad_norm": 8.8125,
      "learning_rate": 2.312857757199422e-05,
      "loss": 3.0529,
      "num_input_tokens_seen": 43797760,
      "step": 20955
    },
    {
      "epoch": 3.4192837915001224,
      "grad_norm": 11.875,
      "learning_rate": 2.3118787784530048e-05,
      "loss": 2.3851,
      "num_input_tokens_seen": 43807680,
      "step": 20960
    },
    {
      "epoch": 3.4200995187209395,
      "grad_norm": 0.1513671875,
      "learning_rate": 2.310899828717243e-05,
      "loss": 1.5343,
      "num_input_tokens_seen": 43818064,
      "step": 20965
    },
    {
      "epoch": 3.420915245941757,
      "grad_norm": 7.28125,
      "learning_rate": 2.309920908143104e-05,
      "loss": 2.3401,
      "num_input_tokens_seen": 43829376,
      "step": 20970
    },
    {
      "epoch": 3.4217309731625742,
      "grad_norm": 5.21875,
      "learning_rate": 2.308942016881551e-05,
      "loss": 2.6402,
      "num_input_tokens_seen": 43839792,
      "step": 20975
    },
    {
      "epoch": 3.422546700383392,
      "grad_norm": 12.4375,
      "learning_rate": 2.307963155083539e-05,
      "loss": 1.7793,
      "num_input_tokens_seen": 43850624,
      "step": 20980
    },
    {
      "epoch": 3.423362427604209,
      "grad_norm": 5.6875,
      "learning_rate": 2.306984322900022e-05,
      "loss": 3.2126,
      "num_input_tokens_seen": 43862416,
      "step": 20985
    },
    {
      "epoch": 3.4241781548250265,
      "grad_norm": 5.625,
      "learning_rate": 2.3060055204819482e-05,
      "loss": 1.8333,
      "num_input_tokens_seen": 43872432,
      "step": 20990
    },
    {
      "epoch": 3.424993882045844,
      "grad_norm": 9.375,
      "learning_rate": 2.3050267479802604e-05,
      "loss": 2.4744,
      "num_input_tokens_seen": 43882976,
      "step": 20995
    },
    {
      "epoch": 3.4258096092666612,
      "grad_norm": 7.0625,
      "learning_rate": 2.304048005545899e-05,
      "loss": 1.9196,
      "num_input_tokens_seen": 43892704,
      "step": 21000
    },
    {
      "epoch": 3.4258096092666612,
      "eval_loss": 2.540254831314087,
      "eval_runtime": 135.074,
      "eval_samples_per_second": 20.174,
      "eval_steps_per_second": 10.091,
      "num_input_tokens_seen": 43892704,
      "step": 21000
    },
    {
      "epoch": 3.4266253364874784,
      "grad_norm": 6.1875,
      "learning_rate": 2.3030692933297972e-05,
      "loss": 2.0137,
      "num_input_tokens_seen": 43903520,
      "step": 21005
    },
    {
      "epoch": 3.427441063708296,
      "grad_norm": 5.71875,
      "learning_rate": 2.3020906114828843e-05,
      "loss": 1.0638,
      "num_input_tokens_seen": 43912960,
      "step": 21010
    },
    {
      "epoch": 3.4282567909291135,
      "grad_norm": 3.59375,
      "learning_rate": 2.301111960156088e-05,
      "loss": 3.6893,
      "num_input_tokens_seen": 43924288,
      "step": 21015
    },
    {
      "epoch": 3.4290725181499306,
      "grad_norm": 8.5,
      "learning_rate": 2.300133339500326e-05,
      "loss": 2.2816,
      "num_input_tokens_seen": 43934240,
      "step": 21020
    },
    {
      "epoch": 3.429888245370748,
      "grad_norm": 3.765625,
      "learning_rate": 2.2991547496665148e-05,
      "loss": 2.6181,
      "num_input_tokens_seen": 43943504,
      "step": 21025
    },
    {
      "epoch": 3.4307039725915653,
      "grad_norm": 13.75,
      "learning_rate": 2.298176190805565e-05,
      "loss": 2.8553,
      "num_input_tokens_seen": 43953264,
      "step": 21030
    },
    {
      "epoch": 3.431519699812383,
      "grad_norm": 3.453125,
      "learning_rate": 2.2971976630683826e-05,
      "loss": 1.7463,
      "num_input_tokens_seen": 43964608,
      "step": 21035
    },
    {
      "epoch": 3.4323354270332,
      "grad_norm": 6.71875,
      "learning_rate": 2.29621916660587e-05,
      "loss": 2.36,
      "num_input_tokens_seen": 43973664,
      "step": 21040
    },
    {
      "epoch": 3.4331511542540176,
      "grad_norm": 8.375,
      "learning_rate": 2.295240701568922e-05,
      "loss": 3.1605,
      "num_input_tokens_seen": 43984960,
      "step": 21045
    },
    {
      "epoch": 3.4339668814748348,
      "grad_norm": 8.5625,
      "learning_rate": 2.2942622681084312e-05,
      "loss": 1.5289,
      "num_input_tokens_seen": 43995424,
      "step": 21050
    },
    {
      "epoch": 3.4347826086956523,
      "grad_norm": 16.5,
      "learning_rate": 2.293283866375284e-05,
      "loss": 3.0836,
      "num_input_tokens_seen": 44006160,
      "step": 21055
    },
    {
      "epoch": 3.4355983359164695,
      "grad_norm": 12.8125,
      "learning_rate": 2.2923054965203627e-05,
      "loss": 2.526,
      "num_input_tokens_seen": 44016960,
      "step": 21060
    },
    {
      "epoch": 3.436414063137287,
      "grad_norm": 4.3125,
      "learning_rate": 2.2913271586945443e-05,
      "loss": 3.2256,
      "num_input_tokens_seen": 44027328,
      "step": 21065
    },
    {
      "epoch": 3.437229790358104,
      "grad_norm": 4.03125,
      "learning_rate": 2.290348853048699e-05,
      "loss": 2.2815,
      "num_input_tokens_seen": 44037824,
      "step": 21070
    },
    {
      "epoch": 3.4380455175789217,
      "grad_norm": 4.75,
      "learning_rate": 2.2893705797336956e-05,
      "loss": 1.7393,
      "num_input_tokens_seen": 44046448,
      "step": 21075
    },
    {
      "epoch": 3.438861244799739,
      "grad_norm": 4.71875,
      "learning_rate": 2.288392338900397e-05,
      "loss": 1.8878,
      "num_input_tokens_seen": 44056896,
      "step": 21080
    },
    {
      "epoch": 3.4396769720205564,
      "grad_norm": 6.625,
      "learning_rate": 2.2874141306996576e-05,
      "loss": 2.6685,
      "num_input_tokens_seen": 44066624,
      "step": 21085
    },
    {
      "epoch": 3.4404926992413736,
      "grad_norm": 9.625,
      "learning_rate": 2.2864359552823312e-05,
      "loss": 2.1314,
      "num_input_tokens_seen": 44076384,
      "step": 21090
    },
    {
      "epoch": 3.441308426462191,
      "grad_norm": 6.375,
      "learning_rate": 2.2854578127992648e-05,
      "loss": 3.8402,
      "num_input_tokens_seen": 44086704,
      "step": 21095
    },
    {
      "epoch": 3.4421241536830083,
      "grad_norm": 13.75,
      "learning_rate": 2.2844797034012988e-05,
      "loss": 2.6663,
      "num_input_tokens_seen": 44097616,
      "step": 21100
    },
    {
      "epoch": 3.442939880903826,
      "grad_norm": 1.3828125,
      "learning_rate": 2.2835016272392722e-05,
      "loss": 1.6125,
      "num_input_tokens_seen": 44108928,
      "step": 21105
    },
    {
      "epoch": 3.443755608124643,
      "grad_norm": 1.890625,
      "learning_rate": 2.2825235844640142e-05,
      "loss": 2.2246,
      "num_input_tokens_seen": 44119888,
      "step": 21110
    },
    {
      "epoch": 3.4445713353454606,
      "grad_norm": 10.5625,
      "learning_rate": 2.2815455752263522e-05,
      "loss": 1.4623,
      "num_input_tokens_seen": 44130656,
      "step": 21115
    },
    {
      "epoch": 3.4453870625662777,
      "grad_norm": 3.6875,
      "learning_rate": 2.2805675996771092e-05,
      "loss": 2.2528,
      "num_input_tokens_seen": 44143360,
      "step": 21120
    },
    {
      "epoch": 3.4462027897870953,
      "grad_norm": 3.0625,
      "learning_rate": 2.2795896579670987e-05,
      "loss": 2.188,
      "num_input_tokens_seen": 44154752,
      "step": 21125
    },
    {
      "epoch": 3.4470185170079124,
      "grad_norm": 5.96875,
      "learning_rate": 2.2786117502471337e-05,
      "loss": 0.6839,
      "num_input_tokens_seen": 44163728,
      "step": 21130
    },
    {
      "epoch": 3.44783424422873,
      "grad_norm": 9.8125,
      "learning_rate": 2.2776338766680185e-05,
      "loss": 3.2804,
      "num_input_tokens_seen": 44174928,
      "step": 21135
    },
    {
      "epoch": 3.448649971449547,
      "grad_norm": 8.125,
      "learning_rate": 2.2766560373805533e-05,
      "loss": 2.6387,
      "num_input_tokens_seen": 44185776,
      "step": 21140
    },
    {
      "epoch": 3.4494656986703647,
      "grad_norm": 10.0,
      "learning_rate": 2.2756782325355353e-05,
      "loss": 3.6729,
      "num_input_tokens_seen": 44195920,
      "step": 21145
    },
    {
      "epoch": 3.450281425891182,
      "grad_norm": 10.875,
      "learning_rate": 2.2747004622837514e-05,
      "loss": 2.3434,
      "num_input_tokens_seen": 44206000,
      "step": 21150
    },
    {
      "epoch": 3.4510971531119994,
      "grad_norm": 3.703125,
      "learning_rate": 2.2737227267759878e-05,
      "loss": 3.1034,
      "num_input_tokens_seen": 44216032,
      "step": 21155
    },
    {
      "epoch": 3.4519128803328165,
      "grad_norm": 4.4375,
      "learning_rate": 2.272745026163024e-05,
      "loss": 2.3247,
      "num_input_tokens_seen": 44226368,
      "step": 21160
    },
    {
      "epoch": 3.452728607553634,
      "grad_norm": 10.1875,
      "learning_rate": 2.271767360595633e-05,
      "loss": 3.0496,
      "num_input_tokens_seen": 44237008,
      "step": 21165
    },
    {
      "epoch": 3.4535443347744517,
      "grad_norm": 5.46875,
      "learning_rate": 2.270789730224583e-05,
      "loss": 1.736,
      "num_input_tokens_seen": 44248144,
      "step": 21170
    },
    {
      "epoch": 3.454360061995269,
      "grad_norm": 10.6875,
      "learning_rate": 2.2698121352006367e-05,
      "loss": 2.5785,
      "num_input_tokens_seen": 44259040,
      "step": 21175
    },
    {
      "epoch": 3.455175789216086,
      "grad_norm": 5.6875,
      "learning_rate": 2.2688345756745517e-05,
      "loss": 1.6497,
      "num_input_tokens_seen": 44269184,
      "step": 21180
    },
    {
      "epoch": 3.4559915164369035,
      "grad_norm": 8.1875,
      "learning_rate": 2.267857051797081e-05,
      "loss": 2.6884,
      "num_input_tokens_seen": 44279808,
      "step": 21185
    },
    {
      "epoch": 3.456807243657721,
      "grad_norm": 10.1875,
      "learning_rate": 2.2668795637189695e-05,
      "loss": 3.2279,
      "num_input_tokens_seen": 44289072,
      "step": 21190
    },
    {
      "epoch": 3.457622970878538,
      "grad_norm": 1.1953125,
      "learning_rate": 2.2659021115909586e-05,
      "loss": 1.9791,
      "num_input_tokens_seen": 44298752,
      "step": 21195
    },
    {
      "epoch": 3.4584386980993553,
      "grad_norm": 3.203125,
      "learning_rate": 2.2649246955637847e-05,
      "loss": 2.9388,
      "num_input_tokens_seen": 44309408,
      "step": 21200
    },
    {
      "epoch": 3.4584386980993553,
      "eval_loss": 2.5528268814086914,
      "eval_runtime": 135.0319,
      "eval_samples_per_second": 20.18,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 44309408,
      "step": 21200
    },
    {
      "epoch": 3.459254425320173,
      "grad_norm": 9.375,
      "learning_rate": 2.2639473157881766e-05,
      "loss": 2.1529,
      "num_input_tokens_seen": 44320160,
      "step": 21205
    },
    {
      "epoch": 3.4600701525409905,
      "grad_norm": 5.40625,
      "learning_rate": 2.2629699724148594e-05,
      "loss": 3.0936,
      "num_input_tokens_seen": 44330528,
      "step": 21210
    },
    {
      "epoch": 3.4608858797618076,
      "grad_norm": 12.25,
      "learning_rate": 2.26199266559455e-05,
      "loss": 3.2488,
      "num_input_tokens_seen": 44340304,
      "step": 21215
    },
    {
      "epoch": 3.461701606982625,
      "grad_norm": 5.5,
      "learning_rate": 2.2610153954779625e-05,
      "loss": 2.3216,
      "num_input_tokens_seen": 44351648,
      "step": 21220
    },
    {
      "epoch": 3.4625173342034423,
      "grad_norm": 4.96875,
      "learning_rate": 2.2600381622158056e-05,
      "loss": 0.8426,
      "num_input_tokens_seen": 44361584,
      "step": 21225
    },
    {
      "epoch": 3.46333306142426,
      "grad_norm": 8.375,
      "learning_rate": 2.2590609659587783e-05,
      "loss": 3.1564,
      "num_input_tokens_seen": 44371200,
      "step": 21230
    },
    {
      "epoch": 3.464148788645077,
      "grad_norm": 8.8125,
      "learning_rate": 2.2580838068575787e-05,
      "loss": 1.7587,
      "num_input_tokens_seen": 44380544,
      "step": 21235
    },
    {
      "epoch": 3.4649645158658946,
      "grad_norm": 9.875,
      "learning_rate": 2.257106685062896e-05,
      "loss": 3.409,
      "num_input_tokens_seen": 44389472,
      "step": 21240
    },
    {
      "epoch": 3.4657802430867117,
      "grad_norm": 9.375,
      "learning_rate": 2.256129600725415e-05,
      "loss": 3.0092,
      "num_input_tokens_seen": 44400704,
      "step": 21245
    },
    {
      "epoch": 3.4665959703075293,
      "grad_norm": 3.578125,
      "learning_rate": 2.2551525539958145e-05,
      "loss": 3.1499,
      "num_input_tokens_seen": 44411120,
      "step": 21250
    },
    {
      "epoch": 3.4674116975283464,
      "grad_norm": 10.3125,
      "learning_rate": 2.2541755450247663e-05,
      "loss": 3.0787,
      "num_input_tokens_seen": 44421040,
      "step": 21255
    },
    {
      "epoch": 3.468227424749164,
      "grad_norm": 10.75,
      "learning_rate": 2.2531985739629382e-05,
      "loss": 3.6333,
      "num_input_tokens_seen": 44431920,
      "step": 21260
    },
    {
      "epoch": 3.469043151969981,
      "grad_norm": 7.1875,
      "learning_rate": 2.2522216409609924e-05,
      "loss": 2.173,
      "num_input_tokens_seen": 44441328,
      "step": 21265
    },
    {
      "epoch": 3.4698588791907987,
      "grad_norm": 3.28125,
      "learning_rate": 2.2512447461695826e-05,
      "loss": 2.8693,
      "num_input_tokens_seen": 44452080,
      "step": 21270
    },
    {
      "epoch": 3.470674606411616,
      "grad_norm": 6.3125,
      "learning_rate": 2.2502678897393593e-05,
      "loss": 2.6314,
      "num_input_tokens_seen": 44463232,
      "step": 21275
    },
    {
      "epoch": 3.4714903336324334,
      "grad_norm": 10.9375,
      "learning_rate": 2.2492910718209665e-05,
      "loss": 2.1098,
      "num_input_tokens_seen": 44473232,
      "step": 21280
    },
    {
      "epoch": 3.4723060608532506,
      "grad_norm": 2.4375,
      "learning_rate": 2.2483142925650398e-05,
      "loss": 2.548,
      "num_input_tokens_seen": 44485232,
      "step": 21285
    },
    {
      "epoch": 3.473121788074068,
      "grad_norm": 9.0625,
      "learning_rate": 2.247337552122213e-05,
      "loss": 3.9409,
      "num_input_tokens_seen": 44493840,
      "step": 21290
    },
    {
      "epoch": 3.4739375152948853,
      "grad_norm": 5.84375,
      "learning_rate": 2.24636085064311e-05,
      "loss": 2.1471,
      "num_input_tokens_seen": 44504560,
      "step": 21295
    },
    {
      "epoch": 3.474753242515703,
      "grad_norm": 4.90625,
      "learning_rate": 2.245384188278351e-05,
      "loss": 2.7483,
      "num_input_tokens_seen": 44515136,
      "step": 21300
    },
    {
      "epoch": 3.47556896973652,
      "grad_norm": 3.0625,
      "learning_rate": 2.2444075651785513e-05,
      "loss": 2.0221,
      "num_input_tokens_seen": 44526256,
      "step": 21305
    },
    {
      "epoch": 3.4763846969573375,
      "grad_norm": 7.875,
      "learning_rate": 2.243430981494316e-05,
      "loss": 2.5828,
      "num_input_tokens_seen": 44535824,
      "step": 21310
    },
    {
      "epoch": 3.4772004241781547,
      "grad_norm": 10.625,
      "learning_rate": 2.2424544373762475e-05,
      "loss": 2.4823,
      "num_input_tokens_seen": 44547472,
      "step": 21315
    },
    {
      "epoch": 3.4780161513989722,
      "grad_norm": 11.875,
      "learning_rate": 2.2414779329749418e-05,
      "loss": 1.8334,
      "num_input_tokens_seen": 44557328,
      "step": 21320
    },
    {
      "epoch": 3.4788318786197894,
      "grad_norm": 5.28125,
      "learning_rate": 2.2405014684409873e-05,
      "loss": 2.3925,
      "num_input_tokens_seen": 44567168,
      "step": 21325
    },
    {
      "epoch": 3.479647605840607,
      "grad_norm": 7.375,
      "learning_rate": 2.239525043924968e-05,
      "loss": 2.0841,
      "num_input_tokens_seen": 44577984,
      "step": 21330
    },
    {
      "epoch": 3.480463333061424,
      "grad_norm": 8.375,
      "learning_rate": 2.2385486595774592e-05,
      "loss": 2.8518,
      "num_input_tokens_seen": 44589184,
      "step": 21335
    },
    {
      "epoch": 3.4812790602822417,
      "grad_norm": 9.25,
      "learning_rate": 2.237572315549033e-05,
      "loss": 3.5003,
      "num_input_tokens_seen": 44598624,
      "step": 21340
    },
    {
      "epoch": 3.482094787503059,
      "grad_norm": 4.71875,
      "learning_rate": 2.2365960119902545e-05,
      "loss": 1.7744,
      "num_input_tokens_seen": 44609584,
      "step": 21345
    },
    {
      "epoch": 3.4829105147238764,
      "grad_norm": 3.0,
      "learning_rate": 2.2356197490516806e-05,
      "loss": 2.1558,
      "num_input_tokens_seen": 44621760,
      "step": 21350
    },
    {
      "epoch": 3.4837262419446935,
      "grad_norm": 11.0,
      "learning_rate": 2.234643526883863e-05,
      "loss": 3.8071,
      "num_input_tokens_seen": 44632832,
      "step": 21355
    },
    {
      "epoch": 3.484541969165511,
      "grad_norm": 9.8125,
      "learning_rate": 2.2336673456373497e-05,
      "loss": 2.6449,
      "num_input_tokens_seen": 44642464,
      "step": 21360
    },
    {
      "epoch": 3.4853576963863286,
      "grad_norm": 10.9375,
      "learning_rate": 2.2326912054626772e-05,
      "loss": 2.911,
      "num_input_tokens_seen": 44652192,
      "step": 21365
    },
    {
      "epoch": 3.4861734236071458,
      "grad_norm": 3.953125,
      "learning_rate": 2.2317151065103813e-05,
      "loss": 1.2649,
      "num_input_tokens_seen": 44663072,
      "step": 21370
    },
    {
      "epoch": 3.486989150827963,
      "grad_norm": 6.8125,
      "learning_rate": 2.2307390489309865e-05,
      "loss": 1.854,
      "num_input_tokens_seen": 44674384,
      "step": 21375
    },
    {
      "epoch": 3.4878048780487805,
      "grad_norm": 12.625,
      "learning_rate": 2.2297630328750146e-05,
      "loss": 2.1862,
      "num_input_tokens_seen": 44684848,
      "step": 21380
    },
    {
      "epoch": 3.488620605269598,
      "grad_norm": 8.5625,
      "learning_rate": 2.228787058492979e-05,
      "loss": 2.4327,
      "num_input_tokens_seen": 44695120,
      "step": 21385
    },
    {
      "epoch": 3.489436332490415,
      "grad_norm": 10.25,
      "learning_rate": 2.2278111259353875e-05,
      "loss": 3.6651,
      "num_input_tokens_seen": 44705456,
      "step": 21390
    },
    {
      "epoch": 3.4902520597112328,
      "grad_norm": 10.0,
      "learning_rate": 2.2268352353527395e-05,
      "loss": 3.7307,
      "num_input_tokens_seen": 44715360,
      "step": 21395
    },
    {
      "epoch": 3.49106778693205,
      "grad_norm": 3.203125,
      "learning_rate": 2.225859386895533e-05,
      "loss": 1.287,
      "num_input_tokens_seen": 44724144,
      "step": 21400
    },
    {
      "epoch": 3.49106778693205,
      "eval_loss": 2.545358419418335,
      "eval_runtime": 134.8349,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.109,
      "num_input_tokens_seen": 44724144,
      "step": 21400
    },
    {
      "epoch": 3.4918835141528675,
      "grad_norm": 3.3125,
      "learning_rate": 2.2248835807142525e-05,
      "loss": 1.6725,
      "num_input_tokens_seen": 44734976,
      "step": 21405
    },
    {
      "epoch": 3.4926992413736846,
      "grad_norm": 8.125,
      "learning_rate": 2.2239078169593826e-05,
      "loss": 1.8006,
      "num_input_tokens_seen": 44744560,
      "step": 21410
    },
    {
      "epoch": 3.493514968594502,
      "grad_norm": 9.25,
      "learning_rate": 2.222932095781396e-05,
      "loss": 1.9292,
      "num_input_tokens_seen": 44753984,
      "step": 21415
    },
    {
      "epoch": 3.4943306958153193,
      "grad_norm": 11.375,
      "learning_rate": 2.221956417330762e-05,
      "loss": 2.6013,
      "num_input_tokens_seen": 44763552,
      "step": 21420
    },
    {
      "epoch": 3.495146423036137,
      "grad_norm": 2.21875,
      "learning_rate": 2.2209807817579438e-05,
      "loss": 1.8122,
      "num_input_tokens_seen": 44774832,
      "step": 21425
    },
    {
      "epoch": 3.495962150256954,
      "grad_norm": 9.4375,
      "learning_rate": 2.220005189213394e-05,
      "loss": 1.8778,
      "num_input_tokens_seen": 44784448,
      "step": 21430
    },
    {
      "epoch": 3.4967778774777716,
      "grad_norm": 6.21875,
      "learning_rate": 2.2190296398475624e-05,
      "loss": 2.6066,
      "num_input_tokens_seen": 44795360,
      "step": 21435
    },
    {
      "epoch": 3.4975936046985887,
      "grad_norm": 4.96875,
      "learning_rate": 2.2180541338108926e-05,
      "loss": 2.3608,
      "num_input_tokens_seen": 44806224,
      "step": 21440
    },
    {
      "epoch": 3.4984093319194063,
      "grad_norm": 4.65625,
      "learning_rate": 2.2170786712538176e-05,
      "loss": 1.852,
      "num_input_tokens_seen": 44816992,
      "step": 21445
    },
    {
      "epoch": 3.4992250591402234,
      "grad_norm": 10.0625,
      "learning_rate": 2.216103252326768e-05,
      "loss": 2.8916,
      "num_input_tokens_seen": 44827984,
      "step": 21450
    },
    {
      "epoch": 3.500040786361041,
      "grad_norm": 8.875,
      "learning_rate": 2.2151278771801635e-05,
      "loss": 2.7281,
      "num_input_tokens_seen": 44838480,
      "step": 21455
    },
    {
      "epoch": 3.500856513581858,
      "grad_norm": 8.875,
      "learning_rate": 2.21415254596442e-05,
      "loss": 2.6169,
      "num_input_tokens_seen": 44847680,
      "step": 21460
    },
    {
      "epoch": 3.5016722408026757,
      "grad_norm": 7.03125,
      "learning_rate": 2.213177258829947e-05,
      "loss": 1.8145,
      "num_input_tokens_seen": 44857872,
      "step": 21465
    },
    {
      "epoch": 3.502487968023493,
      "grad_norm": 5.625,
      "learning_rate": 2.2122020159271445e-05,
      "loss": 2.8653,
      "num_input_tokens_seen": 44868096,
      "step": 21470
    },
    {
      "epoch": 3.5033036952443104,
      "grad_norm": 8.5,
      "learning_rate": 2.2112268174064075e-05,
      "loss": 1.9789,
      "num_input_tokens_seen": 44878592,
      "step": 21475
    },
    {
      "epoch": 3.5041194224651275,
      "grad_norm": 5.34375,
      "learning_rate": 2.2102516634181253e-05,
      "loss": 2.1145,
      "num_input_tokens_seen": 44888304,
      "step": 21480
    },
    {
      "epoch": 3.504935149685945,
      "grad_norm": 6.96875,
      "learning_rate": 2.209276554112677e-05,
      "loss": 2.4559,
      "num_input_tokens_seen": 44897872,
      "step": 21485
    },
    {
      "epoch": 3.5057508769067622,
      "grad_norm": 12.25,
      "learning_rate": 2.2083014896404384e-05,
      "loss": 2.3114,
      "num_input_tokens_seen": 44908912,
      "step": 21490
    },
    {
      "epoch": 3.50656660412758,
      "grad_norm": 9.5625,
      "learning_rate": 2.207326470151775e-05,
      "loss": 1.2112,
      "num_input_tokens_seen": 44919760,
      "step": 21495
    },
    {
      "epoch": 3.507382331348397,
      "grad_norm": 2.484375,
      "learning_rate": 2.2063514957970477e-05,
      "loss": 3.0373,
      "num_input_tokens_seen": 44930912,
      "step": 21500
    },
    {
      "epoch": 3.5081980585692145,
      "grad_norm": 10.875,
      "learning_rate": 2.205376566726611e-05,
      "loss": 3.1224,
      "num_input_tokens_seen": 44942144,
      "step": 21505
    },
    {
      "epoch": 3.5090137857900316,
      "grad_norm": 9.75,
      "learning_rate": 2.204401683090809e-05,
      "loss": 2.7421,
      "num_input_tokens_seen": 44953280,
      "step": 21510
    },
    {
      "epoch": 3.5098295130108492,
      "grad_norm": 3.65625,
      "learning_rate": 2.203426845039982e-05,
      "loss": 1.4317,
      "num_input_tokens_seen": 44963136,
      "step": 21515
    },
    {
      "epoch": 3.510645240231667,
      "grad_norm": 2.75,
      "learning_rate": 2.202452052724464e-05,
      "loss": 4.4474,
      "num_input_tokens_seen": 44973664,
      "step": 21520
    },
    {
      "epoch": 3.511460967452484,
      "grad_norm": 9.0,
      "learning_rate": 2.2014773062945777e-05,
      "loss": 2.5751,
      "num_input_tokens_seen": 44983936,
      "step": 21525
    },
    {
      "epoch": 3.512276694673301,
      "grad_norm": 7.25,
      "learning_rate": 2.2005026059006427e-05,
      "loss": 2.6147,
      "num_input_tokens_seen": 44994272,
      "step": 21530
    },
    {
      "epoch": 3.5130924218941186,
      "grad_norm": 8.5625,
      "learning_rate": 2.1995279516929695e-05,
      "loss": 2.6797,
      "num_input_tokens_seen": 45004848,
      "step": 21535
    },
    {
      "epoch": 3.513908149114936,
      "grad_norm": 0.181640625,
      "learning_rate": 2.1985533438218613e-05,
      "loss": 1.8192,
      "num_input_tokens_seen": 45015872,
      "step": 21540
    },
    {
      "epoch": 3.5147238763357533,
      "grad_norm": 6.75,
      "learning_rate": 2.197578782437617e-05,
      "loss": 1.5126,
      "num_input_tokens_seen": 45027344,
      "step": 21545
    },
    {
      "epoch": 3.5155396035565705,
      "grad_norm": 3.34375,
      "learning_rate": 2.196604267690524e-05,
      "loss": 1.0659,
      "num_input_tokens_seen": 45037744,
      "step": 21550
    },
    {
      "epoch": 3.516355330777388,
      "grad_norm": 4.15625,
      "learning_rate": 2.195629799730865e-05,
      "loss": 2.9049,
      "num_input_tokens_seen": 45048704,
      "step": 21555
    },
    {
      "epoch": 3.5171710579982056,
      "grad_norm": 7.65625,
      "learning_rate": 2.1946553787089173e-05,
      "loss": 2.7401,
      "num_input_tokens_seen": 45058864,
      "step": 21560
    },
    {
      "epoch": 3.5179867852190227,
      "grad_norm": 13.75,
      "learning_rate": 2.193681004774947e-05,
      "loss": 3.1275,
      "num_input_tokens_seen": 45069760,
      "step": 21565
    },
    {
      "epoch": 3.51880251243984,
      "grad_norm": 8.75,
      "learning_rate": 2.1927066780792154e-05,
      "loss": 2.7207,
      "num_input_tokens_seen": 45078912,
      "step": 21570
    },
    {
      "epoch": 3.5196182396606575,
      "grad_norm": 3.890625,
      "learning_rate": 2.191732398771975e-05,
      "loss": 1.6301,
      "num_input_tokens_seen": 45090800,
      "step": 21575
    },
    {
      "epoch": 3.520433966881475,
      "grad_norm": 5.03125,
      "learning_rate": 2.1907581670034725e-05,
      "loss": 1.3946,
      "num_input_tokens_seen": 45101264,
      "step": 21580
    },
    {
      "epoch": 3.521249694102292,
      "grad_norm": 2.953125,
      "learning_rate": 2.189783982923948e-05,
      "loss": 3.0403,
      "num_input_tokens_seen": 45110288,
      "step": 21585
    },
    {
      "epoch": 3.5220654213231093,
      "grad_norm": 5.5625,
      "learning_rate": 2.1888098466836303e-05,
      "loss": 1.342,
      "num_input_tokens_seen": 45121680,
      "step": 21590
    },
    {
      "epoch": 3.522881148543927,
      "grad_norm": 6.75,
      "learning_rate": 2.1878357584327457e-05,
      "loss": 2.1924,
      "num_input_tokens_seen": 45132944,
      "step": 21595
    },
    {
      "epoch": 3.5236968757647444,
      "grad_norm": 6.53125,
      "learning_rate": 2.1868617183215103e-05,
      "loss": 1.2195,
      "num_input_tokens_seen": 45143632,
      "step": 21600
    },
    {
      "epoch": 3.5236968757647444,
      "eval_loss": 2.533942699432373,
      "eval_runtime": 134.9233,
      "eval_samples_per_second": 20.197,
      "eval_steps_per_second": 10.102,
      "num_input_tokens_seen": 45143632,
      "step": 21600
    },
    {
      "epoch": 3.5245126029855616,
      "grad_norm": 4.8125,
      "learning_rate": 2.1858877265001327e-05,
      "loss": 1.9932,
      "num_input_tokens_seen": 45153424,
      "step": 21605
    },
    {
      "epoch": 3.525328330206379,
      "grad_norm": 13.75,
      "learning_rate": 2.184913783118816e-05,
      "loss": 2.5312,
      "num_input_tokens_seen": 45162688,
      "step": 21610
    },
    {
      "epoch": 3.5261440574271963,
      "grad_norm": 8.1875,
      "learning_rate": 2.1839398883277522e-05,
      "loss": 2.6139,
      "num_input_tokens_seen": 45173392,
      "step": 21615
    },
    {
      "epoch": 3.526959784648014,
      "grad_norm": 10.625,
      "learning_rate": 2.182966042277129e-05,
      "loss": 1.7992,
      "num_input_tokens_seen": 45182640,
      "step": 21620
    },
    {
      "epoch": 3.527775511868831,
      "grad_norm": 7.3125,
      "learning_rate": 2.181992245117128e-05,
      "loss": 2.4849,
      "num_input_tokens_seen": 45192848,
      "step": 21625
    },
    {
      "epoch": 3.5285912390896486,
      "grad_norm": 4.59375,
      "learning_rate": 2.181018496997918e-05,
      "loss": 2.4438,
      "num_input_tokens_seen": 45203920,
      "step": 21630
    },
    {
      "epoch": 3.5294069663104657,
      "grad_norm": 3.59375,
      "learning_rate": 2.1800447980696648e-05,
      "loss": 2.9273,
      "num_input_tokens_seen": 45213280,
      "step": 21635
    },
    {
      "epoch": 3.5302226935312833,
      "grad_norm": 10.5,
      "learning_rate": 2.1790711484825248e-05,
      "loss": 3.201,
      "num_input_tokens_seen": 45223712,
      "step": 21640
    },
    {
      "epoch": 3.5310384207521004,
      "grad_norm": 5.875,
      "learning_rate": 2.178097548386646e-05,
      "loss": 2.8379,
      "num_input_tokens_seen": 45234368,
      "step": 21645
    },
    {
      "epoch": 3.531854147972918,
      "grad_norm": 9.6875,
      "learning_rate": 2.1771239979321712e-05,
      "loss": 2.3264,
      "num_input_tokens_seen": 45244656,
      "step": 21650
    },
    {
      "epoch": 3.532669875193735,
      "grad_norm": 8.875,
      "learning_rate": 2.1761504972692327e-05,
      "loss": 2.1105,
      "num_input_tokens_seen": 45255104,
      "step": 21655
    },
    {
      "epoch": 3.5334856024145527,
      "grad_norm": 12.9375,
      "learning_rate": 2.1751770465479572e-05,
      "loss": 2.3326,
      "num_input_tokens_seen": 45266464,
      "step": 21660
    },
    {
      "epoch": 3.53430132963537,
      "grad_norm": 5.4375,
      "learning_rate": 2.174203645918464e-05,
      "loss": 1.3326,
      "num_input_tokens_seen": 45276544,
      "step": 21665
    },
    {
      "epoch": 3.5351170568561874,
      "grad_norm": 4.4375,
      "learning_rate": 2.1732302955308624e-05,
      "loss": 0.8551,
      "num_input_tokens_seen": 45287312,
      "step": 21670
    },
    {
      "epoch": 3.5359327840770045,
      "grad_norm": 4.03125,
      "learning_rate": 2.172256995535255e-05,
      "loss": 2.3107,
      "num_input_tokens_seen": 45298960,
      "step": 21675
    },
    {
      "epoch": 3.536748511297822,
      "grad_norm": 9.6875,
      "learning_rate": 2.171283746081739e-05,
      "loss": 2.5911,
      "num_input_tokens_seen": 45310352,
      "step": 21680
    },
    {
      "epoch": 3.537564238518639,
      "grad_norm": 8.75,
      "learning_rate": 2.1703105473203988e-05,
      "loss": 2.758,
      "num_input_tokens_seen": 45321376,
      "step": 21685
    },
    {
      "epoch": 3.538379965739457,
      "grad_norm": 6.625,
      "learning_rate": 2.1693373994013168e-05,
      "loss": 3.3842,
      "num_input_tokens_seen": 45332192,
      "step": 21690
    },
    {
      "epoch": 3.539195692960274,
      "grad_norm": 8.625,
      "learning_rate": 2.168364302474562e-05,
      "loss": 2.7232,
      "num_input_tokens_seen": 45342800,
      "step": 21695
    },
    {
      "epoch": 3.5400114201810915,
      "grad_norm": 16.75,
      "learning_rate": 2.167391256690199e-05,
      "loss": 1.7412,
      "num_input_tokens_seen": 45354384,
      "step": 21700
    },
    {
      "epoch": 3.5408271474019086,
      "grad_norm": 10.0,
      "learning_rate": 2.1664182621982855e-05,
      "loss": 1.9621,
      "num_input_tokens_seen": 45364752,
      "step": 21705
    },
    {
      "epoch": 3.541642874622726,
      "grad_norm": 5.71875,
      "learning_rate": 2.1654453191488673e-05,
      "loss": 2.2376,
      "num_input_tokens_seen": 45374640,
      "step": 21710
    },
    {
      "epoch": 3.5424586018435438,
      "grad_norm": 9.8125,
      "learning_rate": 2.1644724276919846e-05,
      "loss": 3.0,
      "num_input_tokens_seen": 45385920,
      "step": 21715
    },
    {
      "epoch": 3.543274329064361,
      "grad_norm": 7.875,
      "learning_rate": 2.1634995879776715e-05,
      "loss": 1.405,
      "num_input_tokens_seen": 45396848,
      "step": 21720
    },
    {
      "epoch": 3.544090056285178,
      "grad_norm": 10.0625,
      "learning_rate": 2.162526800155949e-05,
      "loss": 1.943,
      "num_input_tokens_seen": 45405856,
      "step": 21725
    },
    {
      "epoch": 3.5449057835059956,
      "grad_norm": 5.75,
      "learning_rate": 2.1615540643768363e-05,
      "loss": 2.9437,
      "num_input_tokens_seen": 45417600,
      "step": 21730
    },
    {
      "epoch": 3.545721510726813,
      "grad_norm": 11.625,
      "learning_rate": 2.160581380790339e-05,
      "loss": 3.1273,
      "num_input_tokens_seen": 45427856,
      "step": 21735
    },
    {
      "epoch": 3.5465372379476303,
      "grad_norm": 2.953125,
      "learning_rate": 2.1596087495464586e-05,
      "loss": 2.0303,
      "num_input_tokens_seen": 45437600,
      "step": 21740
    },
    {
      "epoch": 3.5473529651684474,
      "grad_norm": 4.96875,
      "learning_rate": 2.1586361707951866e-05,
      "loss": 2.2672,
      "num_input_tokens_seen": 45449312,
      "step": 21745
    },
    {
      "epoch": 3.548168692389265,
      "grad_norm": 7.75,
      "learning_rate": 2.157663644686507e-05,
      "loss": 2.6136,
      "num_input_tokens_seen": 45459520,
      "step": 21750
    },
    {
      "epoch": 3.5489844196100826,
      "grad_norm": 10.6875,
      "learning_rate": 2.156691171370396e-05,
      "loss": 2.3074,
      "num_input_tokens_seen": 45470320,
      "step": 21755
    },
    {
      "epoch": 3.5498001468308997,
      "grad_norm": 4.25,
      "learning_rate": 2.1557187509968195e-05,
      "loss": 3.6416,
      "num_input_tokens_seen": 45480576,
      "step": 21760
    },
    {
      "epoch": 3.550615874051717,
      "grad_norm": 6.78125,
      "learning_rate": 2.1547463837157382e-05,
      "loss": 2.7181,
      "num_input_tokens_seen": 45489344,
      "step": 21765
    },
    {
      "epoch": 3.5514316012725344,
      "grad_norm": 7.8125,
      "learning_rate": 2.1537740696771045e-05,
      "loss": 2.2638,
      "num_input_tokens_seen": 45500592,
      "step": 21770
    },
    {
      "epoch": 3.552247328493352,
      "grad_norm": 3.4375,
      "learning_rate": 2.1528018090308587e-05,
      "loss": 1.5722,
      "num_input_tokens_seen": 45511568,
      "step": 21775
    },
    {
      "epoch": 3.553063055714169,
      "grad_norm": 6.53125,
      "learning_rate": 2.151829601926938e-05,
      "loss": 1.9293,
      "num_input_tokens_seen": 45521840,
      "step": 21780
    },
    {
      "epoch": 3.5538787829349863,
      "grad_norm": 5.34375,
      "learning_rate": 2.1508574485152684e-05,
      "loss": 2.0968,
      "num_input_tokens_seen": 45533632,
      "step": 21785
    },
    {
      "epoch": 3.554694510155804,
      "grad_norm": 7.78125,
      "learning_rate": 2.1498853489457667e-05,
      "loss": 1.8898,
      "num_input_tokens_seen": 45544752,
      "step": 21790
    },
    {
      "epoch": 3.5555102373766214,
      "grad_norm": 8.9375,
      "learning_rate": 2.1489133033683455e-05,
      "loss": 1.9762,
      "num_input_tokens_seen": 45556336,
      "step": 21795
    },
    {
      "epoch": 3.5563259645974385,
      "grad_norm": 8.5625,
      "learning_rate": 2.1479413119329038e-05,
      "loss": 1.8093,
      "num_input_tokens_seen": 45567152,
      "step": 21800
    },
    {
      "epoch": 3.5563259645974385,
      "eval_loss": 2.5370492935180664,
      "eval_runtime": 134.8999,
      "eval_samples_per_second": 20.2,
      "eval_steps_per_second": 10.104,
      "num_input_tokens_seen": 45567152,
      "step": 21800
    },
    {
      "epoch": 3.557141691818256,
      "grad_norm": 5.1875,
      "learning_rate": 2.1469693747893355e-05,
      "loss": 2.0237,
      "num_input_tokens_seen": 45578368,
      "step": 21805
    },
    {
      "epoch": 3.5579574190390733,
      "grad_norm": 3.390625,
      "learning_rate": 2.1459974920875274e-05,
      "loss": 0.9706,
      "num_input_tokens_seen": 45588992,
      "step": 21810
    },
    {
      "epoch": 3.558773146259891,
      "grad_norm": 14.4375,
      "learning_rate": 2.145025663977354e-05,
      "loss": 2.9433,
      "num_input_tokens_seen": 45599840,
      "step": 21815
    },
    {
      "epoch": 3.559588873480708,
      "grad_norm": 5.6875,
      "learning_rate": 2.1440538906086844e-05,
      "loss": 1.8411,
      "num_input_tokens_seen": 45609632,
      "step": 21820
    },
    {
      "epoch": 3.5604046007015255,
      "grad_norm": 9.9375,
      "learning_rate": 2.1430821721313782e-05,
      "loss": 1.7808,
      "num_input_tokens_seen": 45620832,
      "step": 21825
    },
    {
      "epoch": 3.5612203279223427,
      "grad_norm": 8.8125,
      "learning_rate": 2.142110508695286e-05,
      "loss": 2.4799,
      "num_input_tokens_seen": 45631616,
      "step": 21830
    },
    {
      "epoch": 3.5620360551431602,
      "grad_norm": 5.75,
      "learning_rate": 2.1411389004502515e-05,
      "loss": 1.2807,
      "num_input_tokens_seen": 45642048,
      "step": 21835
    },
    {
      "epoch": 3.5628517823639774,
      "grad_norm": 6.84375,
      "learning_rate": 2.140167347546107e-05,
      "loss": 2.209,
      "num_input_tokens_seen": 45653040,
      "step": 21840
    },
    {
      "epoch": 3.563667509584795,
      "grad_norm": 4.21875,
      "learning_rate": 2.1391958501326793e-05,
      "loss": 3.0142,
      "num_input_tokens_seen": 45663344,
      "step": 21845
    },
    {
      "epoch": 3.564483236805612,
      "grad_norm": 6.0,
      "learning_rate": 2.1382244083597873e-05,
      "loss": 1.9283,
      "num_input_tokens_seen": 45672896,
      "step": 21850
    },
    {
      "epoch": 3.5652989640264297,
      "grad_norm": 4.34375,
      "learning_rate": 2.137253022377237e-05,
      "loss": 3.0355,
      "num_input_tokens_seen": 45684048,
      "step": 21855
    },
    {
      "epoch": 3.566114691247247,
      "grad_norm": 4.84375,
      "learning_rate": 2.136281692334829e-05,
      "loss": 0.8018,
      "num_input_tokens_seen": 45695088,
      "step": 21860
    },
    {
      "epoch": 3.5669304184680644,
      "grad_norm": 8.125,
      "learning_rate": 2.135310418382356e-05,
      "loss": 1.307,
      "num_input_tokens_seen": 45705232,
      "step": 21865
    },
    {
      "epoch": 3.5677461456888815,
      "grad_norm": 6.1875,
      "learning_rate": 2.134339200669598e-05,
      "loss": 4.3258,
      "num_input_tokens_seen": 45715856,
      "step": 21870
    },
    {
      "epoch": 3.568561872909699,
      "grad_norm": 5.90625,
      "learning_rate": 2.133368039346331e-05,
      "loss": 2.3396,
      "num_input_tokens_seen": 45725552,
      "step": 21875
    },
    {
      "epoch": 3.569377600130516,
      "grad_norm": 6.375,
      "learning_rate": 2.1323969345623195e-05,
      "loss": 2.3826,
      "num_input_tokens_seen": 45737472,
      "step": 21880
    },
    {
      "epoch": 3.5701933273513338,
      "grad_norm": 3.34375,
      "learning_rate": 2.1314258864673207e-05,
      "loss": 1.5054,
      "num_input_tokens_seen": 45748224,
      "step": 21885
    },
    {
      "epoch": 3.5710090545721513,
      "grad_norm": 5.75,
      "learning_rate": 2.130454895211082e-05,
      "loss": 2.9304,
      "num_input_tokens_seen": 45757696,
      "step": 21890
    },
    {
      "epoch": 3.5718247817929685,
      "grad_norm": 5.84375,
      "learning_rate": 2.129483960943342e-05,
      "loss": 2.1773,
      "num_input_tokens_seen": 45769200,
      "step": 21895
    },
    {
      "epoch": 3.5726405090137856,
      "grad_norm": 4.4375,
      "learning_rate": 2.128513083813831e-05,
      "loss": 1.7327,
      "num_input_tokens_seen": 45779520,
      "step": 21900
    },
    {
      "epoch": 3.573456236234603,
      "grad_norm": 9.625,
      "learning_rate": 2.1275422639722724e-05,
      "loss": 1.6593,
      "num_input_tokens_seen": 45790048,
      "step": 21905
    },
    {
      "epoch": 3.5742719634554208,
      "grad_norm": 4.3125,
      "learning_rate": 2.126571501568376e-05,
      "loss": 1.8003,
      "num_input_tokens_seen": 45800144,
      "step": 21910
    },
    {
      "epoch": 3.575087690676238,
      "grad_norm": 8.25,
      "learning_rate": 2.1256007967518478e-05,
      "loss": 1.1462,
      "num_input_tokens_seen": 45809344,
      "step": 21915
    },
    {
      "epoch": 3.575903417897055,
      "grad_norm": 7.21875,
      "learning_rate": 2.124630149672381e-05,
      "loss": 3.0097,
      "num_input_tokens_seen": 45820096,
      "step": 21920
    },
    {
      "epoch": 3.5767191451178726,
      "grad_norm": 12.6875,
      "learning_rate": 2.1236595604796624e-05,
      "loss": 1.9164,
      "num_input_tokens_seen": 45830448,
      "step": 21925
    },
    {
      "epoch": 3.57753487233869,
      "grad_norm": 5.90625,
      "learning_rate": 2.1226890293233693e-05,
      "loss": 1.6277,
      "num_input_tokens_seen": 45840032,
      "step": 21930
    },
    {
      "epoch": 3.5783505995595073,
      "grad_norm": 10.5625,
      "learning_rate": 2.1217185563531694e-05,
      "loss": 3.0302,
      "num_input_tokens_seen": 45850704,
      "step": 21935
    },
    {
      "epoch": 3.5791663267803244,
      "grad_norm": 7.1875,
      "learning_rate": 2.120748141718721e-05,
      "loss": 3.5045,
      "num_input_tokens_seen": 45861296,
      "step": 21940
    },
    {
      "epoch": 3.579982054001142,
      "grad_norm": 2.5,
      "learning_rate": 2.1197777855696765e-05,
      "loss": 2.5429,
      "num_input_tokens_seen": 45869984,
      "step": 21945
    },
    {
      "epoch": 3.5807977812219596,
      "grad_norm": 2.75,
      "learning_rate": 2.1188074880556746e-05,
      "loss": 1.6309,
      "num_input_tokens_seen": 45879968,
      "step": 21950
    },
    {
      "epoch": 3.5816135084427767,
      "grad_norm": 9.75,
      "learning_rate": 2.1178372493263495e-05,
      "loss": 3.3112,
      "num_input_tokens_seen": 45890608,
      "step": 21955
    },
    {
      "epoch": 3.582429235663594,
      "grad_norm": 3.78125,
      "learning_rate": 2.116867069531322e-05,
      "loss": 1.9357,
      "num_input_tokens_seen": 45901504,
      "step": 21960
    },
    {
      "epoch": 3.5832449628844114,
      "grad_norm": 4.78125,
      "learning_rate": 2.1158969488202073e-05,
      "loss": 2.1044,
      "num_input_tokens_seen": 45911232,
      "step": 21965
    },
    {
      "epoch": 3.584060690105229,
      "grad_norm": 3.375,
      "learning_rate": 2.114926887342611e-05,
      "loss": 1.7192,
      "num_input_tokens_seen": 45922368,
      "step": 21970
    },
    {
      "epoch": 3.584876417326046,
      "grad_norm": 8.1875,
      "learning_rate": 2.113956885248127e-05,
      "loss": 2.5827,
      "num_input_tokens_seen": 45932448,
      "step": 21975
    },
    {
      "epoch": 3.5856921445468637,
      "grad_norm": 7.5,
      "learning_rate": 2.112986942686342e-05,
      "loss": 4.5316,
      "num_input_tokens_seen": 45942560,
      "step": 21980
    },
    {
      "epoch": 3.586507871767681,
      "grad_norm": 6.25,
      "learning_rate": 2.112017059806835e-05,
      "loss": 2.7886,
      "num_input_tokens_seen": 45951824,
      "step": 21985
    },
    {
      "epoch": 3.5873235989884984,
      "grad_norm": 6.75,
      "learning_rate": 2.1110472367591724e-05,
      "loss": 2.4559,
      "num_input_tokens_seen": 45962640,
      "step": 21990
    },
    {
      "epoch": 3.5881393262093155,
      "grad_norm": 8.125,
      "learning_rate": 2.1100774736929145e-05,
      "loss": 3.4588,
      "num_input_tokens_seen": 45972624,
      "step": 21995
    },
    {
      "epoch": 3.588955053430133,
      "grad_norm": 2.84375,
      "learning_rate": 2.10910777075761e-05,
      "loss": 2.1424,
      "num_input_tokens_seen": 45983168,
      "step": 22000
    },
    {
      "epoch": 3.588955053430133,
      "eval_loss": 2.551025152206421,
      "eval_runtime": 134.8504,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 45983168,
      "step": 22000
    },
    {
      "epoch": 3.5897707806509502,
      "grad_norm": 8.75,
      "learning_rate": 2.108138128102799e-05,
      "loss": 2.6027,
      "num_input_tokens_seen": 45994816,
      "step": 22005
    },
    {
      "epoch": 3.590586507871768,
      "grad_norm": 9.625,
      "learning_rate": 2.107168545878014e-05,
      "loss": 2.3617,
      "num_input_tokens_seen": 46004384,
      "step": 22010
    },
    {
      "epoch": 3.591402235092585,
      "grad_norm": 5.5625,
      "learning_rate": 2.106199024232775e-05,
      "loss": 1.235,
      "num_input_tokens_seen": 46015552,
      "step": 22015
    },
    {
      "epoch": 3.5922179623134025,
      "grad_norm": 2.125,
      "learning_rate": 2.105229563316595e-05,
      "loss": 1.2198,
      "num_input_tokens_seen": 46025024,
      "step": 22020
    },
    {
      "epoch": 3.5930336895342196,
      "grad_norm": 7.78125,
      "learning_rate": 2.1042601632789784e-05,
      "loss": 3.5261,
      "num_input_tokens_seen": 46035584,
      "step": 22025
    },
    {
      "epoch": 3.593849416755037,
      "grad_norm": 8.8125,
      "learning_rate": 2.103290824269417e-05,
      "loss": 2.107,
      "num_input_tokens_seen": 46046112,
      "step": 22030
    },
    {
      "epoch": 3.5946651439758543,
      "grad_norm": 10.75,
      "learning_rate": 2.1023215464373965e-05,
      "loss": 3.3555,
      "num_input_tokens_seen": 46057376,
      "step": 22035
    },
    {
      "epoch": 3.595480871196672,
      "grad_norm": 13.1875,
      "learning_rate": 2.1013523299323908e-05,
      "loss": 2.4826,
      "num_input_tokens_seen": 46068944,
      "step": 22040
    },
    {
      "epoch": 3.596296598417489,
      "grad_norm": 5.71875,
      "learning_rate": 2.1003831749038654e-05,
      "loss": 2.3215,
      "num_input_tokens_seen": 46080352,
      "step": 22045
    },
    {
      "epoch": 3.5971123256383066,
      "grad_norm": 13.25,
      "learning_rate": 2.099414081501277e-05,
      "loss": 2.1673,
      "num_input_tokens_seen": 46089872,
      "step": 22050
    },
    {
      "epoch": 3.5979280528591238,
      "grad_norm": 10.8125,
      "learning_rate": 2.09844504987407e-05,
      "loss": 2.5795,
      "num_input_tokens_seen": 46098160,
      "step": 22055
    },
    {
      "epoch": 3.5987437800799413,
      "grad_norm": 9.875,
      "learning_rate": 2.097476080171683e-05,
      "loss": 2.2625,
      "num_input_tokens_seen": 46108704,
      "step": 22060
    },
    {
      "epoch": 3.5995595073007585,
      "grad_norm": 8.4375,
      "learning_rate": 2.0965071725435436e-05,
      "loss": 2.4275,
      "num_input_tokens_seen": 46119664,
      "step": 22065
    },
    {
      "epoch": 3.600375234521576,
      "grad_norm": 11.0625,
      "learning_rate": 2.0955383271390684e-05,
      "loss": 2.9259,
      "num_input_tokens_seen": 46128512,
      "step": 22070
    },
    {
      "epoch": 3.601190961742393,
      "grad_norm": 5.5,
      "learning_rate": 2.094569544107666e-05,
      "loss": 2.2409,
      "num_input_tokens_seen": 46137680,
      "step": 22075
    },
    {
      "epoch": 3.6020066889632107,
      "grad_norm": 2.40625,
      "learning_rate": 2.093600823598735e-05,
      "loss": 3.4845,
      "num_input_tokens_seen": 46148512,
      "step": 22080
    },
    {
      "epoch": 3.6028224161840283,
      "grad_norm": 13.125,
      "learning_rate": 2.092632165761663e-05,
      "loss": 2.5582,
      "num_input_tokens_seen": 46160000,
      "step": 22085
    },
    {
      "epoch": 3.6036381434048455,
      "grad_norm": 6.28125,
      "learning_rate": 2.091663570745832e-05,
      "loss": 4.3453,
      "num_input_tokens_seen": 46170320,
      "step": 22090
    },
    {
      "epoch": 3.6044538706256626,
      "grad_norm": 13.125,
      "learning_rate": 2.0906950387006086e-05,
      "loss": 2.9761,
      "num_input_tokens_seen": 46181552,
      "step": 22095
    },
    {
      "epoch": 3.60526959784648,
      "grad_norm": 4.28125,
      "learning_rate": 2.0897265697753543e-05,
      "loss": 1.2138,
      "num_input_tokens_seen": 46191984,
      "step": 22100
    },
    {
      "epoch": 3.6060853250672977,
      "grad_norm": 1.84375,
      "learning_rate": 2.088758164119419e-05,
      "loss": 1.7952,
      "num_input_tokens_seen": 46201616,
      "step": 22105
    },
    {
      "epoch": 3.606901052288115,
      "grad_norm": 8.625,
      "learning_rate": 2.0877898218821428e-05,
      "loss": 2.1018,
      "num_input_tokens_seen": 46211504,
      "step": 22110
    },
    {
      "epoch": 3.607716779508932,
      "grad_norm": 9.5625,
      "learning_rate": 2.0868215432128565e-05,
      "loss": 1.783,
      "num_input_tokens_seen": 46222768,
      "step": 22115
    },
    {
      "epoch": 3.6085325067297496,
      "grad_norm": 6.84375,
      "learning_rate": 2.0858533282608796e-05,
      "loss": 2.9174,
      "num_input_tokens_seen": 46233168,
      "step": 22120
    },
    {
      "epoch": 3.609348233950567,
      "grad_norm": 5.96875,
      "learning_rate": 2.084885177175524e-05,
      "loss": 2.4737,
      "num_input_tokens_seen": 46243104,
      "step": 22125
    },
    {
      "epoch": 3.6101639611713843,
      "grad_norm": 6.34375,
      "learning_rate": 2.0839170901060917e-05,
      "loss": 3.0332,
      "num_input_tokens_seen": 46254512,
      "step": 22130
    },
    {
      "epoch": 3.6109796883922014,
      "grad_norm": 7.9375,
      "learning_rate": 2.082949067201872e-05,
      "loss": 1.6607,
      "num_input_tokens_seen": 46264352,
      "step": 22135
    },
    {
      "epoch": 3.611795415613019,
      "grad_norm": 9.6875,
      "learning_rate": 2.0819811086121475e-05,
      "loss": 2.3821,
      "num_input_tokens_seen": 46274736,
      "step": 22140
    },
    {
      "epoch": 3.6126111428338366,
      "grad_norm": 0.1318359375,
      "learning_rate": 2.08101321448619e-05,
      "loss": 0.5144,
      "num_input_tokens_seen": 46285680,
      "step": 22145
    },
    {
      "epoch": 3.6134268700546537,
      "grad_norm": 7.5,
      "learning_rate": 2.080045384973259e-05,
      "loss": 2.5745,
      "num_input_tokens_seen": 46296784,
      "step": 22150
    },
    {
      "epoch": 3.614242597275471,
      "grad_norm": 9.1875,
      "learning_rate": 2.0790776202226082e-05,
      "loss": 2.8913,
      "num_input_tokens_seen": 46307888,
      "step": 22155
    },
    {
      "epoch": 3.6150583244962884,
      "grad_norm": 6.65625,
      "learning_rate": 2.078109920383477e-05,
      "loss": 3.1692,
      "num_input_tokens_seen": 46319136,
      "step": 22160
    },
    {
      "epoch": 3.615874051717106,
      "grad_norm": 9.1875,
      "learning_rate": 2.0771422856050978e-05,
      "loss": 2.8868,
      "num_input_tokens_seen": 46329120,
      "step": 22165
    },
    {
      "epoch": 3.616689778937923,
      "grad_norm": 10.5,
      "learning_rate": 2.076174716036693e-05,
      "loss": 1.8302,
      "num_input_tokens_seen": 46339168,
      "step": 22170
    },
    {
      "epoch": 3.6175055061587407,
      "grad_norm": 8.0,
      "learning_rate": 2.075207211827472e-05,
      "loss": 1.7734,
      "num_input_tokens_seen": 46350352,
      "step": 22175
    },
    {
      "epoch": 3.618321233379558,
      "grad_norm": 7.125,
      "learning_rate": 2.074239773126638e-05,
      "loss": 1.8889,
      "num_input_tokens_seen": 46361728,
      "step": 22180
    },
    {
      "epoch": 3.6191369606003754,
      "grad_norm": 7.21875,
      "learning_rate": 2.073272400083382e-05,
      "loss": 3.2668,
      "num_input_tokens_seen": 46372224,
      "step": 22185
    },
    {
      "epoch": 3.6199526878211925,
      "grad_norm": 10.5,
      "learning_rate": 2.072305092846883e-05,
      "loss": 3.6411,
      "num_input_tokens_seen": 46382016,
      "step": 22190
    },
    {
      "epoch": 3.62076841504201,
      "grad_norm": 8.25,
      "learning_rate": 2.0713378515663152e-05,
      "loss": 2.2546,
      "num_input_tokens_seen": 46391904,
      "step": 22195
    },
    {
      "epoch": 3.621584142262827,
      "grad_norm": 9.375,
      "learning_rate": 2.070370676390836e-05,
      "loss": 3.3381,
      "num_input_tokens_seen": 46401184,
      "step": 22200
    },
    {
      "epoch": 3.621584142262827,
      "eval_loss": 2.5417721271514893,
      "eval_runtime": 135.0376,
      "eval_samples_per_second": 20.18,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 46401184,
      "step": 22200
    },
    {
      "epoch": 3.622399869483645,
      "grad_norm": 5.59375,
      "learning_rate": 2.0694035674695974e-05,
      "loss": 1.9254,
      "num_input_tokens_seen": 46411840,
      "step": 22205
    },
    {
      "epoch": 3.623215596704462,
      "grad_norm": 4.78125,
      "learning_rate": 2.0684365249517416e-05,
      "loss": 0.9247,
      "num_input_tokens_seen": 46423152,
      "step": 22210
    },
    {
      "epoch": 3.6240313239252795,
      "grad_norm": 0.890625,
      "learning_rate": 2.067469548986396e-05,
      "loss": 1.7019,
      "num_input_tokens_seen": 46432656,
      "step": 22215
    },
    {
      "epoch": 3.6248470511460966,
      "grad_norm": 7.25,
      "learning_rate": 2.066502639722681e-05,
      "loss": 3.3078,
      "num_input_tokens_seen": 46443392,
      "step": 22220
    },
    {
      "epoch": 3.625662778366914,
      "grad_norm": 15.5,
      "learning_rate": 2.065535797309708e-05,
      "loss": 2.2293,
      "num_input_tokens_seen": 46453296,
      "step": 22225
    },
    {
      "epoch": 3.6264785055877313,
      "grad_norm": 9.3125,
      "learning_rate": 2.0645690218965736e-05,
      "loss": 2.746,
      "num_input_tokens_seen": 46464000,
      "step": 22230
    },
    {
      "epoch": 3.627294232808549,
      "grad_norm": 7.71875,
      "learning_rate": 2.063602313632369e-05,
      "loss": 2.4252,
      "num_input_tokens_seen": 46475120,
      "step": 22235
    },
    {
      "epoch": 3.628109960029366,
      "grad_norm": 7.34375,
      "learning_rate": 2.0626356726661704e-05,
      "loss": 2.3792,
      "num_input_tokens_seen": 46486000,
      "step": 22240
    },
    {
      "epoch": 3.6289256872501836,
      "grad_norm": 7.96875,
      "learning_rate": 2.0616690991470477e-05,
      "loss": 2.2951,
      "num_input_tokens_seen": 46496288,
      "step": 22245
    },
    {
      "epoch": 3.6297414144710007,
      "grad_norm": 5.21875,
      "learning_rate": 2.0607025932240595e-05,
      "loss": 2.6267,
      "num_input_tokens_seen": 46504000,
      "step": 22250
    },
    {
      "epoch": 3.6305571416918183,
      "grad_norm": 9.375,
      "learning_rate": 2.059736155046251e-05,
      "loss": 2.8624,
      "num_input_tokens_seen": 46513056,
      "step": 22255
    },
    {
      "epoch": 3.631372868912636,
      "grad_norm": 9.8125,
      "learning_rate": 2.0587697847626603e-05,
      "loss": 3.029,
      "num_input_tokens_seen": 46523488,
      "step": 22260
    },
    {
      "epoch": 3.632188596133453,
      "grad_norm": 5.96875,
      "learning_rate": 2.057803482522314e-05,
      "loss": 0.8649,
      "num_input_tokens_seen": 46533808,
      "step": 22265
    },
    {
      "epoch": 3.63300432335427,
      "grad_norm": 13.0,
      "learning_rate": 2.056837248474227e-05,
      "loss": 2.7777,
      "num_input_tokens_seen": 46544672,
      "step": 22270
    },
    {
      "epoch": 3.6338200505750877,
      "grad_norm": 2.390625,
      "learning_rate": 2.0558710827674064e-05,
      "loss": 3.1123,
      "num_input_tokens_seen": 46554160,
      "step": 22275
    },
    {
      "epoch": 3.6346357777959053,
      "grad_norm": 4.46875,
      "learning_rate": 2.054904985550845e-05,
      "loss": 1.976,
      "num_input_tokens_seen": 46564672,
      "step": 22280
    },
    {
      "epoch": 3.6354515050167224,
      "grad_norm": 5.5,
      "learning_rate": 2.0539389569735287e-05,
      "loss": 1.4411,
      "num_input_tokens_seen": 46574400,
      "step": 22285
    },
    {
      "epoch": 3.6362672322375396,
      "grad_norm": 5.03125,
      "learning_rate": 2.052972997184431e-05,
      "loss": 2.1281,
      "num_input_tokens_seen": 46583776,
      "step": 22290
    },
    {
      "epoch": 3.637082959458357,
      "grad_norm": 9.8125,
      "learning_rate": 2.0520071063325146e-05,
      "loss": 2.5547,
      "num_input_tokens_seen": 46593552,
      "step": 22295
    },
    {
      "epoch": 3.6378986866791747,
      "grad_norm": 3.171875,
      "learning_rate": 2.051041284566732e-05,
      "loss": 2.0177,
      "num_input_tokens_seen": 46603408,
      "step": 22300
    },
    {
      "epoch": 3.638714413899992,
      "grad_norm": 7.15625,
      "learning_rate": 2.050075532036026e-05,
      "loss": 2.819,
      "num_input_tokens_seen": 46614000,
      "step": 22305
    },
    {
      "epoch": 3.639530141120809,
      "grad_norm": 5.8125,
      "learning_rate": 2.0491098488893264e-05,
      "loss": 1.6924,
      "num_input_tokens_seen": 46624192,
      "step": 22310
    },
    {
      "epoch": 3.6403458683416265,
      "grad_norm": 14.1875,
      "learning_rate": 2.0481442352755546e-05,
      "loss": 3.2055,
      "num_input_tokens_seen": 46635600,
      "step": 22315
    },
    {
      "epoch": 3.641161595562444,
      "grad_norm": 14.5,
      "learning_rate": 2.0471786913436198e-05,
      "loss": 3.002,
      "num_input_tokens_seen": 46646656,
      "step": 22320
    },
    {
      "epoch": 3.6419773227832613,
      "grad_norm": 5.15625,
      "learning_rate": 2.0462132172424218e-05,
      "loss": 3.1638,
      "num_input_tokens_seen": 46657600,
      "step": 22325
    },
    {
      "epoch": 3.6427930500040784,
      "grad_norm": 3.09375,
      "learning_rate": 2.0452478131208484e-05,
      "loss": 3.7448,
      "num_input_tokens_seen": 46669040,
      "step": 22330
    },
    {
      "epoch": 3.643608777224896,
      "grad_norm": 1.7890625,
      "learning_rate": 2.0442824791277765e-05,
      "loss": 1.4292,
      "num_input_tokens_seen": 46679792,
      "step": 22335
    },
    {
      "epoch": 3.6444245044457135,
      "grad_norm": 6.625,
      "learning_rate": 2.0433172154120727e-05,
      "loss": 1.38,
      "num_input_tokens_seen": 46689280,
      "step": 22340
    },
    {
      "epoch": 3.6452402316665307,
      "grad_norm": 7.3125,
      "learning_rate": 2.0423520221225947e-05,
      "loss": 2.2728,
      "num_input_tokens_seen": 46697968,
      "step": 22345
    },
    {
      "epoch": 3.6460559588873482,
      "grad_norm": 5.875,
      "learning_rate": 2.0413868994081848e-05,
      "loss": 2.5299,
      "num_input_tokens_seen": 46708416,
      "step": 22350
    },
    {
      "epoch": 3.6468716861081654,
      "grad_norm": 5.46875,
      "learning_rate": 2.0404218474176795e-05,
      "loss": 2.1664,
      "num_input_tokens_seen": 46718432,
      "step": 22355
    },
    {
      "epoch": 3.647687413328983,
      "grad_norm": 7.71875,
      "learning_rate": 2.0394568662999002e-05,
      "loss": 2.2535,
      "num_input_tokens_seen": 46729008,
      "step": 22360
    },
    {
      "epoch": 3.6485031405498,
      "grad_norm": 9.625,
      "learning_rate": 2.0384919562036593e-05,
      "loss": 2.0641,
      "num_input_tokens_seen": 46740064,
      "step": 22365
    },
    {
      "epoch": 3.6493188677706176,
      "grad_norm": 6.40625,
      "learning_rate": 2.0375271172777593e-05,
      "loss": 2.6467,
      "num_input_tokens_seen": 46749568,
      "step": 22370
    },
    {
      "epoch": 3.6501345949914348,
      "grad_norm": 9.6875,
      "learning_rate": 2.0365623496709885e-05,
      "loss": 2.2089,
      "num_input_tokens_seen": 46760304,
      "step": 22375
    },
    {
      "epoch": 3.6509503222122524,
      "grad_norm": 9.8125,
      "learning_rate": 2.0355976535321283e-05,
      "loss": 1.6465,
      "num_input_tokens_seen": 46771072,
      "step": 22380
    },
    {
      "epoch": 3.6517660494330695,
      "grad_norm": 14.3125,
      "learning_rate": 2.034633029009945e-05,
      "loss": 1.4858,
      "num_input_tokens_seen": 46781856,
      "step": 22385
    },
    {
      "epoch": 3.652581776653887,
      "grad_norm": 13.5,
      "learning_rate": 2.0336684762531972e-05,
      "loss": 1.6981,
      "num_input_tokens_seen": 46792112,
      "step": 22390
    },
    {
      "epoch": 3.653397503874704,
      "grad_norm": 9.75,
      "learning_rate": 2.032703995410631e-05,
      "loss": 1.5857,
      "num_input_tokens_seen": 46801664,
      "step": 22395
    },
    {
      "epoch": 3.6542132310955218,
      "grad_norm": 5.0625,
      "learning_rate": 2.031739586630981e-05,
      "loss": 1.9925,
      "num_input_tokens_seen": 46813008,
      "step": 22400
    },
    {
      "epoch": 3.6542132310955218,
      "eval_loss": 2.5449345111846924,
      "eval_runtime": 134.8181,
      "eval_samples_per_second": 20.212,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 46813008,
      "step": 22400
    },
    {
      "epoch": 3.655028958316339,
      "grad_norm": 11.875,
      "learning_rate": 2.0307752500629707e-05,
      "loss": 2.8708,
      "num_input_tokens_seen": 46824208,
      "step": 22405
    },
    {
      "epoch": 3.6558446855371565,
      "grad_norm": 2.25,
      "learning_rate": 2.0298109858553144e-05,
      "loss": 2.2309,
      "num_input_tokens_seen": 46833312,
      "step": 22410
    },
    {
      "epoch": 3.6566604127579736,
      "grad_norm": 6.5625,
      "learning_rate": 2.028846794156712e-05,
      "loss": 2.2348,
      "num_input_tokens_seen": 46843360,
      "step": 22415
    },
    {
      "epoch": 3.657476139978791,
      "grad_norm": 8.0625,
      "learning_rate": 2.027882675115856e-05,
      "loss": 2.5278,
      "num_input_tokens_seen": 46854240,
      "step": 22420
    },
    {
      "epoch": 3.6582918671996083,
      "grad_norm": 7.34375,
      "learning_rate": 2.026918628881423e-05,
      "loss": 1.7401,
      "num_input_tokens_seen": 46864784,
      "step": 22425
    },
    {
      "epoch": 3.659107594420426,
      "grad_norm": 7.59375,
      "learning_rate": 2.0259546556020833e-05,
      "loss": 1.9327,
      "num_input_tokens_seen": 46876352,
      "step": 22430
    },
    {
      "epoch": 3.659923321641243,
      "grad_norm": 7.71875,
      "learning_rate": 2.024990755426493e-05,
      "loss": 3.1102,
      "num_input_tokens_seen": 46886896,
      "step": 22435
    },
    {
      "epoch": 3.6607390488620606,
      "grad_norm": 9.375,
      "learning_rate": 2.0240269285032975e-05,
      "loss": 1.4497,
      "num_input_tokens_seen": 46898576,
      "step": 22440
    },
    {
      "epoch": 3.6615547760828777,
      "grad_norm": 7.78125,
      "learning_rate": 2.0230631749811306e-05,
      "loss": 1.9124,
      "num_input_tokens_seen": 46908752,
      "step": 22445
    },
    {
      "epoch": 3.6623705033036953,
      "grad_norm": 4.8125,
      "learning_rate": 2.0220994950086162e-05,
      "loss": 2.8595,
      "num_input_tokens_seen": 46918848,
      "step": 22450
    },
    {
      "epoch": 3.663186230524513,
      "grad_norm": 2.59375,
      "learning_rate": 2.021135888734365e-05,
      "loss": 1.4355,
      "num_input_tokens_seen": 46929664,
      "step": 22455
    },
    {
      "epoch": 3.66400195774533,
      "grad_norm": 3.921875,
      "learning_rate": 2.0201723563069783e-05,
      "loss": 1.5092,
      "num_input_tokens_seen": 46939664,
      "step": 22460
    },
    {
      "epoch": 3.664817684966147,
      "grad_norm": 7.15625,
      "learning_rate": 2.0192088978750433e-05,
      "loss": 2.7935,
      "num_input_tokens_seen": 46950912,
      "step": 22465
    },
    {
      "epoch": 3.6656334121869647,
      "grad_norm": 5.6875,
      "learning_rate": 2.0182455135871385e-05,
      "loss": 2.3507,
      "num_input_tokens_seen": 46960272,
      "step": 22470
    },
    {
      "epoch": 3.6664491394077823,
      "grad_norm": 6.84375,
      "learning_rate": 2.0172822035918305e-05,
      "loss": 2.2197,
      "num_input_tokens_seen": 46970144,
      "step": 22475
    },
    {
      "epoch": 3.6672648666285994,
      "grad_norm": 10.875,
      "learning_rate": 2.016318968037671e-05,
      "loss": 1.6904,
      "num_input_tokens_seen": 46981040,
      "step": 22480
    },
    {
      "epoch": 3.6680805938494165,
      "grad_norm": 9.6875,
      "learning_rate": 2.015355807073206e-05,
      "loss": 1.5422,
      "num_input_tokens_seen": 46989744,
      "step": 22485
    },
    {
      "epoch": 3.668896321070234,
      "grad_norm": 7.59375,
      "learning_rate": 2.0143927208469664e-05,
      "loss": 2.2099,
      "num_input_tokens_seen": 47001376,
      "step": 22490
    },
    {
      "epoch": 3.6697120482910517,
      "grad_norm": 20.375,
      "learning_rate": 2.0134297095074708e-05,
      "loss": 1.9581,
      "num_input_tokens_seen": 47012816,
      "step": 22495
    },
    {
      "epoch": 3.670527775511869,
      "grad_norm": 7.34375,
      "learning_rate": 2.0124667732032297e-05,
      "loss": 0.9262,
      "num_input_tokens_seen": 47023488,
      "step": 22500
    },
    {
      "epoch": 3.671343502732686,
      "grad_norm": 4.46875,
      "learning_rate": 2.011503912082738e-05,
      "loss": 1.3541,
      "num_input_tokens_seen": 47033200,
      "step": 22505
    },
    {
      "epoch": 3.6721592299535035,
      "grad_norm": 5.875,
      "learning_rate": 2.0105411262944823e-05,
      "loss": 2.5225,
      "num_input_tokens_seen": 47043936,
      "step": 22510
    },
    {
      "epoch": 3.672974957174321,
      "grad_norm": 1.9375,
      "learning_rate": 2.0095784159869366e-05,
      "loss": 3.2647,
      "num_input_tokens_seen": 47053712,
      "step": 22515
    },
    {
      "epoch": 3.6737906843951382,
      "grad_norm": 7.5,
      "learning_rate": 2.0086157813085608e-05,
      "loss": 1.8715,
      "num_input_tokens_seen": 47063168,
      "step": 22520
    },
    {
      "epoch": 3.6746064116159554,
      "grad_norm": 6.96875,
      "learning_rate": 2.0076532224078068e-05,
      "loss": 1.3041,
      "num_input_tokens_seen": 47073968,
      "step": 22525
    },
    {
      "epoch": 3.675422138836773,
      "grad_norm": 8.8125,
      "learning_rate": 2.0066907394331142e-05,
      "loss": 1.3821,
      "num_input_tokens_seen": 47084944,
      "step": 22530
    },
    {
      "epoch": 3.6762378660575905,
      "grad_norm": 2.34375,
      "learning_rate": 2.0057283325329077e-05,
      "loss": 2.9788,
      "num_input_tokens_seen": 47095888,
      "step": 22535
    },
    {
      "epoch": 3.6770535932784076,
      "grad_norm": 12.875,
      "learning_rate": 2.0047660018556047e-05,
      "loss": 1.3853,
      "num_input_tokens_seen": 47106384,
      "step": 22540
    },
    {
      "epoch": 3.677869320499225,
      "grad_norm": 20.375,
      "learning_rate": 2.0038037475496075e-05,
      "loss": 2.719,
      "num_input_tokens_seen": 47117376,
      "step": 22545
    },
    {
      "epoch": 3.6786850477200423,
      "grad_norm": 8.1875,
      "learning_rate": 2.0028415697633073e-05,
      "loss": 2.7743,
      "num_input_tokens_seen": 47127712,
      "step": 22550
    },
    {
      "epoch": 3.67950077494086,
      "grad_norm": 7.71875,
      "learning_rate": 2.0018794686450858e-05,
      "loss": 2.6223,
      "num_input_tokens_seen": 47136112,
      "step": 22555
    },
    {
      "epoch": 3.680316502161677,
      "grad_norm": 6.75,
      "learning_rate": 2.0009174443433088e-05,
      "loss": 2.0775,
      "num_input_tokens_seen": 47147904,
      "step": 22560
    },
    {
      "epoch": 3.6811322293824946,
      "grad_norm": 7.4375,
      "learning_rate": 1.999955497006334e-05,
      "loss": 2.8767,
      "num_input_tokens_seen": 47158112,
      "step": 22565
    },
    {
      "epoch": 3.6819479566033118,
      "grad_norm": 17.5,
      "learning_rate": 1.9989936267825067e-05,
      "loss": 1.9076,
      "num_input_tokens_seen": 47168512,
      "step": 22570
    },
    {
      "epoch": 3.6827636838241293,
      "grad_norm": 4.09375,
      "learning_rate": 1.9980318338201572e-05,
      "loss": 1.7447,
      "num_input_tokens_seen": 47178176,
      "step": 22575
    },
    {
      "epoch": 3.6835794110449465,
      "grad_norm": 6.4375,
      "learning_rate": 1.997070118267607e-05,
      "loss": 2.6484,
      "num_input_tokens_seen": 47190304,
      "step": 22580
    },
    {
      "epoch": 3.684395138265764,
      "grad_norm": 6.96875,
      "learning_rate": 1.9961084802731654e-05,
      "loss": 2.9103,
      "num_input_tokens_seen": 47201680,
      "step": 22585
    },
    {
      "epoch": 3.685210865486581,
      "grad_norm": 22.625,
      "learning_rate": 1.9951469199851273e-05,
      "loss": 3.0654,
      "num_input_tokens_seen": 47213200,
      "step": 22590
    },
    {
      "epoch": 3.6860265927073987,
      "grad_norm": 10.125,
      "learning_rate": 1.99418543755178e-05,
      "loss": 2.4724,
      "num_input_tokens_seen": 47224736,
      "step": 22595
    },
    {
      "epoch": 3.686842319928216,
      "grad_norm": 6.375,
      "learning_rate": 1.9932240331213936e-05,
      "loss": 1.7949,
      "num_input_tokens_seen": 47233968,
      "step": 22600
    },
    {
      "epoch": 3.686842319928216,
      "eval_loss": 2.540861129760742,
      "eval_runtime": 134.858,
      "eval_samples_per_second": 20.206,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 47233968,
      "step": 22600
    },
    {
      "epoch": 3.6876580471490334,
      "grad_norm": 9.25,
      "learning_rate": 1.9922627068422297e-05,
      "loss": 2.1231,
      "num_input_tokens_seen": 47242496,
      "step": 22605
    },
    {
      "epoch": 3.6884737743698506,
      "grad_norm": 2.9375,
      "learning_rate": 1.991301458862538e-05,
      "loss": 1.5806,
      "num_input_tokens_seen": 47254336,
      "step": 22610
    },
    {
      "epoch": 3.689289501590668,
      "grad_norm": 2.03125,
      "learning_rate": 1.9903402893305536e-05,
      "loss": 1.5829,
      "num_input_tokens_seen": 47264384,
      "step": 22615
    },
    {
      "epoch": 3.6901052288114853,
      "grad_norm": 2.40625,
      "learning_rate": 1.9893791983945016e-05,
      "loss": 1.3518,
      "num_input_tokens_seen": 47276400,
      "step": 22620
    },
    {
      "epoch": 3.690920956032303,
      "grad_norm": 4.8125,
      "learning_rate": 1.988418186202594e-05,
      "loss": 2.9133,
      "num_input_tokens_seen": 47287040,
      "step": 22625
    },
    {
      "epoch": 3.6917366832531204,
      "grad_norm": 7.46875,
      "learning_rate": 1.98745725290303e-05,
      "loss": 3.1148,
      "num_input_tokens_seen": 47298864,
      "step": 22630
    },
    {
      "epoch": 3.6925524104739376,
      "grad_norm": 7.40625,
      "learning_rate": 1.986496398644e-05,
      "loss": 2.4159,
      "num_input_tokens_seen": 47307920,
      "step": 22635
    },
    {
      "epoch": 3.6933681376947547,
      "grad_norm": 5.46875,
      "learning_rate": 1.9855356235736777e-05,
      "loss": 1.9059,
      "num_input_tokens_seen": 47319520,
      "step": 22640
    },
    {
      "epoch": 3.6941838649155723,
      "grad_norm": 6.875,
      "learning_rate": 1.9845749278402277e-05,
      "loss": 3.8034,
      "num_input_tokens_seen": 47330448,
      "step": 22645
    },
    {
      "epoch": 3.69499959213639,
      "grad_norm": 10.5625,
      "learning_rate": 1.9836143115918006e-05,
      "loss": 2.7537,
      "num_input_tokens_seen": 47339728,
      "step": 22650
    },
    {
      "epoch": 3.695815319357207,
      "grad_norm": 3.59375,
      "learning_rate": 1.9826537749765367e-05,
      "loss": 2.2537,
      "num_input_tokens_seen": 47350000,
      "step": 22655
    },
    {
      "epoch": 3.696631046578024,
      "grad_norm": 7.09375,
      "learning_rate": 1.9816933181425625e-05,
      "loss": 3.7077,
      "num_input_tokens_seen": 47360432,
      "step": 22660
    },
    {
      "epoch": 3.6974467737988417,
      "grad_norm": 17.125,
      "learning_rate": 1.9807329412379903e-05,
      "loss": 2.2047,
      "num_input_tokens_seen": 47371264,
      "step": 22665
    },
    {
      "epoch": 3.6982625010196593,
      "grad_norm": 4.8125,
      "learning_rate": 1.9797726444109247e-05,
      "loss": 1.4649,
      "num_input_tokens_seen": 47382144,
      "step": 22670
    },
    {
      "epoch": 3.6990782282404764,
      "grad_norm": 0.5703125,
      "learning_rate": 1.9788124278094557e-05,
      "loss": 2.3226,
      "num_input_tokens_seen": 47391712,
      "step": 22675
    },
    {
      "epoch": 3.6998939554612935,
      "grad_norm": 8.0625,
      "learning_rate": 1.9778522915816594e-05,
      "loss": 2.0422,
      "num_input_tokens_seen": 47402080,
      "step": 22680
    },
    {
      "epoch": 3.700709682682111,
      "grad_norm": 4.59375,
      "learning_rate": 1.9768922358756014e-05,
      "loss": 2.8576,
      "num_input_tokens_seen": 47412720,
      "step": 22685
    },
    {
      "epoch": 3.7015254099029287,
      "grad_norm": 4.15625,
      "learning_rate": 1.9759322608393353e-05,
      "loss": 2.4574,
      "num_input_tokens_seen": 47422368,
      "step": 22690
    },
    {
      "epoch": 3.702341137123746,
      "grad_norm": 4.625,
      "learning_rate": 1.9749723666208992e-05,
      "loss": 2.4857,
      "num_input_tokens_seen": 47433600,
      "step": 22695
    },
    {
      "epoch": 3.703156864344563,
      "grad_norm": 8.0625,
      "learning_rate": 1.9740125533683235e-05,
      "loss": 2.4607,
      "num_input_tokens_seen": 47444624,
      "step": 22700
    },
    {
      "epoch": 3.7039725915653805,
      "grad_norm": 8.5625,
      "learning_rate": 1.9730528212296208e-05,
      "loss": 3.2363,
      "num_input_tokens_seen": 47456112,
      "step": 22705
    },
    {
      "epoch": 3.704788318786198,
      "grad_norm": 8.5625,
      "learning_rate": 1.9720931703527945e-05,
      "loss": 2.4641,
      "num_input_tokens_seen": 47466080,
      "step": 22710
    },
    {
      "epoch": 3.705604046007015,
      "grad_norm": 9.375,
      "learning_rate": 1.9711336008858373e-05,
      "loss": 4.4062,
      "num_input_tokens_seen": 47476688,
      "step": 22715
    },
    {
      "epoch": 3.7064197732278323,
      "grad_norm": 12.125,
      "learning_rate": 1.9701741129767233e-05,
      "loss": 3.5217,
      "num_input_tokens_seen": 47487744,
      "step": 22720
    },
    {
      "epoch": 3.70723550044865,
      "grad_norm": 7.0625,
      "learning_rate": 1.9692147067734202e-05,
      "loss": 2.8436,
      "num_input_tokens_seen": 47497392,
      "step": 22725
    },
    {
      "epoch": 3.7080512276694675,
      "grad_norm": 3.890625,
      "learning_rate": 1.96825538242388e-05,
      "loss": 1.948,
      "num_input_tokens_seen": 47508064,
      "step": 22730
    },
    {
      "epoch": 3.7088669548902846,
      "grad_norm": 8.3125,
      "learning_rate": 1.967296140076041e-05,
      "loss": 2.656,
      "num_input_tokens_seen": 47518080,
      "step": 22735
    },
    {
      "epoch": 3.709682682111102,
      "grad_norm": 6.96875,
      "learning_rate": 1.966336979877833e-05,
      "loss": 2.6002,
      "num_input_tokens_seen": 47529040,
      "step": 22740
    },
    {
      "epoch": 3.7104984093319193,
      "grad_norm": 5.90625,
      "learning_rate": 1.9653779019771678e-05,
      "loss": 2.2671,
      "num_input_tokens_seen": 47537760,
      "step": 22745
    },
    {
      "epoch": 3.711314136552737,
      "grad_norm": 10.1875,
      "learning_rate": 1.9644189065219488e-05,
      "loss": 1.8158,
      "num_input_tokens_seen": 47548544,
      "step": 22750
    },
    {
      "epoch": 3.712129863773554,
      "grad_norm": 7.03125,
      "learning_rate": 1.9634599936600655e-05,
      "loss": 2.112,
      "num_input_tokens_seen": 47558368,
      "step": 22755
    },
    {
      "epoch": 3.7129455909943716,
      "grad_norm": 5.1875,
      "learning_rate": 1.9625011635393935e-05,
      "loss": 1.1854,
      "num_input_tokens_seen": 47567744,
      "step": 22760
    },
    {
      "epoch": 3.7137613182151887,
      "grad_norm": 7.5,
      "learning_rate": 1.9615424163077963e-05,
      "loss": 3.1212,
      "num_input_tokens_seen": 47577488,
      "step": 22765
    },
    {
      "epoch": 3.7145770454360063,
      "grad_norm": 2.78125,
      "learning_rate": 1.9605837521131263e-05,
      "loss": 3.4194,
      "num_input_tokens_seen": 47587760,
      "step": 22770
    },
    {
      "epoch": 3.7153927726568234,
      "grad_norm": 1.546875,
      "learning_rate": 1.9596251711032192e-05,
      "loss": 1.0511,
      "num_input_tokens_seen": 47598960,
      "step": 22775
    },
    {
      "epoch": 3.716208499877641,
      "grad_norm": 3.28125,
      "learning_rate": 1.958666673425903e-05,
      "loss": 1.9501,
      "num_input_tokens_seen": 47608400,
      "step": 22780
    },
    {
      "epoch": 3.717024227098458,
      "grad_norm": 2.640625,
      "learning_rate": 1.957708259228987e-05,
      "loss": 2.2746,
      "num_input_tokens_seen": 47620192,
      "step": 22785
    },
    {
      "epoch": 3.7178399543192757,
      "grad_norm": 6.375,
      "learning_rate": 1.956749928660273e-05,
      "loss": 3.0651,
      "num_input_tokens_seen": 47629360,
      "step": 22790
    },
    {
      "epoch": 3.718655681540093,
      "grad_norm": 12.75,
      "learning_rate": 1.955791681867547e-05,
      "loss": 2.5457,
      "num_input_tokens_seen": 47639456,
      "step": 22795
    },
    {
      "epoch": 3.7194714087609104,
      "grad_norm": 4.46875,
      "learning_rate": 1.9548335189985824e-05,
      "loss": 2.2354,
      "num_input_tokens_seen": 47650016,
      "step": 22800
    },
    {
      "epoch": 3.7194714087609104,
      "eval_loss": 2.5383472442626953,
      "eval_runtime": 134.9191,
      "eval_samples_per_second": 20.197,
      "eval_steps_per_second": 10.102,
      "num_input_tokens_seen": 47650016,
      "step": 22800
    },
    {
      "epoch": 3.7202871359817276,
      "grad_norm": 2.46875,
      "learning_rate": 1.9538754402011396e-05,
      "loss": 3.3677,
      "num_input_tokens_seen": 47661056,
      "step": 22805
    },
    {
      "epoch": 3.721102863202545,
      "grad_norm": 8.25,
      "learning_rate": 1.952917445622968e-05,
      "loss": 2.9967,
      "num_input_tokens_seen": 47671200,
      "step": 22810
    },
    {
      "epoch": 3.7219185904233623,
      "grad_norm": 10.5625,
      "learning_rate": 1.9519595354118005e-05,
      "loss": 3.0766,
      "num_input_tokens_seen": 47681408,
      "step": 22815
    },
    {
      "epoch": 3.72273431764418,
      "grad_norm": 10.125,
      "learning_rate": 1.951001709715361e-05,
      "loss": 2.6799,
      "num_input_tokens_seen": 47691872,
      "step": 22820
    },
    {
      "epoch": 3.7235500448649974,
      "grad_norm": 3.59375,
      "learning_rate": 1.9500439686813556e-05,
      "loss": 2.3022,
      "num_input_tokens_seen": 47703248,
      "step": 22825
    },
    {
      "epoch": 3.7243657720858145,
      "grad_norm": 10.125,
      "learning_rate": 1.949086312457482e-05,
      "loss": 2.3113,
      "num_input_tokens_seen": 47715152,
      "step": 22830
    },
    {
      "epoch": 3.7251814993066317,
      "grad_norm": 12.0,
      "learning_rate": 1.9481287411914223e-05,
      "loss": 1.6907,
      "num_input_tokens_seen": 47725680,
      "step": 22835
    },
    {
      "epoch": 3.7259972265274492,
      "grad_norm": 5.1875,
      "learning_rate": 1.9471712550308457e-05,
      "loss": 2.8205,
      "num_input_tokens_seen": 47737184,
      "step": 22840
    },
    {
      "epoch": 3.726812953748267,
      "grad_norm": 7.40625,
      "learning_rate": 1.946213854123409e-05,
      "loss": 3.477,
      "num_input_tokens_seen": 47747424,
      "step": 22845
    },
    {
      "epoch": 3.727628680969084,
      "grad_norm": 13.875,
      "learning_rate": 1.9452565386167554e-05,
      "loss": 2.4781,
      "num_input_tokens_seen": 47757664,
      "step": 22850
    },
    {
      "epoch": 3.728444408189901,
      "grad_norm": 0.234375,
      "learning_rate": 1.9442993086585142e-05,
      "loss": 2.8343,
      "num_input_tokens_seen": 47767664,
      "step": 22855
    },
    {
      "epoch": 3.7292601354107187,
      "grad_norm": 3.328125,
      "learning_rate": 1.9433421643963043e-05,
      "loss": 1.6406,
      "num_input_tokens_seen": 47778208,
      "step": 22860
    },
    {
      "epoch": 3.7300758626315362,
      "grad_norm": 1.7734375,
      "learning_rate": 1.942385105977727e-05,
      "loss": 1.7675,
      "num_input_tokens_seen": 47789456,
      "step": 22865
    },
    {
      "epoch": 3.7308915898523534,
      "grad_norm": 6.71875,
      "learning_rate": 1.9414281335503743e-05,
      "loss": 2.7943,
      "num_input_tokens_seen": 47800480,
      "step": 22870
    },
    {
      "epoch": 3.7317073170731705,
      "grad_norm": 6.625,
      "learning_rate": 1.9404712472618232e-05,
      "loss": 2.1083,
      "num_input_tokens_seen": 47810048,
      "step": 22875
    },
    {
      "epoch": 3.732523044293988,
      "grad_norm": 8.125,
      "learning_rate": 1.939514447259636e-05,
      "loss": 2.8386,
      "num_input_tokens_seen": 47820320,
      "step": 22880
    },
    {
      "epoch": 3.7333387715148056,
      "grad_norm": 10.5,
      "learning_rate": 1.938557733691365e-05,
      "loss": 2.1758,
      "num_input_tokens_seen": 47830800,
      "step": 22885
    },
    {
      "epoch": 3.7341544987356228,
      "grad_norm": 11.125,
      "learning_rate": 1.9376011067045476e-05,
      "loss": 2.5858,
      "num_input_tokens_seen": 47840464,
      "step": 22890
    },
    {
      "epoch": 3.73497022595644,
      "grad_norm": 5.5625,
      "learning_rate": 1.9366445664467065e-05,
      "loss": 2.0333,
      "num_input_tokens_seen": 47849600,
      "step": 22895
    },
    {
      "epoch": 3.7357859531772575,
      "grad_norm": 7.625,
      "learning_rate": 1.9356881130653533e-05,
      "loss": 2.3987,
      "num_input_tokens_seen": 47859616,
      "step": 22900
    },
    {
      "epoch": 3.736601680398075,
      "grad_norm": 10.625,
      "learning_rate": 1.9347317467079846e-05,
      "loss": 2.7427,
      "num_input_tokens_seen": 47869712,
      "step": 22905
    },
    {
      "epoch": 3.737417407618892,
      "grad_norm": 5.78125,
      "learning_rate": 1.9337754675220836e-05,
      "loss": 2.4678,
      "num_input_tokens_seen": 47879472,
      "step": 22910
    },
    {
      "epoch": 3.7382331348397098,
      "grad_norm": 13.75,
      "learning_rate": 1.9328192756551218e-05,
      "loss": 2.4255,
      "num_input_tokens_seen": 47889008,
      "step": 22915
    },
    {
      "epoch": 3.739048862060527,
      "grad_norm": 6.28125,
      "learning_rate": 1.931863171254555e-05,
      "loss": 2.0429,
      "num_input_tokens_seen": 47899936,
      "step": 22920
    },
    {
      "epoch": 3.7398645892813445,
      "grad_norm": 7.625,
      "learning_rate": 1.930907154467826e-05,
      "loss": 2.0553,
      "num_input_tokens_seen": 47910496,
      "step": 22925
    },
    {
      "epoch": 3.7406803165021616,
      "grad_norm": 7.0625,
      "learning_rate": 1.9299512254423673e-05,
      "loss": 2.8983,
      "num_input_tokens_seen": 47918832,
      "step": 22930
    },
    {
      "epoch": 3.741496043722979,
      "grad_norm": 6.46875,
      "learning_rate": 1.9289953843255914e-05,
      "loss": 2.1715,
      "num_input_tokens_seen": 47929264,
      "step": 22935
    },
    {
      "epoch": 3.7423117709437963,
      "grad_norm": 7.75,
      "learning_rate": 1.9280396312649048e-05,
      "loss": 4.5285,
      "num_input_tokens_seen": 47940032,
      "step": 22940
    },
    {
      "epoch": 3.743127498164614,
      "grad_norm": 10.5625,
      "learning_rate": 1.9270839664076936e-05,
      "loss": 2.8065,
      "num_input_tokens_seen": 47949600,
      "step": 22945
    },
    {
      "epoch": 3.743943225385431,
      "grad_norm": 11.375,
      "learning_rate": 1.9261283899013345e-05,
      "loss": 2.2834,
      "num_input_tokens_seen": 47959792,
      "step": 22950
    },
    {
      "epoch": 3.7447589526062486,
      "grad_norm": 5.5,
      "learning_rate": 1.92517290189319e-05,
      "loss": 1.95,
      "num_input_tokens_seen": 47969680,
      "step": 22955
    },
    {
      "epoch": 3.7455746798270657,
      "grad_norm": 6.65625,
      "learning_rate": 1.924217502530607e-05,
      "loss": 2.2995,
      "num_input_tokens_seen": 47980032,
      "step": 22960
    },
    {
      "epoch": 3.7463904070478833,
      "grad_norm": 3.515625,
      "learning_rate": 1.9232621919609207e-05,
      "loss": 2.1178,
      "num_input_tokens_seen": 47991488,
      "step": 22965
    },
    {
      "epoch": 3.7472061342687004,
      "grad_norm": 3.796875,
      "learning_rate": 1.9223069703314534e-05,
      "loss": 3.0665,
      "num_input_tokens_seen": 48001632,
      "step": 22970
    },
    {
      "epoch": 3.748021861489518,
      "grad_norm": 1.421875,
      "learning_rate": 1.92135183778951e-05,
      "loss": 1.6698,
      "num_input_tokens_seen": 48012864,
      "step": 22975
    },
    {
      "epoch": 3.748837588710335,
      "grad_norm": 13.4375,
      "learning_rate": 1.9203967944823857e-05,
      "loss": 3.7176,
      "num_input_tokens_seen": 48023408,
      "step": 22980
    },
    {
      "epoch": 3.7496533159311527,
      "grad_norm": 2.21875,
      "learning_rate": 1.9194418405573588e-05,
      "loss": 2.5234,
      "num_input_tokens_seen": 48033072,
      "step": 22985
    },
    {
      "epoch": 3.75046904315197,
      "grad_norm": 3.6875,
      "learning_rate": 1.9184869761616954e-05,
      "loss": 2.1939,
      "num_input_tokens_seen": 48043264,
      "step": 22990
    },
    {
      "epoch": 3.7512847703727874,
      "grad_norm": 11.25,
      "learning_rate": 1.9175322014426495e-05,
      "loss": 3.7622,
      "num_input_tokens_seen": 48053616,
      "step": 22995
    },
    {
      "epoch": 3.7521004975936045,
      "grad_norm": 4.78125,
      "learning_rate": 1.9165775165474565e-05,
      "loss": 2.1881,
      "num_input_tokens_seen": 48064160,
      "step": 23000
    },
    {
      "epoch": 3.7521004975936045,
      "eval_loss": 2.535382032394409,
      "eval_runtime": 134.8347,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.109,
      "num_input_tokens_seen": 48064160,
      "step": 23000
    },
    {
      "epoch": 3.752916224814422,
      "grad_norm": 2.421875,
      "learning_rate": 1.9156229216233434e-05,
      "loss": 2.4889,
      "num_input_tokens_seen": 48074288,
      "step": 23005
    },
    {
      "epoch": 3.7537319520352392,
      "grad_norm": 8.0,
      "learning_rate": 1.9146684168175184e-05,
      "loss": 2.9693,
      "num_input_tokens_seen": 48085456,
      "step": 23010
    },
    {
      "epoch": 3.754547679256057,
      "grad_norm": 5.15625,
      "learning_rate": 1.9137140022771796e-05,
      "loss": 1.8511,
      "num_input_tokens_seen": 48096896,
      "step": 23015
    },
    {
      "epoch": 3.7553634064768744,
      "grad_norm": 3.921875,
      "learning_rate": 1.9127596781495103e-05,
      "loss": 2.1199,
      "num_input_tokens_seen": 48108224,
      "step": 23020
    },
    {
      "epoch": 3.7561791336976915,
      "grad_norm": 3.34375,
      "learning_rate": 1.9118054445816767e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 48118192,
      "step": 23025
    },
    {
      "epoch": 3.7569948609185086,
      "grad_norm": 13.0,
      "learning_rate": 1.9108513017208356e-05,
      "loss": 1.7675,
      "num_input_tokens_seen": 48128848,
      "step": 23030
    },
    {
      "epoch": 3.7578105881393262,
      "grad_norm": 6.71875,
      "learning_rate": 1.9098972497141287e-05,
      "loss": 2.165,
      "num_input_tokens_seen": 48139056,
      "step": 23035
    },
    {
      "epoch": 3.758626315360144,
      "grad_norm": 5.1875,
      "learning_rate": 1.9089432887086806e-05,
      "loss": 2.6717,
      "num_input_tokens_seen": 48148928,
      "step": 23040
    },
    {
      "epoch": 3.759442042580961,
      "grad_norm": 10.375,
      "learning_rate": 1.9079894188516056e-05,
      "loss": 2.7976,
      "num_input_tokens_seen": 48159760,
      "step": 23045
    },
    {
      "epoch": 3.760257769801778,
      "grad_norm": 10.875,
      "learning_rate": 1.907035640290002e-05,
      "loss": 2.4682,
      "num_input_tokens_seen": 48169120,
      "step": 23050
    },
    {
      "epoch": 3.7610734970225956,
      "grad_norm": 12.125,
      "learning_rate": 1.9060819531709534e-05,
      "loss": 2.1562,
      "num_input_tokens_seen": 48180832,
      "step": 23055
    },
    {
      "epoch": 3.761889224243413,
      "grad_norm": 1.859375,
      "learning_rate": 1.9051283576415325e-05,
      "loss": 2.6737,
      "num_input_tokens_seen": 48189840,
      "step": 23060
    },
    {
      "epoch": 3.7627049514642303,
      "grad_norm": 15.4375,
      "learning_rate": 1.904174853848793e-05,
      "loss": 3.4441,
      "num_input_tokens_seen": 48200272,
      "step": 23065
    },
    {
      "epoch": 3.7635206786850475,
      "grad_norm": 5.53125,
      "learning_rate": 1.903221441939779e-05,
      "loss": 2.2975,
      "num_input_tokens_seen": 48211040,
      "step": 23070
    },
    {
      "epoch": 3.764336405905865,
      "grad_norm": 6.96875,
      "learning_rate": 1.9022681220615194e-05,
      "loss": 3.1918,
      "num_input_tokens_seen": 48222480,
      "step": 23075
    },
    {
      "epoch": 3.7651521331266826,
      "grad_norm": 13.0,
      "learning_rate": 1.9013148943610255e-05,
      "loss": 2.9126,
      "num_input_tokens_seen": 48233664,
      "step": 23080
    },
    {
      "epoch": 3.7659678603474998,
      "grad_norm": 7.65625,
      "learning_rate": 1.9003617589852998e-05,
      "loss": 2.6854,
      "num_input_tokens_seen": 48243680,
      "step": 23085
    },
    {
      "epoch": 3.766783587568317,
      "grad_norm": 8.75,
      "learning_rate": 1.899408716081326e-05,
      "loss": 2.093,
      "num_input_tokens_seen": 48254128,
      "step": 23090
    },
    {
      "epoch": 3.7675993147891345,
      "grad_norm": 7.1875,
      "learning_rate": 1.898455765796075e-05,
      "loss": 2.514,
      "num_input_tokens_seen": 48263856,
      "step": 23095
    },
    {
      "epoch": 3.768415042009952,
      "grad_norm": 8.9375,
      "learning_rate": 1.8975029082765053e-05,
      "loss": 3.3207,
      "num_input_tokens_seen": 48276144,
      "step": 23100
    },
    {
      "epoch": 3.769230769230769,
      "grad_norm": 8.5625,
      "learning_rate": 1.8965501436695577e-05,
      "loss": 3.4451,
      "num_input_tokens_seen": 48286768,
      "step": 23105
    },
    {
      "epoch": 3.7700464964515867,
      "grad_norm": 7.03125,
      "learning_rate": 1.895597472122161e-05,
      "loss": 3.3405,
      "num_input_tokens_seen": 48296896,
      "step": 23110
    },
    {
      "epoch": 3.770862223672404,
      "grad_norm": 9.4375,
      "learning_rate": 1.894644893781231e-05,
      "loss": 1.7976,
      "num_input_tokens_seen": 48306576,
      "step": 23115
    },
    {
      "epoch": 3.7716779508932214,
      "grad_norm": 3.15625,
      "learning_rate": 1.893692408793665e-05,
      "loss": 2.3311,
      "num_input_tokens_seen": 48317200,
      "step": 23120
    },
    {
      "epoch": 3.7724936781140386,
      "grad_norm": 4.625,
      "learning_rate": 1.8927400173063493e-05,
      "loss": 2.2113,
      "num_input_tokens_seen": 48328080,
      "step": 23125
    },
    {
      "epoch": 3.773309405334856,
      "grad_norm": 8.4375,
      "learning_rate": 1.891787719466154e-05,
      "loss": 3.2427,
      "num_input_tokens_seen": 48338848,
      "step": 23130
    },
    {
      "epoch": 3.7741251325556733,
      "grad_norm": 2.78125,
      "learning_rate": 1.8908355154199346e-05,
      "loss": 2.6816,
      "num_input_tokens_seen": 48347888,
      "step": 23135
    },
    {
      "epoch": 3.774940859776491,
      "grad_norm": 18.75,
      "learning_rate": 1.8898834053145357e-05,
      "loss": 1.594,
      "num_input_tokens_seen": 48357728,
      "step": 23140
    },
    {
      "epoch": 3.775756586997308,
      "grad_norm": 3.375,
      "learning_rate": 1.8889313892967813e-05,
      "loss": 1.8555,
      "num_input_tokens_seen": 48368928,
      "step": 23145
    },
    {
      "epoch": 3.7765723142181256,
      "grad_norm": 8.5625,
      "learning_rate": 1.8879794675134863e-05,
      "loss": 3.3091,
      "num_input_tokens_seen": 48378848,
      "step": 23150
    },
    {
      "epoch": 3.7773880414389427,
      "grad_norm": 10.875,
      "learning_rate": 1.8870276401114494e-05,
      "loss": 1.8459,
      "num_input_tokens_seen": 48389664,
      "step": 23155
    },
    {
      "epoch": 3.7782037686597603,
      "grad_norm": 2.015625,
      "learning_rate": 1.886075907237453e-05,
      "loss": 2.8076,
      "num_input_tokens_seen": 48400928,
      "step": 23160
    },
    {
      "epoch": 3.7790194958805774,
      "grad_norm": 1.6484375,
      "learning_rate": 1.8851242690382672e-05,
      "loss": 1.3478,
      "num_input_tokens_seen": 48412032,
      "step": 23165
    },
    {
      "epoch": 3.779835223101395,
      "grad_norm": 9.25,
      "learning_rate": 1.884172725660645e-05,
      "loss": 1.2957,
      "num_input_tokens_seen": 48422704,
      "step": 23170
    },
    {
      "epoch": 3.780650950322212,
      "grad_norm": 4.03125,
      "learning_rate": 1.8832212772513277e-05,
      "loss": 2.1649,
      "num_input_tokens_seen": 48433104,
      "step": 23175
    },
    {
      "epoch": 3.7814666775430297,
      "grad_norm": 6.1875,
      "learning_rate": 1.8822699239570414e-05,
      "loss": 3.6983,
      "num_input_tokens_seen": 48443632,
      "step": 23180
    },
    {
      "epoch": 3.782282404763847,
      "grad_norm": 7.375,
      "learning_rate": 1.8813186659244943e-05,
      "loss": 4.2684,
      "num_input_tokens_seen": 48453504,
      "step": 23185
    },
    {
      "epoch": 3.7830981319846644,
      "grad_norm": 7.5,
      "learning_rate": 1.880367503300385e-05,
      "loss": 2.2703,
      "num_input_tokens_seen": 48463680,
      "step": 23190
    },
    {
      "epoch": 3.783913859205482,
      "grad_norm": 8.5625,
      "learning_rate": 1.8794164362313927e-05,
      "loss": 2.9701,
      "num_input_tokens_seen": 48473712,
      "step": 23195
    },
    {
      "epoch": 3.784729586426299,
      "grad_norm": 9.0,
      "learning_rate": 1.878465464864185e-05,
      "loss": 3.1696,
      "num_input_tokens_seen": 48484384,
      "step": 23200
    },
    {
      "epoch": 3.784729586426299,
      "eval_loss": 2.534999370574951,
      "eval_runtime": 134.968,
      "eval_samples_per_second": 20.19,
      "eval_steps_per_second": 10.099,
      "num_input_tokens_seen": 48484384,
      "step": 23200
    },
    {
      "epoch": 3.785545313647116,
      "grad_norm": 1.453125,
      "learning_rate": 1.877514589345414e-05,
      "loss": 1.6768,
      "num_input_tokens_seen": 48495936,
      "step": 23205
    },
    {
      "epoch": 3.786361040867934,
      "grad_norm": 9.4375,
      "learning_rate": 1.876563809821715e-05,
      "loss": 3.1496,
      "num_input_tokens_seen": 48506656,
      "step": 23210
    },
    {
      "epoch": 3.7871767680887514,
      "grad_norm": 4.375,
      "learning_rate": 1.8756131264397106e-05,
      "loss": 1.8999,
      "num_input_tokens_seen": 48516624,
      "step": 23215
    },
    {
      "epoch": 3.7879924953095685,
      "grad_norm": 0.0732421875,
      "learning_rate": 1.87466253934601e-05,
      "loss": 2.2128,
      "num_input_tokens_seen": 48526752,
      "step": 23220
    },
    {
      "epoch": 3.7888082225303856,
      "grad_norm": 10.125,
      "learning_rate": 1.8737120486872033e-05,
      "loss": 2.6603,
      "num_input_tokens_seen": 48537664,
      "step": 23225
    },
    {
      "epoch": 3.789623949751203,
      "grad_norm": 8.5,
      "learning_rate": 1.8727616546098696e-05,
      "loss": 1.1937,
      "num_input_tokens_seen": 48548656,
      "step": 23230
    },
    {
      "epoch": 3.7904396769720208,
      "grad_norm": 3.828125,
      "learning_rate": 1.8718113572605716e-05,
      "loss": 2.1051,
      "num_input_tokens_seen": 48559888,
      "step": 23235
    },
    {
      "epoch": 3.791255404192838,
      "grad_norm": 9.1875,
      "learning_rate": 1.8708611567858554e-05,
      "loss": 2.7362,
      "num_input_tokens_seen": 48570768,
      "step": 23240
    },
    {
      "epoch": 3.792071131413655,
      "grad_norm": 12.5625,
      "learning_rate": 1.8699110533322565e-05,
      "loss": 3.1446,
      "num_input_tokens_seen": 48581280,
      "step": 23245
    },
    {
      "epoch": 3.7928868586344726,
      "grad_norm": 8.1875,
      "learning_rate": 1.8689610470462897e-05,
      "loss": 1.6523,
      "num_input_tokens_seen": 48592032,
      "step": 23250
    },
    {
      "epoch": 3.79370258585529,
      "grad_norm": 4.59375,
      "learning_rate": 1.8680111380744604e-05,
      "loss": 2.6382,
      "num_input_tokens_seen": 48601072,
      "step": 23255
    },
    {
      "epoch": 3.7945183130761073,
      "grad_norm": 7.09375,
      "learning_rate": 1.8670613265632564e-05,
      "loss": 3.4684,
      "num_input_tokens_seen": 48611392,
      "step": 23260
    },
    {
      "epoch": 3.7953340402969244,
      "grad_norm": 6.6875,
      "learning_rate": 1.866111612659149e-05,
      "loss": 2.1921,
      "num_input_tokens_seen": 48621984,
      "step": 23265
    },
    {
      "epoch": 3.796149767517742,
      "grad_norm": 8.9375,
      "learning_rate": 1.8651619965085967e-05,
      "loss": 2.4243,
      "num_input_tokens_seen": 48632864,
      "step": 23270
    },
    {
      "epoch": 3.7969654947385596,
      "grad_norm": 7.84375,
      "learning_rate": 1.8642124782580433e-05,
      "loss": 2.5384,
      "num_input_tokens_seen": 48643520,
      "step": 23275
    },
    {
      "epoch": 3.7977812219593767,
      "grad_norm": 3.1875,
      "learning_rate": 1.8632630580539144e-05,
      "loss": 3.2707,
      "num_input_tokens_seen": 48653104,
      "step": 23280
    },
    {
      "epoch": 3.7985969491801943,
      "grad_norm": 6.8125,
      "learning_rate": 1.862313736042625e-05,
      "loss": 2.6351,
      "num_input_tokens_seen": 48663536,
      "step": 23285
    },
    {
      "epoch": 3.7994126764010114,
      "grad_norm": 3.734375,
      "learning_rate": 1.8613645123705703e-05,
      "loss": 1.2241,
      "num_input_tokens_seen": 48674272,
      "step": 23290
    },
    {
      "epoch": 3.800228403621829,
      "grad_norm": 2.21875,
      "learning_rate": 1.8604153871841328e-05,
      "loss": 1.3883,
      "num_input_tokens_seen": 48684848,
      "step": 23295
    },
    {
      "epoch": 3.801044130842646,
      "grad_norm": 8.5625,
      "learning_rate": 1.859466360629682e-05,
      "loss": 3.0263,
      "num_input_tokens_seen": 48694048,
      "step": 23300
    },
    {
      "epoch": 3.8018598580634637,
      "grad_norm": 4.125,
      "learning_rate": 1.8585174328535666e-05,
      "loss": 1.9799,
      "num_input_tokens_seen": 48704640,
      "step": 23305
    },
    {
      "epoch": 3.802675585284281,
      "grad_norm": 6.65625,
      "learning_rate": 1.857568604002124e-05,
      "loss": 3.1137,
      "num_input_tokens_seen": 48716880,
      "step": 23310
    },
    {
      "epoch": 3.8034913125050984,
      "grad_norm": 8.25,
      "learning_rate": 1.8566198742216774e-05,
      "loss": 2.8122,
      "num_input_tokens_seen": 48726336,
      "step": 23315
    },
    {
      "epoch": 3.8043070397259156,
      "grad_norm": 5.15625,
      "learning_rate": 1.85567124365853e-05,
      "loss": 2.8075,
      "num_input_tokens_seen": 48736192,
      "step": 23320
    },
    {
      "epoch": 3.805122766946733,
      "grad_norm": 6.53125,
      "learning_rate": 1.854722712458975e-05,
      "loss": 2.9333,
      "num_input_tokens_seen": 48746864,
      "step": 23325
    },
    {
      "epoch": 3.8059384941675503,
      "grad_norm": 15.5,
      "learning_rate": 1.853774280769286e-05,
      "loss": 2.2157,
      "num_input_tokens_seen": 48756432,
      "step": 23330
    },
    {
      "epoch": 3.806754221388368,
      "grad_norm": 4.75,
      "learning_rate": 1.852825948735724e-05,
      "loss": 1.9392,
      "num_input_tokens_seen": 48766672,
      "step": 23335
    },
    {
      "epoch": 3.807569948609185,
      "grad_norm": 3.765625,
      "learning_rate": 1.851877716504534e-05,
      "loss": 1.3286,
      "num_input_tokens_seen": 48776800,
      "step": 23340
    },
    {
      "epoch": 3.8083856758300025,
      "grad_norm": 5.40625,
      "learning_rate": 1.8509295842219448e-05,
      "loss": 1.7811,
      "num_input_tokens_seen": 48786752,
      "step": 23345
    },
    {
      "epoch": 3.8092014030508197,
      "grad_norm": 3.4375,
      "learning_rate": 1.8499815520341697e-05,
      "loss": 1.5269,
      "num_input_tokens_seen": 48795408,
      "step": 23350
    },
    {
      "epoch": 3.8100171302716372,
      "grad_norm": 7.75,
      "learning_rate": 1.8490336200874094e-05,
      "loss": 3.3678,
      "num_input_tokens_seen": 48806016,
      "step": 23355
    },
    {
      "epoch": 3.8108328574924544,
      "grad_norm": 9.9375,
      "learning_rate": 1.848085788527844e-05,
      "loss": 1.9982,
      "num_input_tokens_seen": 48817088,
      "step": 23360
    },
    {
      "epoch": 3.811648584713272,
      "grad_norm": 5.15625,
      "learning_rate": 1.847138057501644e-05,
      "loss": 3.7848,
      "num_input_tokens_seen": 48828416,
      "step": 23365
    },
    {
      "epoch": 3.812464311934089,
      "grad_norm": 9.5,
      "learning_rate": 1.8461904271549582e-05,
      "loss": 2.5334,
      "num_input_tokens_seen": 48838320,
      "step": 23370
    },
    {
      "epoch": 3.8132800391549067,
      "grad_norm": 10.125,
      "learning_rate": 1.845242897633926e-05,
      "loss": 1.2705,
      "num_input_tokens_seen": 48848288,
      "step": 23375
    },
    {
      "epoch": 3.814095766375724,
      "grad_norm": 6.375,
      "learning_rate": 1.844295469084667e-05,
      "loss": 2.2726,
      "num_input_tokens_seen": 48858672,
      "step": 23380
    },
    {
      "epoch": 3.8149114935965414,
      "grad_norm": 1.7734375,
      "learning_rate": 1.843348141653286e-05,
      "loss": 1.3909,
      "num_input_tokens_seen": 48868720,
      "step": 23385
    },
    {
      "epoch": 3.815727220817359,
      "grad_norm": 11.3125,
      "learning_rate": 1.842400915485874e-05,
      "loss": 2.9399,
      "num_input_tokens_seen": 48878528,
      "step": 23390
    },
    {
      "epoch": 3.816542948038176,
      "grad_norm": 2.875,
      "learning_rate": 1.8414537907285053e-05,
      "loss": 1.2243,
      "num_input_tokens_seen": 48888032,
      "step": 23395
    },
    {
      "epoch": 3.817358675258993,
      "grad_norm": 10.25,
      "learning_rate": 1.840506767527237e-05,
      "loss": 2.9062,
      "num_input_tokens_seen": 48897744,
      "step": 23400
    },
    {
      "epoch": 3.817358675258993,
      "eval_loss": 2.557155132293701,
      "eval_runtime": 135.0065,
      "eval_samples_per_second": 20.184,
      "eval_steps_per_second": 10.096,
      "num_input_tokens_seen": 48897744,
      "step": 23400
    },
    {
      "epoch": 3.8181744024798108,
      "grad_norm": 5.03125,
      "learning_rate": 1.8395598460281137e-05,
      "loss": 3.3745,
      "num_input_tokens_seen": 48907904,
      "step": 23405
    },
    {
      "epoch": 3.8189901297006283,
      "grad_norm": 7.34375,
      "learning_rate": 1.838613026377161e-05,
      "loss": 3.5262,
      "num_input_tokens_seen": 48917888,
      "step": 23410
    },
    {
      "epoch": 3.8198058569214455,
      "grad_norm": 2.390625,
      "learning_rate": 1.8376663087203917e-05,
      "loss": 1.3472,
      "num_input_tokens_seen": 48926976,
      "step": 23415
    },
    {
      "epoch": 3.8206215841422626,
      "grad_norm": 3.921875,
      "learning_rate": 1.8367196932038014e-05,
      "loss": 2.2316,
      "num_input_tokens_seen": 48935440,
      "step": 23420
    },
    {
      "epoch": 3.82143731136308,
      "grad_norm": 6.1875,
      "learning_rate": 1.8357731799733686e-05,
      "loss": 2.7267,
      "num_input_tokens_seen": 48945712,
      "step": 23425
    },
    {
      "epoch": 3.8222530385838978,
      "grad_norm": 10.6875,
      "learning_rate": 1.8348267691750586e-05,
      "loss": 3.2482,
      "num_input_tokens_seen": 48956768,
      "step": 23430
    },
    {
      "epoch": 3.823068765804715,
      "grad_norm": 3.15625,
      "learning_rate": 1.833880460954821e-05,
      "loss": 2.3624,
      "num_input_tokens_seen": 48968144,
      "step": 23435
    },
    {
      "epoch": 3.823884493025532,
      "grad_norm": 8.4375,
      "learning_rate": 1.8329342554585866e-05,
      "loss": 1.9989,
      "num_input_tokens_seen": 48980144,
      "step": 23440
    },
    {
      "epoch": 3.8247002202463496,
      "grad_norm": 7.8125,
      "learning_rate": 1.8319881528322735e-05,
      "loss": 2.293,
      "num_input_tokens_seen": 48991328,
      "step": 23445
    },
    {
      "epoch": 3.825515947467167,
      "grad_norm": 7.65625,
      "learning_rate": 1.8310421532217815e-05,
      "loss": 1.8562,
      "num_input_tokens_seen": 49001984,
      "step": 23450
    },
    {
      "epoch": 3.8263316746879843,
      "grad_norm": 4.40625,
      "learning_rate": 1.8300962567729958e-05,
      "loss": 1.9575,
      "num_input_tokens_seen": 49011888,
      "step": 23455
    },
    {
      "epoch": 3.8271474019088014,
      "grad_norm": 1.5859375,
      "learning_rate": 1.8291504636317866e-05,
      "loss": 1.9032,
      "num_input_tokens_seen": 49021584,
      "step": 23460
    },
    {
      "epoch": 3.827963129129619,
      "grad_norm": 0.8828125,
      "learning_rate": 1.8282047739440055e-05,
      "loss": 2.286,
      "num_input_tokens_seen": 49033184,
      "step": 23465
    },
    {
      "epoch": 3.8287788563504366,
      "grad_norm": 7.28125,
      "learning_rate": 1.8272591878554903e-05,
      "loss": 1.8907,
      "num_input_tokens_seen": 49044064,
      "step": 23470
    },
    {
      "epoch": 3.8295945835712537,
      "grad_norm": 7.34375,
      "learning_rate": 1.8263137055120638e-05,
      "loss": 2.3856,
      "num_input_tokens_seen": 49054320,
      "step": 23475
    },
    {
      "epoch": 3.8304103107920713,
      "grad_norm": 12.5,
      "learning_rate": 1.8253683270595295e-05,
      "loss": 4.0116,
      "num_input_tokens_seen": 49063520,
      "step": 23480
    },
    {
      "epoch": 3.8312260380128884,
      "grad_norm": 5.71875,
      "learning_rate": 1.824423052643677e-05,
      "loss": 2.7427,
      "num_input_tokens_seen": 49074080,
      "step": 23485
    },
    {
      "epoch": 3.832041765233706,
      "grad_norm": 4.65625,
      "learning_rate": 1.82347788241028e-05,
      "loss": 2.8989,
      "num_input_tokens_seen": 49084432,
      "step": 23490
    },
    {
      "epoch": 3.832857492454523,
      "grad_norm": 8.6875,
      "learning_rate": 1.8225328165050942e-05,
      "loss": 2.7773,
      "num_input_tokens_seen": 49094560,
      "step": 23495
    },
    {
      "epoch": 3.8336732196753407,
      "grad_norm": 5.03125,
      "learning_rate": 1.821587855073863e-05,
      "loss": 2.1054,
      "num_input_tokens_seen": 49105216,
      "step": 23500
    },
    {
      "epoch": 3.834488946896158,
      "grad_norm": 5.25,
      "learning_rate": 1.8206429982623086e-05,
      "loss": 1.8277,
      "num_input_tokens_seen": 49115440,
      "step": 23505
    },
    {
      "epoch": 3.8353046741169754,
      "grad_norm": 9.625,
      "learning_rate": 1.8196982462161416e-05,
      "loss": 1.6404,
      "num_input_tokens_seen": 49125200,
      "step": 23510
    },
    {
      "epoch": 3.8361204013377925,
      "grad_norm": 7.4375,
      "learning_rate": 1.818753599081055e-05,
      "loss": 2.7055,
      "num_input_tokens_seen": 49135328,
      "step": 23515
    },
    {
      "epoch": 3.83693612855861,
      "grad_norm": 1.7734375,
      "learning_rate": 1.817809057002724e-05,
      "loss": 1.3624,
      "num_input_tokens_seen": 49146016,
      "step": 23520
    },
    {
      "epoch": 3.8377518557794272,
      "grad_norm": 10.5,
      "learning_rate": 1.8168646201268096e-05,
      "loss": 2.601,
      "num_input_tokens_seen": 49154416,
      "step": 23525
    },
    {
      "epoch": 3.838567583000245,
      "grad_norm": 5.90625,
      "learning_rate": 1.8159202885989557e-05,
      "loss": 1.4826,
      "num_input_tokens_seen": 49163728,
      "step": 23530
    },
    {
      "epoch": 3.839383310221062,
      "grad_norm": 4.625,
      "learning_rate": 1.814976062564789e-05,
      "loss": 0.8096,
      "num_input_tokens_seen": 49174960,
      "step": 23535
    },
    {
      "epoch": 3.8401990374418795,
      "grad_norm": 6.8125,
      "learning_rate": 1.8140319421699234e-05,
      "loss": 2.5002,
      "num_input_tokens_seen": 49185648,
      "step": 23540
    },
    {
      "epoch": 3.8410147646626966,
      "grad_norm": 4.0625,
      "learning_rate": 1.8130879275599515e-05,
      "loss": 1.1655,
      "num_input_tokens_seen": 49196896,
      "step": 23545
    },
    {
      "epoch": 3.841830491883514,
      "grad_norm": 1.28125,
      "learning_rate": 1.8121440188804544e-05,
      "loss": 1.7487,
      "num_input_tokens_seen": 49205728,
      "step": 23550
    },
    {
      "epoch": 3.8426462191043314,
      "grad_norm": 8.75,
      "learning_rate": 1.811200216276993e-05,
      "loss": 2.7157,
      "num_input_tokens_seen": 49214816,
      "step": 23555
    },
    {
      "epoch": 3.843461946325149,
      "grad_norm": 8.4375,
      "learning_rate": 1.810256519895115e-05,
      "loss": 3.6901,
      "num_input_tokens_seen": 49224288,
      "step": 23560
    },
    {
      "epoch": 3.8442776735459665,
      "grad_norm": 7.75,
      "learning_rate": 1.8093129298803494e-05,
      "loss": 2.641,
      "num_input_tokens_seen": 49234000,
      "step": 23565
    },
    {
      "epoch": 3.8450934007667836,
      "grad_norm": 11.4375,
      "learning_rate": 1.808369446378209e-05,
      "loss": 2.5648,
      "num_input_tokens_seen": 49244144,
      "step": 23570
    },
    {
      "epoch": 3.8459091279876008,
      "grad_norm": 3.140625,
      "learning_rate": 1.8074260695341914e-05,
      "loss": 1.4215,
      "num_input_tokens_seen": 49252848,
      "step": 23575
    },
    {
      "epoch": 3.8467248552084183,
      "grad_norm": 10.375,
      "learning_rate": 1.8064827994937782e-05,
      "loss": 2.5475,
      "num_input_tokens_seen": 49263920,
      "step": 23580
    },
    {
      "epoch": 3.847540582429236,
      "grad_norm": 4.71875,
      "learning_rate": 1.8055396364024317e-05,
      "loss": 3.06,
      "num_input_tokens_seen": 49274608,
      "step": 23585
    },
    {
      "epoch": 3.848356309650053,
      "grad_norm": 16.5,
      "learning_rate": 1.804596580405601e-05,
      "loss": 2.7333,
      "num_input_tokens_seen": 49284992,
      "step": 23590
    },
    {
      "epoch": 3.84917203687087,
      "grad_norm": 8.5625,
      "learning_rate": 1.8036536316487174e-05,
      "loss": 2.7093,
      "num_input_tokens_seen": 49297072,
      "step": 23595
    },
    {
      "epoch": 3.8499877640916877,
      "grad_norm": 3.71875,
      "learning_rate": 1.802710790277193e-05,
      "loss": 1.5842,
      "num_input_tokens_seen": 49308304,
      "step": 23600
    },
    {
      "epoch": 3.8499877640916877,
      "eval_loss": 2.5380730628967285,
      "eval_runtime": 135.0144,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 49308304,
      "step": 23600
    },
    {
      "epoch": 3.8508034913125053,
      "grad_norm": 4.78125,
      "learning_rate": 1.801768056436429e-05,
      "loss": 2.4033,
      "num_input_tokens_seen": 49318272,
      "step": 23605
    },
    {
      "epoch": 3.8516192185333225,
      "grad_norm": 7.5625,
      "learning_rate": 1.8008254302718035e-05,
      "loss": 2.6844,
      "num_input_tokens_seen": 49329904,
      "step": 23610
    },
    {
      "epoch": 3.8524349457541396,
      "grad_norm": 2.75,
      "learning_rate": 1.7998829119286837e-05,
      "loss": 1.3121,
      "num_input_tokens_seen": 49339984,
      "step": 23615
    },
    {
      "epoch": 3.853250672974957,
      "grad_norm": 4.84375,
      "learning_rate": 1.798940501552418e-05,
      "loss": 1.8081,
      "num_input_tokens_seen": 49351680,
      "step": 23620
    },
    {
      "epoch": 3.8540664001957747,
      "grad_norm": 7.0625,
      "learning_rate": 1.797998199288336e-05,
      "loss": 1.8598,
      "num_input_tokens_seen": 49360576,
      "step": 23625
    },
    {
      "epoch": 3.854882127416592,
      "grad_norm": 9.875,
      "learning_rate": 1.7970560052817543e-05,
      "loss": 2.517,
      "num_input_tokens_seen": 49371840,
      "step": 23630
    },
    {
      "epoch": 3.855697854637409,
      "grad_norm": 6.96875,
      "learning_rate": 1.7961139196779702e-05,
      "loss": 2.618,
      "num_input_tokens_seen": 49381968,
      "step": 23635
    },
    {
      "epoch": 3.8565135818582266,
      "grad_norm": 7.21875,
      "learning_rate": 1.7951719426222647e-05,
      "loss": 3.6684,
      "num_input_tokens_seen": 49392896,
      "step": 23640
    },
    {
      "epoch": 3.857329309079044,
      "grad_norm": 6.59375,
      "learning_rate": 1.794230074259904e-05,
      "loss": 1.9203,
      "num_input_tokens_seen": 49402352,
      "step": 23645
    },
    {
      "epoch": 3.8581450362998613,
      "grad_norm": 4.90625,
      "learning_rate": 1.7932883147361336e-05,
      "loss": 3.1964,
      "num_input_tokens_seen": 49413504,
      "step": 23650
    },
    {
      "epoch": 3.858960763520679,
      "grad_norm": 7.78125,
      "learning_rate": 1.7923466641961865e-05,
      "loss": 3.1565,
      "num_input_tokens_seen": 49423952,
      "step": 23655
    },
    {
      "epoch": 3.859776490741496,
      "grad_norm": 2.34375,
      "learning_rate": 1.791405122785278e-05,
      "loss": 1.7276,
      "num_input_tokens_seen": 49433808,
      "step": 23660
    },
    {
      "epoch": 3.8605922179623136,
      "grad_norm": 4.4375,
      "learning_rate": 1.7904636906486037e-05,
      "loss": 2.0053,
      "num_input_tokens_seen": 49445152,
      "step": 23665
    },
    {
      "epoch": 3.8614079451831307,
      "grad_norm": 17.25,
      "learning_rate": 1.7895223679313448e-05,
      "loss": 3.456,
      "num_input_tokens_seen": 49456064,
      "step": 23670
    },
    {
      "epoch": 3.8622236724039483,
      "grad_norm": 12.6875,
      "learning_rate": 1.7885811547786653e-05,
      "loss": 2.1831,
      "num_input_tokens_seen": 49467360,
      "step": 23675
    },
    {
      "epoch": 3.8630393996247654,
      "grad_norm": 3.65625,
      "learning_rate": 1.7876400513357115e-05,
      "loss": 2.5175,
      "num_input_tokens_seen": 49478224,
      "step": 23680
    },
    {
      "epoch": 3.863855126845583,
      "grad_norm": 5.65625,
      "learning_rate": 1.7866990577476146e-05,
      "loss": 1.7914,
      "num_input_tokens_seen": 49488960,
      "step": 23685
    },
    {
      "epoch": 3.8646708540664,
      "grad_norm": 1.453125,
      "learning_rate": 1.7857581741594863e-05,
      "loss": 1.8239,
      "num_input_tokens_seen": 49499552,
      "step": 23690
    },
    {
      "epoch": 3.8654865812872177,
      "grad_norm": 0.76953125,
      "learning_rate": 1.7848174007164237e-05,
      "loss": 2.4778,
      "num_input_tokens_seen": 49509696,
      "step": 23695
    },
    {
      "epoch": 3.866302308508035,
      "grad_norm": 8.1875,
      "learning_rate": 1.7838767375635052e-05,
      "loss": 2.1175,
      "num_input_tokens_seen": 49520336,
      "step": 23700
    },
    {
      "epoch": 3.8671180357288524,
      "grad_norm": 4.625,
      "learning_rate": 1.782936184845793e-05,
      "loss": 1.6817,
      "num_input_tokens_seen": 49530320,
      "step": 23705
    },
    {
      "epoch": 3.8679337629496695,
      "grad_norm": 9.5,
      "learning_rate": 1.7819957427083334e-05,
      "loss": 3.01,
      "num_input_tokens_seen": 49540944,
      "step": 23710
    },
    {
      "epoch": 3.868749490170487,
      "grad_norm": 6.25,
      "learning_rate": 1.7810554112961516e-05,
      "loss": 2.4251,
      "num_input_tokens_seen": 49551168,
      "step": 23715
    },
    {
      "epoch": 3.869565217391304,
      "grad_norm": 3.765625,
      "learning_rate": 1.7801151907542607e-05,
      "loss": 1.7166,
      "num_input_tokens_seen": 49561136,
      "step": 23720
    },
    {
      "epoch": 3.870380944612122,
      "grad_norm": 4.75,
      "learning_rate": 1.7791750812276547e-05,
      "loss": 1.8846,
      "num_input_tokens_seen": 49570864,
      "step": 23725
    },
    {
      "epoch": 3.871196671832939,
      "grad_norm": 4.84375,
      "learning_rate": 1.778235082861309e-05,
      "loss": 1.6086,
      "num_input_tokens_seen": 49581824,
      "step": 23730
    },
    {
      "epoch": 3.8720123990537565,
      "grad_norm": 8.6875,
      "learning_rate": 1.777295195800184e-05,
      "loss": 3.4678,
      "num_input_tokens_seen": 49591712,
      "step": 23735
    },
    {
      "epoch": 3.8728281262745736,
      "grad_norm": 4.1875,
      "learning_rate": 1.7763554201892215e-05,
      "loss": 2.5205,
      "num_input_tokens_seen": 49602192,
      "step": 23740
    },
    {
      "epoch": 3.873643853495391,
      "grad_norm": 8.5,
      "learning_rate": 1.7754157561733476e-05,
      "loss": 1.9765,
      "num_input_tokens_seen": 49612704,
      "step": 23745
    },
    {
      "epoch": 3.8744595807162083,
      "grad_norm": 3.9375,
      "learning_rate": 1.7744762038974702e-05,
      "loss": 2.2369,
      "num_input_tokens_seen": 49623184,
      "step": 23750
    },
    {
      "epoch": 3.875275307937026,
      "grad_norm": 4.90625,
      "learning_rate": 1.7735367635064788e-05,
      "loss": 1.5965,
      "num_input_tokens_seen": 49634720,
      "step": 23755
    },
    {
      "epoch": 3.8760910351578435,
      "grad_norm": 5.28125,
      "learning_rate": 1.7725974351452474e-05,
      "loss": 1.8073,
      "num_input_tokens_seen": 49644640,
      "step": 23760
    },
    {
      "epoch": 3.8769067623786606,
      "grad_norm": 7.84375,
      "learning_rate": 1.771658218958634e-05,
      "loss": 1.3121,
      "num_input_tokens_seen": 49654160,
      "step": 23765
    },
    {
      "epoch": 3.8777224895994777,
      "grad_norm": 1.109375,
      "learning_rate": 1.770719115091475e-05,
      "loss": 1.6511,
      "num_input_tokens_seen": 49664816,
      "step": 23770
    },
    {
      "epoch": 3.8785382168202953,
      "grad_norm": 4.5625,
      "learning_rate": 1.7697801236885935e-05,
      "loss": 2.1329,
      "num_input_tokens_seen": 49675968,
      "step": 23775
    },
    {
      "epoch": 3.879353944041113,
      "grad_norm": 16.125,
      "learning_rate": 1.7688412448947944e-05,
      "loss": 2.0832,
      "num_input_tokens_seen": 49686992,
      "step": 23780
    },
    {
      "epoch": 3.88016967126193,
      "grad_norm": 8.75,
      "learning_rate": 1.767902478854862e-05,
      "loss": 2.5624,
      "num_input_tokens_seen": 49697392,
      "step": 23785
    },
    {
      "epoch": 3.880985398482747,
      "grad_norm": 9.6875,
      "learning_rate": 1.766963825713569e-05,
      "loss": 1.3421,
      "num_input_tokens_seen": 49707184,
      "step": 23790
    },
    {
      "epoch": 3.8818011257035647,
      "grad_norm": 3.421875,
      "learning_rate": 1.766025285615665e-05,
      "loss": 1.4656,
      "num_input_tokens_seen": 49717424,
      "step": 23795
    },
    {
      "epoch": 3.8826168529243823,
      "grad_norm": 7.0,
      "learning_rate": 1.7650868587058854e-05,
      "loss": 3.0507,
      "num_input_tokens_seen": 49728368,
      "step": 23800
    },
    {
      "epoch": 3.8826168529243823,
      "eval_loss": 2.532986640930176,
      "eval_runtime": 135.0473,
      "eval_samples_per_second": 20.178,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 49728368,
      "step": 23800
    },
    {
      "epoch": 3.8834325801451994,
      "grad_norm": 9.5,
      "learning_rate": 1.7641485451289484e-05,
      "loss": 3.1382,
      "num_input_tokens_seen": 49738752,
      "step": 23805
    },
    {
      "epoch": 3.8842483073660166,
      "grad_norm": 8.1875,
      "learning_rate": 1.7632103450295534e-05,
      "loss": 3.5653,
      "num_input_tokens_seen": 49746480,
      "step": 23810
    },
    {
      "epoch": 3.885064034586834,
      "grad_norm": 5.625,
      "learning_rate": 1.762272258552381e-05,
      "loss": 2.8683,
      "num_input_tokens_seen": 49758128,
      "step": 23815
    },
    {
      "epoch": 3.8858797618076517,
      "grad_norm": 11.0,
      "learning_rate": 1.7613342858420988e-05,
      "loss": 2.9552,
      "num_input_tokens_seen": 49769072,
      "step": 23820
    },
    {
      "epoch": 3.886695489028469,
      "grad_norm": 3.359375,
      "learning_rate": 1.760396427043351e-05,
      "loss": 1.876,
      "num_input_tokens_seen": 49779104,
      "step": 23825
    },
    {
      "epoch": 3.887511216249286,
      "grad_norm": 2.4375,
      "learning_rate": 1.7594586823007696e-05,
      "loss": 1.8226,
      "num_input_tokens_seen": 49790000,
      "step": 23830
    },
    {
      "epoch": 3.8883269434701035,
      "grad_norm": 2.125,
      "learning_rate": 1.7585210517589646e-05,
      "loss": 1.902,
      "num_input_tokens_seen": 49800256,
      "step": 23835
    },
    {
      "epoch": 3.889142670690921,
      "grad_norm": 5.875,
      "learning_rate": 1.7575835355625314e-05,
      "loss": 2.4173,
      "num_input_tokens_seen": 49810848,
      "step": 23840
    },
    {
      "epoch": 3.8899583979117383,
      "grad_norm": 4.75,
      "learning_rate": 1.756646133856048e-05,
      "loss": 3.121,
      "num_input_tokens_seen": 49821600,
      "step": 23845
    },
    {
      "epoch": 3.890774125132556,
      "grad_norm": 6.75,
      "learning_rate": 1.7557088467840714e-05,
      "loss": 2.3789,
      "num_input_tokens_seen": 49829904,
      "step": 23850
    },
    {
      "epoch": 3.891589852353373,
      "grad_norm": 4.28125,
      "learning_rate": 1.7547716744911438e-05,
      "loss": 2.889,
      "num_input_tokens_seen": 49839024,
      "step": 23855
    },
    {
      "epoch": 3.8924055795741905,
      "grad_norm": 5.4375,
      "learning_rate": 1.7538346171217902e-05,
      "loss": 2.8393,
      "num_input_tokens_seen": 49850304,
      "step": 23860
    },
    {
      "epoch": 3.8932213067950077,
      "grad_norm": 7.59375,
      "learning_rate": 1.7528976748205146e-05,
      "loss": 2.1528,
      "num_input_tokens_seen": 49860608,
      "step": 23865
    },
    {
      "epoch": 3.8940370340158252,
      "grad_norm": 12.75,
      "learning_rate": 1.751960847731807e-05,
      "loss": 2.2034,
      "num_input_tokens_seen": 49870848,
      "step": 23870
    },
    {
      "epoch": 3.8948527612366424,
      "grad_norm": 8.0625,
      "learning_rate": 1.7510241360001362e-05,
      "loss": 2.1133,
      "num_input_tokens_seen": 49882112,
      "step": 23875
    },
    {
      "epoch": 3.89566848845746,
      "grad_norm": 10.75,
      "learning_rate": 1.7500875397699562e-05,
      "loss": 2.1655,
      "num_input_tokens_seen": 49890944,
      "step": 23880
    },
    {
      "epoch": 3.896484215678277,
      "grad_norm": 5.65625,
      "learning_rate": 1.7491510591857015e-05,
      "loss": 2.9392,
      "num_input_tokens_seen": 49901520,
      "step": 23885
    },
    {
      "epoch": 3.8972999428990946,
      "grad_norm": 8.5625,
      "learning_rate": 1.7482146943917896e-05,
      "loss": 2.4056,
      "num_input_tokens_seen": 49911712,
      "step": 23890
    },
    {
      "epoch": 3.898115670119912,
      "grad_norm": 6.75,
      "learning_rate": 1.7472784455326185e-05,
      "loss": 2.0738,
      "num_input_tokens_seen": 49923344,
      "step": 23895
    },
    {
      "epoch": 3.8989313973407294,
      "grad_norm": 9.75,
      "learning_rate": 1.746342312752572e-05,
      "loss": 2.7893,
      "num_input_tokens_seen": 49933520,
      "step": 23900
    },
    {
      "epoch": 3.8997471245615465,
      "grad_norm": 6.6875,
      "learning_rate": 1.74540629619601e-05,
      "loss": 2.4992,
      "num_input_tokens_seen": 49943152,
      "step": 23905
    },
    {
      "epoch": 3.900562851782364,
      "grad_norm": 9.25,
      "learning_rate": 1.7444703960072815e-05,
      "loss": 2.8972,
      "num_input_tokens_seen": 49951840,
      "step": 23910
    },
    {
      "epoch": 3.901378579003181,
      "grad_norm": 4.84375,
      "learning_rate": 1.7435346123307118e-05,
      "loss": 1.2922,
      "num_input_tokens_seen": 49962128,
      "step": 23915
    },
    {
      "epoch": 3.9021943062239988,
      "grad_norm": 6.25,
      "learning_rate": 1.742598945310611e-05,
      "loss": 1.6437,
      "num_input_tokens_seen": 49972528,
      "step": 23920
    },
    {
      "epoch": 3.903010033444816,
      "grad_norm": 8.125,
      "learning_rate": 1.741663395091272e-05,
      "loss": 2.9469,
      "num_input_tokens_seen": 49982096,
      "step": 23925
    },
    {
      "epoch": 3.9038257606656335,
      "grad_norm": 3.5,
      "learning_rate": 1.7407279618169657e-05,
      "loss": 2.8083,
      "num_input_tokens_seen": 49992864,
      "step": 23930
    },
    {
      "epoch": 3.904641487886451,
      "grad_norm": 5.65625,
      "learning_rate": 1.73979264563195e-05,
      "loss": 3.0942,
      "num_input_tokens_seen": 50002864,
      "step": 23935
    },
    {
      "epoch": 3.905457215107268,
      "grad_norm": 0.1884765625,
      "learning_rate": 1.7388574466804625e-05,
      "loss": 1.7955,
      "num_input_tokens_seen": 50014064,
      "step": 23940
    },
    {
      "epoch": 3.9062729423280853,
      "grad_norm": 9.0,
      "learning_rate": 1.7379223651067207e-05,
      "loss": 2.8995,
      "num_input_tokens_seen": 50025088,
      "step": 23945
    },
    {
      "epoch": 3.907088669548903,
      "grad_norm": 10.8125,
      "learning_rate": 1.736987401054928e-05,
      "loss": 2.4534,
      "num_input_tokens_seen": 50034912,
      "step": 23950
    },
    {
      "epoch": 3.9079043967697205,
      "grad_norm": 8.9375,
      "learning_rate": 1.736052554669266e-05,
      "loss": 3.3202,
      "num_input_tokens_seen": 50044912,
      "step": 23955
    },
    {
      "epoch": 3.9087201239905376,
      "grad_norm": 12.25,
      "learning_rate": 1.7351178260939007e-05,
      "loss": 2.3299,
      "num_input_tokens_seen": 50055728,
      "step": 23960
    },
    {
      "epoch": 3.9095358512113547,
      "grad_norm": 3.578125,
      "learning_rate": 1.7341832154729794e-05,
      "loss": 2.7093,
      "num_input_tokens_seen": 50065168,
      "step": 23965
    },
    {
      "epoch": 3.9103515784321723,
      "grad_norm": 5.625,
      "learning_rate": 1.7332487229506286e-05,
      "loss": 2.3807,
      "num_input_tokens_seen": 50076352,
      "step": 23970
    },
    {
      "epoch": 3.91116730565299,
      "grad_norm": 4.65625,
      "learning_rate": 1.732314348670961e-05,
      "loss": 2.8095,
      "num_input_tokens_seen": 50088432,
      "step": 23975
    },
    {
      "epoch": 3.911983032873807,
      "grad_norm": 3.25,
      "learning_rate": 1.7313800927780686e-05,
      "loss": 2.4315,
      "num_input_tokens_seen": 50100112,
      "step": 23980
    },
    {
      "epoch": 3.912798760094624,
      "grad_norm": 8.9375,
      "learning_rate": 1.7304459554160245e-05,
      "loss": 2.5146,
      "num_input_tokens_seen": 50110320,
      "step": 23985
    },
    {
      "epoch": 3.9136144873154417,
      "grad_norm": 9.5625,
      "learning_rate": 1.7295119367288853e-05,
      "loss": 2.723,
      "num_input_tokens_seen": 50120464,
      "step": 23990
    },
    {
      "epoch": 3.9144302145362593,
      "grad_norm": 4.5,
      "learning_rate": 1.728578036860688e-05,
      "loss": 2.7446,
      "num_input_tokens_seen": 50129680,
      "step": 23995
    },
    {
      "epoch": 3.9152459417570764,
      "grad_norm": 11.375,
      "learning_rate": 1.7276442559554513e-05,
      "loss": 3.2435,
      "num_input_tokens_seen": 50140272,
      "step": 24000
    },
    {
      "epoch": 3.9152459417570764,
      "eval_loss": 2.5345358848571777,
      "eval_runtime": 135.0261,
      "eval_samples_per_second": 20.181,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 50140272,
      "step": 24000
    },
    {
      "epoch": 3.9160616689778935,
      "grad_norm": 17.0,
      "learning_rate": 1.726710594157177e-05,
      "loss": 1.9235,
      "num_input_tokens_seen": 50149872,
      "step": 24005
    },
    {
      "epoch": 3.916877396198711,
      "grad_norm": 2.53125,
      "learning_rate": 1.725777051609846e-05,
      "loss": 1.5808,
      "num_input_tokens_seen": 50161136,
      "step": 24010
    },
    {
      "epoch": 3.9176931234195287,
      "grad_norm": 4.90625,
      "learning_rate": 1.7248436284574228e-05,
      "loss": 2.2378,
      "num_input_tokens_seen": 50169104,
      "step": 24015
    },
    {
      "epoch": 3.918508850640346,
      "grad_norm": 8.6875,
      "learning_rate": 1.723910324843855e-05,
      "loss": 1.5061,
      "num_input_tokens_seen": 50179104,
      "step": 24020
    },
    {
      "epoch": 3.919324577861163,
      "grad_norm": 4.625,
      "learning_rate": 1.722977140913067e-05,
      "loss": 2.3459,
      "num_input_tokens_seen": 50189920,
      "step": 24025
    },
    {
      "epoch": 3.9201403050819805,
      "grad_norm": 6.5,
      "learning_rate": 1.7220440768089688e-05,
      "loss": 3.9588,
      "num_input_tokens_seen": 50200608,
      "step": 24030
    },
    {
      "epoch": 3.920956032302798,
      "grad_norm": 4.375,
      "learning_rate": 1.7211111326754505e-05,
      "loss": 2.3685,
      "num_input_tokens_seen": 50210560,
      "step": 24035
    },
    {
      "epoch": 3.9217717595236152,
      "grad_norm": 9.0625,
      "learning_rate": 1.720178308656383e-05,
      "loss": 2.5645,
      "num_input_tokens_seen": 50221568,
      "step": 24040
    },
    {
      "epoch": 3.922587486744433,
      "grad_norm": 6.96875,
      "learning_rate": 1.719245604895621e-05,
      "loss": 3.063,
      "num_input_tokens_seen": 50230848,
      "step": 24045
    },
    {
      "epoch": 3.92340321396525,
      "grad_norm": 3.375,
      "learning_rate": 1.7183130215369972e-05,
      "loss": 2.5264,
      "num_input_tokens_seen": 50241840,
      "step": 24050
    },
    {
      "epoch": 3.9242189411860675,
      "grad_norm": 5.65625,
      "learning_rate": 1.7173805587243292e-05,
      "loss": 1.5753,
      "num_input_tokens_seen": 50252432,
      "step": 24055
    },
    {
      "epoch": 3.9250346684068846,
      "grad_norm": 7.71875,
      "learning_rate": 1.7164482166014147e-05,
      "loss": 3.072,
      "num_input_tokens_seen": 50262704,
      "step": 24060
    },
    {
      "epoch": 3.925850395627702,
      "grad_norm": 12.625,
      "learning_rate": 1.7155159953120313e-05,
      "loss": 2.0696,
      "num_input_tokens_seen": 50273184,
      "step": 24065
    },
    {
      "epoch": 3.9266661228485193,
      "grad_norm": 7.9375,
      "learning_rate": 1.714583894999941e-05,
      "loss": 2.933,
      "num_input_tokens_seen": 50283296,
      "step": 24070
    },
    {
      "epoch": 3.927481850069337,
      "grad_norm": 8.4375,
      "learning_rate": 1.7136519158088826e-05,
      "loss": 1.298,
      "num_input_tokens_seen": 50293536,
      "step": 24075
    },
    {
      "epoch": 3.928297577290154,
      "grad_norm": 7.03125,
      "learning_rate": 1.712720057882581e-05,
      "loss": 3.1418,
      "num_input_tokens_seen": 50305104,
      "step": 24080
    },
    {
      "epoch": 3.9291133045109716,
      "grad_norm": 10.75,
      "learning_rate": 1.7117883213647413e-05,
      "loss": 3.1658,
      "num_input_tokens_seen": 50315648,
      "step": 24085
    },
    {
      "epoch": 3.9299290317317888,
      "grad_norm": 6.15625,
      "learning_rate": 1.710856706399046e-05,
      "loss": 2.8169,
      "num_input_tokens_seen": 50326352,
      "step": 24090
    },
    {
      "epoch": 3.9307447589526063,
      "grad_norm": 4.65625,
      "learning_rate": 1.7099252131291648e-05,
      "loss": 1.7872,
      "num_input_tokens_seen": 50337520,
      "step": 24095
    },
    {
      "epoch": 3.9315604861734235,
      "grad_norm": 6.5625,
      "learning_rate": 1.708993841698744e-05,
      "loss": 1.7968,
      "num_input_tokens_seen": 50348880,
      "step": 24100
    },
    {
      "epoch": 3.932376213394241,
      "grad_norm": 4.1875,
      "learning_rate": 1.7080625922514132e-05,
      "loss": 2.7252,
      "num_input_tokens_seen": 50359840,
      "step": 24105
    },
    {
      "epoch": 3.933191940615058,
      "grad_norm": 4.625,
      "learning_rate": 1.7071314649307836e-05,
      "loss": 3.2273,
      "num_input_tokens_seen": 50371248,
      "step": 24110
    },
    {
      "epoch": 3.9340076678358757,
      "grad_norm": 8.625,
      "learning_rate": 1.7062004598804448e-05,
      "loss": 2.5219,
      "num_input_tokens_seen": 50382672,
      "step": 24115
    },
    {
      "epoch": 3.934823395056693,
      "grad_norm": 6.40625,
      "learning_rate": 1.7052695772439702e-05,
      "loss": 1.8529,
      "num_input_tokens_seen": 50393584,
      "step": 24120
    },
    {
      "epoch": 3.9356391222775104,
      "grad_norm": 17.75,
      "learning_rate": 1.7043388171649154e-05,
      "loss": 2.5868,
      "num_input_tokens_seen": 50404016,
      "step": 24125
    },
    {
      "epoch": 3.936454849498328,
      "grad_norm": 10.0,
      "learning_rate": 1.7034081797868127e-05,
      "loss": 2.3666,
      "num_input_tokens_seen": 50415488,
      "step": 24130
    },
    {
      "epoch": 3.937270576719145,
      "grad_norm": 9.4375,
      "learning_rate": 1.70247766525318e-05,
      "loss": 2.8201,
      "num_input_tokens_seen": 50426464,
      "step": 24135
    },
    {
      "epoch": 3.9380863039399623,
      "grad_norm": 6.4375,
      "learning_rate": 1.701547273707514e-05,
      "loss": 1.9513,
      "num_input_tokens_seen": 50435312,
      "step": 24140
    },
    {
      "epoch": 3.93890203116078,
      "grad_norm": 7.8125,
      "learning_rate": 1.7006170052932916e-05,
      "loss": 2.5555,
      "num_input_tokens_seen": 50445792,
      "step": 24145
    },
    {
      "epoch": 3.9397177583815974,
      "grad_norm": 9.0,
      "learning_rate": 1.6996868601539735e-05,
      "loss": 3.1563,
      "num_input_tokens_seen": 50455808,
      "step": 24150
    },
    {
      "epoch": 3.9405334856024146,
      "grad_norm": 4.5625,
      "learning_rate": 1.6987568384329977e-05,
      "loss": 2.6695,
      "num_input_tokens_seen": 50464960,
      "step": 24155
    },
    {
      "epoch": 3.9413492128232317,
      "grad_norm": 8.3125,
      "learning_rate": 1.6978269402737866e-05,
      "loss": 1.6504,
      "num_input_tokens_seen": 50475760,
      "step": 24160
    },
    {
      "epoch": 3.9421649400440493,
      "grad_norm": 4.90625,
      "learning_rate": 1.696897165819743e-05,
      "loss": 1.9029,
      "num_input_tokens_seen": 50487600,
      "step": 24165
    },
    {
      "epoch": 3.942980667264867,
      "grad_norm": 4.90625,
      "learning_rate": 1.6959675152142487e-05,
      "loss": 1.84,
      "num_input_tokens_seen": 50497200,
      "step": 24170
    },
    {
      "epoch": 3.943796394485684,
      "grad_norm": 6.3125,
      "learning_rate": 1.6950379886006667e-05,
      "loss": 2.2768,
      "num_input_tokens_seen": 50508416,
      "step": 24175
    },
    {
      "epoch": 3.944612121706501,
      "grad_norm": 10.875,
      "learning_rate": 1.6941085861223438e-05,
      "loss": 2.633,
      "num_input_tokens_seen": 50518992,
      "step": 24180
    },
    {
      "epoch": 3.9454278489273187,
      "grad_norm": 6.65625,
      "learning_rate": 1.6931793079226034e-05,
      "loss": 2.492,
      "num_input_tokens_seen": 50529664,
      "step": 24185
    },
    {
      "epoch": 3.9462435761481363,
      "grad_norm": 2.84375,
      "learning_rate": 1.692250154144754e-05,
      "loss": 1.5206,
      "num_input_tokens_seen": 50540192,
      "step": 24190
    },
    {
      "epoch": 3.9470593033689534,
      "grad_norm": 8.4375,
      "learning_rate": 1.6913211249320807e-05,
      "loss": 3.3063,
      "num_input_tokens_seen": 50549568,
      "step": 24195
    },
    {
      "epoch": 3.9478750305897705,
      "grad_norm": 9.0,
      "learning_rate": 1.6903922204278522e-05,
      "loss": 2.597,
      "num_input_tokens_seen": 50557680,
      "step": 24200
    },
    {
      "epoch": 3.9478750305897705,
      "eval_loss": 2.538020133972168,
      "eval_runtime": 134.8482,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 50557680,
      "step": 24200
    },
    {
      "epoch": 3.948690757810588,
      "grad_norm": 8.25,
      "learning_rate": 1.6894634407753186e-05,
      "loss": 1.9356,
      "num_input_tokens_seen": 50569904,
      "step": 24205
    },
    {
      "epoch": 3.9495064850314057,
      "grad_norm": 8.75,
      "learning_rate": 1.6885347861177077e-05,
      "loss": 2.0969,
      "num_input_tokens_seen": 50580304,
      "step": 24210
    },
    {
      "epoch": 3.950322212252223,
      "grad_norm": 9.1875,
      "learning_rate": 1.6876062565982298e-05,
      "loss": 2.8843,
      "num_input_tokens_seen": 50591760,
      "step": 24215
    },
    {
      "epoch": 3.9511379394730404,
      "grad_norm": 8.8125,
      "learning_rate": 1.6866778523600774e-05,
      "loss": 1.3718,
      "num_input_tokens_seen": 50602928,
      "step": 24220
    },
    {
      "epoch": 3.9519536666938575,
      "grad_norm": 5.25,
      "learning_rate": 1.6857495735464195e-05,
      "loss": 2.2752,
      "num_input_tokens_seen": 50613200,
      "step": 24225
    },
    {
      "epoch": 3.952769393914675,
      "grad_norm": 10.0625,
      "learning_rate": 1.6848214203004115e-05,
      "loss": 2.8655,
      "num_input_tokens_seen": 50623984,
      "step": 24230
    },
    {
      "epoch": 3.953585121135492,
      "grad_norm": 6.0625,
      "learning_rate": 1.6838933927651835e-05,
      "loss": 1.7066,
      "num_input_tokens_seen": 50634160,
      "step": 24235
    },
    {
      "epoch": 3.95440084835631,
      "grad_norm": 6.84375,
      "learning_rate": 1.6829654910838506e-05,
      "loss": 1.5895,
      "num_input_tokens_seen": 50644512,
      "step": 24240
    },
    {
      "epoch": 3.955216575577127,
      "grad_norm": 4.9375,
      "learning_rate": 1.6820377153995065e-05,
      "loss": 2.3008,
      "num_input_tokens_seen": 50655456,
      "step": 24245
    },
    {
      "epoch": 3.9560323027979445,
      "grad_norm": 2.515625,
      "learning_rate": 1.681110065855226e-05,
      "loss": 2.5212,
      "num_input_tokens_seen": 50665584,
      "step": 24250
    },
    {
      "epoch": 3.9568480300187616,
      "grad_norm": 8.125,
      "learning_rate": 1.6801825425940642e-05,
      "loss": 2.7638,
      "num_input_tokens_seen": 50675328,
      "step": 24255
    },
    {
      "epoch": 3.957663757239579,
      "grad_norm": 6.0,
      "learning_rate": 1.679255145759056e-05,
      "loss": 2.979,
      "num_input_tokens_seen": 50685232,
      "step": 24260
    },
    {
      "epoch": 3.9584794844603963,
      "grad_norm": 8.4375,
      "learning_rate": 1.6783278754932187e-05,
      "loss": 2.7366,
      "num_input_tokens_seen": 50696144,
      "step": 24265
    },
    {
      "epoch": 3.959295211681214,
      "grad_norm": 8.375,
      "learning_rate": 1.6774007319395496e-05,
      "loss": 1.5716,
      "num_input_tokens_seen": 50706224,
      "step": 24270
    },
    {
      "epoch": 3.960110938902031,
      "grad_norm": 1.71875,
      "learning_rate": 1.6764737152410243e-05,
      "loss": 0.8679,
      "num_input_tokens_seen": 50716144,
      "step": 24275
    },
    {
      "epoch": 3.9609266661228486,
      "grad_norm": 7.09375,
      "learning_rate": 1.6755468255406016e-05,
      "loss": 3.1037,
      "num_input_tokens_seen": 50726448,
      "step": 24280
    },
    {
      "epoch": 3.9617423933436657,
      "grad_norm": 9.0625,
      "learning_rate": 1.674620062981219e-05,
      "loss": 3.5542,
      "num_input_tokens_seen": 50736144,
      "step": 24285
    },
    {
      "epoch": 3.9625581205644833,
      "grad_norm": 8.9375,
      "learning_rate": 1.6736934277057947e-05,
      "loss": 2.3121,
      "num_input_tokens_seen": 50746080,
      "step": 24290
    },
    {
      "epoch": 3.9633738477853004,
      "grad_norm": 4.875,
      "learning_rate": 1.6727669198572286e-05,
      "loss": 2.1518,
      "num_input_tokens_seen": 50756960,
      "step": 24295
    },
    {
      "epoch": 3.964189575006118,
      "grad_norm": 8.0,
      "learning_rate": 1.6718405395783984e-05,
      "loss": 2.8379,
      "num_input_tokens_seen": 50767632,
      "step": 24300
    },
    {
      "epoch": 3.965005302226935,
      "grad_norm": 5.5,
      "learning_rate": 1.6709142870121643e-05,
      "loss": 3.4078,
      "num_input_tokens_seen": 50777072,
      "step": 24305
    },
    {
      "epoch": 3.9658210294477527,
      "grad_norm": 4.46875,
      "learning_rate": 1.669988162301367e-05,
      "loss": 2.0472,
      "num_input_tokens_seen": 50788672,
      "step": 24310
    },
    {
      "epoch": 3.96663675666857,
      "grad_norm": 4.84375,
      "learning_rate": 1.6690621655888243e-05,
      "loss": 1.5713,
      "num_input_tokens_seen": 50799888,
      "step": 24315
    },
    {
      "epoch": 3.9674524838893874,
      "grad_norm": 4.09375,
      "learning_rate": 1.6681362970173386e-05,
      "loss": 1.7874,
      "num_input_tokens_seen": 50809600,
      "step": 24320
    },
    {
      "epoch": 3.968268211110205,
      "grad_norm": 8.125,
      "learning_rate": 1.6672105567296904e-05,
      "loss": 4.0698,
      "num_input_tokens_seen": 50817744,
      "step": 24325
    },
    {
      "epoch": 3.969083938331022,
      "grad_norm": 6.28125,
      "learning_rate": 1.666284944868639e-05,
      "loss": 1.6144,
      "num_input_tokens_seen": 50830240,
      "step": 24330
    },
    {
      "epoch": 3.9698996655518393,
      "grad_norm": 2.125,
      "learning_rate": 1.665359461576927e-05,
      "loss": 1.6948,
      "num_input_tokens_seen": 50840800,
      "step": 24335
    },
    {
      "epoch": 3.970715392772657,
      "grad_norm": 7.78125,
      "learning_rate": 1.6644341069972736e-05,
      "loss": 2.3517,
      "num_input_tokens_seen": 50851440,
      "step": 24340
    },
    {
      "epoch": 3.9715311199934744,
      "grad_norm": 6.125,
      "learning_rate": 1.6635088812723813e-05,
      "loss": 2.0406,
      "num_input_tokens_seen": 50861856,
      "step": 24345
    },
    {
      "epoch": 3.9723468472142915,
      "grad_norm": 6.625,
      "learning_rate": 1.6625837845449328e-05,
      "loss": 1.773,
      "num_input_tokens_seen": 50872528,
      "step": 24350
    },
    {
      "epoch": 3.9731625744351087,
      "grad_norm": 5.4375,
      "learning_rate": 1.6616588169575874e-05,
      "loss": 0.7588,
      "num_input_tokens_seen": 50882928,
      "step": 24355
    },
    {
      "epoch": 3.9739783016559262,
      "grad_norm": 5.90625,
      "learning_rate": 1.6607339786529878e-05,
      "loss": 2.7793,
      "num_input_tokens_seen": 50894336,
      "step": 24360
    },
    {
      "epoch": 3.974794028876744,
      "grad_norm": 9.375,
      "learning_rate": 1.659809269773756e-05,
      "loss": 1.2777,
      "num_input_tokens_seen": 50904112,
      "step": 24365
    },
    {
      "epoch": 3.975609756097561,
      "grad_norm": 6.0,
      "learning_rate": 1.658884690462493e-05,
      "loss": 2.5333,
      "num_input_tokens_seen": 50914752,
      "step": 24370
    },
    {
      "epoch": 3.976425483318378,
      "grad_norm": 4.59375,
      "learning_rate": 1.6579602408617813e-05,
      "loss": 3.1664,
      "num_input_tokens_seen": 50926304,
      "step": 24375
    },
    {
      "epoch": 3.9772412105391957,
      "grad_norm": 7.09375,
      "learning_rate": 1.657035921114181e-05,
      "loss": 2.6584,
      "num_input_tokens_seen": 50936400,
      "step": 24380
    },
    {
      "epoch": 3.9780569377600132,
      "grad_norm": 10.375,
      "learning_rate": 1.656111731362236e-05,
      "loss": 2.8105,
      "num_input_tokens_seen": 50946176,
      "step": 24385
    },
    {
      "epoch": 3.9788726649808304,
      "grad_norm": 7.375,
      "learning_rate": 1.6551876717484666e-05,
      "loss": 1.6455,
      "num_input_tokens_seen": 50957424,
      "step": 24390
    },
    {
      "epoch": 3.9796883922016475,
      "grad_norm": 6.65625,
      "learning_rate": 1.6542637424153752e-05,
      "loss": 2.5444,
      "num_input_tokens_seen": 50968656,
      "step": 24395
    },
    {
      "epoch": 3.980504119422465,
      "grad_norm": 6.40625,
      "learning_rate": 1.6533399435054418e-05,
      "loss": 2.3862,
      "num_input_tokens_seen": 50978512,
      "step": 24400
    },
    {
      "epoch": 3.980504119422465,
      "eval_loss": 2.557262420654297,
      "eval_runtime": 134.9833,
      "eval_samples_per_second": 20.188,
      "eval_steps_per_second": 10.098,
      "num_input_tokens_seen": 50978512,
      "step": 24400
    },
    {
      "epoch": 3.9813198466432826,
      "grad_norm": 2.5,
      "learning_rate": 1.6524162751611304e-05,
      "loss": 2.895,
      "num_input_tokens_seen": 50990544,
      "step": 24405
    },
    {
      "epoch": 3.9821355738640998,
      "grad_norm": 7.25,
      "learning_rate": 1.6514927375248796e-05,
      "loss": 1.7926,
      "num_input_tokens_seen": 50999664,
      "step": 24410
    },
    {
      "epoch": 3.9829513010849174,
      "grad_norm": 9.0,
      "learning_rate": 1.6505693307391127e-05,
      "loss": 1.9168,
      "num_input_tokens_seen": 51009920,
      "step": 24415
    },
    {
      "epoch": 3.9837670283057345,
      "grad_norm": 8.1875,
      "learning_rate": 1.6496460549462288e-05,
      "loss": 1.8072,
      "num_input_tokens_seen": 51020544,
      "step": 24420
    },
    {
      "epoch": 3.984582755526552,
      "grad_norm": 15.6875,
      "learning_rate": 1.6487229102886097e-05,
      "loss": 1.8038,
      "num_input_tokens_seen": 51030672,
      "step": 24425
    },
    {
      "epoch": 3.985398482747369,
      "grad_norm": 4.28125,
      "learning_rate": 1.6477998969086155e-05,
      "loss": 2.4318,
      "num_input_tokens_seen": 51040544,
      "step": 24430
    },
    {
      "epoch": 3.9862142099681868,
      "grad_norm": 10.25,
      "learning_rate": 1.646877014948587e-05,
      "loss": 1.2334,
      "num_input_tokens_seen": 51050672,
      "step": 24435
    },
    {
      "epoch": 3.987029937189004,
      "grad_norm": 10.0625,
      "learning_rate": 1.6459542645508433e-05,
      "loss": 2.9823,
      "num_input_tokens_seen": 51061808,
      "step": 24440
    },
    {
      "epoch": 3.9878456644098215,
      "grad_norm": 4.09375,
      "learning_rate": 1.6450316458576852e-05,
      "loss": 2.9226,
      "num_input_tokens_seen": 51071872,
      "step": 24445
    },
    {
      "epoch": 3.9886613916306386,
      "grad_norm": 9.6875,
      "learning_rate": 1.6441091590113912e-05,
      "loss": 2.1689,
      "num_input_tokens_seen": 51081600,
      "step": 24450
    },
    {
      "epoch": 3.989477118851456,
      "grad_norm": 5.34375,
      "learning_rate": 1.6431868041542213e-05,
      "loss": 1.9945,
      "num_input_tokens_seen": 51091584,
      "step": 24455
    },
    {
      "epoch": 3.9902928460722733,
      "grad_norm": 6.375,
      "learning_rate": 1.6422645814284123e-05,
      "loss": 1.2654,
      "num_input_tokens_seen": 51101568,
      "step": 24460
    },
    {
      "epoch": 3.991108573293091,
      "grad_norm": 2.953125,
      "learning_rate": 1.6413424909761846e-05,
      "loss": 2.4927,
      "num_input_tokens_seen": 51112080,
      "step": 24465
    },
    {
      "epoch": 3.991924300513908,
      "grad_norm": 9.75,
      "learning_rate": 1.640420532939736e-05,
      "loss": 2.1908,
      "num_input_tokens_seen": 51122960,
      "step": 24470
    },
    {
      "epoch": 3.9927400277347256,
      "grad_norm": 14.5,
      "learning_rate": 1.639498707461242e-05,
      "loss": 2.2578,
      "num_input_tokens_seen": 51133232,
      "step": 24475
    },
    {
      "epoch": 3.9935557549555427,
      "grad_norm": 9.8125,
      "learning_rate": 1.6385770146828614e-05,
      "loss": 2.4918,
      "num_input_tokens_seen": 51143872,
      "step": 24480
    },
    {
      "epoch": 3.9943714821763603,
      "grad_norm": 0.2255859375,
      "learning_rate": 1.637655454746731e-05,
      "loss": 1.2688,
      "num_input_tokens_seen": 51153024,
      "step": 24485
    },
    {
      "epoch": 3.9951872093971774,
      "grad_norm": 9.0,
      "learning_rate": 1.6367340277949658e-05,
      "loss": 2.4167,
      "num_input_tokens_seen": 51163504,
      "step": 24490
    },
    {
      "epoch": 3.996002936617995,
      "grad_norm": 13.125,
      "learning_rate": 1.635812733969663e-05,
      "loss": 3.6797,
      "num_input_tokens_seen": 51174416,
      "step": 24495
    },
    {
      "epoch": 3.9968186638388126,
      "grad_norm": 6.78125,
      "learning_rate": 1.634891573412896e-05,
      "loss": 2.5743,
      "num_input_tokens_seen": 51185584,
      "step": 24500
    },
    {
      "epoch": 3.9976343910596297,
      "grad_norm": 2.859375,
      "learning_rate": 1.6339705462667196e-05,
      "loss": 2.6267,
      "num_input_tokens_seen": 51195856,
      "step": 24505
    },
    {
      "epoch": 3.998450118280447,
      "grad_norm": 1.609375,
      "learning_rate": 1.633049652673169e-05,
      "loss": 1.6336,
      "num_input_tokens_seen": 51205760,
      "step": 24510
    },
    {
      "epoch": 3.9992658455012644,
      "grad_norm": 9.875,
      "learning_rate": 1.632128892774256e-05,
      "loss": 2.6896,
      "num_input_tokens_seen": 51216512,
      "step": 24515
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.375,
      "learning_rate": 1.6312082667119737e-05,
      "loss": 1.7332,
      "num_input_tokens_seen": 51225984,
      "step": 24520
    },
    {
      "epoch": 4.000815727220817,
      "grad_norm": 10.0625,
      "learning_rate": 1.630287774628296e-05,
      "loss": 2.7793,
      "num_input_tokens_seen": 51237728,
      "step": 24525
    },
    {
      "epoch": 4.001631454441635,
      "grad_norm": 5.0,
      "learning_rate": 1.6293674166651718e-05,
      "loss": 2.4229,
      "num_input_tokens_seen": 51249056,
      "step": 24530
    },
    {
      "epoch": 4.002447181662452,
      "grad_norm": 3.09375,
      "learning_rate": 1.6284471929645338e-05,
      "loss": 2.0425,
      "num_input_tokens_seen": 51257376,
      "step": 24535
    },
    {
      "epoch": 4.003262908883269,
      "grad_norm": 5.0625,
      "learning_rate": 1.627527103668291e-05,
      "loss": 1.8085,
      "num_input_tokens_seen": 51265952,
      "step": 24540
    },
    {
      "epoch": 4.0040786361040865,
      "grad_norm": 10.6875,
      "learning_rate": 1.6266071489183327e-05,
      "loss": 3.2492,
      "num_input_tokens_seen": 51277120,
      "step": 24545
    },
    {
      "epoch": 4.004894363324905,
      "grad_norm": 11.25,
      "learning_rate": 1.6256873288565283e-05,
      "loss": 2.8756,
      "num_input_tokens_seen": 51287936,
      "step": 24550
    },
    {
      "epoch": 4.005710090545722,
      "grad_norm": 2.0,
      "learning_rate": 1.6247676436247245e-05,
      "loss": 1.9261,
      "num_input_tokens_seen": 51298240,
      "step": 24555
    },
    {
      "epoch": 4.006525817766539,
      "grad_norm": 11.1875,
      "learning_rate": 1.6238480933647486e-05,
      "loss": 3.6083,
      "num_input_tokens_seen": 51307616,
      "step": 24560
    },
    {
      "epoch": 4.007341544987356,
      "grad_norm": 5.25,
      "learning_rate": 1.6229286782184083e-05,
      "loss": 1.1911,
      "num_input_tokens_seen": 51318096,
      "step": 24565
    },
    {
      "epoch": 4.008157272208174,
      "grad_norm": 14.0625,
      "learning_rate": 1.622009398327487e-05,
      "loss": 2.1849,
      "num_input_tokens_seen": 51329808,
      "step": 24570
    },
    {
      "epoch": 4.008972999428991,
      "grad_norm": 6.8125,
      "learning_rate": 1.6210902538337502e-05,
      "loss": 2.5064,
      "num_input_tokens_seen": 51341056,
      "step": 24575
    },
    {
      "epoch": 4.009788726649808,
      "grad_norm": 5.53125,
      "learning_rate": 1.6201712448789413e-05,
      "loss": 2.088,
      "num_input_tokens_seen": 51351568,
      "step": 24580
    },
    {
      "epoch": 4.010604453870625,
      "grad_norm": 4.375,
      "learning_rate": 1.6192523716047827e-05,
      "loss": 2.0302,
      "num_input_tokens_seen": 51362640,
      "step": 24585
    },
    {
      "epoch": 4.011420181091443,
      "grad_norm": 6.40625,
      "learning_rate": 1.6183336341529776e-05,
      "loss": 1.9117,
      "num_input_tokens_seen": 51372368,
      "step": 24590
    },
    {
      "epoch": 4.0122359083122605,
      "grad_norm": 10.0625,
      "learning_rate": 1.6174150326652047e-05,
      "loss": 2.4326,
      "num_input_tokens_seen": 51382528,
      "step": 24595
    },
    {
      "epoch": 4.013051635533078,
      "grad_norm": 3.109375,
      "learning_rate": 1.6164965672831256e-05,
      "loss": 2.1603,
      "num_input_tokens_seen": 51394160,
      "step": 24600
    },
    {
      "epoch": 4.013051635533078,
      "eval_loss": 2.540343999862671,
      "eval_runtime": 135.0117,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 51394160,
      "step": 24600
    },
    {
      "epoch": 4.013867362753895,
      "grad_norm": 4.6875,
      "learning_rate": 1.6155782381483784e-05,
      "loss": 3.2689,
      "num_input_tokens_seen": 51405728,
      "step": 24605
    },
    {
      "epoch": 4.014683089974713,
      "grad_norm": 4.625,
      "learning_rate": 1.6146600454025813e-05,
      "loss": 2.0811,
      "num_input_tokens_seen": 51415616,
      "step": 24610
    },
    {
      "epoch": 4.01549881719553,
      "grad_norm": 8.9375,
      "learning_rate": 1.6137419891873317e-05,
      "loss": 3.4216,
      "num_input_tokens_seen": 51426240,
      "step": 24615
    },
    {
      "epoch": 4.016314544416347,
      "grad_norm": 10.5,
      "learning_rate": 1.6128240696442038e-05,
      "loss": 2.3852,
      "num_input_tokens_seen": 51436048,
      "step": 24620
    },
    {
      "epoch": 4.017130271637164,
      "grad_norm": 4.40625,
      "learning_rate": 1.611906286914753e-05,
      "loss": 1.864,
      "num_input_tokens_seen": 51446320,
      "step": 24625
    },
    {
      "epoch": 4.017945998857982,
      "grad_norm": 10.1875,
      "learning_rate": 1.6109886411405144e-05,
      "loss": 3.3988,
      "num_input_tokens_seen": 51457104,
      "step": 24630
    },
    {
      "epoch": 4.018761726078799,
      "grad_norm": 6.625,
      "learning_rate": 1.6100711324629985e-05,
      "loss": 3.9376,
      "num_input_tokens_seen": 51469088,
      "step": 24635
    },
    {
      "epoch": 4.0195774532996165,
      "grad_norm": 7.0,
      "learning_rate": 1.609153761023698e-05,
      "loss": 2.0672,
      "num_input_tokens_seen": 51480320,
      "step": 24640
    },
    {
      "epoch": 4.020393180520434,
      "grad_norm": 10.5,
      "learning_rate": 1.608236526964083e-05,
      "loss": 4.404,
      "num_input_tokens_seen": 51490976,
      "step": 24645
    },
    {
      "epoch": 4.021208907741252,
      "grad_norm": 8.4375,
      "learning_rate": 1.607319430425601e-05,
      "loss": 2.3763,
      "num_input_tokens_seen": 51500784,
      "step": 24650
    },
    {
      "epoch": 4.022024634962069,
      "grad_norm": 5.1875,
      "learning_rate": 1.606402471549682e-05,
      "loss": 1.6848,
      "num_input_tokens_seen": 51510432,
      "step": 24655
    },
    {
      "epoch": 4.022840362182886,
      "grad_norm": 12.4375,
      "learning_rate": 1.6054856504777312e-05,
      "loss": 2.2264,
      "num_input_tokens_seen": 51520896,
      "step": 24660
    },
    {
      "epoch": 4.023656089403703,
      "grad_norm": 4.90625,
      "learning_rate": 1.6045689673511334e-05,
      "loss": 1.9036,
      "num_input_tokens_seen": 51531872,
      "step": 24665
    },
    {
      "epoch": 4.024471816624521,
      "grad_norm": 10.375,
      "learning_rate": 1.6036524223112548e-05,
      "loss": 2.106,
      "num_input_tokens_seen": 51542816,
      "step": 24670
    },
    {
      "epoch": 4.025287543845338,
      "grad_norm": 17.0,
      "learning_rate": 1.602736015499436e-05,
      "loss": 2.0568,
      "num_input_tokens_seen": 51553568,
      "step": 24675
    },
    {
      "epoch": 4.026103271066155,
      "grad_norm": 6.6875,
      "learning_rate": 1.601819747057e-05,
      "loss": 1.9333,
      "num_input_tokens_seen": 51563920,
      "step": 24680
    },
    {
      "epoch": 4.026918998286972,
      "grad_norm": 4.28125,
      "learning_rate": 1.6009036171252465e-05,
      "loss": 1.7162,
      "num_input_tokens_seen": 51575584,
      "step": 24685
    },
    {
      "epoch": 4.02773472550779,
      "grad_norm": 5.15625,
      "learning_rate": 1.599987625845453e-05,
      "loss": 2.5217,
      "num_input_tokens_seen": 51587680,
      "step": 24690
    },
    {
      "epoch": 4.028550452728608,
      "grad_norm": 4.40625,
      "learning_rate": 1.599071773358879e-05,
      "loss": 2.7604,
      "num_input_tokens_seen": 51598160,
      "step": 24695
    },
    {
      "epoch": 4.029366179949425,
      "grad_norm": 6.15625,
      "learning_rate": 1.598156059806758e-05,
      "loss": 2.2855,
      "num_input_tokens_seen": 51609072,
      "step": 24700
    },
    {
      "epoch": 4.030181907170242,
      "grad_norm": 3.46875,
      "learning_rate": 1.5972404853303062e-05,
      "loss": 3.2393,
      "num_input_tokens_seen": 51619664,
      "step": 24705
    },
    {
      "epoch": 4.03099763439106,
      "grad_norm": 6.6875,
      "learning_rate": 1.5963250500707172e-05,
      "loss": 3.5823,
      "num_input_tokens_seen": 51630016,
      "step": 24710
    },
    {
      "epoch": 4.031813361611877,
      "grad_norm": 6.71875,
      "learning_rate": 1.5954097541691612e-05,
      "loss": 3.8073,
      "num_input_tokens_seen": 51639664,
      "step": 24715
    },
    {
      "epoch": 4.032629088832694,
      "grad_norm": 5.96875,
      "learning_rate": 1.5944945977667884e-05,
      "loss": 2.0225,
      "num_input_tokens_seen": 51650096,
      "step": 24720
    },
    {
      "epoch": 4.033444816053512,
      "grad_norm": 7.5,
      "learning_rate": 1.593579581004729e-05,
      "loss": 3.432,
      "num_input_tokens_seen": 51659952,
      "step": 24725
    },
    {
      "epoch": 4.034260543274329,
      "grad_norm": 13.5,
      "learning_rate": 1.592664704024088e-05,
      "loss": 3.7906,
      "num_input_tokens_seen": 51669488,
      "step": 24730
    },
    {
      "epoch": 4.035076270495146,
      "grad_norm": 5.4375,
      "learning_rate": 1.591749966965953e-05,
      "loss": 2.3926,
      "num_input_tokens_seen": 51681232,
      "step": 24735
    },
    {
      "epoch": 4.0358919977159635,
      "grad_norm": 7.8125,
      "learning_rate": 1.5908353699713856e-05,
      "loss": 1.8183,
      "num_input_tokens_seen": 51692112,
      "step": 24740
    },
    {
      "epoch": 4.0367077249367815,
      "grad_norm": 12.1875,
      "learning_rate": 1.5899209131814298e-05,
      "loss": 2.2185,
      "num_input_tokens_seen": 51702096,
      "step": 24745
    },
    {
      "epoch": 4.037523452157599,
      "grad_norm": 10.9375,
      "learning_rate": 1.5890065967371067e-05,
      "loss": 3.4912,
      "num_input_tokens_seen": 51713936,
      "step": 24750
    },
    {
      "epoch": 4.038339179378416,
      "grad_norm": 10.0,
      "learning_rate": 1.5880924207794144e-05,
      "loss": 3.7162,
      "num_input_tokens_seen": 51722512,
      "step": 24755
    },
    {
      "epoch": 4.039154906599233,
      "grad_norm": 4.53125,
      "learning_rate": 1.5871783854493298e-05,
      "loss": 1.8996,
      "num_input_tokens_seen": 51734720,
      "step": 24760
    },
    {
      "epoch": 4.039970633820051,
      "grad_norm": 9.4375,
      "learning_rate": 1.5862644908878106e-05,
      "loss": 2.7278,
      "num_input_tokens_seen": 51745600,
      "step": 24765
    },
    {
      "epoch": 4.040786361040868,
      "grad_norm": 9.75,
      "learning_rate": 1.5853507372357885e-05,
      "loss": 1.9228,
      "num_input_tokens_seen": 51757568,
      "step": 24770
    },
    {
      "epoch": 4.041602088261685,
      "grad_norm": 7.09375,
      "learning_rate": 1.5844371246341776e-05,
      "loss": 2.1256,
      "num_input_tokens_seen": 51767328,
      "step": 24775
    },
    {
      "epoch": 4.042417815482502,
      "grad_norm": 6.78125,
      "learning_rate": 1.5835236532238674e-05,
      "loss": 2.9969,
      "num_input_tokens_seen": 51777536,
      "step": 24780
    },
    {
      "epoch": 4.04323354270332,
      "grad_norm": 5.03125,
      "learning_rate": 1.582610323145727e-05,
      "loss": 3.3975,
      "num_input_tokens_seen": 51788096,
      "step": 24785
    },
    {
      "epoch": 4.0440492699241375,
      "grad_norm": 8.8125,
      "learning_rate": 1.5816971345406035e-05,
      "loss": 2.3444,
      "num_input_tokens_seen": 51799056,
      "step": 24790
    },
    {
      "epoch": 4.044864997144955,
      "grad_norm": 7.03125,
      "learning_rate": 1.5807840875493225e-05,
      "loss": 2.5925,
      "num_input_tokens_seen": 51810320,
      "step": 24795
    },
    {
      "epoch": 4.045680724365772,
      "grad_norm": 6.4375,
      "learning_rate": 1.5798711823126854e-05,
      "loss": 2.1706,
      "num_input_tokens_seen": 51821712,
      "step": 24800
    },
    {
      "epoch": 4.045680724365772,
      "eval_loss": 2.545102119445801,
      "eval_runtime": 134.848,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 51821712,
      "step": 24800
    },
    {
      "epoch": 4.04649645158659,
      "grad_norm": 4.75,
      "learning_rate": 1.578958418971477e-05,
      "loss": 2.2454,
      "num_input_tokens_seen": 51831136,
      "step": 24805
    },
    {
      "epoch": 4.047312178807407,
      "grad_norm": 7.46875,
      "learning_rate": 1.578045797666453e-05,
      "loss": 1.7889,
      "num_input_tokens_seen": 51841488,
      "step": 24810
    },
    {
      "epoch": 4.048127906028224,
      "grad_norm": 10.0,
      "learning_rate": 1.5771333185383548e-05,
      "loss": 2.5769,
      "num_input_tokens_seen": 51852528,
      "step": 24815
    },
    {
      "epoch": 4.048943633249041,
      "grad_norm": 1.9375,
      "learning_rate": 1.576220981727895e-05,
      "loss": 1.8891,
      "num_input_tokens_seen": 51864240,
      "step": 24820
    },
    {
      "epoch": 4.049759360469859,
      "grad_norm": 15.625,
      "learning_rate": 1.575308787375769e-05,
      "loss": 2.218,
      "num_input_tokens_seen": 51873168,
      "step": 24825
    },
    {
      "epoch": 4.050575087690676,
      "grad_norm": 8.9375,
      "learning_rate": 1.5743967356226492e-05,
      "loss": 2.3698,
      "num_input_tokens_seen": 51884496,
      "step": 24830
    },
    {
      "epoch": 4.051390814911493,
      "grad_norm": 0.177734375,
      "learning_rate": 1.5734848266091835e-05,
      "loss": 2.6955,
      "num_input_tokens_seen": 51896304,
      "step": 24835
    },
    {
      "epoch": 4.052206542132311,
      "grad_norm": 12.0,
      "learning_rate": 1.572573060476001e-05,
      "loss": 1.8319,
      "num_input_tokens_seen": 51906608,
      "step": 24840
    },
    {
      "epoch": 4.053022269353129,
      "grad_norm": 5.46875,
      "learning_rate": 1.5716614373637085e-05,
      "loss": 3.6039,
      "num_input_tokens_seen": 51916880,
      "step": 24845
    },
    {
      "epoch": 4.053837996573946,
      "grad_norm": 7.9375,
      "learning_rate": 1.570749957412887e-05,
      "loss": 2.1091,
      "num_input_tokens_seen": 51926672,
      "step": 24850
    },
    {
      "epoch": 4.054653723794763,
      "grad_norm": 1.1953125,
      "learning_rate": 1.5698386207641013e-05,
      "loss": 2.4642,
      "num_input_tokens_seen": 51937280,
      "step": 24855
    },
    {
      "epoch": 4.05546945101558,
      "grad_norm": 8.5625,
      "learning_rate": 1.5689274275578884e-05,
      "loss": 3.1211,
      "num_input_tokens_seen": 51948752,
      "step": 24860
    },
    {
      "epoch": 4.056285178236398,
      "grad_norm": 10.625,
      "learning_rate": 1.5680163779347667e-05,
      "loss": 3.1941,
      "num_input_tokens_seen": 51959760,
      "step": 24865
    },
    {
      "epoch": 4.057100905457215,
      "grad_norm": 5.25,
      "learning_rate": 1.5671054720352327e-05,
      "loss": 1.3567,
      "num_input_tokens_seen": 51970000,
      "step": 24870
    },
    {
      "epoch": 4.057916632678032,
      "grad_norm": 5.59375,
      "learning_rate": 1.566194709999757e-05,
      "loss": 3.4178,
      "num_input_tokens_seen": 51981456,
      "step": 24875
    },
    {
      "epoch": 4.058732359898849,
      "grad_norm": 9.125,
      "learning_rate": 1.5652840919687933e-05,
      "loss": 3.3472,
      "num_input_tokens_seen": 51991584,
      "step": 24880
    },
    {
      "epoch": 4.059548087119667,
      "grad_norm": 4.03125,
      "learning_rate": 1.5643736180827676e-05,
      "loss": 1.9951,
      "num_input_tokens_seen": 52001152,
      "step": 24885
    },
    {
      "epoch": 4.0603638143404845,
      "grad_norm": 6.34375,
      "learning_rate": 1.5634632884820878e-05,
      "loss": 1.7925,
      "num_input_tokens_seen": 52011600,
      "step": 24890
    },
    {
      "epoch": 4.061179541561302,
      "grad_norm": 2.296875,
      "learning_rate": 1.5625531033071395e-05,
      "loss": 1.8111,
      "num_input_tokens_seen": 52022496,
      "step": 24895
    },
    {
      "epoch": 4.06199526878212,
      "grad_norm": 4.375,
      "learning_rate": 1.5616430626982828e-05,
      "loss": 2.8478,
      "num_input_tokens_seen": 52033472,
      "step": 24900
    },
    {
      "epoch": 4.062810996002937,
      "grad_norm": 6.53125,
      "learning_rate": 1.5607331667958575e-05,
      "loss": 2.8529,
      "num_input_tokens_seen": 52043280,
      "step": 24905
    },
    {
      "epoch": 4.063626723223754,
      "grad_norm": 2.8125,
      "learning_rate": 1.5598234157401824e-05,
      "loss": 1.6373,
      "num_input_tokens_seen": 52053680,
      "step": 24910
    },
    {
      "epoch": 4.064442450444571,
      "grad_norm": 10.0,
      "learning_rate": 1.5589138096715503e-05,
      "loss": 3.0509,
      "num_input_tokens_seen": 52063584,
      "step": 24915
    },
    {
      "epoch": 4.065258177665389,
      "grad_norm": 6.84375,
      "learning_rate": 1.5580043487302365e-05,
      "loss": 1.8038,
      "num_input_tokens_seen": 52073488,
      "step": 24920
    },
    {
      "epoch": 4.066073904886206,
      "grad_norm": 11.25,
      "learning_rate": 1.5570950330564888e-05,
      "loss": 1.5999,
      "num_input_tokens_seen": 52082688,
      "step": 24925
    },
    {
      "epoch": 4.066889632107023,
      "grad_norm": 4.71875,
      "learning_rate": 1.5561858627905367e-05,
      "loss": 2.4747,
      "num_input_tokens_seen": 52094176,
      "step": 24930
    },
    {
      "epoch": 4.0677053593278405,
      "grad_norm": 8.8125,
      "learning_rate": 1.5552768380725857e-05,
      "loss": 2.0008,
      "num_input_tokens_seen": 52105200,
      "step": 24935
    },
    {
      "epoch": 4.0685210865486585,
      "grad_norm": 4.46875,
      "learning_rate": 1.5543679590428183e-05,
      "loss": 2.2917,
      "num_input_tokens_seen": 52115024,
      "step": 24940
    },
    {
      "epoch": 4.069336813769476,
      "grad_norm": 7.25,
      "learning_rate": 1.5534592258413943e-05,
      "loss": 2.7056,
      "num_input_tokens_seen": 52125360,
      "step": 24945
    },
    {
      "epoch": 4.070152540990293,
      "grad_norm": 10.5,
      "learning_rate": 1.5525506386084538e-05,
      "loss": 3.5798,
      "num_input_tokens_seen": 52135376,
      "step": 24950
    },
    {
      "epoch": 4.07096826821111,
      "grad_norm": 11.0,
      "learning_rate": 1.55164219748411e-05,
      "loss": 1.877,
      "num_input_tokens_seen": 52147040,
      "step": 24955
    },
    {
      "epoch": 4.071783995431928,
      "grad_norm": 2.390625,
      "learning_rate": 1.550733902608459e-05,
      "loss": 1.7353,
      "num_input_tokens_seen": 52157680,
      "step": 24960
    },
    {
      "epoch": 4.072599722652745,
      "grad_norm": 9.0625,
      "learning_rate": 1.549825754121568e-05,
      "loss": 2.5551,
      "num_input_tokens_seen": 52167888,
      "step": 24965
    },
    {
      "epoch": 4.073415449873562,
      "grad_norm": 7.90625,
      "learning_rate": 1.5489177521634864e-05,
      "loss": 2.9202,
      "num_input_tokens_seen": 52178752,
      "step": 24970
    },
    {
      "epoch": 4.074231177094379,
      "grad_norm": 6.0625,
      "learning_rate": 1.5480098968742402e-05,
      "loss": 1.5285,
      "num_input_tokens_seen": 52189200,
      "step": 24975
    },
    {
      "epoch": 4.075046904315197,
      "grad_norm": 9.1875,
      "learning_rate": 1.5471021883938304e-05,
      "loss": 2.4032,
      "num_input_tokens_seen": 52199984,
      "step": 24980
    },
    {
      "epoch": 4.0758626315360145,
      "grad_norm": 11.375,
      "learning_rate": 1.546194626862238e-05,
      "loss": 1.8813,
      "num_input_tokens_seen": 52211584,
      "step": 24985
    },
    {
      "epoch": 4.076678358756832,
      "grad_norm": 2.203125,
      "learning_rate": 1.5452872124194216e-05,
      "loss": 1.5998,
      "num_input_tokens_seen": 52222816,
      "step": 24990
    },
    {
      "epoch": 4.077494085977649,
      "grad_norm": 5.5625,
      "learning_rate": 1.5443799452053136e-05,
      "loss": 2.8754,
      "num_input_tokens_seen": 52234016,
      "step": 24995
    },
    {
      "epoch": 4.078309813198467,
      "grad_norm": 8.375,
      "learning_rate": 1.543472825359828e-05,
      "loss": 1.6054,
      "num_input_tokens_seen": 52244608,
      "step": 25000
    },
    {
      "epoch": 4.078309813198467,
      "eval_loss": 2.5288426876068115,
      "eval_runtime": 134.8652,
      "eval_samples_per_second": 20.205,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 52244608,
      "step": 25000
    },
    {
      "epoch": 4.079125540419284,
      "grad_norm": 7.09375,
      "learning_rate": 1.5425658530228522e-05,
      "loss": 3.1642,
      "num_input_tokens_seen": 52254960,
      "step": 25005
    },
    {
      "epoch": 4.079941267640101,
      "grad_norm": 5.34375,
      "learning_rate": 1.5416590283342546e-05,
      "loss": 2.4792,
      "num_input_tokens_seen": 52265136,
      "step": 25010
    },
    {
      "epoch": 4.080756994860918,
      "grad_norm": 5.75,
      "learning_rate": 1.5407523514338783e-05,
      "loss": 0.9519,
      "num_input_tokens_seen": 52276048,
      "step": 25015
    },
    {
      "epoch": 4.081572722081736,
      "grad_norm": 9.5625,
      "learning_rate": 1.539845822461543e-05,
      "loss": 2.0118,
      "num_input_tokens_seen": 52286256,
      "step": 25020
    },
    {
      "epoch": 4.082388449302553,
      "grad_norm": 5.375,
      "learning_rate": 1.538939441557048e-05,
      "loss": 2.3859,
      "num_input_tokens_seen": 52295872,
      "step": 25025
    },
    {
      "epoch": 4.08320417652337,
      "grad_norm": 6.4375,
      "learning_rate": 1.5380332088601696e-05,
      "loss": 4.7387,
      "num_input_tokens_seen": 52305616,
      "step": 25030
    },
    {
      "epoch": 4.0840199037441876,
      "grad_norm": 7.1875,
      "learning_rate": 1.537127124510658e-05,
      "loss": 2.3978,
      "num_input_tokens_seen": 52317376,
      "step": 25035
    },
    {
      "epoch": 4.084835630965006,
      "grad_norm": 6.71875,
      "learning_rate": 1.5362211886482457e-05,
      "loss": 3.6264,
      "num_input_tokens_seen": 52327792,
      "step": 25040
    },
    {
      "epoch": 4.085651358185823,
      "grad_norm": 3.5,
      "learning_rate": 1.5353154014126363e-05,
      "loss": 2.7063,
      "num_input_tokens_seen": 52338112,
      "step": 25045
    },
    {
      "epoch": 4.08646708540664,
      "grad_norm": 6.59375,
      "learning_rate": 1.534409762943515e-05,
      "loss": 3.0695,
      "num_input_tokens_seen": 52347840,
      "step": 25050
    },
    {
      "epoch": 4.087282812627457,
      "grad_norm": 5.90625,
      "learning_rate": 1.5335042733805438e-05,
      "loss": 1.9829,
      "num_input_tokens_seen": 52358000,
      "step": 25055
    },
    {
      "epoch": 4.088098539848275,
      "grad_norm": 3.890625,
      "learning_rate": 1.532598932863358e-05,
      "loss": 2.0714,
      "num_input_tokens_seen": 52368512,
      "step": 25060
    },
    {
      "epoch": 4.088914267069092,
      "grad_norm": 4.6875,
      "learning_rate": 1.531693741531574e-05,
      "loss": 2.4597,
      "num_input_tokens_seen": 52379568,
      "step": 25065
    },
    {
      "epoch": 4.089729994289909,
      "grad_norm": 17.0,
      "learning_rate": 1.5307886995247844e-05,
      "loss": 3.4794,
      "num_input_tokens_seen": 52388832,
      "step": 25070
    },
    {
      "epoch": 4.090545721510727,
      "grad_norm": 9.125,
      "learning_rate": 1.529883806982557e-05,
      "loss": 3.1409,
      "num_input_tokens_seen": 52399056,
      "step": 25075
    },
    {
      "epoch": 4.091361448731544,
      "grad_norm": 2.8125,
      "learning_rate": 1.5289790640444376e-05,
      "loss": 2.4856,
      "num_input_tokens_seen": 52411456,
      "step": 25080
    },
    {
      "epoch": 4.0921771759523615,
      "grad_norm": 12.625,
      "learning_rate": 1.5280744708499494e-05,
      "loss": 2.5573,
      "num_input_tokens_seen": 52422176,
      "step": 25085
    },
    {
      "epoch": 4.092992903173179,
      "grad_norm": 2.46875,
      "learning_rate": 1.527170027538591e-05,
      "loss": 1.6164,
      "num_input_tokens_seen": 52432880,
      "step": 25090
    },
    {
      "epoch": 4.093808630393997,
      "grad_norm": 6.96875,
      "learning_rate": 1.5262657342498407e-05,
      "loss": 1.7433,
      "num_input_tokens_seen": 52444256,
      "step": 25095
    },
    {
      "epoch": 4.094624357614814,
      "grad_norm": 6.09375,
      "learning_rate": 1.52536159112315e-05,
      "loss": 1.8,
      "num_input_tokens_seen": 52453856,
      "step": 25100
    },
    {
      "epoch": 4.095440084835631,
      "grad_norm": 11.6875,
      "learning_rate": 1.5244575982979497e-05,
      "loss": 1.95,
      "num_input_tokens_seen": 52463568,
      "step": 25105
    },
    {
      "epoch": 4.096255812056448,
      "grad_norm": 9.6875,
      "learning_rate": 1.5235537559136487e-05,
      "loss": 2.9884,
      "num_input_tokens_seen": 52473232,
      "step": 25110
    },
    {
      "epoch": 4.097071539277266,
      "grad_norm": 8.3125,
      "learning_rate": 1.5226500641096286e-05,
      "loss": 1.6695,
      "num_input_tokens_seen": 52482672,
      "step": 25115
    },
    {
      "epoch": 4.097887266498083,
      "grad_norm": 0.177734375,
      "learning_rate": 1.5217465230252509e-05,
      "loss": 1.0523,
      "num_input_tokens_seen": 52494656,
      "step": 25120
    },
    {
      "epoch": 4.0987029937189,
      "grad_norm": 6.9375,
      "learning_rate": 1.5208431327998523e-05,
      "loss": 1.3795,
      "num_input_tokens_seen": 52504064,
      "step": 25125
    },
    {
      "epoch": 4.0995187209397175,
      "grad_norm": 10.5,
      "learning_rate": 1.5199398935727477e-05,
      "loss": 2.6356,
      "num_input_tokens_seen": 52515760,
      "step": 25130
    },
    {
      "epoch": 4.1003344481605355,
      "grad_norm": 5.75,
      "learning_rate": 1.5190368054832282e-05,
      "loss": 2.2369,
      "num_input_tokens_seen": 52526400,
      "step": 25135
    },
    {
      "epoch": 4.101150175381353,
      "grad_norm": 3.5625,
      "learning_rate": 1.5181338686705601e-05,
      "loss": 1.7902,
      "num_input_tokens_seen": 52537456,
      "step": 25140
    },
    {
      "epoch": 4.10196590260217,
      "grad_norm": 5.5,
      "learning_rate": 1.5172310832739889e-05,
      "loss": 1.669,
      "num_input_tokens_seen": 52547792,
      "step": 25145
    },
    {
      "epoch": 4.102781629822987,
      "grad_norm": 5.3125,
      "learning_rate": 1.5163284494327346e-05,
      "loss": 2.0257,
      "num_input_tokens_seen": 52558288,
      "step": 25150
    },
    {
      "epoch": 4.103597357043805,
      "grad_norm": 13.6875,
      "learning_rate": 1.5154259672859952e-05,
      "loss": 3.0248,
      "num_input_tokens_seen": 52568000,
      "step": 25155
    },
    {
      "epoch": 4.104413084264622,
      "grad_norm": 8.8125,
      "learning_rate": 1.5145236369729452e-05,
      "loss": 2.9275,
      "num_input_tokens_seen": 52577920,
      "step": 25160
    },
    {
      "epoch": 4.105228811485439,
      "grad_norm": 9.0625,
      "learning_rate": 1.5136214586327335e-05,
      "loss": 1.569,
      "num_input_tokens_seen": 52587536,
      "step": 25165
    },
    {
      "epoch": 4.106044538706256,
      "grad_norm": 7.65625,
      "learning_rate": 1.5127194324044885e-05,
      "loss": 3.0632,
      "num_input_tokens_seen": 52597008,
      "step": 25170
    },
    {
      "epoch": 4.106860265927074,
      "grad_norm": 9.375,
      "learning_rate": 1.5118175584273148e-05,
      "loss": 2.833,
      "num_input_tokens_seen": 52606624,
      "step": 25175
    },
    {
      "epoch": 4.1076759931478914,
      "grad_norm": 7.625,
      "learning_rate": 1.5109158368402909e-05,
      "loss": 2.0158,
      "num_input_tokens_seen": 52617376,
      "step": 25180
    },
    {
      "epoch": 4.108491720368709,
      "grad_norm": 12.1875,
      "learning_rate": 1.5100142677824753e-05,
      "loss": 2.7997,
      "num_input_tokens_seen": 52628400,
      "step": 25185
    },
    {
      "epoch": 4.109307447589526,
      "grad_norm": 5.5,
      "learning_rate": 1.509112851392901e-05,
      "loss": 2.3182,
      "num_input_tokens_seen": 52638192,
      "step": 25190
    },
    {
      "epoch": 4.110123174810344,
      "grad_norm": 8.5,
      "learning_rate": 1.5082115878105763e-05,
      "loss": 4.5375,
      "num_input_tokens_seen": 52649008,
      "step": 25195
    },
    {
      "epoch": 4.110938902031161,
      "grad_norm": 6.34375,
      "learning_rate": 1.5073104771744892e-05,
      "loss": 2.0526,
      "num_input_tokens_seen": 52659888,
      "step": 25200
    },
    {
      "epoch": 4.110938902031161,
      "eval_loss": 2.551703691482544,
      "eval_runtime": 134.8052,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 52659888,
      "step": 25200
    },
    {
      "epoch": 4.111754629251978,
      "grad_norm": 1.4375,
      "learning_rate": 1.5064095196236006e-05,
      "loss": 2.6566,
      "num_input_tokens_seen": 52668528,
      "step": 25205
    },
    {
      "epoch": 4.112570356472795,
      "grad_norm": 5.03125,
      "learning_rate": 1.50550871529685e-05,
      "loss": 2.7985,
      "num_input_tokens_seen": 52679824,
      "step": 25210
    },
    {
      "epoch": 4.113386083693613,
      "grad_norm": 7.15625,
      "learning_rate": 1.5046080643331546e-05,
      "loss": 1.5852,
      "num_input_tokens_seen": 52688496,
      "step": 25215
    },
    {
      "epoch": 4.11420181091443,
      "grad_norm": 10.9375,
      "learning_rate": 1.5037075668714028e-05,
      "loss": 2.7034,
      "num_input_tokens_seen": 52700544,
      "step": 25220
    },
    {
      "epoch": 4.115017538135247,
      "grad_norm": 0.1689453125,
      "learning_rate": 1.5028072230504656e-05,
      "loss": 1.9227,
      "num_input_tokens_seen": 52710304,
      "step": 25225
    },
    {
      "epoch": 4.1158332653560645,
      "grad_norm": 8.375,
      "learning_rate": 1.5019070330091861e-05,
      "loss": 2.2664,
      "num_input_tokens_seen": 52720528,
      "step": 25230
    },
    {
      "epoch": 4.1166489925768825,
      "grad_norm": 8.5625,
      "learning_rate": 1.5010069968863843e-05,
      "loss": 1.9659,
      "num_input_tokens_seen": 52731280,
      "step": 25235
    },
    {
      "epoch": 4.1174647197977,
      "grad_norm": 6.21875,
      "learning_rate": 1.5001071148208584e-05,
      "loss": 2.0341,
      "num_input_tokens_seen": 52742448,
      "step": 25240
    },
    {
      "epoch": 4.118280447018517,
      "grad_norm": 9.5,
      "learning_rate": 1.49920738695138e-05,
      "loss": 2.2142,
      "num_input_tokens_seen": 52753328,
      "step": 25245
    },
    {
      "epoch": 4.119096174239334,
      "grad_norm": 6.6875,
      "learning_rate": 1.4983078134166995e-05,
      "loss": 2.4535,
      "num_input_tokens_seen": 52764416,
      "step": 25250
    },
    {
      "epoch": 4.119911901460152,
      "grad_norm": 3.828125,
      "learning_rate": 1.4974083943555428e-05,
      "loss": 2.6979,
      "num_input_tokens_seen": 52775616,
      "step": 25255
    },
    {
      "epoch": 4.120727628680969,
      "grad_norm": 5.125,
      "learning_rate": 1.496509129906611e-05,
      "loss": 3.6935,
      "num_input_tokens_seen": 52787056,
      "step": 25260
    },
    {
      "epoch": 4.121543355901786,
      "grad_norm": 11.0,
      "learning_rate": 1.4956100202085809e-05,
      "loss": 2.3477,
      "num_input_tokens_seen": 52796016,
      "step": 25265
    },
    {
      "epoch": 4.122359083122603,
      "grad_norm": 6.75,
      "learning_rate": 1.4947110654001093e-05,
      "loss": 3.1076,
      "num_input_tokens_seen": 52807248,
      "step": 25270
    },
    {
      "epoch": 4.123174810343421,
      "grad_norm": 8.5625,
      "learning_rate": 1.4938122656198234e-05,
      "loss": 2.2347,
      "num_input_tokens_seen": 52816848,
      "step": 25275
    },
    {
      "epoch": 4.1239905375642385,
      "grad_norm": 4.34375,
      "learning_rate": 1.4929136210063316e-05,
      "loss": 2.4091,
      "num_input_tokens_seen": 52826816,
      "step": 25280
    },
    {
      "epoch": 4.124806264785056,
      "grad_norm": 4.3125,
      "learning_rate": 1.4920151316982146e-05,
      "loss": 1.2216,
      "num_input_tokens_seen": 52836528,
      "step": 25285
    },
    {
      "epoch": 4.125621992005874,
      "grad_norm": 5.40625,
      "learning_rate": 1.4911167978340312e-05,
      "loss": 2.4897,
      "num_input_tokens_seen": 52846576,
      "step": 25290
    },
    {
      "epoch": 4.126437719226691,
      "grad_norm": 9.375,
      "learning_rate": 1.4902186195523166e-05,
      "loss": 2.5142,
      "num_input_tokens_seen": 52857712,
      "step": 25295
    },
    {
      "epoch": 4.127253446447508,
      "grad_norm": 8.625,
      "learning_rate": 1.4893205969915805e-05,
      "loss": 1.9978,
      "num_input_tokens_seen": 52868576,
      "step": 25300
    },
    {
      "epoch": 4.128069173668325,
      "grad_norm": 8.25,
      "learning_rate": 1.4884227302903086e-05,
      "loss": 2.9423,
      "num_input_tokens_seen": 52879552,
      "step": 25305
    },
    {
      "epoch": 4.128884900889143,
      "grad_norm": 5.5625,
      "learning_rate": 1.4875250195869653e-05,
      "loss": 2.0911,
      "num_input_tokens_seen": 52888144,
      "step": 25310
    },
    {
      "epoch": 4.12970062810996,
      "grad_norm": 6.8125,
      "learning_rate": 1.4866274650199862e-05,
      "loss": 2.4658,
      "num_input_tokens_seen": 52898208,
      "step": 25315
    },
    {
      "epoch": 4.130516355330777,
      "grad_norm": 12.4375,
      "learning_rate": 1.485730066727788e-05,
      "loss": 2.7093,
      "num_input_tokens_seen": 52909136,
      "step": 25320
    },
    {
      "epoch": 4.1313320825515945,
      "grad_norm": 9.25,
      "learning_rate": 1.4848328248487586e-05,
      "loss": 2.6395,
      "num_input_tokens_seen": 52918720,
      "step": 25325
    },
    {
      "epoch": 4.1321478097724125,
      "grad_norm": 4.09375,
      "learning_rate": 1.4839357395212656e-05,
      "loss": 2.8061,
      "num_input_tokens_seen": 52928464,
      "step": 25330
    },
    {
      "epoch": 4.13296353699323,
      "grad_norm": 9.1875,
      "learning_rate": 1.4830388108836502e-05,
      "loss": 2.9722,
      "num_input_tokens_seen": 52939008,
      "step": 25335
    },
    {
      "epoch": 4.133779264214047,
      "grad_norm": 8.9375,
      "learning_rate": 1.4821420390742299e-05,
      "loss": 2.0187,
      "num_input_tokens_seen": 52949664,
      "step": 25340
    },
    {
      "epoch": 4.134594991434864,
      "grad_norm": 15.125,
      "learning_rate": 1.4812454242312979e-05,
      "loss": 3.8748,
      "num_input_tokens_seen": 52959760,
      "step": 25345
    },
    {
      "epoch": 4.135410718655682,
      "grad_norm": 4.9375,
      "learning_rate": 1.4803489664931253e-05,
      "loss": 1.4902,
      "num_input_tokens_seen": 52969664,
      "step": 25350
    },
    {
      "epoch": 4.136226445876499,
      "grad_norm": 5.5625,
      "learning_rate": 1.4794526659979544e-05,
      "loss": 2.2104,
      "num_input_tokens_seen": 52979200,
      "step": 25355
    },
    {
      "epoch": 4.137042173097316,
      "grad_norm": 2.140625,
      "learning_rate": 1.4785565228840086e-05,
      "loss": 1.1141,
      "num_input_tokens_seen": 52988976,
      "step": 25360
    },
    {
      "epoch": 4.137857900318133,
      "grad_norm": 6.78125,
      "learning_rate": 1.4776605372894819e-05,
      "loss": 2.4724,
      "num_input_tokens_seen": 52999936,
      "step": 25365
    },
    {
      "epoch": 4.138673627538951,
      "grad_norm": 9.9375,
      "learning_rate": 1.4767647093525488e-05,
      "loss": 1.0567,
      "num_input_tokens_seen": 53010512,
      "step": 25370
    },
    {
      "epoch": 4.139489354759768,
      "grad_norm": 9.0,
      "learning_rate": 1.4758690392113566e-05,
      "loss": 1.6404,
      "num_input_tokens_seen": 53020880,
      "step": 25375
    },
    {
      "epoch": 4.1403050819805856,
      "grad_norm": 8.9375,
      "learning_rate": 1.4749735270040276e-05,
      "loss": 3.3843,
      "num_input_tokens_seen": 53031280,
      "step": 25380
    },
    {
      "epoch": 4.141120809201403,
      "grad_norm": 2.203125,
      "learning_rate": 1.4740781728686623e-05,
      "loss": 2.0223,
      "num_input_tokens_seen": 53042176,
      "step": 25385
    },
    {
      "epoch": 4.141936536422221,
      "grad_norm": 10.0,
      "learning_rate": 1.4731829769433358e-05,
      "loss": 1.7687,
      "num_input_tokens_seen": 53051424,
      "step": 25390
    },
    {
      "epoch": 4.142752263643038,
      "grad_norm": 5.5625,
      "learning_rate": 1.4722879393660976e-05,
      "loss": 1.5099,
      "num_input_tokens_seen": 53062576,
      "step": 25395
    },
    {
      "epoch": 4.143567990863855,
      "grad_norm": 6.625,
      "learning_rate": 1.4713930602749748e-05,
      "loss": 1.658,
      "num_input_tokens_seen": 53073648,
      "step": 25400
    },
    {
      "epoch": 4.143567990863855,
      "eval_loss": 2.54581880569458,
      "eval_runtime": 135.0407,
      "eval_samples_per_second": 20.179,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 53073648,
      "step": 25400
    },
    {
      "epoch": 4.144383718084672,
      "grad_norm": 11.625,
      "learning_rate": 1.470498339807968e-05,
      "loss": 3.5884,
      "num_input_tokens_seen": 53083344,
      "step": 25405
    },
    {
      "epoch": 4.14519944530549,
      "grad_norm": 4.78125,
      "learning_rate": 1.4696037781030542e-05,
      "loss": 2.7814,
      "num_input_tokens_seen": 53093280,
      "step": 25410
    },
    {
      "epoch": 4.146015172526307,
      "grad_norm": 3.453125,
      "learning_rate": 1.4687093752981876e-05,
      "loss": 1.6041,
      "num_input_tokens_seen": 53103952,
      "step": 25415
    },
    {
      "epoch": 4.146830899747124,
      "grad_norm": 3.703125,
      "learning_rate": 1.4678151315312943e-05,
      "loss": 2.2638,
      "num_input_tokens_seen": 53114256,
      "step": 25420
    },
    {
      "epoch": 4.1476466269679415,
      "grad_norm": 5.21875,
      "learning_rate": 1.4669210469402789e-05,
      "loss": 2.6153,
      "num_input_tokens_seen": 53125136,
      "step": 25425
    },
    {
      "epoch": 4.1484623541887595,
      "grad_norm": 7.5625,
      "learning_rate": 1.4660271216630218e-05,
      "loss": 3.0091,
      "num_input_tokens_seen": 53136720,
      "step": 25430
    },
    {
      "epoch": 4.149278081409577,
      "grad_norm": 7.6875,
      "learning_rate": 1.4651333558373748e-05,
      "loss": 1.6617,
      "num_input_tokens_seen": 53147456,
      "step": 25435
    },
    {
      "epoch": 4.150093808630394,
      "grad_norm": 7.1875,
      "learning_rate": 1.4642397496011707e-05,
      "loss": 1.922,
      "num_input_tokens_seen": 53156704,
      "step": 25440
    },
    {
      "epoch": 4.150909535851211,
      "grad_norm": 7.96875,
      "learning_rate": 1.4633463030922129e-05,
      "loss": 2.7945,
      "num_input_tokens_seen": 53166528,
      "step": 25445
    },
    {
      "epoch": 4.151725263072029,
      "grad_norm": 5.75,
      "learning_rate": 1.462453016448282e-05,
      "loss": 1.8874,
      "num_input_tokens_seen": 53176992,
      "step": 25450
    },
    {
      "epoch": 4.152540990292846,
      "grad_norm": 0.10791015625,
      "learning_rate": 1.4615598898071354e-05,
      "loss": 1.6731,
      "num_input_tokens_seen": 53188688,
      "step": 25455
    },
    {
      "epoch": 4.153356717513663,
      "grad_norm": 7.59375,
      "learning_rate": 1.4606669233065026e-05,
      "loss": 3.3342,
      "num_input_tokens_seen": 53198672,
      "step": 25460
    },
    {
      "epoch": 4.154172444734481,
      "grad_norm": 5.1875,
      "learning_rate": 1.4597741170840914e-05,
      "loss": 1.5395,
      "num_input_tokens_seen": 53208640,
      "step": 25465
    },
    {
      "epoch": 4.154988171955298,
      "grad_norm": 3.328125,
      "learning_rate": 1.4588814712775853e-05,
      "loss": 2.2881,
      "num_input_tokens_seen": 53219648,
      "step": 25470
    },
    {
      "epoch": 4.1558038991761155,
      "grad_norm": 5.34375,
      "learning_rate": 1.4579889860246382e-05,
      "loss": 2.3771,
      "num_input_tokens_seen": 53231152,
      "step": 25475
    },
    {
      "epoch": 4.156619626396933,
      "grad_norm": 12.625,
      "learning_rate": 1.457096661462885e-05,
      "loss": 1.9645,
      "num_input_tokens_seen": 53242784,
      "step": 25480
    },
    {
      "epoch": 4.157435353617751,
      "grad_norm": 14.25,
      "learning_rate": 1.4562044977299322e-05,
      "loss": 3.9255,
      "num_input_tokens_seen": 53252704,
      "step": 25485
    },
    {
      "epoch": 4.158251080838568,
      "grad_norm": 6.5625,
      "learning_rate": 1.4553124949633623e-05,
      "loss": 2.4001,
      "num_input_tokens_seen": 53263424,
      "step": 25490
    },
    {
      "epoch": 4.159066808059385,
      "grad_norm": 5.84375,
      "learning_rate": 1.4544206533007354e-05,
      "loss": 2.8383,
      "num_input_tokens_seen": 53272240,
      "step": 25495
    },
    {
      "epoch": 4.159882535280202,
      "grad_norm": 6.375,
      "learning_rate": 1.4535289728795821e-05,
      "loss": 3.5275,
      "num_input_tokens_seen": 53283120,
      "step": 25500
    },
    {
      "epoch": 4.16069826250102,
      "grad_norm": 2.234375,
      "learning_rate": 1.4526374538374132e-05,
      "loss": 1.6568,
      "num_input_tokens_seen": 53294464,
      "step": 25505
    },
    {
      "epoch": 4.161513989721837,
      "grad_norm": 6.0625,
      "learning_rate": 1.4517460963117097e-05,
      "loss": 2.4369,
      "num_input_tokens_seen": 53304464,
      "step": 25510
    },
    {
      "epoch": 4.162329716942654,
      "grad_norm": 7.8125,
      "learning_rate": 1.4508549004399314e-05,
      "loss": 2.7863,
      "num_input_tokens_seen": 53314688,
      "step": 25515
    },
    {
      "epoch": 4.163145444163471,
      "grad_norm": 8.6875,
      "learning_rate": 1.449963866359513e-05,
      "loss": 1.7282,
      "num_input_tokens_seen": 53325152,
      "step": 25520
    },
    {
      "epoch": 4.1639611713842895,
      "grad_norm": 6.15625,
      "learning_rate": 1.4490729942078607e-05,
      "loss": 1.2701,
      "num_input_tokens_seen": 53336240,
      "step": 25525
    },
    {
      "epoch": 4.164776898605107,
      "grad_norm": 10.125,
      "learning_rate": 1.4481822841223608e-05,
      "loss": 3.1916,
      "num_input_tokens_seen": 53346800,
      "step": 25530
    },
    {
      "epoch": 4.165592625825924,
      "grad_norm": 4.53125,
      "learning_rate": 1.4472917362403704e-05,
      "loss": 2.2709,
      "num_input_tokens_seen": 53357168,
      "step": 25535
    },
    {
      "epoch": 4.166408353046741,
      "grad_norm": 8.875,
      "learning_rate": 1.4464013506992224e-05,
      "loss": 1.7058,
      "num_input_tokens_seen": 53367120,
      "step": 25540
    },
    {
      "epoch": 4.167224080267559,
      "grad_norm": 2.625,
      "learning_rate": 1.4455111276362277e-05,
      "loss": 3.6354,
      "num_input_tokens_seen": 53376864,
      "step": 25545
    },
    {
      "epoch": 4.168039807488376,
      "grad_norm": 5.84375,
      "learning_rate": 1.4446210671886676e-05,
      "loss": 2.0646,
      "num_input_tokens_seen": 53386192,
      "step": 25550
    },
    {
      "epoch": 4.168855534709193,
      "grad_norm": 3.625,
      "learning_rate": 1.4437311694938015e-05,
      "loss": 1.8337,
      "num_input_tokens_seen": 53397664,
      "step": 25555
    },
    {
      "epoch": 4.16967126193001,
      "grad_norm": 2.25,
      "learning_rate": 1.442841434688864e-05,
      "loss": 2.0724,
      "num_input_tokens_seen": 53408752,
      "step": 25560
    },
    {
      "epoch": 4.170486989150828,
      "grad_norm": 7.96875,
      "learning_rate": 1.4419518629110615e-05,
      "loss": 2.7139,
      "num_input_tokens_seen": 53420560,
      "step": 25565
    },
    {
      "epoch": 4.171302716371645,
      "grad_norm": 5.03125,
      "learning_rate": 1.4410624542975778e-05,
      "loss": 1.7932,
      "num_input_tokens_seen": 53429136,
      "step": 25570
    },
    {
      "epoch": 4.1721184435924625,
      "grad_norm": 15.25,
      "learning_rate": 1.4401732089855724e-05,
      "loss": 2.9515,
      "num_input_tokens_seen": 53440016,
      "step": 25575
    },
    {
      "epoch": 4.17293417081328,
      "grad_norm": 4.84375,
      "learning_rate": 1.4392841271121754e-05,
      "loss": 1.5418,
      "num_input_tokens_seen": 53450816,
      "step": 25580
    },
    {
      "epoch": 4.173749898034098,
      "grad_norm": 6.90625,
      "learning_rate": 1.438395208814497e-05,
      "loss": 2.2517,
      "num_input_tokens_seen": 53461024,
      "step": 25585
    },
    {
      "epoch": 4.174565625254915,
      "grad_norm": 7.28125,
      "learning_rate": 1.4375064542296174e-05,
      "loss": 2.0148,
      "num_input_tokens_seen": 53470912,
      "step": 25590
    },
    {
      "epoch": 4.175381352475732,
      "grad_norm": 6.75,
      "learning_rate": 1.4366178634945946e-05,
      "loss": 2.6814,
      "num_input_tokens_seen": 53482736,
      "step": 25595
    },
    {
      "epoch": 4.176197079696549,
      "grad_norm": 8.5625,
      "learning_rate": 1.4357294367464616e-05,
      "loss": 1.2321,
      "num_input_tokens_seen": 53493696,
      "step": 25600
    },
    {
      "epoch": 4.176197079696549,
      "eval_loss": 2.5427770614624023,
      "eval_runtime": 135.0388,
      "eval_samples_per_second": 20.179,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 53493696,
      "step": 25600
    },
    {
      "epoch": 4.177012806917367,
      "grad_norm": 11.4375,
      "learning_rate": 1.434841174122224e-05,
      "loss": 2.7373,
      "num_input_tokens_seen": 53503328,
      "step": 25605
    },
    {
      "epoch": 4.177828534138184,
      "grad_norm": 12.25,
      "learning_rate": 1.4339530757588615e-05,
      "loss": 2.5927,
      "num_input_tokens_seen": 53514032,
      "step": 25610
    },
    {
      "epoch": 4.178644261359001,
      "grad_norm": 3.234375,
      "learning_rate": 1.433065141793333e-05,
      "loss": 2.3291,
      "num_input_tokens_seen": 53524368,
      "step": 25615
    },
    {
      "epoch": 4.1794599885798185,
      "grad_norm": 4.65625,
      "learning_rate": 1.4321773723625665e-05,
      "loss": 3.2613,
      "num_input_tokens_seen": 53534976,
      "step": 25620
    },
    {
      "epoch": 4.1802757158006365,
      "grad_norm": 1.9140625,
      "learning_rate": 1.4312897676034693e-05,
      "loss": 2.1765,
      "num_input_tokens_seen": 53545472,
      "step": 25625
    },
    {
      "epoch": 4.181091443021454,
      "grad_norm": 8.1875,
      "learning_rate": 1.4304023276529188e-05,
      "loss": 2.1716,
      "num_input_tokens_seen": 53557344,
      "step": 25630
    },
    {
      "epoch": 4.181907170242271,
      "grad_norm": 7.96875,
      "learning_rate": 1.4295150526477712e-05,
      "loss": 2.7278,
      "num_input_tokens_seen": 53567248,
      "step": 25635
    },
    {
      "epoch": 4.182722897463089,
      "grad_norm": 5.5,
      "learning_rate": 1.4286279427248562e-05,
      "loss": 2.4444,
      "num_input_tokens_seen": 53576720,
      "step": 25640
    },
    {
      "epoch": 4.183538624683906,
      "grad_norm": 5.1875,
      "learning_rate": 1.4277409980209747e-05,
      "loss": 0.9651,
      "num_input_tokens_seen": 53586848,
      "step": 25645
    },
    {
      "epoch": 4.184354351904723,
      "grad_norm": 7.09375,
      "learning_rate": 1.4268542186729061e-05,
      "loss": 3.0174,
      "num_input_tokens_seen": 53596352,
      "step": 25650
    },
    {
      "epoch": 4.18517007912554,
      "grad_norm": 4.59375,
      "learning_rate": 1.4259676048174043e-05,
      "loss": 1.6685,
      "num_input_tokens_seen": 53606768,
      "step": 25655
    },
    {
      "epoch": 4.185985806346358,
      "grad_norm": 5.65625,
      "learning_rate": 1.4250811565911937e-05,
      "loss": 2.0507,
      "num_input_tokens_seen": 53616768,
      "step": 25660
    },
    {
      "epoch": 4.186801533567175,
      "grad_norm": 2.953125,
      "learning_rate": 1.4241948741309782e-05,
      "loss": 2.2424,
      "num_input_tokens_seen": 53626944,
      "step": 25665
    },
    {
      "epoch": 4.1876172607879925,
      "grad_norm": 9.9375,
      "learning_rate": 1.4233087575734317e-05,
      "loss": 3.4302,
      "num_input_tokens_seen": 53636016,
      "step": 25670
    },
    {
      "epoch": 4.18843298800881,
      "grad_norm": 10.875,
      "learning_rate": 1.422422807055206e-05,
      "loss": 2.0502,
      "num_input_tokens_seen": 53648112,
      "step": 25675
    },
    {
      "epoch": 4.189248715229628,
      "grad_norm": 7.28125,
      "learning_rate": 1.4215370227129243e-05,
      "loss": 1.939,
      "num_input_tokens_seen": 53658112,
      "step": 25680
    },
    {
      "epoch": 4.190064442450445,
      "grad_norm": 12.9375,
      "learning_rate": 1.4206514046831876e-05,
      "loss": 3.2173,
      "num_input_tokens_seen": 53669696,
      "step": 25685
    },
    {
      "epoch": 4.190880169671262,
      "grad_norm": 6.34375,
      "learning_rate": 1.419765953102567e-05,
      "loss": 2.4448,
      "num_input_tokens_seen": 53679856,
      "step": 25690
    },
    {
      "epoch": 4.191695896892079,
      "grad_norm": 7.21875,
      "learning_rate": 1.4188806681076125e-05,
      "loss": 2.607,
      "num_input_tokens_seen": 53690176,
      "step": 25695
    },
    {
      "epoch": 4.192511624112897,
      "grad_norm": 3.140625,
      "learning_rate": 1.4179955498348443e-05,
      "loss": 4.4754,
      "num_input_tokens_seen": 53699744,
      "step": 25700
    },
    {
      "epoch": 4.193327351333714,
      "grad_norm": 4.6875,
      "learning_rate": 1.4171105984207605e-05,
      "loss": 2.3131,
      "num_input_tokens_seen": 53710144,
      "step": 25705
    },
    {
      "epoch": 4.194143078554531,
      "grad_norm": 4.90625,
      "learning_rate": 1.4162258140018304e-05,
      "loss": 2.7778,
      "num_input_tokens_seen": 53721808,
      "step": 25710
    },
    {
      "epoch": 4.194958805775348,
      "grad_norm": 7.0,
      "learning_rate": 1.4153411967144986e-05,
      "loss": 3.0982,
      "num_input_tokens_seen": 53732848,
      "step": 25715
    },
    {
      "epoch": 4.195774532996166,
      "grad_norm": 8.4375,
      "learning_rate": 1.4144567466951864e-05,
      "loss": 2.6465,
      "num_input_tokens_seen": 53742944,
      "step": 25720
    },
    {
      "epoch": 4.196590260216984,
      "grad_norm": 5.125,
      "learning_rate": 1.4135724640802844e-05,
      "loss": 2.036,
      "num_input_tokens_seen": 53753152,
      "step": 25725
    },
    {
      "epoch": 4.197405987437801,
      "grad_norm": 9.5,
      "learning_rate": 1.4126883490061615e-05,
      "loss": 3.3645,
      "num_input_tokens_seen": 53764384,
      "step": 25730
    },
    {
      "epoch": 4.198221714658618,
      "grad_norm": 4.28125,
      "learning_rate": 1.4118044016091603e-05,
      "loss": 1.4475,
      "num_input_tokens_seen": 53775360,
      "step": 25735
    },
    {
      "epoch": 4.199037441879436,
      "grad_norm": 6.46875,
      "learning_rate": 1.410920622025594e-05,
      "loss": 2.3715,
      "num_input_tokens_seen": 53784400,
      "step": 25740
    },
    {
      "epoch": 4.199853169100253,
      "grad_norm": 6.28125,
      "learning_rate": 1.4100370103917554e-05,
      "loss": 2.3947,
      "num_input_tokens_seen": 53794048,
      "step": 25745
    },
    {
      "epoch": 4.20066889632107,
      "grad_norm": 2.84375,
      "learning_rate": 1.409153566843907e-05,
      "loss": 1.6669,
      "num_input_tokens_seen": 53804560,
      "step": 25750
    },
    {
      "epoch": 4.201484623541887,
      "grad_norm": 0.94140625,
      "learning_rate": 1.408270291518286e-05,
      "loss": 1.1062,
      "num_input_tokens_seen": 53815536,
      "step": 25755
    },
    {
      "epoch": 4.202300350762705,
      "grad_norm": 4.46875,
      "learning_rate": 1.407387184551107e-05,
      "loss": 2.0715,
      "num_input_tokens_seen": 53826400,
      "step": 25760
    },
    {
      "epoch": 4.203116077983522,
      "grad_norm": 6.1875,
      "learning_rate": 1.4065042460785532e-05,
      "loss": 2.4419,
      "num_input_tokens_seen": 53835808,
      "step": 25765
    },
    {
      "epoch": 4.2039318052043395,
      "grad_norm": 5.375,
      "learning_rate": 1.405621476236787e-05,
      "loss": 1.1933,
      "num_input_tokens_seen": 53844880,
      "step": 25770
    },
    {
      "epoch": 4.204747532425157,
      "grad_norm": 12.1875,
      "learning_rate": 1.4047388751619423e-05,
      "loss": 3.1226,
      "num_input_tokens_seen": 53856352,
      "step": 25775
    },
    {
      "epoch": 4.205563259645975,
      "grad_norm": 5.25,
      "learning_rate": 1.4038564429901264e-05,
      "loss": 1.8335,
      "num_input_tokens_seen": 53867232,
      "step": 25780
    },
    {
      "epoch": 4.206378986866792,
      "grad_norm": 11.6875,
      "learning_rate": 1.4029741798574227e-05,
      "loss": 2.8169,
      "num_input_tokens_seen": 53877248,
      "step": 25785
    },
    {
      "epoch": 4.207194714087609,
      "grad_norm": 7.25,
      "learning_rate": 1.402092085899886e-05,
      "loss": 3.2848,
      "num_input_tokens_seen": 53887328,
      "step": 25790
    },
    {
      "epoch": 4.208010441308426,
      "grad_norm": 8.875,
      "learning_rate": 1.4012101612535464e-05,
      "loss": 2.9211,
      "num_input_tokens_seen": 53897504,
      "step": 25795
    },
    {
      "epoch": 4.208826168529244,
      "grad_norm": 6.1875,
      "learning_rate": 1.4003284060544092e-05,
      "loss": 2.4438,
      "num_input_tokens_seen": 53907648,
      "step": 25800
    },
    {
      "epoch": 4.208826168529244,
      "eval_loss": 2.5512468814849854,
      "eval_runtime": 134.8314,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.109,
      "num_input_tokens_seen": 53907648,
      "step": 25800
    },
    {
      "epoch": 4.209641895750061,
      "grad_norm": 0.2294921875,
      "learning_rate": 1.3994468204384504e-05,
      "loss": 2.0562,
      "num_input_tokens_seen": 53917760,
      "step": 25805
    },
    {
      "epoch": 4.210457622970878,
      "grad_norm": 10.8125,
      "learning_rate": 1.398565404541622e-05,
      "loss": 2.6092,
      "num_input_tokens_seen": 53928512,
      "step": 25810
    },
    {
      "epoch": 4.211273350191696,
      "grad_norm": 5.96875,
      "learning_rate": 1.3976841584998513e-05,
      "loss": 1.9569,
      "num_input_tokens_seen": 53939424,
      "step": 25815
    },
    {
      "epoch": 4.2120890774125135,
      "grad_norm": 6.4375,
      "learning_rate": 1.3968030824490352e-05,
      "loss": 1.6437,
      "num_input_tokens_seen": 53949408,
      "step": 25820
    },
    {
      "epoch": 4.212904804633331,
      "grad_norm": 6.0625,
      "learning_rate": 1.3959221765250469e-05,
      "loss": 2.976,
      "num_input_tokens_seen": 53960928,
      "step": 25825
    },
    {
      "epoch": 4.213720531854148,
      "grad_norm": 2.375,
      "learning_rate": 1.3950414408637343e-05,
      "loss": 3.0501,
      "num_input_tokens_seen": 53970144,
      "step": 25830
    },
    {
      "epoch": 4.214536259074965,
      "grad_norm": 8.3125,
      "learning_rate": 1.3941608756009166e-05,
      "loss": 2.853,
      "num_input_tokens_seen": 53979808,
      "step": 25835
    },
    {
      "epoch": 4.215351986295783,
      "grad_norm": 8.9375,
      "learning_rate": 1.3932804808723898e-05,
      "loss": 2.5677,
      "num_input_tokens_seen": 53990992,
      "step": 25840
    },
    {
      "epoch": 4.2161677135166,
      "grad_norm": 3.375,
      "learning_rate": 1.3924002568139194e-05,
      "loss": 2.016,
      "num_input_tokens_seen": 54001200,
      "step": 25845
    },
    {
      "epoch": 4.216983440737417,
      "grad_norm": 4.53125,
      "learning_rate": 1.3915202035612485e-05,
      "loss": 1.4556,
      "num_input_tokens_seen": 54011120,
      "step": 25850
    },
    {
      "epoch": 4.217799167958235,
      "grad_norm": 6.9375,
      "learning_rate": 1.3906403212500935e-05,
      "loss": 1.6477,
      "num_input_tokens_seen": 54021888,
      "step": 25855
    },
    {
      "epoch": 4.218614895179052,
      "grad_norm": 8.0625,
      "learning_rate": 1.3897606100161409e-05,
      "loss": 2.5998,
      "num_input_tokens_seen": 54032800,
      "step": 25860
    },
    {
      "epoch": 4.219430622399869,
      "grad_norm": 4.46875,
      "learning_rate": 1.388881069995055e-05,
      "loss": 2.1501,
      "num_input_tokens_seen": 54044784,
      "step": 25865
    },
    {
      "epoch": 4.220246349620687,
      "grad_norm": 10.25,
      "learning_rate": 1.3880017013224708e-05,
      "loss": 3.1494,
      "num_input_tokens_seen": 54055664,
      "step": 25870
    },
    {
      "epoch": 4.221062076841505,
      "grad_norm": 11.5625,
      "learning_rate": 1.3871225041339984e-05,
      "loss": 1.8781,
      "num_input_tokens_seen": 54066416,
      "step": 25875
    },
    {
      "epoch": 4.221877804062322,
      "grad_norm": 9.125,
      "learning_rate": 1.386243478565222e-05,
      "loss": 3.6046,
      "num_input_tokens_seen": 54077792,
      "step": 25880
    },
    {
      "epoch": 4.222693531283139,
      "grad_norm": 7.65625,
      "learning_rate": 1.3853646247516966e-05,
      "loss": 1.0474,
      "num_input_tokens_seen": 54088192,
      "step": 25885
    },
    {
      "epoch": 4.223509258503956,
      "grad_norm": 5.1875,
      "learning_rate": 1.3844859428289545e-05,
      "loss": 2.4658,
      "num_input_tokens_seen": 54098160,
      "step": 25890
    },
    {
      "epoch": 4.224324985724774,
      "grad_norm": 11.875,
      "learning_rate": 1.3836074329324984e-05,
      "loss": 4.0572,
      "num_input_tokens_seen": 54108704,
      "step": 25895
    },
    {
      "epoch": 4.225140712945591,
      "grad_norm": 5.78125,
      "learning_rate": 1.3827290951978044e-05,
      "loss": 2.9406,
      "num_input_tokens_seen": 54120736,
      "step": 25900
    },
    {
      "epoch": 4.225956440166408,
      "grad_norm": 7.8125,
      "learning_rate": 1.381850929760326e-05,
      "loss": 2.3327,
      "num_input_tokens_seen": 54131696,
      "step": 25905
    },
    {
      "epoch": 4.226772167387225,
      "grad_norm": 9.0625,
      "learning_rate": 1.3809729367554842e-05,
      "loss": 2.7324,
      "num_input_tokens_seen": 54141552,
      "step": 25910
    },
    {
      "epoch": 4.227587894608043,
      "grad_norm": 5.375,
      "learning_rate": 1.3800951163186784e-05,
      "loss": 2.4908,
      "num_input_tokens_seen": 54152112,
      "step": 25915
    },
    {
      "epoch": 4.2284036218288605,
      "grad_norm": 9.3125,
      "learning_rate": 1.3792174685852801e-05,
      "loss": 1.7665,
      "num_input_tokens_seen": 54162400,
      "step": 25920
    },
    {
      "epoch": 4.229219349049678,
      "grad_norm": 8.1875,
      "learning_rate": 1.378339993690632e-05,
      "loss": 1.7352,
      "num_input_tokens_seen": 54173472,
      "step": 25925
    },
    {
      "epoch": 4.230035076270495,
      "grad_norm": 8.9375,
      "learning_rate": 1.3774626917700523e-05,
      "loss": 2.8969,
      "num_input_tokens_seen": 54185312,
      "step": 25930
    },
    {
      "epoch": 4.230850803491313,
      "grad_norm": 7.40625,
      "learning_rate": 1.3765855629588334e-05,
      "loss": 2.0314,
      "num_input_tokens_seen": 54196304,
      "step": 25935
    },
    {
      "epoch": 4.23166653071213,
      "grad_norm": 6.875,
      "learning_rate": 1.3757086073922374e-05,
      "loss": 2.1068,
      "num_input_tokens_seen": 54206400,
      "step": 25940
    },
    {
      "epoch": 4.232482257932947,
      "grad_norm": 8.6875,
      "learning_rate": 1.3748318252055038e-05,
      "loss": 1.3149,
      "num_input_tokens_seen": 54215760,
      "step": 25945
    },
    {
      "epoch": 4.233297985153764,
      "grad_norm": 7.90625,
      "learning_rate": 1.3739552165338416e-05,
      "loss": 2.1171,
      "num_input_tokens_seen": 54226768,
      "step": 25950
    },
    {
      "epoch": 4.234113712374582,
      "grad_norm": 5.40625,
      "learning_rate": 1.3730787815124354e-05,
      "loss": 2.33,
      "num_input_tokens_seen": 54237120,
      "step": 25955
    },
    {
      "epoch": 4.234929439595399,
      "grad_norm": 2.34375,
      "learning_rate": 1.3722025202764443e-05,
      "loss": 3.0722,
      "num_input_tokens_seen": 54247664,
      "step": 25960
    },
    {
      "epoch": 4.2357451668162165,
      "grad_norm": 9.5,
      "learning_rate": 1.371326432960997e-05,
      "loss": 1.6266,
      "num_input_tokens_seen": 54257248,
      "step": 25965
    },
    {
      "epoch": 4.236560894037034,
      "grad_norm": 10.0625,
      "learning_rate": 1.3704505197011969e-05,
      "loss": 2.0549,
      "num_input_tokens_seen": 54265808,
      "step": 25970
    },
    {
      "epoch": 4.237376621257852,
      "grad_norm": 2.296875,
      "learning_rate": 1.3695747806321224e-05,
      "loss": 2.7867,
      "num_input_tokens_seen": 54277552,
      "step": 25975
    },
    {
      "epoch": 4.238192348478669,
      "grad_norm": 3.484375,
      "learning_rate": 1.3686992158888212e-05,
      "loss": 1.8499,
      "num_input_tokens_seen": 54288288,
      "step": 25980
    },
    {
      "epoch": 4.239008075699486,
      "grad_norm": 9.8125,
      "learning_rate": 1.367823825606319e-05,
      "loss": 3.647,
      "num_input_tokens_seen": 54297600,
      "step": 25985
    },
    {
      "epoch": 4.239823802920303,
      "grad_norm": 7.75,
      "learning_rate": 1.36694860991961e-05,
      "loss": 1.6617,
      "num_input_tokens_seen": 54307440,
      "step": 25990
    },
    {
      "epoch": 4.240639530141121,
      "grad_norm": 2.96875,
      "learning_rate": 1.3660735689636636e-05,
      "loss": 2.4633,
      "num_input_tokens_seen": 54317024,
      "step": 25995
    },
    {
      "epoch": 4.241455257361938,
      "grad_norm": 9.4375,
      "learning_rate": 1.365198702873424e-05,
      "loss": 1.8597,
      "num_input_tokens_seen": 54327568,
      "step": 26000
    },
    {
      "epoch": 4.241455257361938,
      "eval_loss": 2.5442895889282227,
      "eval_runtime": 134.7854,
      "eval_samples_per_second": 20.217,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 54327568,
      "step": 26000
    },
    {
      "epoch": 4.242270984582755,
      "grad_norm": 8.0625,
      "learning_rate": 1.364324011783804e-05,
      "loss": 2.2468,
      "num_input_tokens_seen": 54337136,
      "step": 26005
    },
    {
      "epoch": 4.243086711803572,
      "grad_norm": 9.3125,
      "learning_rate": 1.3634494958296934e-05,
      "loss": 1.6152,
      "num_input_tokens_seen": 54348752,
      "step": 26010
    },
    {
      "epoch": 4.2439024390243905,
      "grad_norm": 13.0625,
      "learning_rate": 1.3625751551459542e-05,
      "loss": 2.6483,
      "num_input_tokens_seen": 54359840,
      "step": 26015
    },
    {
      "epoch": 4.244718166245208,
      "grad_norm": 7.28125,
      "learning_rate": 1.3617009898674188e-05,
      "loss": 2.9703,
      "num_input_tokens_seen": 54370656,
      "step": 26020
    },
    {
      "epoch": 4.245533893466025,
      "grad_norm": 7.71875,
      "learning_rate": 1.3608270001288967e-05,
      "loss": 2.9144,
      "num_input_tokens_seen": 54380768,
      "step": 26025
    },
    {
      "epoch": 4.246349620686843,
      "grad_norm": 6.25,
      "learning_rate": 1.359953186065166e-05,
      "loss": 3.5743,
      "num_input_tokens_seen": 54390688,
      "step": 26030
    },
    {
      "epoch": 4.24716534790766,
      "grad_norm": 1.84375,
      "learning_rate": 1.3590795478109814e-05,
      "loss": 1.4763,
      "num_input_tokens_seen": 54400368,
      "step": 26035
    },
    {
      "epoch": 4.247981075128477,
      "grad_norm": 5.375,
      "learning_rate": 1.3582060855010675e-05,
      "loss": 2.2997,
      "num_input_tokens_seen": 54411168,
      "step": 26040
    },
    {
      "epoch": 4.248796802349294,
      "grad_norm": 2.0625,
      "learning_rate": 1.3573327992701245e-05,
      "loss": 2.3867,
      "num_input_tokens_seen": 54421344,
      "step": 26045
    },
    {
      "epoch": 4.249612529570112,
      "grad_norm": 10.125,
      "learning_rate": 1.356459689252823e-05,
      "loss": 2.6096,
      "num_input_tokens_seen": 54433344,
      "step": 26050
    },
    {
      "epoch": 4.250428256790929,
      "grad_norm": 6.25,
      "learning_rate": 1.3555867555838087e-05,
      "loss": 1.7033,
      "num_input_tokens_seen": 54444256,
      "step": 26055
    },
    {
      "epoch": 4.251243984011746,
      "grad_norm": 9.375,
      "learning_rate": 1.3547139983976975e-05,
      "loss": 1.2953,
      "num_input_tokens_seen": 54454384,
      "step": 26060
    },
    {
      "epoch": 4.2520597112325635,
      "grad_norm": 0.298828125,
      "learning_rate": 1.3538414178290815e-05,
      "loss": 1.5873,
      "num_input_tokens_seen": 54466032,
      "step": 26065
    },
    {
      "epoch": 4.252875438453382,
      "grad_norm": 4.6875,
      "learning_rate": 1.3529690140125209e-05,
      "loss": 1.652,
      "num_input_tokens_seen": 54476704,
      "step": 26070
    },
    {
      "epoch": 4.253691165674199,
      "grad_norm": 6.5625,
      "learning_rate": 1.352096787082553e-05,
      "loss": 4.1459,
      "num_input_tokens_seen": 54488320,
      "step": 26075
    },
    {
      "epoch": 4.254506892895016,
      "grad_norm": 6.71875,
      "learning_rate": 1.3512247371736871e-05,
      "loss": 2.4114,
      "num_input_tokens_seen": 54497456,
      "step": 26080
    },
    {
      "epoch": 4.255322620115833,
      "grad_norm": 6.625,
      "learning_rate": 1.3503528644204022e-05,
      "loss": 2.1498,
      "num_input_tokens_seen": 54508944,
      "step": 26085
    },
    {
      "epoch": 4.256138347336651,
      "grad_norm": 10.375,
      "learning_rate": 1.349481168957153e-05,
      "loss": 2.4941,
      "num_input_tokens_seen": 54518960,
      "step": 26090
    },
    {
      "epoch": 4.256954074557468,
      "grad_norm": 7.1875,
      "learning_rate": 1.3486096509183665e-05,
      "loss": 1.5968,
      "num_input_tokens_seen": 54527920,
      "step": 26095
    },
    {
      "epoch": 4.257769801778285,
      "grad_norm": 7.4375,
      "learning_rate": 1.3477383104384406e-05,
      "loss": 2.293,
      "num_input_tokens_seen": 54539904,
      "step": 26100
    },
    {
      "epoch": 4.258585528999102,
      "grad_norm": 4.09375,
      "learning_rate": 1.3468671476517481e-05,
      "loss": 1.5148,
      "num_input_tokens_seen": 54550416,
      "step": 26105
    },
    {
      "epoch": 4.25940125621992,
      "grad_norm": 9.9375,
      "learning_rate": 1.3459961626926326e-05,
      "loss": 2.6581,
      "num_input_tokens_seen": 54560640,
      "step": 26110
    },
    {
      "epoch": 4.2602169834407375,
      "grad_norm": 15.25,
      "learning_rate": 1.3451253556954101e-05,
      "loss": 3.6614,
      "num_input_tokens_seen": 54570752,
      "step": 26115
    },
    {
      "epoch": 4.261032710661555,
      "grad_norm": 9.125,
      "learning_rate": 1.3442547267943717e-05,
      "loss": 1.876,
      "num_input_tokens_seen": 54580224,
      "step": 26120
    },
    {
      "epoch": 4.261848437882372,
      "grad_norm": 7.125,
      "learning_rate": 1.3433842761237774e-05,
      "loss": 3.3499,
      "num_input_tokens_seen": 54589488,
      "step": 26125
    },
    {
      "epoch": 4.26266416510319,
      "grad_norm": 10.0625,
      "learning_rate": 1.3425140038178639e-05,
      "loss": 2.2879,
      "num_input_tokens_seen": 54599072,
      "step": 26130
    },
    {
      "epoch": 4.263479892324007,
      "grad_norm": 6.90625,
      "learning_rate": 1.3416439100108358e-05,
      "loss": 3.4869,
      "num_input_tokens_seen": 54609312,
      "step": 26135
    },
    {
      "epoch": 4.264295619544824,
      "grad_norm": 10.6875,
      "learning_rate": 1.3407739948368734e-05,
      "loss": 3.3091,
      "num_input_tokens_seen": 54619472,
      "step": 26140
    },
    {
      "epoch": 4.265111346765641,
      "grad_norm": 5.96875,
      "learning_rate": 1.3399042584301298e-05,
      "loss": 2.0133,
      "num_input_tokens_seen": 54627840,
      "step": 26145
    },
    {
      "epoch": 4.265927073986459,
      "grad_norm": 3.203125,
      "learning_rate": 1.3390347009247272e-05,
      "loss": 2.734,
      "num_input_tokens_seen": 54638848,
      "step": 26150
    },
    {
      "epoch": 4.266742801207276,
      "grad_norm": 2.203125,
      "learning_rate": 1.3381653224547635e-05,
      "loss": 1.3911,
      "num_input_tokens_seen": 54649680,
      "step": 26155
    },
    {
      "epoch": 4.2675585284280935,
      "grad_norm": 8.1875,
      "learning_rate": 1.3372961231543086e-05,
      "loss": 2.3937,
      "num_input_tokens_seen": 54658608,
      "step": 26160
    },
    {
      "epoch": 4.268374255648911,
      "grad_norm": 6.0,
      "learning_rate": 1.3364271031574016e-05,
      "loss": 2.2395,
      "num_input_tokens_seen": 54670272,
      "step": 26165
    },
    {
      "epoch": 4.269189982869729,
      "grad_norm": 8.6875,
      "learning_rate": 1.335558262598059e-05,
      "loss": 1.8139,
      "num_input_tokens_seen": 54682048,
      "step": 26170
    },
    {
      "epoch": 4.270005710090546,
      "grad_norm": 7.6875,
      "learning_rate": 1.3346896016102645e-05,
      "loss": 1.4124,
      "num_input_tokens_seen": 54691856,
      "step": 26175
    },
    {
      "epoch": 4.270821437311363,
      "grad_norm": 3.875,
      "learning_rate": 1.3338211203279788e-05,
      "loss": 2.2178,
      "num_input_tokens_seen": 54701552,
      "step": 26180
    },
    {
      "epoch": 4.27163716453218,
      "grad_norm": 5.84375,
      "learning_rate": 1.3329528188851303e-05,
      "loss": 3.1556,
      "num_input_tokens_seen": 54712528,
      "step": 26185
    },
    {
      "epoch": 4.272452891752998,
      "grad_norm": 5.125,
      "learning_rate": 1.3320846974156242e-05,
      "loss": 2.9252,
      "num_input_tokens_seen": 54723360,
      "step": 26190
    },
    {
      "epoch": 4.273268618973815,
      "grad_norm": 5.84375,
      "learning_rate": 1.3312167560533337e-05,
      "loss": 1.7974,
      "num_input_tokens_seen": 54733856,
      "step": 26195
    },
    {
      "epoch": 4.274084346194632,
      "grad_norm": 5.59375,
      "learning_rate": 1.3303489949321082e-05,
      "loss": 2.9831,
      "num_input_tokens_seen": 54743840,
      "step": 26200
    },
    {
      "epoch": 4.274084346194632,
      "eval_loss": 2.538890838623047,
      "eval_runtime": 134.9681,
      "eval_samples_per_second": 20.19,
      "eval_steps_per_second": 10.099,
      "num_input_tokens_seen": 54743840,
      "step": 26200
    },
    {
      "epoch": 4.27490007341545,
      "grad_norm": 4.03125,
      "learning_rate": 1.3294814141857653e-05,
      "loss": 2.2755,
      "num_input_tokens_seen": 54753920,
      "step": 26205
    },
    {
      "epoch": 4.275715800636267,
      "grad_norm": 5.4375,
      "learning_rate": 1.3286140139480992e-05,
      "loss": 2.7358,
      "num_input_tokens_seen": 54763408,
      "step": 26210
    },
    {
      "epoch": 4.276531527857085,
      "grad_norm": 5.28125,
      "learning_rate": 1.3277467943528719e-05,
      "loss": 2.5934,
      "num_input_tokens_seen": 54774096,
      "step": 26215
    },
    {
      "epoch": 4.277347255077902,
      "grad_norm": 8.125,
      "learning_rate": 1.3268797555338203e-05,
      "loss": 1.948,
      "num_input_tokens_seen": 54784560,
      "step": 26220
    },
    {
      "epoch": 4.27816298229872,
      "grad_norm": 11.9375,
      "learning_rate": 1.3260128976246533e-05,
      "loss": 2.1775,
      "num_input_tokens_seen": 54795056,
      "step": 26225
    },
    {
      "epoch": 4.278978709519537,
      "grad_norm": 9.5,
      "learning_rate": 1.32514622075905e-05,
      "loss": 3.1502,
      "num_input_tokens_seen": 54804816,
      "step": 26230
    },
    {
      "epoch": 4.279794436740354,
      "grad_norm": 9.0625,
      "learning_rate": 1.3242797250706638e-05,
      "loss": 3.2743,
      "num_input_tokens_seen": 54815184,
      "step": 26235
    },
    {
      "epoch": 4.280610163961171,
      "grad_norm": 8.8125,
      "learning_rate": 1.3234134106931195e-05,
      "loss": 1.6456,
      "num_input_tokens_seen": 54826192,
      "step": 26240
    },
    {
      "epoch": 4.281425891181989,
      "grad_norm": 11.0625,
      "learning_rate": 1.322547277760013e-05,
      "loss": 2.8276,
      "num_input_tokens_seen": 54837392,
      "step": 26245
    },
    {
      "epoch": 4.282241618402806,
      "grad_norm": 6.5,
      "learning_rate": 1.3216813264049132e-05,
      "loss": 1.6048,
      "num_input_tokens_seen": 54847600,
      "step": 26250
    },
    {
      "epoch": 4.283057345623623,
      "grad_norm": 3.78125,
      "learning_rate": 1.32081555676136e-05,
      "loss": 1.2416,
      "num_input_tokens_seen": 54858128,
      "step": 26255
    },
    {
      "epoch": 4.2838730728444405,
      "grad_norm": 10.1875,
      "learning_rate": 1.3199499689628674e-05,
      "loss": 3.753,
      "num_input_tokens_seen": 54868896,
      "step": 26260
    },
    {
      "epoch": 4.2846888000652585,
      "grad_norm": 8.3125,
      "learning_rate": 1.3190845631429192e-05,
      "loss": 2.3611,
      "num_input_tokens_seen": 54879280,
      "step": 26265
    },
    {
      "epoch": 4.285504527286076,
      "grad_norm": 7.5,
      "learning_rate": 1.3182193394349704e-05,
      "loss": 2.7388,
      "num_input_tokens_seen": 54890368,
      "step": 26270
    },
    {
      "epoch": 4.286320254506893,
      "grad_norm": 5.1875,
      "learning_rate": 1.3173542979724507e-05,
      "loss": 3.2111,
      "num_input_tokens_seen": 54900016,
      "step": 26275
    },
    {
      "epoch": 4.28713598172771,
      "grad_norm": 6.3125,
      "learning_rate": 1.3164894388887617e-05,
      "loss": 2.3468,
      "num_input_tokens_seen": 54910112,
      "step": 26280
    },
    {
      "epoch": 4.287951708948528,
      "grad_norm": 2.65625,
      "learning_rate": 1.3156247623172727e-05,
      "loss": 1.3178,
      "num_input_tokens_seen": 54921200,
      "step": 26285
    },
    {
      "epoch": 4.288767436169345,
      "grad_norm": 5.53125,
      "learning_rate": 1.3147602683913302e-05,
      "loss": 2.2775,
      "num_input_tokens_seen": 54932144,
      "step": 26290
    },
    {
      "epoch": 4.289583163390162,
      "grad_norm": 8.5,
      "learning_rate": 1.3138959572442481e-05,
      "loss": 3.49,
      "num_input_tokens_seen": 54941744,
      "step": 26295
    },
    {
      "epoch": 4.290398890610979,
      "grad_norm": 4.3125,
      "learning_rate": 1.3130318290093146e-05,
      "loss": 2.0101,
      "num_input_tokens_seen": 54950800,
      "step": 26300
    },
    {
      "epoch": 4.291214617831797,
      "grad_norm": 7.59375,
      "learning_rate": 1.3121678838197909e-05,
      "loss": 2.4744,
      "num_input_tokens_seen": 54961936,
      "step": 26305
    },
    {
      "epoch": 4.2920303450526145,
      "grad_norm": 4.09375,
      "learning_rate": 1.3113041218089056e-05,
      "loss": 1.1062,
      "num_input_tokens_seen": 54972512,
      "step": 26310
    },
    {
      "epoch": 4.292846072273432,
      "grad_norm": 4.21875,
      "learning_rate": 1.3104405431098626e-05,
      "loss": 2.0107,
      "num_input_tokens_seen": 54983040,
      "step": 26315
    },
    {
      "epoch": 4.293661799494249,
      "grad_norm": 4.09375,
      "learning_rate": 1.3095771478558377e-05,
      "loss": 2.337,
      "num_input_tokens_seen": 54993328,
      "step": 26320
    },
    {
      "epoch": 4.294477526715067,
      "grad_norm": 0.169921875,
      "learning_rate": 1.3087139361799766e-05,
      "loss": 2.1072,
      "num_input_tokens_seen": 55003680,
      "step": 26325
    },
    {
      "epoch": 4.295293253935884,
      "grad_norm": 6.46875,
      "learning_rate": 1.3078509082153964e-05,
      "loss": 1.4382,
      "num_input_tokens_seen": 55014032,
      "step": 26330
    },
    {
      "epoch": 4.296108981156701,
      "grad_norm": 7.0625,
      "learning_rate": 1.3069880640951885e-05,
      "loss": 1.3562,
      "num_input_tokens_seen": 55023472,
      "step": 26335
    },
    {
      "epoch": 4.296924708377518,
      "grad_norm": 2.359375,
      "learning_rate": 1.3061254039524123e-05,
      "loss": 2.9152,
      "num_input_tokens_seen": 55033664,
      "step": 26340
    },
    {
      "epoch": 4.297740435598336,
      "grad_norm": 5.15625,
      "learning_rate": 1.3052629279201028e-05,
      "loss": 1.9892,
      "num_input_tokens_seen": 55044480,
      "step": 26345
    },
    {
      "epoch": 4.298556162819153,
      "grad_norm": 3.984375,
      "learning_rate": 1.3044006361312633e-05,
      "loss": 2.2614,
      "num_input_tokens_seen": 55055472,
      "step": 26350
    },
    {
      "epoch": 4.2993718900399704,
      "grad_norm": 7.46875,
      "learning_rate": 1.30353852871887e-05,
      "loss": 3.2531,
      "num_input_tokens_seen": 55064512,
      "step": 26355
    },
    {
      "epoch": 4.300187617260788,
      "grad_norm": 8.625,
      "learning_rate": 1.302676605815873e-05,
      "loss": 2.5758,
      "num_input_tokens_seen": 55075408,
      "step": 26360
    },
    {
      "epoch": 4.301003344481606,
      "grad_norm": 4.3125,
      "learning_rate": 1.3018148675551884e-05,
      "loss": 3.4504,
      "num_input_tokens_seen": 55085760,
      "step": 26365
    },
    {
      "epoch": 4.301819071702423,
      "grad_norm": 8.1875,
      "learning_rate": 1.3009533140697094e-05,
      "loss": 2.8685,
      "num_input_tokens_seen": 55095936,
      "step": 26370
    },
    {
      "epoch": 4.30263479892324,
      "grad_norm": 6.5,
      "learning_rate": 1.3000919454922966e-05,
      "loss": 2.498,
      "num_input_tokens_seen": 55106064,
      "step": 26375
    },
    {
      "epoch": 4.303450526144058,
      "grad_norm": 8.0,
      "learning_rate": 1.299230761955785e-05,
      "loss": 2.0844,
      "num_input_tokens_seen": 55118080,
      "step": 26380
    },
    {
      "epoch": 4.304266253364875,
      "grad_norm": 4.5,
      "learning_rate": 1.2983697635929807e-05,
      "loss": 2.586,
      "num_input_tokens_seen": 55128464,
      "step": 26385
    },
    {
      "epoch": 4.305081980585692,
      "grad_norm": 14.4375,
      "learning_rate": 1.2975089505366584e-05,
      "loss": 3.7998,
      "num_input_tokens_seen": 55137632,
      "step": 26390
    },
    {
      "epoch": 4.305897707806509,
      "grad_norm": 2.109375,
      "learning_rate": 1.2966483229195683e-05,
      "loss": 2.7398,
      "num_input_tokens_seen": 55148192,
      "step": 26395
    },
    {
      "epoch": 4.306713435027326,
      "grad_norm": 6.34375,
      "learning_rate": 1.2957878808744283e-05,
      "loss": 2.0809,
      "num_input_tokens_seen": 55158912,
      "step": 26400
    },
    {
      "epoch": 4.306713435027326,
      "eval_loss": 2.5430989265441895,
      "eval_runtime": 134.9563,
      "eval_samples_per_second": 20.192,
      "eval_steps_per_second": 10.1,
      "num_input_tokens_seen": 55158912,
      "step": 26400
    },
    {
      "epoch": 4.307529162248144,
      "grad_norm": 8.8125,
      "learning_rate": 1.294927624533931e-05,
      "loss": 2.1283,
      "num_input_tokens_seen": 55169312,
      "step": 26405
    },
    {
      "epoch": 4.3083448894689615,
      "grad_norm": 7.53125,
      "learning_rate": 1.2940675540307378e-05,
      "loss": 2.6489,
      "num_input_tokens_seen": 55180208,
      "step": 26410
    },
    {
      "epoch": 4.309160616689779,
      "grad_norm": 8.375,
      "learning_rate": 1.2932076694974814e-05,
      "loss": 3.074,
      "num_input_tokens_seen": 55190016,
      "step": 26415
    },
    {
      "epoch": 4.309976343910597,
      "grad_norm": 9.9375,
      "learning_rate": 1.2923479710667682e-05,
      "loss": 2.5658,
      "num_input_tokens_seen": 55199984,
      "step": 26420
    },
    {
      "epoch": 4.310792071131414,
      "grad_norm": 4.25,
      "learning_rate": 1.2914884588711751e-05,
      "loss": 2.3926,
      "num_input_tokens_seen": 55209968,
      "step": 26425
    },
    {
      "epoch": 4.311607798352231,
      "grad_norm": 7.1875,
      "learning_rate": 1.2906291330432475e-05,
      "loss": 2.2025,
      "num_input_tokens_seen": 55220064,
      "step": 26430
    },
    {
      "epoch": 4.312423525573048,
      "grad_norm": 5.0625,
      "learning_rate": 1.2897699937155055e-05,
      "loss": 2.3129,
      "num_input_tokens_seen": 55231184,
      "step": 26435
    },
    {
      "epoch": 4.313239252793866,
      "grad_norm": 4.65625,
      "learning_rate": 1.2889110410204403e-05,
      "loss": 2.3949,
      "num_input_tokens_seen": 55241264,
      "step": 26440
    },
    {
      "epoch": 4.314054980014683,
      "grad_norm": 7.28125,
      "learning_rate": 1.2880522750905111e-05,
      "loss": 2.4862,
      "num_input_tokens_seen": 55252336,
      "step": 26445
    },
    {
      "epoch": 4.3148707072355,
      "grad_norm": 2.53125,
      "learning_rate": 1.2871936960581523e-05,
      "loss": 2.3529,
      "num_input_tokens_seen": 55262736,
      "step": 26450
    },
    {
      "epoch": 4.3156864344563175,
      "grad_norm": 7.5625,
      "learning_rate": 1.2863353040557658e-05,
      "loss": 3.0663,
      "num_input_tokens_seen": 55272512,
      "step": 26455
    },
    {
      "epoch": 4.3165021616771355,
      "grad_norm": 7.875,
      "learning_rate": 1.2854770992157273e-05,
      "loss": 2.7455,
      "num_input_tokens_seen": 55283200,
      "step": 26460
    },
    {
      "epoch": 4.317317888897953,
      "grad_norm": 7.875,
      "learning_rate": 1.2846190816703835e-05,
      "loss": 3.5423,
      "num_input_tokens_seen": 55293536,
      "step": 26465
    },
    {
      "epoch": 4.31813361611877,
      "grad_norm": 3.625,
      "learning_rate": 1.2837612515520498e-05,
      "loss": 1.0637,
      "num_input_tokens_seen": 55303952,
      "step": 26470
    },
    {
      "epoch": 4.318949343339587,
      "grad_norm": 3.984375,
      "learning_rate": 1.2829036089930163e-05,
      "loss": 1.6944,
      "num_input_tokens_seen": 55315136,
      "step": 26475
    },
    {
      "epoch": 4.319765070560405,
      "grad_norm": 5.8125,
      "learning_rate": 1.2820461541255412e-05,
      "loss": 2.3862,
      "num_input_tokens_seen": 55325264,
      "step": 26480
    },
    {
      "epoch": 4.320580797781222,
      "grad_norm": 3.671875,
      "learning_rate": 1.2811888870818543e-05,
      "loss": 1.4603,
      "num_input_tokens_seen": 55335040,
      "step": 26485
    },
    {
      "epoch": 4.321396525002039,
      "grad_norm": 1.71875,
      "learning_rate": 1.2803318079941581e-05,
      "loss": 1.3448,
      "num_input_tokens_seen": 55345648,
      "step": 26490
    },
    {
      "epoch": 4.322212252222856,
      "grad_norm": 7.34375,
      "learning_rate": 1.2794749169946235e-05,
      "loss": 2.214,
      "num_input_tokens_seen": 55356592,
      "step": 26495
    },
    {
      "epoch": 4.323027979443674,
      "grad_norm": 12.1875,
      "learning_rate": 1.2786182142153952e-05,
      "loss": 3.4936,
      "num_input_tokens_seen": 55368272,
      "step": 26500
    },
    {
      "epoch": 4.3238437066644915,
      "grad_norm": 10.9375,
      "learning_rate": 1.2777616997885878e-05,
      "loss": 2.3098,
      "num_input_tokens_seen": 55379008,
      "step": 26505
    },
    {
      "epoch": 4.324659433885309,
      "grad_norm": 7.4375,
      "learning_rate": 1.2769053738462847e-05,
      "loss": 1.1381,
      "num_input_tokens_seen": 55389216,
      "step": 26510
    },
    {
      "epoch": 4.325475161106126,
      "grad_norm": 5.65625,
      "learning_rate": 1.2760492365205434e-05,
      "loss": 1.8611,
      "num_input_tokens_seen": 55401024,
      "step": 26515
    },
    {
      "epoch": 4.326290888326944,
      "grad_norm": 8.875,
      "learning_rate": 1.2751932879433919e-05,
      "loss": 2.5534,
      "num_input_tokens_seen": 55410432,
      "step": 26520
    },
    {
      "epoch": 4.327106615547761,
      "grad_norm": 8.75,
      "learning_rate": 1.2743375282468267e-05,
      "loss": 3.6348,
      "num_input_tokens_seen": 55420480,
      "step": 26525
    },
    {
      "epoch": 4.327922342768578,
      "grad_norm": 7.5625,
      "learning_rate": 1.2734819575628182e-05,
      "loss": 2.3713,
      "num_input_tokens_seen": 55429792,
      "step": 26530
    },
    {
      "epoch": 4.328738069989395,
      "grad_norm": 11.125,
      "learning_rate": 1.2726265760233039e-05,
      "loss": 2.7636,
      "num_input_tokens_seen": 55439600,
      "step": 26535
    },
    {
      "epoch": 4.329553797210213,
      "grad_norm": 9.1875,
      "learning_rate": 1.271771383760197e-05,
      "loss": 1.8167,
      "num_input_tokens_seen": 55450320,
      "step": 26540
    },
    {
      "epoch": 4.33036952443103,
      "grad_norm": 10.0,
      "learning_rate": 1.2709163809053764e-05,
      "loss": 1.9796,
      "num_input_tokens_seen": 55462160,
      "step": 26545
    },
    {
      "epoch": 4.331185251651847,
      "grad_norm": 3.6875,
      "learning_rate": 1.2700615675906963e-05,
      "loss": 1.4681,
      "num_input_tokens_seen": 55472288,
      "step": 26550
    },
    {
      "epoch": 4.332000978872665,
      "grad_norm": 9.9375,
      "learning_rate": 1.269206943947978e-05,
      "loss": 2.764,
      "num_input_tokens_seen": 55482848,
      "step": 26555
    },
    {
      "epoch": 4.332816706093483,
      "grad_norm": 7.4375,
      "learning_rate": 1.2683525101090177e-05,
      "loss": 2.2095,
      "num_input_tokens_seen": 55492480,
      "step": 26560
    },
    {
      "epoch": 4.3336324333143,
      "grad_norm": 9.375,
      "learning_rate": 1.2674982662055765e-05,
      "loss": 1.9035,
      "num_input_tokens_seen": 55503328,
      "step": 26565
    },
    {
      "epoch": 4.334448160535117,
      "grad_norm": 7.875,
      "learning_rate": 1.2666442123693922e-05,
      "loss": 1.9483,
      "num_input_tokens_seen": 55513392,
      "step": 26570
    },
    {
      "epoch": 4.335263887755934,
      "grad_norm": 4.0,
      "learning_rate": 1.265790348732169e-05,
      "loss": 3.4337,
      "num_input_tokens_seen": 55523808,
      "step": 26575
    },
    {
      "epoch": 4.336079614976752,
      "grad_norm": 10.125,
      "learning_rate": 1.264936675425584e-05,
      "loss": 3.2104,
      "num_input_tokens_seen": 55534256,
      "step": 26580
    },
    {
      "epoch": 4.336895342197569,
      "grad_norm": 11.75,
      "learning_rate": 1.2640831925812852e-05,
      "loss": 2.6843,
      "num_input_tokens_seen": 55545360,
      "step": 26585
    },
    {
      "epoch": 4.337711069418386,
      "grad_norm": 6.46875,
      "learning_rate": 1.263229900330889e-05,
      "loss": 0.9418,
      "num_input_tokens_seen": 55554896,
      "step": 26590
    },
    {
      "epoch": 4.338526796639204,
      "grad_norm": 6.53125,
      "learning_rate": 1.2623767988059843e-05,
      "loss": 1.2915,
      "num_input_tokens_seen": 55565904,
      "step": 26595
    },
    {
      "epoch": 4.339342523860021,
      "grad_norm": 8.5,
      "learning_rate": 1.2615238881381309e-05,
      "loss": 2.8693,
      "num_input_tokens_seen": 55575232,
      "step": 26600
    },
    {
      "epoch": 4.339342523860021,
      "eval_loss": 2.5383238792419434,
      "eval_runtime": 134.8546,
      "eval_samples_per_second": 20.207,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 55575232,
      "step": 26600
    },
    {
      "epoch": 4.3401582510808385,
      "grad_norm": 6.15625,
      "learning_rate": 1.2606711684588568e-05,
      "loss": 2.6183,
      "num_input_tokens_seen": 55586864,
      "step": 26605
    },
    {
      "epoch": 4.340973978301656,
      "grad_norm": 7.1875,
      "learning_rate": 1.2598186398996636e-05,
      "loss": 1.1416,
      "num_input_tokens_seen": 55597040,
      "step": 26610
    },
    {
      "epoch": 4.341789705522474,
      "grad_norm": 6.4375,
      "learning_rate": 1.2589663025920207e-05,
      "loss": 2.8979,
      "num_input_tokens_seen": 55607040,
      "step": 26615
    },
    {
      "epoch": 4.342605432743291,
      "grad_norm": 3.03125,
      "learning_rate": 1.2581141566673705e-05,
      "loss": 2.6637,
      "num_input_tokens_seen": 55617872,
      "step": 26620
    },
    {
      "epoch": 4.343421159964108,
      "grad_norm": 7.59375,
      "learning_rate": 1.257262202257124e-05,
      "loss": 2.0756,
      "num_input_tokens_seen": 55629680,
      "step": 26625
    },
    {
      "epoch": 4.344236887184925,
      "grad_norm": 11.4375,
      "learning_rate": 1.2564104394926618e-05,
      "loss": 2.9739,
      "num_input_tokens_seen": 55641296,
      "step": 26630
    },
    {
      "epoch": 4.345052614405743,
      "grad_norm": 8.125,
      "learning_rate": 1.2555588685053383e-05,
      "loss": 1.4902,
      "num_input_tokens_seen": 55652576,
      "step": 26635
    },
    {
      "epoch": 4.34586834162656,
      "grad_norm": 7.84375,
      "learning_rate": 1.2547074894264762e-05,
      "loss": 1.9406,
      "num_input_tokens_seen": 55662992,
      "step": 26640
    },
    {
      "epoch": 4.346684068847377,
      "grad_norm": 9.5,
      "learning_rate": 1.2538563023873679e-05,
      "loss": 1.5873,
      "num_input_tokens_seen": 55673120,
      "step": 26645
    },
    {
      "epoch": 4.3474997960681945,
      "grad_norm": 10.0,
      "learning_rate": 1.2530053075192789e-05,
      "loss": 2.4964,
      "num_input_tokens_seen": 55683552,
      "step": 26650
    },
    {
      "epoch": 4.3483155232890125,
      "grad_norm": 12.0625,
      "learning_rate": 1.252154504953441e-05,
      "loss": 1.6233,
      "num_input_tokens_seen": 55695008,
      "step": 26655
    },
    {
      "epoch": 4.34913125050983,
      "grad_norm": 2.984375,
      "learning_rate": 1.25130389482106e-05,
      "loss": 2.1137,
      "num_input_tokens_seen": 55705360,
      "step": 26660
    },
    {
      "epoch": 4.349946977730647,
      "grad_norm": 7.65625,
      "learning_rate": 1.2504534772533116e-05,
      "loss": 3.5857,
      "num_input_tokens_seen": 55715680,
      "step": 26665
    },
    {
      "epoch": 4.350762704951464,
      "grad_norm": 2.203125,
      "learning_rate": 1.2496032523813387e-05,
      "loss": 3.2373,
      "num_input_tokens_seen": 55726368,
      "step": 26670
    },
    {
      "epoch": 4.351578432172282,
      "grad_norm": 6.90625,
      "learning_rate": 1.2487532203362576e-05,
      "loss": 2.1156,
      "num_input_tokens_seen": 55736416,
      "step": 26675
    },
    {
      "epoch": 4.352394159393099,
      "grad_norm": 10.9375,
      "learning_rate": 1.247903381249155e-05,
      "loss": 3.6187,
      "num_input_tokens_seen": 55747552,
      "step": 26680
    },
    {
      "epoch": 4.353209886613916,
      "grad_norm": 7.5625,
      "learning_rate": 1.2470537352510853e-05,
      "loss": 2.6886,
      "num_input_tokens_seen": 55758752,
      "step": 26685
    },
    {
      "epoch": 4.354025613834733,
      "grad_norm": 5.34375,
      "learning_rate": 1.2462042824730758e-05,
      "loss": 2.4493,
      "num_input_tokens_seen": 55769520,
      "step": 26690
    },
    {
      "epoch": 4.354841341055551,
      "grad_norm": 6.625,
      "learning_rate": 1.245355023046122e-05,
      "loss": 1.459,
      "num_input_tokens_seen": 55779296,
      "step": 26695
    },
    {
      "epoch": 4.3556570682763684,
      "grad_norm": 1.1875,
      "learning_rate": 1.2445059571011896e-05,
      "loss": 1.3119,
      "num_input_tokens_seen": 55790112,
      "step": 26700
    },
    {
      "epoch": 4.356472795497186,
      "grad_norm": 4.6875,
      "learning_rate": 1.2436570847692173e-05,
      "loss": 3.4496,
      "num_input_tokens_seen": 55800848,
      "step": 26705
    },
    {
      "epoch": 4.357288522718003,
      "grad_norm": 0.142578125,
      "learning_rate": 1.2428084061811096e-05,
      "loss": 1.8793,
      "num_input_tokens_seen": 55810736,
      "step": 26710
    },
    {
      "epoch": 4.358104249938821,
      "grad_norm": 7.0,
      "learning_rate": 1.2419599214677447e-05,
      "loss": 2.6779,
      "num_input_tokens_seen": 55821168,
      "step": 26715
    },
    {
      "epoch": 4.358919977159638,
      "grad_norm": 3.0625,
      "learning_rate": 1.2411116307599702e-05,
      "loss": 2.681,
      "num_input_tokens_seen": 55831744,
      "step": 26720
    },
    {
      "epoch": 4.359735704380455,
      "grad_norm": 14.0,
      "learning_rate": 1.2402635341886016e-05,
      "loss": 1.896,
      "num_input_tokens_seen": 55840800,
      "step": 26725
    },
    {
      "epoch": 4.360551431601272,
      "grad_norm": 12.5,
      "learning_rate": 1.2394156318844278e-05,
      "loss": 2.4084,
      "num_input_tokens_seen": 55851472,
      "step": 26730
    },
    {
      "epoch": 4.36136715882209,
      "grad_norm": 5.46875,
      "learning_rate": 1.2385679239782039e-05,
      "loss": 2.0209,
      "num_input_tokens_seen": 55861216,
      "step": 26735
    },
    {
      "epoch": 4.362182886042907,
      "grad_norm": 4.15625,
      "learning_rate": 1.2377204106006585e-05,
      "loss": 1.6956,
      "num_input_tokens_seen": 55872448,
      "step": 26740
    },
    {
      "epoch": 4.362998613263724,
      "grad_norm": 4.46875,
      "learning_rate": 1.2368730918824891e-05,
      "loss": 1.7769,
      "num_input_tokens_seen": 55883360,
      "step": 26745
    },
    {
      "epoch": 4.3638143404845415,
      "grad_norm": 5.875,
      "learning_rate": 1.236025967954362e-05,
      "loss": 1.3866,
      "num_input_tokens_seen": 55892480,
      "step": 26750
    },
    {
      "epoch": 4.3646300677053596,
      "grad_norm": 3.796875,
      "learning_rate": 1.2351790389469153e-05,
      "loss": 1.7371,
      "num_input_tokens_seen": 55903360,
      "step": 26755
    },
    {
      "epoch": 4.365445794926177,
      "grad_norm": 7.46875,
      "learning_rate": 1.234332304990755e-05,
      "loss": 2.3808,
      "num_input_tokens_seen": 55912720,
      "step": 26760
    },
    {
      "epoch": 4.366261522146994,
      "grad_norm": 11.9375,
      "learning_rate": 1.2334857662164593e-05,
      "loss": 3.2799,
      "num_input_tokens_seen": 55922912,
      "step": 26765
    },
    {
      "epoch": 4.367077249367812,
      "grad_norm": 7.625,
      "learning_rate": 1.2326394227545743e-05,
      "loss": 3.0729,
      "num_input_tokens_seen": 55933952,
      "step": 26770
    },
    {
      "epoch": 4.367892976588629,
      "grad_norm": 8.75,
      "learning_rate": 1.2317932747356162e-05,
      "loss": 3.0676,
      "num_input_tokens_seen": 55943728,
      "step": 26775
    },
    {
      "epoch": 4.368708703809446,
      "grad_norm": 8.375,
      "learning_rate": 1.2309473222900726e-05,
      "loss": 2.0989,
      "num_input_tokens_seen": 55955536,
      "step": 26780
    },
    {
      "epoch": 4.369524431030263,
      "grad_norm": 4.15625,
      "learning_rate": 1.2301015655484006e-05,
      "loss": 1.9864,
      "num_input_tokens_seen": 55966016,
      "step": 26785
    },
    {
      "epoch": 4.370340158251081,
      "grad_norm": 7.4375,
      "learning_rate": 1.2292560046410245e-05,
      "loss": 2.984,
      "num_input_tokens_seen": 55975376,
      "step": 26790
    },
    {
      "epoch": 4.371155885471898,
      "grad_norm": 5.90625,
      "learning_rate": 1.228410639698343e-05,
      "loss": 2.8631,
      "num_input_tokens_seen": 55984848,
      "step": 26795
    },
    {
      "epoch": 4.3719716126927155,
      "grad_norm": 6.59375,
      "learning_rate": 1.2275654708507195e-05,
      "loss": 1.9899,
      "num_input_tokens_seen": 55994160,
      "step": 26800
    },
    {
      "epoch": 4.3719716126927155,
      "eval_loss": 2.5557234287261963,
      "eval_runtime": 134.8218,
      "eval_samples_per_second": 20.212,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 55994160,
      "step": 26800
    },
    {
      "epoch": 4.372787339913533,
      "grad_norm": 4.59375,
      "learning_rate": 1.2267204982284908e-05,
      "loss": 1.653,
      "num_input_tokens_seen": 56004288,
      "step": 26805
    },
    {
      "epoch": 4.373603067134351,
      "grad_norm": 7.90625,
      "learning_rate": 1.2258757219619635e-05,
      "loss": 1.7344,
      "num_input_tokens_seen": 56015504,
      "step": 26810
    },
    {
      "epoch": 4.374418794355168,
      "grad_norm": 14.0,
      "learning_rate": 1.2250311421814104e-05,
      "loss": 2.3234,
      "num_input_tokens_seen": 56025184,
      "step": 26815
    },
    {
      "epoch": 4.375234521575985,
      "grad_norm": 8.125,
      "learning_rate": 1.2241867590170772e-05,
      "loss": 3.2723,
      "num_input_tokens_seen": 56036560,
      "step": 26820
    },
    {
      "epoch": 4.376050248796802,
      "grad_norm": 5.96875,
      "learning_rate": 1.2233425725991799e-05,
      "loss": 2.1071,
      "num_input_tokens_seen": 56046208,
      "step": 26825
    },
    {
      "epoch": 4.37686597601762,
      "grad_norm": 7.8125,
      "learning_rate": 1.2224985830579003e-05,
      "loss": 2.6846,
      "num_input_tokens_seen": 56056560,
      "step": 26830
    },
    {
      "epoch": 4.377681703238437,
      "grad_norm": 2.234375,
      "learning_rate": 1.2216547905233944e-05,
      "loss": 2.0788,
      "num_input_tokens_seen": 56067680,
      "step": 26835
    },
    {
      "epoch": 4.378497430459254,
      "grad_norm": 0.0830078125,
      "learning_rate": 1.2208111951257842e-05,
      "loss": 2.6445,
      "num_input_tokens_seen": 56077904,
      "step": 26840
    },
    {
      "epoch": 4.3793131576800715,
      "grad_norm": 4.03125,
      "learning_rate": 1.2199677969951622e-05,
      "loss": 2.3529,
      "num_input_tokens_seen": 56089232,
      "step": 26845
    },
    {
      "epoch": 4.3801288849008895,
      "grad_norm": 11.0625,
      "learning_rate": 1.2191245962615927e-05,
      "loss": 3.2528,
      "num_input_tokens_seen": 56100208,
      "step": 26850
    },
    {
      "epoch": 4.380944612121707,
      "grad_norm": 4.0625,
      "learning_rate": 1.218281593055106e-05,
      "loss": 1.6481,
      "num_input_tokens_seen": 56109968,
      "step": 26855
    },
    {
      "epoch": 4.381760339342524,
      "grad_norm": 4.46875,
      "learning_rate": 1.217438787505705e-05,
      "loss": 1.4445,
      "num_input_tokens_seen": 56120176,
      "step": 26860
    },
    {
      "epoch": 4.382576066563341,
      "grad_norm": 6.9375,
      "learning_rate": 1.2165961797433615e-05,
      "loss": 3.0079,
      "num_input_tokens_seen": 56131232,
      "step": 26865
    },
    {
      "epoch": 4.383391793784159,
      "grad_norm": 10.4375,
      "learning_rate": 1.215753769898014e-05,
      "loss": 1.6381,
      "num_input_tokens_seen": 56141776,
      "step": 26870
    },
    {
      "epoch": 4.384207521004976,
      "grad_norm": 8.0,
      "learning_rate": 1.2149115580995755e-05,
      "loss": 2.7466,
      "num_input_tokens_seen": 56151232,
      "step": 26875
    },
    {
      "epoch": 4.385023248225793,
      "grad_norm": 6.53125,
      "learning_rate": 1.2140695444779227e-05,
      "loss": 1.6192,
      "num_input_tokens_seen": 56161152,
      "step": 26880
    },
    {
      "epoch": 4.38583897544661,
      "grad_norm": 9.9375,
      "learning_rate": 1.2132277291629066e-05,
      "loss": 2.1711,
      "num_input_tokens_seen": 56172384,
      "step": 26885
    },
    {
      "epoch": 4.386654702667428,
      "grad_norm": 4.375,
      "learning_rate": 1.2123861122843458e-05,
      "loss": 1.457,
      "num_input_tokens_seen": 56182960,
      "step": 26890
    },
    {
      "epoch": 4.387470429888245,
      "grad_norm": 9.9375,
      "learning_rate": 1.2115446939720271e-05,
      "loss": 2.6034,
      "num_input_tokens_seen": 56194512,
      "step": 26895
    },
    {
      "epoch": 4.388286157109063,
      "grad_norm": 9.375,
      "learning_rate": 1.210703474355708e-05,
      "loss": 2.9705,
      "num_input_tokens_seen": 56204816,
      "step": 26900
    },
    {
      "epoch": 4.38910188432988,
      "grad_norm": 9.9375,
      "learning_rate": 1.2098624535651164e-05,
      "loss": 2.42,
      "num_input_tokens_seen": 56216240,
      "step": 26905
    },
    {
      "epoch": 4.389917611550698,
      "grad_norm": 5.0,
      "learning_rate": 1.2090216317299477e-05,
      "loss": 3.1761,
      "num_input_tokens_seen": 56226304,
      "step": 26910
    },
    {
      "epoch": 4.390733338771515,
      "grad_norm": 4.46875,
      "learning_rate": 1.2081810089798668e-05,
      "loss": 1.9562,
      "num_input_tokens_seen": 56237264,
      "step": 26915
    },
    {
      "epoch": 4.391549065992332,
      "grad_norm": 5.96875,
      "learning_rate": 1.2073405854445072e-05,
      "loss": 2.3134,
      "num_input_tokens_seen": 56246848,
      "step": 26920
    },
    {
      "epoch": 4.392364793213149,
      "grad_norm": 6.59375,
      "learning_rate": 1.206500361253474e-05,
      "loss": 1.8387,
      "num_input_tokens_seen": 56256800,
      "step": 26925
    },
    {
      "epoch": 4.393180520433967,
      "grad_norm": 6.0,
      "learning_rate": 1.2056603365363409e-05,
      "loss": 3.0284,
      "num_input_tokens_seen": 56266320,
      "step": 26930
    },
    {
      "epoch": 4.393996247654784,
      "grad_norm": 9.625,
      "learning_rate": 1.2048205114226487e-05,
      "loss": 3.3402,
      "num_input_tokens_seen": 56276896,
      "step": 26935
    },
    {
      "epoch": 4.394811974875601,
      "grad_norm": 13.625,
      "learning_rate": 1.2039808860419102e-05,
      "loss": 2.9582,
      "num_input_tokens_seen": 56287424,
      "step": 26940
    },
    {
      "epoch": 4.395627702096419,
      "grad_norm": 8.1875,
      "learning_rate": 1.2031414605236066e-05,
      "loss": 1.9413,
      "num_input_tokens_seen": 56298624,
      "step": 26945
    },
    {
      "epoch": 4.3964434293172365,
      "grad_norm": 2.84375,
      "learning_rate": 1.2023022349971862e-05,
      "loss": 2.5914,
      "num_input_tokens_seen": 56307536,
      "step": 26950
    },
    {
      "epoch": 4.397259156538054,
      "grad_norm": 6.125,
      "learning_rate": 1.20146320959207e-05,
      "loss": 2.9134,
      "num_input_tokens_seen": 56317504,
      "step": 26955
    },
    {
      "epoch": 4.398074883758871,
      "grad_norm": 7.90625,
      "learning_rate": 1.2006243844376445e-05,
      "loss": 1.5769,
      "num_input_tokens_seen": 56328816,
      "step": 26960
    },
    {
      "epoch": 4.398890610979688,
      "grad_norm": 9.25,
      "learning_rate": 1.1997857596632678e-05,
      "loss": 1.6905,
      "num_input_tokens_seen": 56339744,
      "step": 26965
    },
    {
      "epoch": 4.399706338200506,
      "grad_norm": 9.875,
      "learning_rate": 1.1989473353982672e-05,
      "loss": 3.4085,
      "num_input_tokens_seen": 56348320,
      "step": 26970
    },
    {
      "epoch": 4.400522065421323,
      "grad_norm": 7.75,
      "learning_rate": 1.198109111771937e-05,
      "loss": 3.682,
      "num_input_tokens_seen": 56358512,
      "step": 26975
    },
    {
      "epoch": 4.40133779264214,
      "grad_norm": 10.875,
      "learning_rate": 1.197271088913543e-05,
      "loss": 1.9749,
      "num_input_tokens_seen": 56369104,
      "step": 26980
    },
    {
      "epoch": 4.402153519862958,
      "grad_norm": 3.625,
      "learning_rate": 1.1964332669523182e-05,
      "loss": 3.1665,
      "num_input_tokens_seen": 56379824,
      "step": 26985
    },
    {
      "epoch": 4.402969247083775,
      "grad_norm": 3.1875,
      "learning_rate": 1.1955956460174645e-05,
      "loss": 2.5987,
      "num_input_tokens_seen": 56389936,
      "step": 26990
    },
    {
      "epoch": 4.4037849743045925,
      "grad_norm": 2.703125,
      "learning_rate": 1.1947582262381552e-05,
      "loss": 3.062,
      "num_input_tokens_seen": 56400336,
      "step": 26995
    },
    {
      "epoch": 4.40460070152541,
      "grad_norm": 11.375,
      "learning_rate": 1.1939210077435293e-05,
      "loss": 2.8289,
      "num_input_tokens_seen": 56410736,
      "step": 27000
    },
    {
      "epoch": 4.40460070152541,
      "eval_loss": 2.542567253112793,
      "eval_runtime": 135.0559,
      "eval_samples_per_second": 20.177,
      "eval_steps_per_second": 10.092,
      "num_input_tokens_seen": 56410736,
      "step": 27000
    },
    {
      "epoch": 4.405416428746228,
      "grad_norm": 4.8125,
      "learning_rate": 1.193083990662697e-05,
      "loss": 2.6392,
      "num_input_tokens_seen": 56421248,
      "step": 27005
    },
    {
      "epoch": 4.406232155967045,
      "grad_norm": 4.0,
      "learning_rate": 1.192247175124738e-05,
      "loss": 2.2897,
      "num_input_tokens_seen": 56432560,
      "step": 27010
    },
    {
      "epoch": 4.407047883187862,
      "grad_norm": 8.6875,
      "learning_rate": 1.191410561258698e-05,
      "loss": 2.3512,
      "num_input_tokens_seen": 56444432,
      "step": 27015
    },
    {
      "epoch": 4.407863610408679,
      "grad_norm": 10.25,
      "learning_rate": 1.1905741491935944e-05,
      "loss": 2.0979,
      "num_input_tokens_seen": 56455584,
      "step": 27020
    },
    {
      "epoch": 4.408679337629497,
      "grad_norm": 6.78125,
      "learning_rate": 1.1897379390584129e-05,
      "loss": 3.0597,
      "num_input_tokens_seen": 56465632,
      "step": 27025
    },
    {
      "epoch": 4.409495064850314,
      "grad_norm": 9.0625,
      "learning_rate": 1.1889019309821062e-05,
      "loss": 2.1277,
      "num_input_tokens_seen": 56475120,
      "step": 27030
    },
    {
      "epoch": 4.410310792071131,
      "grad_norm": 8.625,
      "learning_rate": 1.188066125093599e-05,
      "loss": 2.3598,
      "num_input_tokens_seen": 56486160,
      "step": 27035
    },
    {
      "epoch": 4.411126519291948,
      "grad_norm": 7.125,
      "learning_rate": 1.1872305215217811e-05,
      "loss": 2.3317,
      "num_input_tokens_seen": 56495968,
      "step": 27040
    },
    {
      "epoch": 4.4119422465127665,
      "grad_norm": 8.8125,
      "learning_rate": 1.186395120395514e-05,
      "loss": 3.8431,
      "num_input_tokens_seen": 56506464,
      "step": 27045
    },
    {
      "epoch": 4.412757973733584,
      "grad_norm": 11.6875,
      "learning_rate": 1.1855599218436283e-05,
      "loss": 1.8081,
      "num_input_tokens_seen": 56515680,
      "step": 27050
    },
    {
      "epoch": 4.413573700954401,
      "grad_norm": 4.59375,
      "learning_rate": 1.1847249259949209e-05,
      "loss": 2.9268,
      "num_input_tokens_seen": 56526528,
      "step": 27055
    },
    {
      "epoch": 4.414389428175218,
      "grad_norm": 15.4375,
      "learning_rate": 1.1838901329781574e-05,
      "loss": 2.4452,
      "num_input_tokens_seen": 56536720,
      "step": 27060
    },
    {
      "epoch": 4.415205155396036,
      "grad_norm": 3.359375,
      "learning_rate": 1.1830555429220758e-05,
      "loss": 1.697,
      "num_input_tokens_seen": 56546432,
      "step": 27065
    },
    {
      "epoch": 4.416020882616853,
      "grad_norm": 0.265625,
      "learning_rate": 1.1822211559553784e-05,
      "loss": 1.8248,
      "num_input_tokens_seen": 56557248,
      "step": 27070
    },
    {
      "epoch": 4.41683660983767,
      "grad_norm": 15.0,
      "learning_rate": 1.18138697220674e-05,
      "loss": 3.2158,
      "num_input_tokens_seen": 56568784,
      "step": 27075
    },
    {
      "epoch": 4.417652337058487,
      "grad_norm": 4.25,
      "learning_rate": 1.1805529918048e-05,
      "loss": 3.1081,
      "num_input_tokens_seen": 56580256,
      "step": 27080
    },
    {
      "epoch": 4.418468064279305,
      "grad_norm": 10.625,
      "learning_rate": 1.1797192148781702e-05,
      "loss": 2.3387,
      "num_input_tokens_seen": 56590144,
      "step": 27085
    },
    {
      "epoch": 4.419283791500122,
      "grad_norm": 12.125,
      "learning_rate": 1.1788856415554297e-05,
      "loss": 4.6517,
      "num_input_tokens_seen": 56601408,
      "step": 27090
    },
    {
      "epoch": 4.4200995187209395,
      "grad_norm": 1.859375,
      "learning_rate": 1.1780522719651249e-05,
      "loss": 2.5073,
      "num_input_tokens_seen": 56611200,
      "step": 27095
    },
    {
      "epoch": 4.420915245941757,
      "grad_norm": 2.9375,
      "learning_rate": 1.1772191062357721e-05,
      "loss": 1.3269,
      "num_input_tokens_seen": 56622416,
      "step": 27100
    },
    {
      "epoch": 4.421730973162575,
      "grad_norm": 13.75,
      "learning_rate": 1.1763861444958573e-05,
      "loss": 2.1386,
      "num_input_tokens_seen": 56633504,
      "step": 27105
    },
    {
      "epoch": 4.422546700383392,
      "grad_norm": 9.8125,
      "learning_rate": 1.1755533868738317e-05,
      "loss": 2.663,
      "num_input_tokens_seen": 56645328,
      "step": 27110
    },
    {
      "epoch": 4.423362427604209,
      "grad_norm": 4.875,
      "learning_rate": 1.1747208334981185e-05,
      "loss": 2.8413,
      "num_input_tokens_seen": 56655616,
      "step": 27115
    },
    {
      "epoch": 4.424178154825027,
      "grad_norm": 7.46875,
      "learning_rate": 1.1738884844971067e-05,
      "loss": 3.8827,
      "num_input_tokens_seen": 56666256,
      "step": 27120
    },
    {
      "epoch": 4.424993882045844,
      "grad_norm": 4.40625,
      "learning_rate": 1.1730563399991563e-05,
      "loss": 1.9422,
      "num_input_tokens_seen": 56676848,
      "step": 27125
    },
    {
      "epoch": 4.425809609266661,
      "grad_norm": 6.0625,
      "learning_rate": 1.1722244001325938e-05,
      "loss": 2.3056,
      "num_input_tokens_seen": 56687472,
      "step": 27130
    },
    {
      "epoch": 4.426625336487478,
      "grad_norm": 7.3125,
      "learning_rate": 1.1713926650257137e-05,
      "loss": 1.5109,
      "num_input_tokens_seen": 56697760,
      "step": 27135
    },
    {
      "epoch": 4.4274410637082955,
      "grad_norm": 7.375,
      "learning_rate": 1.170561134806781e-05,
      "loss": 1.982,
      "num_input_tokens_seen": 56708736,
      "step": 27140
    },
    {
      "epoch": 4.4282567909291135,
      "grad_norm": 9.125,
      "learning_rate": 1.1697298096040287e-05,
      "loss": 2.8317,
      "num_input_tokens_seen": 56719504,
      "step": 27145
    },
    {
      "epoch": 4.429072518149931,
      "grad_norm": 6.5625,
      "learning_rate": 1.1688986895456567e-05,
      "loss": 2.0861,
      "num_input_tokens_seen": 56728608,
      "step": 27150
    },
    {
      "epoch": 4.429888245370748,
      "grad_norm": 13.4375,
      "learning_rate": 1.1680677747598349e-05,
      "loss": 2.3026,
      "num_input_tokens_seen": 56738656,
      "step": 27155
    },
    {
      "epoch": 4.430703972591566,
      "grad_norm": 6.90625,
      "learning_rate": 1.1672370653746995e-05,
      "loss": 1.1011,
      "num_input_tokens_seen": 56749344,
      "step": 27160
    },
    {
      "epoch": 4.431519699812383,
      "grad_norm": 3.171875,
      "learning_rate": 1.166406561518357e-05,
      "loss": 2.1379,
      "num_input_tokens_seen": 56759008,
      "step": 27165
    },
    {
      "epoch": 4.4323354270332,
      "grad_norm": 4.59375,
      "learning_rate": 1.1655762633188826e-05,
      "loss": 1.803,
      "num_input_tokens_seen": 56769616,
      "step": 27170
    },
    {
      "epoch": 4.433151154254017,
      "grad_norm": 8.0,
      "learning_rate": 1.1647461709043172e-05,
      "loss": 2.0809,
      "num_input_tokens_seen": 56781232,
      "step": 27175
    },
    {
      "epoch": 4.433966881474835,
      "grad_norm": 8.25,
      "learning_rate": 1.1639162844026722e-05,
      "loss": 1.4889,
      "num_input_tokens_seen": 56793568,
      "step": 27180
    },
    {
      "epoch": 4.434782608695652,
      "grad_norm": 3.90625,
      "learning_rate": 1.163086603941927e-05,
      "loss": 1.2822,
      "num_input_tokens_seen": 56805280,
      "step": 27185
    },
    {
      "epoch": 4.4355983359164695,
      "grad_norm": 7.03125,
      "learning_rate": 1.1622571296500273e-05,
      "loss": 2.8293,
      "num_input_tokens_seen": 56816720,
      "step": 27190
    },
    {
      "epoch": 4.436414063137287,
      "grad_norm": 11.5,
      "learning_rate": 1.1614278616548904e-05,
      "loss": 2.4947,
      "num_input_tokens_seen": 56828112,
      "step": 27195
    },
    {
      "epoch": 4.437229790358105,
      "grad_norm": 6.21875,
      "learning_rate": 1.1605988000843986e-05,
      "loss": 2.7897,
      "num_input_tokens_seen": 56838864,
      "step": 27200
    },
    {
      "epoch": 4.437229790358105,
      "eval_loss": 2.5423049926757812,
      "eval_runtime": 134.8218,
      "eval_samples_per_second": 20.212,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 56838864,
      "step": 27200
    },
    {
      "epoch": 4.438045517578922,
      "grad_norm": 8.8125,
      "learning_rate": 1.1597699450664028e-05,
      "loss": 3.2075,
      "num_input_tokens_seen": 56848512,
      "step": 27205
    },
    {
      "epoch": 4.438861244799739,
      "grad_norm": 5.46875,
      "learning_rate": 1.1589412967287252e-05,
      "loss": 3.2601,
      "num_input_tokens_seen": 56857600,
      "step": 27210
    },
    {
      "epoch": 4.439676972020556,
      "grad_norm": 6.03125,
      "learning_rate": 1.1581128551991514e-05,
      "loss": 1.0898,
      "num_input_tokens_seen": 56867824,
      "step": 27215
    },
    {
      "epoch": 4.440492699241374,
      "grad_norm": 8.25,
      "learning_rate": 1.1572846206054383e-05,
      "loss": 2.5979,
      "num_input_tokens_seen": 56877632,
      "step": 27220
    },
    {
      "epoch": 4.441308426462191,
      "grad_norm": 4.65625,
      "learning_rate": 1.1564565930753113e-05,
      "loss": 2.7843,
      "num_input_tokens_seen": 56889136,
      "step": 27225
    },
    {
      "epoch": 4.442124153683008,
      "grad_norm": 7.65625,
      "learning_rate": 1.1556287727364606e-05,
      "loss": 2.6444,
      "num_input_tokens_seen": 56900160,
      "step": 27230
    },
    {
      "epoch": 4.442939880903825,
      "grad_norm": 3.9375,
      "learning_rate": 1.1548011597165489e-05,
      "loss": 1.407,
      "num_input_tokens_seen": 56910752,
      "step": 27235
    },
    {
      "epoch": 4.443755608124643,
      "grad_norm": 5.96875,
      "learning_rate": 1.1539737541432019e-05,
      "loss": 1.0073,
      "num_input_tokens_seen": 56920096,
      "step": 27240
    },
    {
      "epoch": 4.444571335345461,
      "grad_norm": 2.671875,
      "learning_rate": 1.1531465561440174e-05,
      "loss": 1.4787,
      "num_input_tokens_seen": 56929808,
      "step": 27245
    },
    {
      "epoch": 4.445387062566278,
      "grad_norm": 7.9375,
      "learning_rate": 1.1523195658465605e-05,
      "loss": 2.346,
      "num_input_tokens_seen": 56940016,
      "step": 27250
    },
    {
      "epoch": 4.446202789787095,
      "grad_norm": 2.15625,
      "learning_rate": 1.1514927833783618e-05,
      "loss": 1.5121,
      "num_input_tokens_seen": 56950176,
      "step": 27255
    },
    {
      "epoch": 4.447018517007913,
      "grad_norm": 9.0625,
      "learning_rate": 1.150666208866922e-05,
      "loss": 3.3724,
      "num_input_tokens_seen": 56960656,
      "step": 27260
    },
    {
      "epoch": 4.44783424422873,
      "grad_norm": 10.625,
      "learning_rate": 1.1498398424397106e-05,
      "loss": 3.8869,
      "num_input_tokens_seen": 56970800,
      "step": 27265
    },
    {
      "epoch": 4.448649971449547,
      "grad_norm": 5.84375,
      "learning_rate": 1.1490136842241628e-05,
      "loss": 2.2372,
      "num_input_tokens_seen": 56981984,
      "step": 27270
    },
    {
      "epoch": 4.449465698670364,
      "grad_norm": 10.1875,
      "learning_rate": 1.1481877343476813e-05,
      "loss": 2.0308,
      "num_input_tokens_seen": 56991312,
      "step": 27275
    },
    {
      "epoch": 4.450281425891182,
      "grad_norm": 11.25,
      "learning_rate": 1.14736199293764e-05,
      "loss": 2.7058,
      "num_input_tokens_seen": 57002032,
      "step": 27280
    },
    {
      "epoch": 4.451097153111999,
      "grad_norm": 8.9375,
      "learning_rate": 1.1465364601213771e-05,
      "loss": 3.0221,
      "num_input_tokens_seen": 57013296,
      "step": 27285
    },
    {
      "epoch": 4.4519128803328165,
      "grad_norm": 8.0,
      "learning_rate": 1.1457111360262012e-05,
      "loss": 2.1033,
      "num_input_tokens_seen": 57023104,
      "step": 27290
    },
    {
      "epoch": 4.4527286075536345,
      "grad_norm": 5.75,
      "learning_rate": 1.1448860207793869e-05,
      "loss": 1.8154,
      "num_input_tokens_seen": 57033696,
      "step": 27295
    },
    {
      "epoch": 4.453544334774452,
      "grad_norm": 5.09375,
      "learning_rate": 1.144061114508177e-05,
      "loss": 2.3669,
      "num_input_tokens_seen": 57044320,
      "step": 27300
    },
    {
      "epoch": 4.454360061995269,
      "grad_norm": 16.625,
      "learning_rate": 1.1432364173397842e-05,
      "loss": 2.4075,
      "num_input_tokens_seen": 57053712,
      "step": 27305
    },
    {
      "epoch": 4.455175789216086,
      "grad_norm": 4.0625,
      "learning_rate": 1.1424119294013852e-05,
      "loss": 1.3778,
      "num_input_tokens_seen": 57063376,
      "step": 27310
    },
    {
      "epoch": 4.455991516436903,
      "grad_norm": 3.046875,
      "learning_rate": 1.1415876508201279e-05,
      "loss": 1.5237,
      "num_input_tokens_seen": 57074512,
      "step": 27315
    },
    {
      "epoch": 4.456807243657721,
      "grad_norm": 5.6875,
      "learning_rate": 1.140763581723125e-05,
      "loss": 2.4006,
      "num_input_tokens_seen": 57083952,
      "step": 27320
    },
    {
      "epoch": 4.457622970878538,
      "grad_norm": 13.5625,
      "learning_rate": 1.1399397222374588e-05,
      "loss": 2.9348,
      "num_input_tokens_seen": 57094256,
      "step": 27325
    },
    {
      "epoch": 4.458438698099355,
      "grad_norm": 2.875,
      "learning_rate": 1.1391160724901804e-05,
      "loss": 1.8388,
      "num_input_tokens_seen": 57106000,
      "step": 27330
    },
    {
      "epoch": 4.459254425320173,
      "grad_norm": 7.90625,
      "learning_rate": 1.138292632608304e-05,
      "loss": 2.464,
      "num_input_tokens_seen": 57115504,
      "step": 27335
    },
    {
      "epoch": 4.4600701525409905,
      "grad_norm": 8.0,
      "learning_rate": 1.1374694027188174e-05,
      "loss": 3.3646,
      "num_input_tokens_seen": 57125808,
      "step": 27340
    },
    {
      "epoch": 4.460885879761808,
      "grad_norm": 8.0,
      "learning_rate": 1.1366463829486711e-05,
      "loss": 1.9847,
      "num_input_tokens_seen": 57134688,
      "step": 27345
    },
    {
      "epoch": 4.461701606982625,
      "grad_norm": 5.71875,
      "learning_rate": 1.1358235734247849e-05,
      "loss": 1.9088,
      "num_input_tokens_seen": 57144032,
      "step": 27350
    },
    {
      "epoch": 4.462517334203443,
      "grad_norm": 5.46875,
      "learning_rate": 1.1350009742740478e-05,
      "loss": 1.8997,
      "num_input_tokens_seen": 57153712,
      "step": 27355
    },
    {
      "epoch": 4.46333306142426,
      "grad_norm": 9.0,
      "learning_rate": 1.134178585623313e-05,
      "loss": 2.6563,
      "num_input_tokens_seen": 57164560,
      "step": 27360
    },
    {
      "epoch": 4.464148788645077,
      "grad_norm": 4.8125,
      "learning_rate": 1.1333564075994047e-05,
      "loss": 2.5535,
      "num_input_tokens_seen": 57174048,
      "step": 27365
    },
    {
      "epoch": 4.464964515865894,
      "grad_norm": 5.90625,
      "learning_rate": 1.1325344403291133e-05,
      "loss": 3.1231,
      "num_input_tokens_seen": 57184304,
      "step": 27370
    },
    {
      "epoch": 4.465780243086712,
      "grad_norm": 17.875,
      "learning_rate": 1.1317126839391951e-05,
      "loss": 1.8863,
      "num_input_tokens_seen": 57194240,
      "step": 27375
    },
    {
      "epoch": 4.466595970307529,
      "grad_norm": 0.08642578125,
      "learning_rate": 1.1308911385563766e-05,
      "loss": 2.3274,
      "num_input_tokens_seen": 57202928,
      "step": 27380
    },
    {
      "epoch": 4.467411697528346,
      "grad_norm": 11.125,
      "learning_rate": 1.1300698043073494e-05,
      "loss": 3.1338,
      "num_input_tokens_seen": 57212272,
      "step": 27385
    },
    {
      "epoch": 4.468227424749164,
      "grad_norm": 7.8125,
      "learning_rate": 1.1292486813187736e-05,
      "loss": 1.3611,
      "num_input_tokens_seen": 57223424,
      "step": 27390
    },
    {
      "epoch": 4.469043151969982,
      "grad_norm": 10.375,
      "learning_rate": 1.1284277697172782e-05,
      "loss": 3.5807,
      "num_input_tokens_seen": 57234608,
      "step": 27395
    },
    {
      "epoch": 4.469858879190799,
      "grad_norm": 7.25,
      "learning_rate": 1.127607069629456e-05,
      "loss": 1.8468,
      "num_input_tokens_seen": 57245776,
      "step": 27400
    },
    {
      "epoch": 4.469858879190799,
      "eval_loss": 2.5330114364624023,
      "eval_runtime": 134.9853,
      "eval_samples_per_second": 20.187,
      "eval_steps_per_second": 10.097,
      "num_input_tokens_seen": 57245776,
      "step": 27400
    },
    {
      "epoch": 4.470674606411616,
      "grad_norm": 13.4375,
      "learning_rate": 1.1267865811818701e-05,
      "loss": 1.8755,
      "num_input_tokens_seen": 57256528,
      "step": 27405
    },
    {
      "epoch": 4.471490333632433,
      "grad_norm": 9.1875,
      "learning_rate": 1.1259663045010513e-05,
      "loss": 2.423,
      "num_input_tokens_seen": 57266272,
      "step": 27410
    },
    {
      "epoch": 4.472306060853251,
      "grad_norm": 15.3125,
      "learning_rate": 1.1251462397134957e-05,
      "loss": 2.6156,
      "num_input_tokens_seen": 57277008,
      "step": 27415
    },
    {
      "epoch": 4.473121788074068,
      "grad_norm": 0.13671875,
      "learning_rate": 1.1243263869456664e-05,
      "loss": 0.857,
      "num_input_tokens_seen": 57287632,
      "step": 27420
    },
    {
      "epoch": 4.473937515294885,
      "grad_norm": 6.53125,
      "learning_rate": 1.1235067463239967e-05,
      "loss": 2.1285,
      "num_input_tokens_seen": 57297104,
      "step": 27425
    },
    {
      "epoch": 4.474753242515702,
      "grad_norm": 10.0,
      "learning_rate": 1.122687317974884e-05,
      "loss": 3.6802,
      "num_input_tokens_seen": 57305968,
      "step": 27430
    },
    {
      "epoch": 4.47556896973652,
      "grad_norm": 3.9375,
      "learning_rate": 1.1218681020246963e-05,
      "loss": 1.8872,
      "num_input_tokens_seen": 57316400,
      "step": 27435
    },
    {
      "epoch": 4.4763846969573375,
      "grad_norm": 3.65625,
      "learning_rate": 1.1210490985997652e-05,
      "loss": 2.7732,
      "num_input_tokens_seen": 57327376,
      "step": 27440
    },
    {
      "epoch": 4.477200424178155,
      "grad_norm": 5.75,
      "learning_rate": 1.1202303078263917e-05,
      "loss": 1.554,
      "num_input_tokens_seen": 57336704,
      "step": 27445
    },
    {
      "epoch": 4.478016151398972,
      "grad_norm": 6.96875,
      "learning_rate": 1.1194117298308451e-05,
      "loss": 2.4769,
      "num_input_tokens_seen": 57347056,
      "step": 27450
    },
    {
      "epoch": 4.47883187861979,
      "grad_norm": 4.90625,
      "learning_rate": 1.1185933647393585e-05,
      "loss": 3.1823,
      "num_input_tokens_seen": 57357792,
      "step": 27455
    },
    {
      "epoch": 4.479647605840607,
      "grad_norm": 3.90625,
      "learning_rate": 1.1177752126781354e-05,
      "loss": 2.6147,
      "num_input_tokens_seen": 57368576,
      "step": 27460
    },
    {
      "epoch": 4.480463333061424,
      "grad_norm": 4.15625,
      "learning_rate": 1.1169572737733441e-05,
      "loss": 2.5051,
      "num_input_tokens_seen": 57379616,
      "step": 27465
    },
    {
      "epoch": 4.481279060282241,
      "grad_norm": 7.0625,
      "learning_rate": 1.1161395481511216e-05,
      "loss": 1.526,
      "num_input_tokens_seen": 57389840,
      "step": 27470
    },
    {
      "epoch": 4.482094787503059,
      "grad_norm": 1.390625,
      "learning_rate": 1.1153220359375722e-05,
      "loss": 2.0087,
      "num_input_tokens_seen": 57399840,
      "step": 27475
    },
    {
      "epoch": 4.482910514723876,
      "grad_norm": 7.125,
      "learning_rate": 1.114504737258765e-05,
      "loss": 3.0372,
      "num_input_tokens_seen": 57409568,
      "step": 27480
    },
    {
      "epoch": 4.4837262419446935,
      "grad_norm": 3.109375,
      "learning_rate": 1.1136876522407393e-05,
      "loss": 1.6737,
      "num_input_tokens_seen": 57418560,
      "step": 27485
    },
    {
      "epoch": 4.484541969165511,
      "grad_norm": 10.0625,
      "learning_rate": 1.1128707810094985e-05,
      "loss": 3.1618,
      "num_input_tokens_seen": 57427904,
      "step": 27490
    },
    {
      "epoch": 4.485357696386329,
      "grad_norm": 6.84375,
      "learning_rate": 1.1120541236910157e-05,
      "loss": 2.0874,
      "num_input_tokens_seen": 57437408,
      "step": 27495
    },
    {
      "epoch": 4.486173423607146,
      "grad_norm": 3.609375,
      "learning_rate": 1.111237680411229e-05,
      "loss": 2.8262,
      "num_input_tokens_seen": 57447664,
      "step": 27500
    },
    {
      "epoch": 4.486989150827963,
      "grad_norm": 5.875,
      "learning_rate": 1.1104214512960433e-05,
      "loss": 1.4062,
      "num_input_tokens_seen": 57457424,
      "step": 27505
    },
    {
      "epoch": 4.487804878048781,
      "grad_norm": 4.96875,
      "learning_rate": 1.1096054364713327e-05,
      "loss": 2.5104,
      "num_input_tokens_seen": 57466784,
      "step": 27510
    },
    {
      "epoch": 4.488620605269598,
      "grad_norm": 4.59375,
      "learning_rate": 1.1087896360629371e-05,
      "loss": 1.7033,
      "num_input_tokens_seen": 57475296,
      "step": 27515
    },
    {
      "epoch": 4.489436332490415,
      "grad_norm": 10.9375,
      "learning_rate": 1.107974050196662e-05,
      "loss": 2.4584,
      "num_input_tokens_seen": 57486400,
      "step": 27520
    },
    {
      "epoch": 4.490252059711232,
      "grad_norm": 6.4375,
      "learning_rate": 1.1071586789982816e-05,
      "loss": 2.2201,
      "num_input_tokens_seen": 57498304,
      "step": 27525
    },
    {
      "epoch": 4.49106778693205,
      "grad_norm": 11.375,
      "learning_rate": 1.1063435225935373e-05,
      "loss": 2.7589,
      "num_input_tokens_seen": 57507664,
      "step": 27530
    },
    {
      "epoch": 4.4918835141528675,
      "grad_norm": 5.65625,
      "learning_rate": 1.1055285811081348e-05,
      "loss": 2.1304,
      "num_input_tokens_seen": 57518720,
      "step": 27535
    },
    {
      "epoch": 4.492699241373685,
      "grad_norm": 7.59375,
      "learning_rate": 1.1047138546677499e-05,
      "loss": 2.0693,
      "num_input_tokens_seen": 57528496,
      "step": 27540
    },
    {
      "epoch": 4.493514968594502,
      "grad_norm": 7.75,
      "learning_rate": 1.1038993433980219e-05,
      "loss": 2.4205,
      "num_input_tokens_seen": 57538416,
      "step": 27545
    },
    {
      "epoch": 4.49433069581532,
      "grad_norm": 7.59375,
      "learning_rate": 1.1030850474245597e-05,
      "loss": 4.3318,
      "num_input_tokens_seen": 57548560,
      "step": 27550
    },
    {
      "epoch": 4.495146423036137,
      "grad_norm": 13.25,
      "learning_rate": 1.102270966872939e-05,
      "loss": 2.3987,
      "num_input_tokens_seen": 57558400,
      "step": 27555
    },
    {
      "epoch": 4.495962150256954,
      "grad_norm": 6.1875,
      "learning_rate": 1.1014571018687e-05,
      "loss": 2.2123,
      "num_input_tokens_seen": 57568496,
      "step": 27560
    },
    {
      "epoch": 4.496777877477771,
      "grad_norm": 8.75,
      "learning_rate": 1.1006434525373502e-05,
      "loss": 2.4697,
      "num_input_tokens_seen": 57578704,
      "step": 27565
    },
    {
      "epoch": 4.497593604698589,
      "grad_norm": 9.375,
      "learning_rate": 1.0998300190043664e-05,
      "loss": 1.7434,
      "num_input_tokens_seen": 57590128,
      "step": 27570
    },
    {
      "epoch": 4.498409331919406,
      "grad_norm": 1.6796875,
      "learning_rate": 1.0990168013951882e-05,
      "loss": 2.6784,
      "num_input_tokens_seen": 57601472,
      "step": 27575
    },
    {
      "epoch": 4.499225059140223,
      "grad_norm": 7.59375,
      "learning_rate": 1.0982037998352263e-05,
      "loss": 1.3403,
      "num_input_tokens_seen": 57611648,
      "step": 27580
    },
    {
      "epoch": 4.5000407863610405,
      "grad_norm": 9.125,
      "learning_rate": 1.0973910144498534e-05,
      "loss": 3.3324,
      "num_input_tokens_seen": 57622560,
      "step": 27585
    },
    {
      "epoch": 4.500856513581859,
      "grad_norm": 6.96875,
      "learning_rate": 1.0965784453644123e-05,
      "loss": 1.8981,
      "num_input_tokens_seen": 57632656,
      "step": 27590
    },
    {
      "epoch": 4.501672240802676,
      "grad_norm": 9.8125,
      "learning_rate": 1.0957660927042127e-05,
      "loss": 3.1912,
      "num_input_tokens_seen": 57642704,
      "step": 27595
    },
    {
      "epoch": 4.502487968023493,
      "grad_norm": 5.125,
      "learning_rate": 1.094953956594527e-05,
      "loss": 2.2285,
      "num_input_tokens_seen": 57651824,
      "step": 27600
    },
    {
      "epoch": 4.502487968023493,
      "eval_loss": 2.5300850868225098,
      "eval_runtime": 134.9579,
      "eval_samples_per_second": 20.191,
      "eval_steps_per_second": 10.099,
      "num_input_tokens_seen": 57651824,
      "step": 27600
    },
    {
      "epoch": 4.50330369524431,
      "grad_norm": 4.875,
      "learning_rate": 1.0941420371605981e-05,
      "loss": 2.9662,
      "num_input_tokens_seen": 57662976,
      "step": 27605
    },
    {
      "epoch": 4.504119422465128,
      "grad_norm": 7.46875,
      "learning_rate": 1.0933303345276354e-05,
      "loss": 2.599,
      "num_input_tokens_seen": 57673952,
      "step": 27610
    },
    {
      "epoch": 4.504935149685945,
      "grad_norm": 8.0625,
      "learning_rate": 1.0925188488208112e-05,
      "loss": 2.3898,
      "num_input_tokens_seen": 57684704,
      "step": 27615
    },
    {
      "epoch": 4.505750876906762,
      "grad_norm": 13.3125,
      "learning_rate": 1.0917075801652694e-05,
      "loss": 2.9725,
      "num_input_tokens_seen": 57695008,
      "step": 27620
    },
    {
      "epoch": 4.506566604127579,
      "grad_norm": 4.84375,
      "learning_rate": 1.0908965286861151e-05,
      "loss": 1.3661,
      "num_input_tokens_seen": 57704624,
      "step": 27625
    },
    {
      "epoch": 4.507382331348397,
      "grad_norm": 9.3125,
      "learning_rate": 1.090085694508425e-05,
      "loss": 3.6842,
      "num_input_tokens_seen": 57714224,
      "step": 27630
    },
    {
      "epoch": 4.5081980585692145,
      "grad_norm": 8.625,
      "learning_rate": 1.089275077757238e-05,
      "loss": 3.1895,
      "num_input_tokens_seen": 57723680,
      "step": 27635
    },
    {
      "epoch": 4.509013785790032,
      "grad_norm": 10.1875,
      "learning_rate": 1.0884646785575633e-05,
      "loss": 1.9219,
      "num_input_tokens_seen": 57732576,
      "step": 27640
    },
    {
      "epoch": 4.50982951301085,
      "grad_norm": 7.40625,
      "learning_rate": 1.0876544970343728e-05,
      "loss": 2.692,
      "num_input_tokens_seen": 57741872,
      "step": 27645
    },
    {
      "epoch": 4.510645240231667,
      "grad_norm": 1.8671875,
      "learning_rate": 1.0868445333126082e-05,
      "loss": 2.8436,
      "num_input_tokens_seen": 57752272,
      "step": 27650
    },
    {
      "epoch": 4.511460967452484,
      "grad_norm": 5.1875,
      "learning_rate": 1.0860347875171745e-05,
      "loss": 1.2589,
      "num_input_tokens_seen": 57761488,
      "step": 27655
    },
    {
      "epoch": 4.512276694673301,
      "grad_norm": 3.578125,
      "learning_rate": 1.0852252597729465e-05,
      "loss": 1.6207,
      "num_input_tokens_seen": 57770704,
      "step": 27660
    },
    {
      "epoch": 4.513092421894118,
      "grad_norm": 9.4375,
      "learning_rate": 1.0844159502047615e-05,
      "loss": 2.4946,
      "num_input_tokens_seen": 57781632,
      "step": 27665
    },
    {
      "epoch": 4.513908149114936,
      "grad_norm": 4.09375,
      "learning_rate": 1.0836068589374265e-05,
      "loss": 1.6343,
      "num_input_tokens_seen": 57791264,
      "step": 27670
    },
    {
      "epoch": 4.514723876335753,
      "grad_norm": 12.6875,
      "learning_rate": 1.0827979860957144e-05,
      "loss": 4.1241,
      "num_input_tokens_seen": 57803632,
      "step": 27675
    },
    {
      "epoch": 4.5155396035565705,
      "grad_norm": 8.0625,
      "learning_rate": 1.0819893318043615e-05,
      "loss": 2.1605,
      "num_input_tokens_seen": 57813504,
      "step": 27680
    },
    {
      "epoch": 4.5163553307773885,
      "grad_norm": 5.25,
      "learning_rate": 1.0811808961880734e-05,
      "loss": 1.9107,
      "num_input_tokens_seen": 57824816,
      "step": 27685
    },
    {
      "epoch": 4.517171057998206,
      "grad_norm": 7.65625,
      "learning_rate": 1.080372679371522e-05,
      "loss": 2.653,
      "num_input_tokens_seen": 57833760,
      "step": 27690
    },
    {
      "epoch": 4.517986785219023,
      "grad_norm": 10.0,
      "learning_rate": 1.0795646814793428e-05,
      "loss": 2.8833,
      "num_input_tokens_seen": 57843328,
      "step": 27695
    },
    {
      "epoch": 4.51880251243984,
      "grad_norm": 3.203125,
      "learning_rate": 1.078756902636141e-05,
      "loss": 2.2048,
      "num_input_tokens_seen": 57853504,
      "step": 27700
    },
    {
      "epoch": 4.519618239660657,
      "grad_norm": 9.625,
      "learning_rate": 1.077949342966485e-05,
      "loss": 2.8611,
      "num_input_tokens_seen": 57863632,
      "step": 27705
    },
    {
      "epoch": 4.520433966881475,
      "grad_norm": 5.84375,
      "learning_rate": 1.0771420025949103e-05,
      "loss": 2.8812,
      "num_input_tokens_seen": 57872704,
      "step": 27710
    },
    {
      "epoch": 4.521249694102292,
      "grad_norm": 4.875,
      "learning_rate": 1.0763348816459204e-05,
      "loss": 1.882,
      "num_input_tokens_seen": 57882272,
      "step": 27715
    },
    {
      "epoch": 4.522065421323109,
      "grad_norm": 4.65625,
      "learning_rate": 1.0755279802439816e-05,
      "loss": 2.4799,
      "num_input_tokens_seen": 57893712,
      "step": 27720
    },
    {
      "epoch": 4.522881148543927,
      "grad_norm": 10.9375,
      "learning_rate": 1.0747212985135293e-05,
      "loss": 2.1825,
      "num_input_tokens_seen": 57903520,
      "step": 27725
    },
    {
      "epoch": 4.523696875764744,
      "grad_norm": 8.875,
      "learning_rate": 1.073914836578965e-05,
      "loss": 2.1964,
      "num_input_tokens_seen": 57913808,
      "step": 27730
    },
    {
      "epoch": 4.524512602985562,
      "grad_norm": 5.65625,
      "learning_rate": 1.0731085945646529e-05,
      "loss": 1.7204,
      "num_input_tokens_seen": 57924064,
      "step": 27735
    },
    {
      "epoch": 4.525328330206379,
      "grad_norm": 2.078125,
      "learning_rate": 1.0723025725949285e-05,
      "loss": 0.9123,
      "num_input_tokens_seen": 57935104,
      "step": 27740
    },
    {
      "epoch": 4.526144057427197,
      "grad_norm": 0.056640625,
      "learning_rate": 1.0714967707940875e-05,
      "loss": 2.1329,
      "num_input_tokens_seen": 57946800,
      "step": 27745
    },
    {
      "epoch": 4.526959784648014,
      "grad_norm": 6.25,
      "learning_rate": 1.0706911892863963e-05,
      "loss": 1.6029,
      "num_input_tokens_seen": 57956608,
      "step": 27750
    },
    {
      "epoch": 4.527775511868831,
      "grad_norm": 12.5625,
      "learning_rate": 1.0698858281960866e-05,
      "loss": 2.3899,
      "num_input_tokens_seen": 57966464,
      "step": 27755
    },
    {
      "epoch": 4.528591239089648,
      "grad_norm": 10.875,
      "learning_rate": 1.069080687647353e-05,
      "loss": 1.6803,
      "num_input_tokens_seen": 57977344,
      "step": 27760
    },
    {
      "epoch": 4.529406966310466,
      "grad_norm": 5.46875,
      "learning_rate": 1.0682757677643596e-05,
      "loss": 2.0569,
      "num_input_tokens_seen": 57986752,
      "step": 27765
    },
    {
      "epoch": 4.530222693531283,
      "grad_norm": 6.09375,
      "learning_rate": 1.0674710686712359e-05,
      "loss": 2.6496,
      "num_input_tokens_seen": 57997104,
      "step": 27770
    },
    {
      "epoch": 4.5310384207521,
      "grad_norm": 5.9375,
      "learning_rate": 1.0666665904920756e-05,
      "loss": 2.3924,
      "num_input_tokens_seen": 58008464,
      "step": 27775
    },
    {
      "epoch": 4.5318541479729175,
      "grad_norm": 4.53125,
      "learning_rate": 1.0658623333509385e-05,
      "loss": 2.9048,
      "num_input_tokens_seen": 58019488,
      "step": 27780
    },
    {
      "epoch": 4.5326698751937355,
      "grad_norm": 5.84375,
      "learning_rate": 1.0650582973718532e-05,
      "loss": 2.031,
      "num_input_tokens_seen": 58029504,
      "step": 27785
    },
    {
      "epoch": 4.533485602414553,
      "grad_norm": 5.28125,
      "learning_rate": 1.0642544826788098e-05,
      "loss": 2.1859,
      "num_input_tokens_seen": 58039744,
      "step": 27790
    },
    {
      "epoch": 4.53430132963537,
      "grad_norm": 0.31640625,
      "learning_rate": 1.063450889395769e-05,
      "loss": 1.9924,
      "num_input_tokens_seen": 58050432,
      "step": 27795
    },
    {
      "epoch": 4.535117056856187,
      "grad_norm": 7.59375,
      "learning_rate": 1.062647517646653e-05,
      "loss": 2.8184,
      "num_input_tokens_seen": 58060672,
      "step": 27800
    },
    {
      "epoch": 4.535117056856187,
      "eval_loss": 2.5348377227783203,
      "eval_runtime": 134.6924,
      "eval_samples_per_second": 20.231,
      "eval_steps_per_second": 10.119,
      "num_input_tokens_seen": 58060672,
      "step": 27800
    },
    {
      "epoch": 4.535932784077005,
      "grad_norm": 7.4375,
      "learning_rate": 1.0618443675553527e-05,
      "loss": 2.6065,
      "num_input_tokens_seen": 58070560,
      "step": 27805
    },
    {
      "epoch": 4.536748511297822,
      "grad_norm": 9.5,
      "learning_rate": 1.0610414392457247e-05,
      "loss": 2.2254,
      "num_input_tokens_seen": 58079824,
      "step": 27810
    },
    {
      "epoch": 4.537564238518639,
      "grad_norm": 6.09375,
      "learning_rate": 1.0602387328415888e-05,
      "loss": 1.8626,
      "num_input_tokens_seen": 58092144,
      "step": 27815
    },
    {
      "epoch": 4.538379965739456,
      "grad_norm": 11.625,
      "learning_rate": 1.0594362484667347e-05,
      "loss": 2.4524,
      "num_input_tokens_seen": 58103568,
      "step": 27820
    },
    {
      "epoch": 4.539195692960274,
      "grad_norm": 5.71875,
      "learning_rate": 1.0586339862449132e-05,
      "loss": 2.6355,
      "num_input_tokens_seen": 58113744,
      "step": 27825
    },
    {
      "epoch": 4.5400114201810915,
      "grad_norm": 6.65625,
      "learning_rate": 1.0578319462998445e-05,
      "loss": 1.3788,
      "num_input_tokens_seen": 58123520,
      "step": 27830
    },
    {
      "epoch": 4.540827147401909,
      "grad_norm": 8.75,
      "learning_rate": 1.057030128755214e-05,
      "loss": 2.1747,
      "num_input_tokens_seen": 58134720,
      "step": 27835
    },
    {
      "epoch": 4.541642874622726,
      "grad_norm": 4.21875,
      "learning_rate": 1.0562285337346703e-05,
      "loss": 2.6409,
      "num_input_tokens_seen": 58143776,
      "step": 27840
    },
    {
      "epoch": 4.542458601843544,
      "grad_norm": 8.25,
      "learning_rate": 1.0554271613618308e-05,
      "loss": 1.1453,
      "num_input_tokens_seen": 58153520,
      "step": 27845
    },
    {
      "epoch": 4.543274329064361,
      "grad_norm": 8.8125,
      "learning_rate": 1.054626011760276e-05,
      "loss": 2.139,
      "num_input_tokens_seen": 58163696,
      "step": 27850
    },
    {
      "epoch": 4.544090056285178,
      "grad_norm": 11.6875,
      "learning_rate": 1.0538250850535549e-05,
      "loss": 2.1314,
      "num_input_tokens_seen": 58174080,
      "step": 27855
    },
    {
      "epoch": 4.544905783505996,
      "grad_norm": 4.34375,
      "learning_rate": 1.0530243813651794e-05,
      "loss": 0.8476,
      "num_input_tokens_seen": 58185104,
      "step": 27860
    },
    {
      "epoch": 4.545721510726813,
      "grad_norm": 7.15625,
      "learning_rate": 1.0522239008186271e-05,
      "loss": 1.7967,
      "num_input_tokens_seen": 58195200,
      "step": 27865
    },
    {
      "epoch": 4.54653723794763,
      "grad_norm": 9.5625,
      "learning_rate": 1.0514236435373434e-05,
      "loss": 3.1794,
      "num_input_tokens_seen": 58205632,
      "step": 27870
    },
    {
      "epoch": 4.5473529651684474,
      "grad_norm": 4.59375,
      "learning_rate": 1.0506236096447386e-05,
      "loss": 1.8804,
      "num_input_tokens_seen": 58218112,
      "step": 27875
    },
    {
      "epoch": 4.548168692389265,
      "grad_norm": 8.75,
      "learning_rate": 1.049823799264186e-05,
      "loss": 1.4054,
      "num_input_tokens_seen": 58226848,
      "step": 27880
    },
    {
      "epoch": 4.548984419610083,
      "grad_norm": 6.3125,
      "learning_rate": 1.049024212519028e-05,
      "loss": 1.9463,
      "num_input_tokens_seen": 58238080,
      "step": 27885
    },
    {
      "epoch": 4.5498001468309,
      "grad_norm": 11.25,
      "learning_rate": 1.0482248495325713e-05,
      "loss": 2.6347,
      "num_input_tokens_seen": 58248400,
      "step": 27890
    },
    {
      "epoch": 4.550615874051717,
      "grad_norm": 7.25,
      "learning_rate": 1.047425710428086e-05,
      "loss": 2.4034,
      "num_input_tokens_seen": 58258560,
      "step": 27895
    },
    {
      "epoch": 4.551431601272535,
      "grad_norm": 12.0625,
      "learning_rate": 1.0466267953288114e-05,
      "loss": 2.6312,
      "num_input_tokens_seen": 58268800,
      "step": 27900
    },
    {
      "epoch": 4.552247328493352,
      "grad_norm": 6.625,
      "learning_rate": 1.0458281043579482e-05,
      "loss": 2.5601,
      "num_input_tokens_seen": 58278592,
      "step": 27905
    },
    {
      "epoch": 4.553063055714169,
      "grad_norm": 2.734375,
      "learning_rate": 1.0450296376386657e-05,
      "loss": 2.6781,
      "num_input_tokens_seen": 58289936,
      "step": 27910
    },
    {
      "epoch": 4.553878782934986,
      "grad_norm": 5.5,
      "learning_rate": 1.044231395294098e-05,
      "loss": 2.2968,
      "num_input_tokens_seen": 58300576,
      "step": 27915
    },
    {
      "epoch": 4.554694510155803,
      "grad_norm": 8.0625,
      "learning_rate": 1.0434333774473435e-05,
      "loss": 1.814,
      "num_input_tokens_seen": 58310160,
      "step": 27920
    },
    {
      "epoch": 4.555510237376621,
      "grad_norm": 6.09375,
      "learning_rate": 1.0426355842214657e-05,
      "loss": 1.8297,
      "num_input_tokens_seen": 58319808,
      "step": 27925
    },
    {
      "epoch": 4.5563259645974385,
      "grad_norm": 11.875,
      "learning_rate": 1.0418380157394963e-05,
      "loss": 1.9359,
      "num_input_tokens_seen": 58331360,
      "step": 27930
    },
    {
      "epoch": 4.557141691818256,
      "grad_norm": 3.125,
      "learning_rate": 1.0410406721244281e-05,
      "loss": 2.0197,
      "num_input_tokens_seen": 58341152,
      "step": 27935
    },
    {
      "epoch": 4.557957419039074,
      "grad_norm": 6.28125,
      "learning_rate": 1.0402435534992238e-05,
      "loss": 2.9594,
      "num_input_tokens_seen": 58351472,
      "step": 27940
    },
    {
      "epoch": 4.558773146259891,
      "grad_norm": 6.46875,
      "learning_rate": 1.0394466599868071e-05,
      "loss": 2.102,
      "num_input_tokens_seen": 58359664,
      "step": 27945
    },
    {
      "epoch": 4.559588873480708,
      "grad_norm": 1.796875,
      "learning_rate": 1.0386499917100697e-05,
      "loss": 2.8509,
      "num_input_tokens_seen": 58370112,
      "step": 27950
    },
    {
      "epoch": 4.560404600701525,
      "grad_norm": 6.4375,
      "learning_rate": 1.0378535487918692e-05,
      "loss": 2.457,
      "num_input_tokens_seen": 58380560,
      "step": 27955
    },
    {
      "epoch": 4.561220327922343,
      "grad_norm": 7.4375,
      "learning_rate": 1.037057331355025e-05,
      "loss": 5.1613,
      "num_input_tokens_seen": 58391968,
      "step": 27960
    },
    {
      "epoch": 4.56203605514316,
      "grad_norm": 4.9375,
      "learning_rate": 1.0362613395223247e-05,
      "loss": 2.9612,
      "num_input_tokens_seen": 58401984,
      "step": 27965
    },
    {
      "epoch": 4.562851782363977,
      "grad_norm": 4.9375,
      "learning_rate": 1.0354655734165212e-05,
      "loss": 1.8174,
      "num_input_tokens_seen": 58412048,
      "step": 27970
    },
    {
      "epoch": 4.5636675095847945,
      "grad_norm": 13.625,
      "learning_rate": 1.03467003316033e-05,
      "loss": 2.6338,
      "num_input_tokens_seen": 58422864,
      "step": 27975
    },
    {
      "epoch": 4.5644832368056125,
      "grad_norm": 4.46875,
      "learning_rate": 1.033874718876435e-05,
      "loss": 2.5256,
      "num_input_tokens_seen": 58433808,
      "step": 27980
    },
    {
      "epoch": 4.56529896402643,
      "grad_norm": 7.625,
      "learning_rate": 1.0330796306874818e-05,
      "loss": 2.8845,
      "num_input_tokens_seen": 58443376,
      "step": 27985
    },
    {
      "epoch": 4.566114691247247,
      "grad_norm": 4.96875,
      "learning_rate": 1.032284768716085e-05,
      "loss": 2.2115,
      "num_input_tokens_seen": 58453936,
      "step": 27990
    },
    {
      "epoch": 4.566930418468064,
      "grad_norm": 9.0,
      "learning_rate": 1.0314901330848206e-05,
      "loss": 1.7832,
      "num_input_tokens_seen": 58464416,
      "step": 27995
    },
    {
      "epoch": 4.567746145688882,
      "grad_norm": 12.125,
      "learning_rate": 1.030695723916233e-05,
      "loss": 3.3468,
      "num_input_tokens_seen": 58475488,
      "step": 28000
    },
    {
      "epoch": 4.567746145688882,
      "eval_loss": 2.545409679412842,
      "eval_runtime": 134.6944,
      "eval_samples_per_second": 20.231,
      "eval_steps_per_second": 10.119,
      "num_input_tokens_seen": 58475488,
      "step": 28000
    },
    {
      "epoch": 4.568561872909699,
      "grad_norm": 4.28125,
      "learning_rate": 1.0299015413328289e-05,
      "loss": 1.2479,
      "num_input_tokens_seen": 58484720,
      "step": 28005
    },
    {
      "epoch": 4.569377600130516,
      "grad_norm": 3.390625,
      "learning_rate": 1.0291075854570809e-05,
      "loss": 4.0472,
      "num_input_tokens_seen": 58495184,
      "step": 28010
    },
    {
      "epoch": 4.570193327351333,
      "grad_norm": 2.90625,
      "learning_rate": 1.0283138564114275e-05,
      "loss": 4.1318,
      "num_input_tokens_seen": 58506336,
      "step": 28015
    },
    {
      "epoch": 4.571009054572151,
      "grad_norm": 13.125,
      "learning_rate": 1.027520354318273e-05,
      "loss": 2.3654,
      "num_input_tokens_seen": 58516768,
      "step": 28020
    },
    {
      "epoch": 4.5718247817929685,
      "grad_norm": 6.09375,
      "learning_rate": 1.0267270792999828e-05,
      "loss": 2.7872,
      "num_input_tokens_seen": 58528272,
      "step": 28025
    },
    {
      "epoch": 4.572640509013786,
      "grad_norm": 12.4375,
      "learning_rate": 1.0259340314788919e-05,
      "loss": 2.4599,
      "num_input_tokens_seen": 58539232,
      "step": 28030
    },
    {
      "epoch": 4.573456236234604,
      "grad_norm": 2.328125,
      "learning_rate": 1.0251412109772979e-05,
      "loss": 1.3142,
      "num_input_tokens_seen": 58550400,
      "step": 28035
    },
    {
      "epoch": 4.574271963455421,
      "grad_norm": 8.8125,
      "learning_rate": 1.0243486179174627e-05,
      "loss": 1.3613,
      "num_input_tokens_seen": 58560288,
      "step": 28040
    },
    {
      "epoch": 4.575087690676238,
      "grad_norm": 10.25,
      "learning_rate": 1.0235562524216158e-05,
      "loss": 2.5354,
      "num_input_tokens_seen": 58570048,
      "step": 28045
    },
    {
      "epoch": 4.575903417897055,
      "grad_norm": 1.5390625,
      "learning_rate": 1.022764114611948e-05,
      "loss": 1.8604,
      "num_input_tokens_seen": 58580240,
      "step": 28050
    },
    {
      "epoch": 4.576719145117872,
      "grad_norm": 10.25,
      "learning_rate": 1.0219722046106178e-05,
      "loss": 2.3342,
      "num_input_tokens_seen": 58589200,
      "step": 28055
    },
    {
      "epoch": 4.57753487233869,
      "grad_norm": 14.5625,
      "learning_rate": 1.0211805225397486e-05,
      "loss": 3.5172,
      "num_input_tokens_seen": 58600768,
      "step": 28060
    },
    {
      "epoch": 4.578350599559507,
      "grad_norm": 5.59375,
      "learning_rate": 1.020389068521426e-05,
      "loss": 3.196,
      "num_input_tokens_seen": 58611376,
      "step": 28065
    },
    {
      "epoch": 4.579166326780324,
      "grad_norm": 7.3125,
      "learning_rate": 1.0195978426777039e-05,
      "loss": 1.9882,
      "num_input_tokens_seen": 58620496,
      "step": 28070
    },
    {
      "epoch": 4.5799820540011424,
      "grad_norm": 4.625,
      "learning_rate": 1.0188068451305982e-05,
      "loss": 2.1651,
      "num_input_tokens_seen": 58631264,
      "step": 28075
    },
    {
      "epoch": 4.58079778122196,
      "grad_norm": 3.40625,
      "learning_rate": 1.0180160760020902e-05,
      "loss": 2.1447,
      "num_input_tokens_seen": 58640272,
      "step": 28080
    },
    {
      "epoch": 4.581613508442777,
      "grad_norm": 2.625,
      "learning_rate": 1.0172255354141278e-05,
      "loss": 2.6813,
      "num_input_tokens_seen": 58649280,
      "step": 28085
    },
    {
      "epoch": 4.582429235663594,
      "grad_norm": 8.0,
      "learning_rate": 1.0164352234886205e-05,
      "loss": 2.1149,
      "num_input_tokens_seen": 58658960,
      "step": 28090
    },
    {
      "epoch": 4.583244962884411,
      "grad_norm": 2.546875,
      "learning_rate": 1.0156451403474454e-05,
      "loss": 1.4688,
      "num_input_tokens_seen": 58671856,
      "step": 28095
    },
    {
      "epoch": 4.584060690105229,
      "grad_norm": 11.25,
      "learning_rate": 1.0148552861124443e-05,
      "loss": 2.6753,
      "num_input_tokens_seen": 58681456,
      "step": 28100
    },
    {
      "epoch": 4.584876417326046,
      "grad_norm": 3.140625,
      "learning_rate": 1.0140656609054205e-05,
      "loss": 2.3744,
      "num_input_tokens_seen": 58692080,
      "step": 28105
    },
    {
      "epoch": 4.585692144546863,
      "grad_norm": 4.25,
      "learning_rate": 1.0132762648481455e-05,
      "loss": 2.0643,
      "num_input_tokens_seen": 58703024,
      "step": 28110
    },
    {
      "epoch": 4.586507871767681,
      "grad_norm": 3.984375,
      "learning_rate": 1.0124870980623543e-05,
      "loss": 2.5376,
      "num_input_tokens_seen": 58714960,
      "step": 28115
    },
    {
      "epoch": 4.587323598988498,
      "grad_norm": 6.84375,
      "learning_rate": 1.0116981606697453e-05,
      "loss": 1.5032,
      "num_input_tokens_seen": 58725920,
      "step": 28120
    },
    {
      "epoch": 4.5881393262093155,
      "grad_norm": 6.09375,
      "learning_rate": 1.0109094527919838e-05,
      "loss": 2.4988,
      "num_input_tokens_seen": 58737584,
      "step": 28125
    },
    {
      "epoch": 4.588955053430133,
      "grad_norm": 0.059814453125,
      "learning_rate": 1.010120974550697e-05,
      "loss": 2.5805,
      "num_input_tokens_seen": 58748016,
      "step": 28130
    },
    {
      "epoch": 4.589770780650951,
      "grad_norm": 7.34375,
      "learning_rate": 1.0093327260674795e-05,
      "loss": 3.4765,
      "num_input_tokens_seen": 58758800,
      "step": 28135
    },
    {
      "epoch": 4.590586507871768,
      "grad_norm": 11.9375,
      "learning_rate": 1.0085447074638878e-05,
      "loss": 2.0011,
      "num_input_tokens_seen": 58769904,
      "step": 28140
    },
    {
      "epoch": 4.591402235092585,
      "grad_norm": 15.25,
      "learning_rate": 1.0077569188614461e-05,
      "loss": 2.9371,
      "num_input_tokens_seen": 58779696,
      "step": 28145
    },
    {
      "epoch": 4.592217962313402,
      "grad_norm": 5.84375,
      "learning_rate": 1.0069693603816393e-05,
      "loss": 2.0271,
      "num_input_tokens_seen": 58790816,
      "step": 28150
    },
    {
      "epoch": 4.59303368953422,
      "grad_norm": 4.5,
      "learning_rate": 1.0061820321459204e-05,
      "loss": 2.6568,
      "num_input_tokens_seen": 58802208,
      "step": 28155
    },
    {
      "epoch": 4.593849416755037,
      "grad_norm": 5.96875,
      "learning_rate": 1.0053949342757038e-05,
      "loss": 2.2766,
      "num_input_tokens_seen": 58813168,
      "step": 28160
    },
    {
      "epoch": 4.594665143975854,
      "grad_norm": 4.96875,
      "learning_rate": 1.0046080668923717e-05,
      "loss": 4.1358,
      "num_input_tokens_seen": 58824880,
      "step": 28165
    },
    {
      "epoch": 4.5954808711966715,
      "grad_norm": 5.75,
      "learning_rate": 1.003821430117267e-05,
      "loss": 2.2151,
      "num_input_tokens_seen": 58836016,
      "step": 28170
    },
    {
      "epoch": 4.5962965984174895,
      "grad_norm": 8.625,
      "learning_rate": 1.0030350240716999e-05,
      "loss": 2.6984,
      "num_input_tokens_seen": 58846768,
      "step": 28175
    },
    {
      "epoch": 4.597112325638307,
      "grad_norm": 7.65625,
      "learning_rate": 1.0022488488769449e-05,
      "loss": 2.6146,
      "num_input_tokens_seen": 58856816,
      "step": 28180
    },
    {
      "epoch": 4.597928052859124,
      "grad_norm": 8.6875,
      "learning_rate": 1.0014629046542387e-05,
      "loss": 2.7373,
      "num_input_tokens_seen": 58867648,
      "step": 28185
    },
    {
      "epoch": 4.598743780079941,
      "grad_norm": 10.4375,
      "learning_rate": 1.0006771915247842e-05,
      "loss": 2.5584,
      "num_input_tokens_seen": 58877792,
      "step": 28190
    },
    {
      "epoch": 4.599559507300759,
      "grad_norm": 6.375,
      "learning_rate": 9.998917096097495e-06,
      "loss": 2.6661,
      "num_input_tokens_seen": 58888176,
      "step": 28195
    },
    {
      "epoch": 4.600375234521576,
      "grad_norm": 3.65625,
      "learning_rate": 9.991064590302638e-06,
      "loss": 3.1295,
      "num_input_tokens_seen": 58898896,
      "step": 28200
    },
    {
      "epoch": 4.600375234521576,
      "eval_loss": 2.5404255390167236,
      "eval_runtime": 134.8721,
      "eval_samples_per_second": 20.204,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 58898896,
      "step": 28200
    },
    {
      "epoch": 4.601190961742393,
      "grad_norm": 7.21875,
      "learning_rate": 9.983214399074241e-06,
      "loss": 1.968,
      "num_input_tokens_seen": 58909616,
      "step": 28205
    },
    {
      "epoch": 4.602006688963211,
      "grad_norm": 10.3125,
      "learning_rate": 9.975366523622893e-06,
      "loss": 2.455,
      "num_input_tokens_seen": 58919776,
      "step": 28210
    },
    {
      "epoch": 4.602822416184028,
      "grad_norm": 6.875,
      "learning_rate": 9.967520965158841e-06,
      "loss": 2.1679,
      "num_input_tokens_seen": 58930704,
      "step": 28215
    },
    {
      "epoch": 4.6036381434048455,
      "grad_norm": 8.5,
      "learning_rate": 9.95967772489197e-06,
      "loss": 2.1011,
      "num_input_tokens_seen": 58940608,
      "step": 28220
    },
    {
      "epoch": 4.604453870625663,
      "grad_norm": 2.203125,
      "learning_rate": 9.951836804031794e-06,
      "loss": 1.7826,
      "num_input_tokens_seen": 58952352,
      "step": 28225
    },
    {
      "epoch": 4.60526959784648,
      "grad_norm": 10.125,
      "learning_rate": 9.943998203787489e-06,
      "loss": 3.2396,
      "num_input_tokens_seen": 58962528,
      "step": 28230
    },
    {
      "epoch": 4.606085325067298,
      "grad_norm": 12.0,
      "learning_rate": 9.936161925367874e-06,
      "loss": 3.3115,
      "num_input_tokens_seen": 58971472,
      "step": 28235
    },
    {
      "epoch": 4.606901052288115,
      "grad_norm": 7.5,
      "learning_rate": 9.928327969981386e-06,
      "loss": 3.3945,
      "num_input_tokens_seen": 58982672,
      "step": 28240
    },
    {
      "epoch": 4.607716779508932,
      "grad_norm": 1.8125,
      "learning_rate": 9.920496338836135e-06,
      "loss": 2.9928,
      "num_input_tokens_seen": 58993904,
      "step": 28245
    },
    {
      "epoch": 4.60853250672975,
      "grad_norm": 2.890625,
      "learning_rate": 9.912667033139844e-06,
      "loss": 1.4599,
      "num_input_tokens_seen": 59004352,
      "step": 28250
    },
    {
      "epoch": 4.609348233950567,
      "grad_norm": 6.1875,
      "learning_rate": 9.904840054099893e-06,
      "loss": 2.5722,
      "num_input_tokens_seen": 59015344,
      "step": 28255
    },
    {
      "epoch": 4.610163961171384,
      "grad_norm": 4.15625,
      "learning_rate": 9.897015402923312e-06,
      "loss": 2.591,
      "num_input_tokens_seen": 59024640,
      "step": 28260
    },
    {
      "epoch": 4.610979688392201,
      "grad_norm": 3.734375,
      "learning_rate": 9.889193080816744e-06,
      "loss": 1.5616,
      "num_input_tokens_seen": 59034672,
      "step": 28265
    },
    {
      "epoch": 4.6117954156130185,
      "grad_norm": 10.0625,
      "learning_rate": 9.881373088986498e-06,
      "loss": 1.8154,
      "num_input_tokens_seen": 59044112,
      "step": 28270
    },
    {
      "epoch": 4.6126111428338366,
      "grad_norm": 6.53125,
      "learning_rate": 9.873555428638523e-06,
      "loss": 2.0809,
      "num_input_tokens_seen": 59053808,
      "step": 28275
    },
    {
      "epoch": 4.613426870054654,
      "grad_norm": 8.1875,
      "learning_rate": 9.865740100978383e-06,
      "loss": 3.9938,
      "num_input_tokens_seen": 59064640,
      "step": 28280
    },
    {
      "epoch": 4.614242597275471,
      "grad_norm": 10.5625,
      "learning_rate": 9.857927107211315e-06,
      "loss": 3.3046,
      "num_input_tokens_seen": 59076512,
      "step": 28285
    },
    {
      "epoch": 4.615058324496289,
      "grad_norm": 4.90625,
      "learning_rate": 9.850116448542177e-06,
      "loss": 2.0611,
      "num_input_tokens_seen": 59086432,
      "step": 28290
    },
    {
      "epoch": 4.615874051717106,
      "grad_norm": 10.1875,
      "learning_rate": 9.842308126175457e-06,
      "loss": 1.7429,
      "num_input_tokens_seen": 59096240,
      "step": 28295
    },
    {
      "epoch": 4.616689778937923,
      "grad_norm": 9.375,
      "learning_rate": 9.834502141315315e-06,
      "loss": 2.4445,
      "num_input_tokens_seen": 59106080,
      "step": 28300
    },
    {
      "epoch": 4.61750550615874,
      "grad_norm": 5.5625,
      "learning_rate": 9.82669849516552e-06,
      "loss": 2.6314,
      "num_input_tokens_seen": 59116816,
      "step": 28305
    },
    {
      "epoch": 4.618321233379558,
      "grad_norm": 4.46875,
      "learning_rate": 9.818897188929493e-06,
      "loss": 1.6128,
      "num_input_tokens_seen": 59128208,
      "step": 28310
    },
    {
      "epoch": 4.619136960600375,
      "grad_norm": 1.46875,
      "learning_rate": 9.811098223810309e-06,
      "loss": 1.9283,
      "num_input_tokens_seen": 59138208,
      "step": 28315
    },
    {
      "epoch": 4.6199526878211925,
      "grad_norm": 5.9375,
      "learning_rate": 9.803301601010641e-06,
      "loss": 2.3968,
      "num_input_tokens_seen": 59149728,
      "step": 28320
    },
    {
      "epoch": 4.62076841504201,
      "grad_norm": 4.75,
      "learning_rate": 9.795507321732853e-06,
      "loss": 1.5971,
      "num_input_tokens_seen": 59160272,
      "step": 28325
    },
    {
      "epoch": 4.621584142262828,
      "grad_norm": 10.1875,
      "learning_rate": 9.787715387178898e-06,
      "loss": 2.1268,
      "num_input_tokens_seen": 59172320,
      "step": 28330
    },
    {
      "epoch": 4.622399869483645,
      "grad_norm": 12.375,
      "learning_rate": 9.779925798550399e-06,
      "loss": 2.7148,
      "num_input_tokens_seen": 59182816,
      "step": 28335
    },
    {
      "epoch": 4.623215596704462,
      "grad_norm": 10.6875,
      "learning_rate": 9.772138557048619e-06,
      "loss": 2.5023,
      "num_input_tokens_seen": 59193328,
      "step": 28340
    },
    {
      "epoch": 4.624031323925279,
      "grad_norm": 4.1875,
      "learning_rate": 9.764353663874426e-06,
      "loss": 1.3306,
      "num_input_tokens_seen": 59203392,
      "step": 28345
    },
    {
      "epoch": 4.624847051146097,
      "grad_norm": 6.625,
      "learning_rate": 9.756571120228375e-06,
      "loss": 3.0225,
      "num_input_tokens_seen": 59212800,
      "step": 28350
    },
    {
      "epoch": 4.625662778366914,
      "grad_norm": 14.9375,
      "learning_rate": 9.748790927310605e-06,
      "loss": 3.9625,
      "num_input_tokens_seen": 59222464,
      "step": 28355
    },
    {
      "epoch": 4.626478505587731,
      "grad_norm": 12.8125,
      "learning_rate": 9.741013086320946e-06,
      "loss": 2.2866,
      "num_input_tokens_seen": 59233248,
      "step": 28360
    },
    {
      "epoch": 4.6272942328085485,
      "grad_norm": 4.75,
      "learning_rate": 9.733237598458821e-06,
      "loss": 2.2495,
      "num_input_tokens_seen": 59243392,
      "step": 28365
    },
    {
      "epoch": 4.6281099600293665,
      "grad_norm": 6.1875,
      "learning_rate": 9.725464464923308e-06,
      "loss": 3.1369,
      "num_input_tokens_seen": 59254432,
      "step": 28370
    },
    {
      "epoch": 4.628925687250184,
      "grad_norm": 9.9375,
      "learning_rate": 9.717693686913123e-06,
      "loss": 1.7599,
      "num_input_tokens_seen": 59265408,
      "step": 28375
    },
    {
      "epoch": 4.629741414471001,
      "grad_norm": 6.34375,
      "learning_rate": 9.709925265626632e-06,
      "loss": 2.4338,
      "num_input_tokens_seen": 59276224,
      "step": 28380
    },
    {
      "epoch": 4.630557141691818,
      "grad_norm": 6.0,
      "learning_rate": 9.702159202261801e-06,
      "loss": 1.6574,
      "num_input_tokens_seen": 59286560,
      "step": 28385
    },
    {
      "epoch": 4.631372868912636,
      "grad_norm": 7.34375,
      "learning_rate": 9.694395498016268e-06,
      "loss": 1.4097,
      "num_input_tokens_seen": 59297920,
      "step": 28390
    },
    {
      "epoch": 4.632188596133453,
      "grad_norm": 4.46875,
      "learning_rate": 9.686634154087298e-06,
      "loss": 1.3523,
      "num_input_tokens_seen": 59308704,
      "step": 28395
    },
    {
      "epoch": 4.63300432335427,
      "grad_norm": 2.09375,
      "learning_rate": 9.678875171671776e-06,
      "loss": 1.6093,
      "num_input_tokens_seen": 59318976,
      "step": 28400
    },
    {
      "epoch": 4.63300432335427,
      "eval_loss": 2.539818048477173,
      "eval_runtime": 134.8784,
      "eval_samples_per_second": 20.203,
      "eval_steps_per_second": 10.105,
      "num_input_tokens_seen": 59318976,
      "step": 28400
    },
    {
      "epoch": 4.633820050575087,
      "grad_norm": 8.1875,
      "learning_rate": 9.671118551966246e-06,
      "loss": 2.7529,
      "num_input_tokens_seen": 59328672,
      "step": 28405
    },
    {
      "epoch": 4.634635777795905,
      "grad_norm": 5.09375,
      "learning_rate": 9.66336429616686e-06,
      "loss": 2.3953,
      "num_input_tokens_seen": 59338848,
      "step": 28410
    },
    {
      "epoch": 4.635451505016722,
      "grad_norm": 1.8984375,
      "learning_rate": 9.655612405469436e-06,
      "loss": 3.4484,
      "num_input_tokens_seen": 59348848,
      "step": 28415
    },
    {
      "epoch": 4.63626723223754,
      "grad_norm": 3.0,
      "learning_rate": 9.647862881069413e-06,
      "loss": 1.2669,
      "num_input_tokens_seen": 59359568,
      "step": 28420
    },
    {
      "epoch": 4.637082959458358,
      "grad_norm": 7.75,
      "learning_rate": 9.640115724161855e-06,
      "loss": 2.1765,
      "num_input_tokens_seen": 59371504,
      "step": 28425
    },
    {
      "epoch": 4.637898686679175,
      "grad_norm": 3.21875,
      "learning_rate": 9.632370935941483e-06,
      "loss": 1.8536,
      "num_input_tokens_seen": 59381904,
      "step": 28430
    },
    {
      "epoch": 4.638714413899992,
      "grad_norm": 6.0625,
      "learning_rate": 9.624628517602634e-06,
      "loss": 3.1756,
      "num_input_tokens_seen": 59392896,
      "step": 28435
    },
    {
      "epoch": 4.639530141120809,
      "grad_norm": 10.3125,
      "learning_rate": 9.61688847033928e-06,
      "loss": 2.7083,
      "num_input_tokens_seen": 59403712,
      "step": 28440
    },
    {
      "epoch": 4.640345868341626,
      "grad_norm": 7.3125,
      "learning_rate": 9.609150795345051e-06,
      "loss": 1.4618,
      "num_input_tokens_seen": 59413664,
      "step": 28445
    },
    {
      "epoch": 4.641161595562444,
      "grad_norm": 7.21875,
      "learning_rate": 9.601415493813171e-06,
      "loss": 2.306,
      "num_input_tokens_seen": 59423408,
      "step": 28450
    },
    {
      "epoch": 4.641977322783261,
      "grad_norm": 10.1875,
      "learning_rate": 9.593682566936533e-06,
      "loss": 1.4951,
      "num_input_tokens_seen": 59433808,
      "step": 28455
    },
    {
      "epoch": 4.642793050004078,
      "grad_norm": 9.125,
      "learning_rate": 9.58595201590766e-06,
      "loss": 3.7171,
      "num_input_tokens_seen": 59442560,
      "step": 28460
    },
    {
      "epoch": 4.643608777224896,
      "grad_norm": 9.75,
      "learning_rate": 9.578223841918681e-06,
      "loss": 2.8257,
      "num_input_tokens_seen": 59453696,
      "step": 28465
    },
    {
      "epoch": 4.6444245044457135,
      "grad_norm": 3.84375,
      "learning_rate": 9.570498046161389e-06,
      "loss": 1.9643,
      "num_input_tokens_seen": 59464592,
      "step": 28470
    },
    {
      "epoch": 4.645240231666531,
      "grad_norm": 6.6875,
      "learning_rate": 9.562774629827206e-06,
      "loss": 1.8316,
      "num_input_tokens_seen": 59475568,
      "step": 28475
    },
    {
      "epoch": 4.646055958887348,
      "grad_norm": 9.5625,
      "learning_rate": 9.555053594107163e-06,
      "loss": 4.0345,
      "num_input_tokens_seen": 59486560,
      "step": 28480
    },
    {
      "epoch": 4.646871686108166,
      "grad_norm": 7.0625,
      "learning_rate": 9.547334940191957e-06,
      "loss": 2.0991,
      "num_input_tokens_seen": 59495312,
      "step": 28485
    },
    {
      "epoch": 4.647687413328983,
      "grad_norm": 14.5625,
      "learning_rate": 9.539618669271886e-06,
      "loss": 2.1454,
      "num_input_tokens_seen": 59505408,
      "step": 28490
    },
    {
      "epoch": 4.6485031405498,
      "grad_norm": 6.03125,
      "learning_rate": 9.531904782536904e-06,
      "loss": 1.3281,
      "num_input_tokens_seen": 59516096,
      "step": 28495
    },
    {
      "epoch": 4.649318867770617,
      "grad_norm": 4.71875,
      "learning_rate": 9.524193281176597e-06,
      "loss": 1.9429,
      "num_input_tokens_seen": 59527264,
      "step": 28500
    },
    {
      "epoch": 4.650134594991435,
      "grad_norm": 5.34375,
      "learning_rate": 9.516484166380165e-06,
      "loss": 1.2131,
      "num_input_tokens_seen": 59537808,
      "step": 28505
    },
    {
      "epoch": 4.650950322212252,
      "grad_norm": 8.375,
      "learning_rate": 9.508777439336447e-06,
      "loss": 2.1242,
      "num_input_tokens_seen": 59549408,
      "step": 28510
    },
    {
      "epoch": 4.6517660494330695,
      "grad_norm": 4.71875,
      "learning_rate": 9.50107310123393e-06,
      "loss": 2.0554,
      "num_input_tokens_seen": 59560496,
      "step": 28515
    },
    {
      "epoch": 4.652581776653887,
      "grad_norm": 9.8125,
      "learning_rate": 9.493371153260702e-06,
      "loss": 2.671,
      "num_input_tokens_seen": 59570624,
      "step": 28520
    },
    {
      "epoch": 4.653397503874705,
      "grad_norm": 0.07666015625,
      "learning_rate": 9.485671596604523e-06,
      "loss": 2.5152,
      "num_input_tokens_seen": 59580192,
      "step": 28525
    },
    {
      "epoch": 4.654213231095522,
      "grad_norm": 6.90625,
      "learning_rate": 9.477974432452738e-06,
      "loss": 3.4834,
      "num_input_tokens_seen": 59590752,
      "step": 28530
    },
    {
      "epoch": 4.655028958316339,
      "grad_norm": 4.1875,
      "learning_rate": 9.470279661992356e-06,
      "loss": 1.4426,
      "num_input_tokens_seen": 59601440,
      "step": 28535
    },
    {
      "epoch": 4.655844685537156,
      "grad_norm": 3.9375,
      "learning_rate": 9.462587286410021e-06,
      "loss": 1.2188,
      "num_input_tokens_seen": 59611824,
      "step": 28540
    },
    {
      "epoch": 4.656660412757974,
      "grad_norm": 5.65625,
      "learning_rate": 9.454897306891972e-06,
      "loss": 2.1165,
      "num_input_tokens_seen": 59622880,
      "step": 28545
    },
    {
      "epoch": 4.657476139978791,
      "grad_norm": 10.75,
      "learning_rate": 9.44720972462411e-06,
      "loss": 3.6317,
      "num_input_tokens_seen": 59633904,
      "step": 28550
    },
    {
      "epoch": 4.658291867199608,
      "grad_norm": 14.0,
      "learning_rate": 9.439524540791964e-06,
      "loss": 3.1657,
      "num_input_tokens_seen": 59644992,
      "step": 28555
    },
    {
      "epoch": 4.659107594420425,
      "grad_norm": 11.0,
      "learning_rate": 9.431841756580673e-06,
      "loss": 3.0301,
      "num_input_tokens_seen": 59655392,
      "step": 28560
    },
    {
      "epoch": 4.6599233216412435,
      "grad_norm": 12.125,
      "learning_rate": 9.42416137317503e-06,
      "loss": 2.9359,
      "num_input_tokens_seen": 59665728,
      "step": 28565
    },
    {
      "epoch": 4.660739048862061,
      "grad_norm": 5.03125,
      "learning_rate": 9.416483391759437e-06,
      "loss": 2.0589,
      "num_input_tokens_seen": 59676656,
      "step": 28570
    },
    {
      "epoch": 4.661554776082878,
      "grad_norm": 7.65625,
      "learning_rate": 9.408807813517945e-06,
      "loss": 3.3369,
      "num_input_tokens_seen": 59686800,
      "step": 28575
    },
    {
      "epoch": 4.662370503303695,
      "grad_norm": 4.75,
      "learning_rate": 9.401134639634221e-06,
      "loss": 1.7582,
      "num_input_tokens_seen": 59696768,
      "step": 28580
    },
    {
      "epoch": 4.663186230524513,
      "grad_norm": 9.0,
      "learning_rate": 9.393463871291555e-06,
      "loss": 2.3603,
      "num_input_tokens_seen": 59707760,
      "step": 28585
    },
    {
      "epoch": 4.66400195774533,
      "grad_norm": 7.25,
      "learning_rate": 9.385795509672881e-06,
      "loss": 2.195,
      "num_input_tokens_seen": 59717952,
      "step": 28590
    },
    {
      "epoch": 4.664817684966147,
      "grad_norm": 0.29296875,
      "learning_rate": 9.378129555960771e-06,
      "loss": 2.0939,
      "num_input_tokens_seen": 59728832,
      "step": 28595
    },
    {
      "epoch": 4.665633412186965,
      "grad_norm": 13.375,
      "learning_rate": 9.370466011337392e-06,
      "loss": 1.9266,
      "num_input_tokens_seen": 59739680,
      "step": 28600
    },
    {
      "epoch": 4.665633412186965,
      "eval_loss": 2.5382657051086426,
      "eval_runtime": 135.0974,
      "eval_samples_per_second": 20.171,
      "eval_steps_per_second": 10.089,
      "num_input_tokens_seen": 59739680,
      "step": 28600
    },
    {
      "epoch": 4.666449139407782,
      "grad_norm": 9.25,
      "learning_rate": 9.362804876984573e-06,
      "loss": 2.6995,
      "num_input_tokens_seen": 59751520,
      "step": 28605
    },
    {
      "epoch": 4.667264866628599,
      "grad_norm": 3.71875,
      "learning_rate": 9.355146154083747e-06,
      "loss": 3.2112,
      "num_input_tokens_seen": 59762512,
      "step": 28610
    },
    {
      "epoch": 4.6680805938494165,
      "grad_norm": 7.59375,
      "learning_rate": 9.347489843815987e-06,
      "loss": 1.9752,
      "num_input_tokens_seen": 59773120,
      "step": 28615
    },
    {
      "epoch": 4.668896321070234,
      "grad_norm": 6.65625,
      "learning_rate": 9.339835947362002e-06,
      "loss": 2.55,
      "num_input_tokens_seen": 59783056,
      "step": 28620
    },
    {
      "epoch": 4.669712048291052,
      "grad_norm": 6.15625,
      "learning_rate": 9.332184465902105e-06,
      "loss": 2.4322,
      "num_input_tokens_seen": 59793744,
      "step": 28625
    },
    {
      "epoch": 4.670527775511869,
      "grad_norm": 4.34375,
      "learning_rate": 9.324535400616266e-06,
      "loss": 1.9307,
      "num_input_tokens_seen": 59804752,
      "step": 28630
    },
    {
      "epoch": 4.671343502732686,
      "grad_norm": 6.21875,
      "learning_rate": 9.31688875268405e-06,
      "loss": 4.8734,
      "num_input_tokens_seen": 59815392,
      "step": 28635
    },
    {
      "epoch": 4.672159229953504,
      "grad_norm": 6.8125,
      "learning_rate": 9.309244523284674e-06,
      "loss": 2.2389,
      "num_input_tokens_seen": 59826784,
      "step": 28640
    },
    {
      "epoch": 4.672974957174321,
      "grad_norm": 6.3125,
      "learning_rate": 9.301602713596982e-06,
      "loss": 2.1255,
      "num_input_tokens_seen": 59836560,
      "step": 28645
    },
    {
      "epoch": 4.673790684395138,
      "grad_norm": 4.15625,
      "learning_rate": 9.293963324799432e-06,
      "loss": 2.0453,
      "num_input_tokens_seen": 59846848,
      "step": 28650
    },
    {
      "epoch": 4.674606411615955,
      "grad_norm": 9.5,
      "learning_rate": 9.286326358070104e-06,
      "loss": 2.6656,
      "num_input_tokens_seen": 59857152,
      "step": 28655
    },
    {
      "epoch": 4.6754221388367725,
      "grad_norm": 7.375,
      "learning_rate": 9.278691814586729e-06,
      "loss": 3.6563,
      "num_input_tokens_seen": 59867328,
      "step": 28660
    },
    {
      "epoch": 4.6762378660575905,
      "grad_norm": 11.625,
      "learning_rate": 9.271059695526635e-06,
      "loss": 5.2712,
      "num_input_tokens_seen": 59877776,
      "step": 28665
    },
    {
      "epoch": 4.677053593278408,
      "grad_norm": 7.0625,
      "learning_rate": 9.263430002066805e-06,
      "loss": 1.384,
      "num_input_tokens_seen": 59887696,
      "step": 28670
    },
    {
      "epoch": 4.677869320499225,
      "grad_norm": 11.8125,
      "learning_rate": 9.25580273538382e-06,
      "loss": 3.4245,
      "num_input_tokens_seen": 59897952,
      "step": 28675
    },
    {
      "epoch": 4.678685047720043,
      "grad_norm": 7.46875,
      "learning_rate": 9.248177896653907e-06,
      "loss": 2.0009,
      "num_input_tokens_seen": 59908192,
      "step": 28680
    },
    {
      "epoch": 4.67950077494086,
      "grad_norm": 6.03125,
      "learning_rate": 9.240555487052918e-06,
      "loss": 2.277,
      "num_input_tokens_seen": 59918096,
      "step": 28685
    },
    {
      "epoch": 4.680316502161677,
      "grad_norm": 4.5625,
      "learning_rate": 9.232935507756313e-06,
      "loss": 2.2817,
      "num_input_tokens_seen": 59928672,
      "step": 28690
    },
    {
      "epoch": 4.681132229382494,
      "grad_norm": 2.671875,
      "learning_rate": 9.225317959939193e-06,
      "loss": 1.9533,
      "num_input_tokens_seen": 59938976,
      "step": 28695
    },
    {
      "epoch": 4.681947956603312,
      "grad_norm": 4.9375,
      "learning_rate": 9.217702844776287e-06,
      "loss": 1.4298,
      "num_input_tokens_seen": 59948224,
      "step": 28700
    },
    {
      "epoch": 4.682763683824129,
      "grad_norm": 6.46875,
      "learning_rate": 9.210090163441929e-06,
      "loss": 2.7252,
      "num_input_tokens_seen": 59956928,
      "step": 28705
    },
    {
      "epoch": 4.6835794110449465,
      "grad_norm": 7.9375,
      "learning_rate": 9.202479917110105e-06,
      "loss": 2.6277,
      "num_input_tokens_seen": 59967552,
      "step": 28710
    },
    {
      "epoch": 4.684395138265764,
      "grad_norm": 6.09375,
      "learning_rate": 9.194872106954392e-06,
      "loss": 2.2262,
      "num_input_tokens_seen": 59979072,
      "step": 28715
    },
    {
      "epoch": 4.685210865486582,
      "grad_norm": 4.8125,
      "learning_rate": 9.187266734148029e-06,
      "loss": 1.8501,
      "num_input_tokens_seen": 59989504,
      "step": 28720
    },
    {
      "epoch": 4.686026592707399,
      "grad_norm": 8.875,
      "learning_rate": 9.179663799863849e-06,
      "loss": 2.8195,
      "num_input_tokens_seen": 59999856,
      "step": 28725
    },
    {
      "epoch": 4.686842319928216,
      "grad_norm": 4.8125,
      "learning_rate": 9.172063305274317e-06,
      "loss": 2.5312,
      "num_input_tokens_seen": 60011200,
      "step": 28730
    },
    {
      "epoch": 4.687658047149033,
      "grad_norm": 4.25,
      "learning_rate": 9.164465251551527e-06,
      "loss": 4.282,
      "num_input_tokens_seen": 60021824,
      "step": 28735
    },
    {
      "epoch": 4.688473774369851,
      "grad_norm": 8.125,
      "learning_rate": 9.156869639867205e-06,
      "loss": 1.9375,
      "num_input_tokens_seen": 60031136,
      "step": 28740
    },
    {
      "epoch": 4.689289501590668,
      "grad_norm": 3.5,
      "learning_rate": 9.149276471392677e-06,
      "loss": 1.5486,
      "num_input_tokens_seen": 60041344,
      "step": 28745
    },
    {
      "epoch": 4.690105228811485,
      "grad_norm": 6.8125,
      "learning_rate": 9.141685747298914e-06,
      "loss": 2.6044,
      "num_input_tokens_seen": 60052272,
      "step": 28750
    },
    {
      "epoch": 4.690920956032302,
      "grad_norm": 7.03125,
      "learning_rate": 9.13409746875649e-06,
      "loss": 2.8864,
      "num_input_tokens_seen": 60063376,
      "step": 28755
    },
    {
      "epoch": 4.69173668325312,
      "grad_norm": 1.953125,
      "learning_rate": 9.12651163693562e-06,
      "loss": 2.2091,
      "num_input_tokens_seen": 60073776,
      "step": 28760
    },
    {
      "epoch": 4.692552410473938,
      "grad_norm": 14.6875,
      "learning_rate": 9.11892825300614e-06,
      "loss": 3.2949,
      "num_input_tokens_seen": 60084032,
      "step": 28765
    },
    {
      "epoch": 4.693368137694755,
      "grad_norm": 3.28125,
      "learning_rate": 9.111347318137491e-06,
      "loss": 2.9996,
      "num_input_tokens_seen": 60094944,
      "step": 28770
    },
    {
      "epoch": 4.694183864915573,
      "grad_norm": 2.15625,
      "learning_rate": 9.103768833498755e-06,
      "loss": 2.9596,
      "num_input_tokens_seen": 60106592,
      "step": 28775
    },
    {
      "epoch": 4.69499959213639,
      "grad_norm": 2.375,
      "learning_rate": 9.096192800258639e-06,
      "loss": 2.5903,
      "num_input_tokens_seen": 60116976,
      "step": 28780
    },
    {
      "epoch": 4.695815319357207,
      "grad_norm": 7.28125,
      "learning_rate": 9.088619219585443e-06,
      "loss": 1.9037,
      "num_input_tokens_seen": 60128064,
      "step": 28785
    },
    {
      "epoch": 4.696631046578024,
      "grad_norm": 10.9375,
      "learning_rate": 9.081048092647127e-06,
      "loss": 2.7493,
      "num_input_tokens_seen": 60139104,
      "step": 28790
    },
    {
      "epoch": 4.697446773798841,
      "grad_norm": 9.0,
      "learning_rate": 9.073479420611245e-06,
      "loss": 1.703,
      "num_input_tokens_seen": 60148912,
      "step": 28795
    },
    {
      "epoch": 4.698262501019659,
      "grad_norm": 14.6875,
      "learning_rate": 9.065913204644974e-06,
      "loss": 2.7749,
      "num_input_tokens_seen": 60159984,
      "step": 28800
    },
    {
      "epoch": 4.698262501019659,
      "eval_loss": 2.548642873764038,
      "eval_runtime": 134.8455,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 60159984,
      "step": 28800
    },
    {
      "epoch": 4.699078228240476,
      "grad_norm": 5.90625,
      "learning_rate": 9.058349445915135e-06,
      "loss": 1.5363,
      "num_input_tokens_seen": 60169904,
      "step": 28805
    },
    {
      "epoch": 4.6998939554612935,
      "grad_norm": 9.5,
      "learning_rate": 9.050788145588138e-06,
      "loss": 2.4724,
      "num_input_tokens_seen": 60181232,
      "step": 28810
    },
    {
      "epoch": 4.7007096826821115,
      "grad_norm": 4.71875,
      "learning_rate": 9.043229304830039e-06,
      "loss": 2.4214,
      "num_input_tokens_seen": 60192144,
      "step": 28815
    },
    {
      "epoch": 4.701525409902929,
      "grad_norm": 15.625,
      "learning_rate": 9.035672924806515e-06,
      "loss": 3.5206,
      "num_input_tokens_seen": 60202992,
      "step": 28820
    },
    {
      "epoch": 4.702341137123746,
      "grad_norm": 0.04150390625,
      "learning_rate": 9.028119006682839e-06,
      "loss": 3.0826,
      "num_input_tokens_seen": 60212768,
      "step": 28825
    },
    {
      "epoch": 4.703156864344563,
      "grad_norm": 9.8125,
      "learning_rate": 9.020567551623935e-06,
      "loss": 2.9875,
      "num_input_tokens_seen": 60224016,
      "step": 28830
    },
    {
      "epoch": 4.70397259156538,
      "grad_norm": 5.4375,
      "learning_rate": 9.013018560794318e-06,
      "loss": 2.3459,
      "num_input_tokens_seen": 60234128,
      "step": 28835
    },
    {
      "epoch": 4.704788318786198,
      "grad_norm": 10.8125,
      "learning_rate": 9.005472035358139e-06,
      "loss": 2.9559,
      "num_input_tokens_seen": 60244272,
      "step": 28840
    },
    {
      "epoch": 4.705604046007015,
      "grad_norm": 5.34375,
      "learning_rate": 8.997927976479185e-06,
      "loss": 2.1347,
      "num_input_tokens_seen": 60255376,
      "step": 28845
    },
    {
      "epoch": 4.706419773227832,
      "grad_norm": 9.6875,
      "learning_rate": 8.99038638532082e-06,
      "loss": 1.9097,
      "num_input_tokens_seen": 60265616,
      "step": 28850
    },
    {
      "epoch": 4.70723550044865,
      "grad_norm": 4.96875,
      "learning_rate": 8.982847263046065e-06,
      "loss": 1.4363,
      "num_input_tokens_seen": 60276656,
      "step": 28855
    },
    {
      "epoch": 4.7080512276694675,
      "grad_norm": 12.25,
      "learning_rate": 8.975310610817555e-06,
      "loss": 2.7443,
      "num_input_tokens_seen": 60286624,
      "step": 28860
    },
    {
      "epoch": 4.708866954890285,
      "grad_norm": 8.0625,
      "learning_rate": 8.967776429797528e-06,
      "loss": 1.5421,
      "num_input_tokens_seen": 60297120,
      "step": 28865
    },
    {
      "epoch": 4.709682682111102,
      "grad_norm": 8.0,
      "learning_rate": 8.960244721147842e-06,
      "loss": 1.9063,
      "num_input_tokens_seen": 60309456,
      "step": 28870
    },
    {
      "epoch": 4.71049840933192,
      "grad_norm": 5.84375,
      "learning_rate": 8.952715486029995e-06,
      "loss": 2.4174,
      "num_input_tokens_seen": 60319680,
      "step": 28875
    },
    {
      "epoch": 4.711314136552737,
      "grad_norm": 8.0625,
      "learning_rate": 8.945188725605075e-06,
      "loss": 2.0819,
      "num_input_tokens_seen": 60331120,
      "step": 28880
    },
    {
      "epoch": 4.712129863773554,
      "grad_norm": 10.5625,
      "learning_rate": 8.937664441033817e-06,
      "loss": 2.2217,
      "num_input_tokens_seen": 60340560,
      "step": 28885
    },
    {
      "epoch": 4.712945590994371,
      "grad_norm": 11.125,
      "learning_rate": 8.930142633476549e-06,
      "loss": 2.3769,
      "num_input_tokens_seen": 60349776,
      "step": 28890
    },
    {
      "epoch": 4.713761318215189,
      "grad_norm": 4.71875,
      "learning_rate": 8.92262330409323e-06,
      "loss": 2.496,
      "num_input_tokens_seen": 60359824,
      "step": 28895
    },
    {
      "epoch": 4.714577045436006,
      "grad_norm": 4.65625,
      "learning_rate": 8.915106454043448e-06,
      "loss": 1.6167,
      "num_input_tokens_seen": 60369952,
      "step": 28900
    },
    {
      "epoch": 4.715392772656823,
      "grad_norm": 6.1875,
      "learning_rate": 8.90759208448638e-06,
      "loss": 2.4359,
      "num_input_tokens_seen": 60380368,
      "step": 28905
    },
    {
      "epoch": 4.716208499877641,
      "grad_norm": 9.3125,
      "learning_rate": 8.900080196580848e-06,
      "loss": 3.6153,
      "num_input_tokens_seen": 60390400,
      "step": 28910
    },
    {
      "epoch": 4.717024227098459,
      "grad_norm": 8.75,
      "learning_rate": 8.892570791485267e-06,
      "loss": 1.9987,
      "num_input_tokens_seen": 60402256,
      "step": 28915
    },
    {
      "epoch": 4.717839954319276,
      "grad_norm": 6.1875,
      "learning_rate": 8.885063870357688e-06,
      "loss": 2.3064,
      "num_input_tokens_seen": 60412128,
      "step": 28920
    },
    {
      "epoch": 4.718655681540093,
      "grad_norm": 6.03125,
      "learning_rate": 8.87755943435578e-06,
      "loss": 2.5007,
      "num_input_tokens_seen": 60421648,
      "step": 28925
    },
    {
      "epoch": 4.71947140876091,
      "grad_norm": 8.5,
      "learning_rate": 8.87005748463681e-06,
      "loss": 1.904,
      "num_input_tokens_seen": 60432384,
      "step": 28930
    },
    {
      "epoch": 4.720287135981728,
      "grad_norm": 8.9375,
      "learning_rate": 8.862558022357681e-06,
      "loss": 2.3383,
      "num_input_tokens_seen": 60441088,
      "step": 28935
    },
    {
      "epoch": 4.721102863202545,
      "grad_norm": 8.625,
      "learning_rate": 8.855061048674903e-06,
      "loss": 3.0606,
      "num_input_tokens_seen": 60451392,
      "step": 28940
    },
    {
      "epoch": 4.721918590423362,
      "grad_norm": 7.375,
      "learning_rate": 8.847566564744595e-06,
      "loss": 3.7979,
      "num_input_tokens_seen": 60461024,
      "step": 28945
    },
    {
      "epoch": 4.72273431764418,
      "grad_norm": 2.90625,
      "learning_rate": 8.840074571722512e-06,
      "loss": 1.8077,
      "num_input_tokens_seen": 60471680,
      "step": 28950
    },
    {
      "epoch": 4.723550044864997,
      "grad_norm": 7.4375,
      "learning_rate": 8.832585070764002e-06,
      "loss": 3.1042,
      "num_input_tokens_seen": 60483488,
      "step": 28955
    },
    {
      "epoch": 4.7243657720858145,
      "grad_norm": 5.53125,
      "learning_rate": 8.825098063024045e-06,
      "loss": 1.3728,
      "num_input_tokens_seen": 60494080,
      "step": 28960
    },
    {
      "epoch": 4.725181499306632,
      "grad_norm": 6.96875,
      "learning_rate": 8.817613549657244e-06,
      "loss": 3.4277,
      "num_input_tokens_seen": 60504592,
      "step": 28965
    },
    {
      "epoch": 4.725997226527449,
      "grad_norm": 6.6875,
      "learning_rate": 8.810131531817783e-06,
      "loss": 2.8987,
      "num_input_tokens_seen": 60515984,
      "step": 28970
    },
    {
      "epoch": 4.726812953748267,
      "grad_norm": 5.125,
      "learning_rate": 8.802652010659496e-06,
      "loss": 2.2613,
      "num_input_tokens_seen": 60526352,
      "step": 28975
    },
    {
      "epoch": 4.727628680969084,
      "grad_norm": 9.875,
      "learning_rate": 8.795174987335827e-06,
      "loss": 2.1431,
      "num_input_tokens_seen": 60535920,
      "step": 28980
    },
    {
      "epoch": 4.728444408189901,
      "grad_norm": 6.125,
      "learning_rate": 8.787700462999807e-06,
      "loss": 1.9851,
      "num_input_tokens_seen": 60546976,
      "step": 28985
    },
    {
      "epoch": 4.729260135410719,
      "grad_norm": 5.625,
      "learning_rate": 8.780228438804122e-06,
      "loss": 1.8261,
      "num_input_tokens_seen": 60557520,
      "step": 28990
    },
    {
      "epoch": 4.730075862631536,
      "grad_norm": 5.15625,
      "learning_rate": 8.772758915901032e-06,
      "loss": 4.3183,
      "num_input_tokens_seen": 60568208,
      "step": 28995
    },
    {
      "epoch": 4.730891589852353,
      "grad_norm": 5.34375,
      "learning_rate": 8.765291895442443e-06,
      "loss": 2.3452,
      "num_input_tokens_seen": 60579344,
      "step": 29000
    },
    {
      "epoch": 4.730891589852353,
      "eval_loss": 2.53888201713562,
      "eval_runtime": 134.8359,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.109,
      "num_input_tokens_seen": 60579344,
      "step": 29000
    },
    {
      "epoch": 4.7317073170731705,
      "grad_norm": 5.6875,
      "learning_rate": 8.75782737857987e-06,
      "loss": 1.9156,
      "num_input_tokens_seen": 60590496,
      "step": 29005
    },
    {
      "epoch": 4.732523044293988,
      "grad_norm": 7.78125,
      "learning_rate": 8.750365366464425e-06,
      "loss": 3.7307,
      "num_input_tokens_seen": 60600928,
      "step": 29010
    },
    {
      "epoch": 4.733338771514806,
      "grad_norm": 10.0,
      "learning_rate": 8.742905860246838e-06,
      "loss": 2.5871,
      "num_input_tokens_seen": 60609712,
      "step": 29015
    },
    {
      "epoch": 4.734154498735623,
      "grad_norm": 5.0625,
      "learning_rate": 8.735448861077478e-06,
      "loss": 2.4364,
      "num_input_tokens_seen": 60620160,
      "step": 29020
    },
    {
      "epoch": 4.73497022595644,
      "grad_norm": 6.09375,
      "learning_rate": 8.727994370106288e-06,
      "loss": 3.3468,
      "num_input_tokens_seen": 60630288,
      "step": 29025
    },
    {
      "epoch": 4.735785953177258,
      "grad_norm": 1.5859375,
      "learning_rate": 8.720542388482861e-06,
      "loss": 1.0243,
      "num_input_tokens_seen": 60640128,
      "step": 29030
    },
    {
      "epoch": 4.736601680398075,
      "grad_norm": 3.125,
      "learning_rate": 8.71309291735637e-06,
      "loss": 2.5953,
      "num_input_tokens_seen": 60649488,
      "step": 29035
    },
    {
      "epoch": 4.737417407618892,
      "grad_norm": 0.40234375,
      "learning_rate": 8.705645957875621e-06,
      "loss": 1.3528,
      "num_input_tokens_seen": 60660448,
      "step": 29040
    },
    {
      "epoch": 4.738233134839709,
      "grad_norm": 8.0625,
      "learning_rate": 8.698201511189048e-06,
      "loss": 2.8239,
      "num_input_tokens_seen": 60671168,
      "step": 29045
    },
    {
      "epoch": 4.739048862060527,
      "grad_norm": 5.0,
      "learning_rate": 8.690759578444649e-06,
      "loss": 2.2809,
      "num_input_tokens_seen": 60679776,
      "step": 29050
    },
    {
      "epoch": 4.7398645892813445,
      "grad_norm": 11.9375,
      "learning_rate": 8.68332016079008e-06,
      "loss": 2.9079,
      "num_input_tokens_seen": 60689792,
      "step": 29055
    },
    {
      "epoch": 4.740680316502162,
      "grad_norm": 7.3125,
      "learning_rate": 8.6758832593726e-06,
      "loss": 2.9563,
      "num_input_tokens_seen": 60701568,
      "step": 29060
    },
    {
      "epoch": 4.741496043722979,
      "grad_norm": 1.0703125,
      "learning_rate": 8.668448875339053e-06,
      "loss": 2.2458,
      "num_input_tokens_seen": 60711200,
      "step": 29065
    },
    {
      "epoch": 4.742311770943797,
      "grad_norm": 7.28125,
      "learning_rate": 8.661017009835933e-06,
      "loss": 2.1967,
      "num_input_tokens_seen": 60721456,
      "step": 29070
    },
    {
      "epoch": 4.743127498164614,
      "grad_norm": 6.8125,
      "learning_rate": 8.653587664009311e-06,
      "loss": 2.2493,
      "num_input_tokens_seen": 60730432,
      "step": 29075
    },
    {
      "epoch": 4.743943225385431,
      "grad_norm": 15.9375,
      "learning_rate": 8.646160839004902e-06,
      "loss": 3.0611,
      "num_input_tokens_seen": 60740768,
      "step": 29080
    },
    {
      "epoch": 4.744758952606248,
      "grad_norm": 7.375,
      "learning_rate": 8.638736535967998e-06,
      "loss": 1.5774,
      "num_input_tokens_seen": 60751680,
      "step": 29085
    },
    {
      "epoch": 4.745574679827066,
      "grad_norm": 12.1875,
      "learning_rate": 8.631314756043535e-06,
      "loss": 2.3536,
      "num_input_tokens_seen": 60762928,
      "step": 29090
    },
    {
      "epoch": 4.746390407047883,
      "grad_norm": 5.78125,
      "learning_rate": 8.62389550037603e-06,
      "loss": 1.4272,
      "num_input_tokens_seen": 60772320,
      "step": 29095
    },
    {
      "epoch": 4.7472061342687,
      "grad_norm": 5.59375,
      "learning_rate": 8.616478770109646e-06,
      "loss": 2.0383,
      "num_input_tokens_seen": 60782304,
      "step": 29100
    },
    {
      "epoch": 4.7480218614895175,
      "grad_norm": 6.71875,
      "learning_rate": 8.609064566388111e-06,
      "loss": 1.3792,
      "num_input_tokens_seen": 60793824,
      "step": 29105
    },
    {
      "epoch": 4.748837588710336,
      "grad_norm": 7.34375,
      "learning_rate": 8.601652890354815e-06,
      "loss": 2.6113,
      "num_input_tokens_seen": 60803936,
      "step": 29110
    },
    {
      "epoch": 4.749653315931153,
      "grad_norm": 4.34375,
      "learning_rate": 8.594243743152705e-06,
      "loss": 1.39,
      "num_input_tokens_seen": 60814064,
      "step": 29115
    },
    {
      "epoch": 4.75046904315197,
      "grad_norm": 8.4375,
      "learning_rate": 8.58683712592438e-06,
      "loss": 3.6088,
      "num_input_tokens_seen": 60825728,
      "step": 29120
    },
    {
      "epoch": 4.751284770372787,
      "grad_norm": 5.78125,
      "learning_rate": 8.579433039812037e-06,
      "loss": 3.157,
      "num_input_tokens_seen": 60836592,
      "step": 29125
    },
    {
      "epoch": 4.752100497593605,
      "grad_norm": 5.0625,
      "learning_rate": 8.572031485957466e-06,
      "loss": 2.4529,
      "num_input_tokens_seen": 60846224,
      "step": 29130
    },
    {
      "epoch": 4.752916224814422,
      "grad_norm": 5.40625,
      "learning_rate": 8.564632465502084e-06,
      "loss": 2.7171,
      "num_input_tokens_seen": 60856576,
      "step": 29135
    },
    {
      "epoch": 4.753731952035239,
      "grad_norm": 2.578125,
      "learning_rate": 8.557235979586928e-06,
      "loss": 1.907,
      "num_input_tokens_seen": 60866896,
      "step": 29140
    },
    {
      "epoch": 4.754547679256056,
      "grad_norm": 6.59375,
      "learning_rate": 8.549842029352606e-06,
      "loss": 2.772,
      "num_input_tokens_seen": 60878880,
      "step": 29145
    },
    {
      "epoch": 4.755363406476874,
      "grad_norm": 2.859375,
      "learning_rate": 8.542450615939376e-06,
      "loss": 1.6168,
      "num_input_tokens_seen": 60890768,
      "step": 29150
    },
    {
      "epoch": 4.7561791336976915,
      "grad_norm": 13.5,
      "learning_rate": 8.535061740487082e-06,
      "loss": 2.4291,
      "num_input_tokens_seen": 60900944,
      "step": 29155
    },
    {
      "epoch": 4.756994860918509,
      "grad_norm": 9.8125,
      "learning_rate": 8.527675404135168e-06,
      "loss": 1.6582,
      "num_input_tokens_seen": 60911200,
      "step": 29160
    },
    {
      "epoch": 4.757810588139327,
      "grad_norm": 3.65625,
      "learning_rate": 8.520291608022724e-06,
      "loss": 1.7639,
      "num_input_tokens_seen": 60919824,
      "step": 29165
    },
    {
      "epoch": 4.758626315360144,
      "grad_norm": 5.15625,
      "learning_rate": 8.512910353288398e-06,
      "loss": 1.6367,
      "num_input_tokens_seen": 60930640,
      "step": 29170
    },
    {
      "epoch": 4.759442042580961,
      "grad_norm": 3.921875,
      "learning_rate": 8.505531641070486e-06,
      "loss": 2.7253,
      "num_input_tokens_seen": 60941984,
      "step": 29175
    },
    {
      "epoch": 4.760257769801778,
      "grad_norm": 5.75,
      "learning_rate": 8.498155472506885e-06,
      "loss": 1.9723,
      "num_input_tokens_seen": 60952032,
      "step": 29180
    },
    {
      "epoch": 4.761073497022595,
      "grad_norm": 8.0625,
      "learning_rate": 8.49078184873508e-06,
      "loss": 2.5172,
      "num_input_tokens_seen": 60962800,
      "step": 29185
    },
    {
      "epoch": 4.761889224243413,
      "grad_norm": 5.0,
      "learning_rate": 8.483410770892188e-06,
      "loss": 3.5589,
      "num_input_tokens_seen": 60972320,
      "step": 29190
    },
    {
      "epoch": 4.76270495146423,
      "grad_norm": 3.234375,
      "learning_rate": 8.476042240114909e-06,
      "loss": 2.6208,
      "num_input_tokens_seen": 60983072,
      "step": 29195
    },
    {
      "epoch": 4.7635206786850475,
      "grad_norm": 9.6875,
      "learning_rate": 8.468676257539568e-06,
      "loss": 3.2174,
      "num_input_tokens_seen": 60992736,
      "step": 29200
    },
    {
      "epoch": 4.7635206786850475,
      "eval_loss": 2.533663034439087,
      "eval_runtime": 135.6807,
      "eval_samples_per_second": 20.084,
      "eval_steps_per_second": 10.046,
      "num_input_tokens_seen": 60992736,
      "step": 29200
    },
    {
      "epoch": 4.7643364059058655,
      "grad_norm": 5.90625,
      "learning_rate": 8.4613128243021e-06,
      "loss": 1.4508,
      "num_input_tokens_seen": 61003536,
      "step": 29205
    },
    {
      "epoch": 4.765152133126683,
      "grad_norm": 3.96875,
      "learning_rate": 8.453951941538028e-06,
      "loss": 2.5904,
      "num_input_tokens_seen": 61013056,
      "step": 29210
    },
    {
      "epoch": 4.7659678603475,
      "grad_norm": 6.21875,
      "learning_rate": 8.446593610382495e-06,
      "loss": 3.1013,
      "num_input_tokens_seen": 61022816,
      "step": 29215
    },
    {
      "epoch": 4.766783587568317,
      "grad_norm": 10.75,
      "learning_rate": 8.439237831970259e-06,
      "loss": 1.8063,
      "num_input_tokens_seen": 61033536,
      "step": 29220
    },
    {
      "epoch": 4.767599314789135,
      "grad_norm": 10.8125,
      "learning_rate": 8.431884607435667e-06,
      "loss": 3.6055,
      "num_input_tokens_seen": 61043920,
      "step": 29225
    },
    {
      "epoch": 4.768415042009952,
      "grad_norm": 15.1875,
      "learning_rate": 8.424533937912665e-06,
      "loss": 1.6303,
      "num_input_tokens_seen": 61053872,
      "step": 29230
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 8.6875,
      "learning_rate": 8.41718582453484e-06,
      "loss": 3.2607,
      "num_input_tokens_seen": 61065536,
      "step": 29235
    },
    {
      "epoch": 4.770046496451586,
      "grad_norm": 6.21875,
      "learning_rate": 8.409840268435346e-06,
      "loss": 2.2722,
      "num_input_tokens_seen": 61075280,
      "step": 29240
    },
    {
      "epoch": 4.770862223672404,
      "grad_norm": 11.6875,
      "learning_rate": 8.402497270746976e-06,
      "loss": 3.0279,
      "num_input_tokens_seen": 61085776,
      "step": 29245
    },
    {
      "epoch": 4.771677950893221,
      "grad_norm": 11.625,
      "learning_rate": 8.395156832602095e-06,
      "loss": 2.3957,
      "num_input_tokens_seen": 61095328,
      "step": 29250
    },
    {
      "epoch": 4.772493678114039,
      "grad_norm": 8.375,
      "learning_rate": 8.387818955132707e-06,
      "loss": 2.5582,
      "num_input_tokens_seen": 61105744,
      "step": 29255
    },
    {
      "epoch": 4.773309405334856,
      "grad_norm": 7.09375,
      "learning_rate": 8.38048363947039e-06,
      "loss": 2.2532,
      "num_input_tokens_seen": 61115200,
      "step": 29260
    },
    {
      "epoch": 4.774125132555674,
      "grad_norm": 1.0,
      "learning_rate": 8.373150886746351e-06,
      "loss": 1.7752,
      "num_input_tokens_seen": 61125024,
      "step": 29265
    },
    {
      "epoch": 4.774940859776491,
      "grad_norm": 0.69140625,
      "learning_rate": 8.365820698091397e-06,
      "loss": 1.7667,
      "num_input_tokens_seen": 61136960,
      "step": 29270
    },
    {
      "epoch": 4.775756586997308,
      "grad_norm": 10.9375,
      "learning_rate": 8.358493074635922e-06,
      "loss": 2.0961,
      "num_input_tokens_seen": 61147888,
      "step": 29275
    },
    {
      "epoch": 4.776572314218125,
      "grad_norm": 13.25,
      "learning_rate": 8.351168017509948e-06,
      "loss": 1.7394,
      "num_input_tokens_seen": 61159184,
      "step": 29280
    },
    {
      "epoch": 4.777388041438943,
      "grad_norm": 7.65625,
      "learning_rate": 8.343845527843094e-06,
      "loss": 2.5913,
      "num_input_tokens_seen": 61169680,
      "step": 29285
    },
    {
      "epoch": 4.77820376865976,
      "grad_norm": 6.3125,
      "learning_rate": 8.336525606764566e-06,
      "loss": 2.5112,
      "num_input_tokens_seen": 61179376,
      "step": 29290
    },
    {
      "epoch": 4.779019495880577,
      "grad_norm": 6.0,
      "learning_rate": 8.329208255403204e-06,
      "loss": 1.542,
      "num_input_tokens_seen": 61189840,
      "step": 29295
    },
    {
      "epoch": 4.7798352231013945,
      "grad_norm": 9.0,
      "learning_rate": 8.321893474887426e-06,
      "loss": 2.6158,
      "num_input_tokens_seen": 61199856,
      "step": 29300
    },
    {
      "epoch": 4.7806509503222125,
      "grad_norm": 9.375,
      "learning_rate": 8.31458126634526e-06,
      "loss": 3.2177,
      "num_input_tokens_seen": 61209936,
      "step": 29305
    },
    {
      "epoch": 4.78146667754303,
      "grad_norm": 8.75,
      "learning_rate": 8.30727163090435e-06,
      "loss": 3.0107,
      "num_input_tokens_seen": 61220880,
      "step": 29310
    },
    {
      "epoch": 4.782282404763847,
      "grad_norm": 9.0,
      "learning_rate": 8.29996456969192e-06,
      "loss": 2.0529,
      "num_input_tokens_seen": 61232064,
      "step": 29315
    },
    {
      "epoch": 4.783098131984664,
      "grad_norm": 19.5,
      "learning_rate": 8.292660083834818e-06,
      "loss": 2.5228,
      "num_input_tokens_seen": 61241216,
      "step": 29320
    },
    {
      "epoch": 4.783913859205482,
      "grad_norm": 0.171875,
      "learning_rate": 8.2853581744595e-06,
      "loss": 1.9976,
      "num_input_tokens_seen": 61250272,
      "step": 29325
    },
    {
      "epoch": 4.784729586426299,
      "grad_norm": 8.25,
      "learning_rate": 8.278058842691991e-06,
      "loss": 1.9719,
      "num_input_tokens_seen": 61261872,
      "step": 29330
    },
    {
      "epoch": 4.785545313647116,
      "grad_norm": 6.375,
      "learning_rate": 8.27076208965796e-06,
      "loss": 1.9284,
      "num_input_tokens_seen": 61272016,
      "step": 29335
    },
    {
      "epoch": 4.786361040867934,
      "grad_norm": 8.8125,
      "learning_rate": 8.263467916482637e-06,
      "loss": 2.9884,
      "num_input_tokens_seen": 61283808,
      "step": 29340
    },
    {
      "epoch": 4.787176768088751,
      "grad_norm": 11.625,
      "learning_rate": 8.256176324290885e-06,
      "loss": 2.9141,
      "num_input_tokens_seen": 61295744,
      "step": 29345
    },
    {
      "epoch": 4.7879924953095685,
      "grad_norm": 4.0625,
      "learning_rate": 8.248887314207168e-06,
      "loss": 2.5051,
      "num_input_tokens_seen": 61307520,
      "step": 29350
    },
    {
      "epoch": 4.788808222530386,
      "grad_norm": 7.53125,
      "learning_rate": 8.24160088735553e-06,
      "loss": 3.0049,
      "num_input_tokens_seen": 61318064,
      "step": 29355
    },
    {
      "epoch": 4.789623949751203,
      "grad_norm": 5.125,
      "learning_rate": 8.234317044859629e-06,
      "loss": 1.2895,
      "num_input_tokens_seen": 61329200,
      "step": 29360
    },
    {
      "epoch": 4.790439676972021,
      "grad_norm": 4.75,
      "learning_rate": 8.227035787842744e-06,
      "loss": 2.4549,
      "num_input_tokens_seen": 61338976,
      "step": 29365
    },
    {
      "epoch": 4.791255404192838,
      "grad_norm": 6.84375,
      "learning_rate": 8.219757117427721e-06,
      "loss": 2.316,
      "num_input_tokens_seen": 61350368,
      "step": 29370
    },
    {
      "epoch": 4.792071131413655,
      "grad_norm": 7.15625,
      "learning_rate": 8.212481034737014e-06,
      "loss": 3.7944,
      "num_input_tokens_seen": 61360192,
      "step": 29375
    },
    {
      "epoch": 4.792886858634473,
      "grad_norm": 4.8125,
      "learning_rate": 8.205207540892707e-06,
      "loss": 1.5103,
      "num_input_tokens_seen": 61370912,
      "step": 29380
    },
    {
      "epoch": 4.79370258585529,
      "grad_norm": 1.2890625,
      "learning_rate": 8.197936637016442e-06,
      "loss": 0.4701,
      "num_input_tokens_seen": 61381712,
      "step": 29385
    },
    {
      "epoch": 4.794518313076107,
      "grad_norm": 5.96875,
      "learning_rate": 8.190668324229508e-06,
      "loss": 2.2722,
      "num_input_tokens_seen": 61392752,
      "step": 29390
    },
    {
      "epoch": 4.7953340402969244,
      "grad_norm": 7.53125,
      "learning_rate": 8.183402603652749e-06,
      "loss": 2.5569,
      "num_input_tokens_seen": 61405296,
      "step": 29395
    },
    {
      "epoch": 4.796149767517742,
      "grad_norm": 8.0,
      "learning_rate": 8.176139476406635e-06,
      "loss": 2.856,
      "num_input_tokens_seen": 61414080,
      "step": 29400
    },
    {
      "epoch": 4.796149767517742,
      "eval_loss": 2.5530948638916016,
      "eval_runtime": 134.8735,
      "eval_samples_per_second": 20.204,
      "eval_steps_per_second": 10.106,
      "num_input_tokens_seen": 61414080,
      "step": 29400
    },
    {
      "epoch": 4.79696549473856,
      "grad_norm": 10.1875,
      "learning_rate": 8.16887894361125e-06,
      "loss": 2.3682,
      "num_input_tokens_seen": 61423616,
      "step": 29405
    },
    {
      "epoch": 4.797781221959377,
      "grad_norm": 6.71875,
      "learning_rate": 8.161621006386233e-06,
      "loss": 2.9715,
      "num_input_tokens_seen": 61434048,
      "step": 29410
    },
    {
      "epoch": 4.798596949180194,
      "grad_norm": 8.5625,
      "learning_rate": 8.154365665850869e-06,
      "loss": 1.4095,
      "num_input_tokens_seen": 61442848,
      "step": 29415
    },
    {
      "epoch": 4.799412676401012,
      "grad_norm": 13.4375,
      "learning_rate": 8.147112923124005e-06,
      "loss": 3.8017,
      "num_input_tokens_seen": 61453664,
      "step": 29420
    },
    {
      "epoch": 4.800228403621829,
      "grad_norm": 4.625,
      "learning_rate": 8.13986277932412e-06,
      "loss": 2.6763,
      "num_input_tokens_seen": 61464880,
      "step": 29425
    },
    {
      "epoch": 4.801044130842646,
      "grad_norm": 6.25,
      "learning_rate": 8.132615235569277e-06,
      "loss": 1.9304,
      "num_input_tokens_seen": 61474464,
      "step": 29430
    },
    {
      "epoch": 4.801859858063463,
      "grad_norm": 13.4375,
      "learning_rate": 8.125370292977124e-06,
      "loss": 3.1859,
      "num_input_tokens_seen": 61483520,
      "step": 29435
    },
    {
      "epoch": 4.802675585284281,
      "grad_norm": 4.8125,
      "learning_rate": 8.118127952664944e-06,
      "loss": 2.3823,
      "num_input_tokens_seen": 61494000,
      "step": 29440
    },
    {
      "epoch": 4.803491312505098,
      "grad_norm": 7.28125,
      "learning_rate": 8.110888215749574e-06,
      "loss": 2.3261,
      "num_input_tokens_seen": 61503728,
      "step": 29445
    },
    {
      "epoch": 4.8043070397259156,
      "grad_norm": 5.46875,
      "learning_rate": 8.10365108334749e-06,
      "loss": 3.4818,
      "num_input_tokens_seen": 61514912,
      "step": 29450
    },
    {
      "epoch": 4.805122766946733,
      "grad_norm": 3.0625,
      "learning_rate": 8.096416556574743e-06,
      "loss": 2.3805,
      "num_input_tokens_seen": 61526080,
      "step": 29455
    },
    {
      "epoch": 4.805938494167551,
      "grad_norm": 6.4375,
      "learning_rate": 8.08918463654698e-06,
      "loss": 2.6585,
      "num_input_tokens_seen": 61535920,
      "step": 29460
    },
    {
      "epoch": 4.806754221388368,
      "grad_norm": 6.78125,
      "learning_rate": 8.081955324379458e-06,
      "loss": 2.1325,
      "num_input_tokens_seen": 61547280,
      "step": 29465
    },
    {
      "epoch": 4.807569948609185,
      "grad_norm": 10.75,
      "learning_rate": 8.074728621187039e-06,
      "loss": 3.5367,
      "num_input_tokens_seen": 61559104,
      "step": 29470
    },
    {
      "epoch": 4.808385675830002,
      "grad_norm": 11.3125,
      "learning_rate": 8.067504528084158e-06,
      "loss": 3.5439,
      "num_input_tokens_seen": 61569760,
      "step": 29475
    },
    {
      "epoch": 4.80920140305082,
      "grad_norm": 8.1875,
      "learning_rate": 8.060283046184861e-06,
      "loss": 2.936,
      "num_input_tokens_seen": 61580144,
      "step": 29480
    },
    {
      "epoch": 4.810017130271637,
      "grad_norm": 9.875,
      "learning_rate": 8.053064176602806e-06,
      "loss": 2.8643,
      "num_input_tokens_seen": 61591488,
      "step": 29485
    },
    {
      "epoch": 4.810832857492454,
      "grad_norm": 2.625,
      "learning_rate": 8.045847920451216e-06,
      "loss": 2.1768,
      "num_input_tokens_seen": 61601280,
      "step": 29490
    },
    {
      "epoch": 4.8116485847132715,
      "grad_norm": 3.40625,
      "learning_rate": 8.038634278842944e-06,
      "loss": 2.0438,
      "num_input_tokens_seen": 61612080,
      "step": 29495
    },
    {
      "epoch": 4.8124643119340895,
      "grad_norm": 7.25,
      "learning_rate": 8.031423252890408e-06,
      "loss": 2.5265,
      "num_input_tokens_seen": 61622896,
      "step": 29500
    },
    {
      "epoch": 4.813280039154907,
      "grad_norm": 6.1875,
      "learning_rate": 8.024214843705646e-06,
      "loss": 1.7685,
      "num_input_tokens_seen": 61633136,
      "step": 29505
    },
    {
      "epoch": 4.814095766375724,
      "grad_norm": 2.609375,
      "learning_rate": 8.017009052400295e-06,
      "loss": 1.5523,
      "num_input_tokens_seen": 61643440,
      "step": 29510
    },
    {
      "epoch": 4.814911493596542,
      "grad_norm": 8.8125,
      "learning_rate": 8.00980588008557e-06,
      "loss": 3.2456,
      "num_input_tokens_seen": 61654592,
      "step": 29515
    },
    {
      "epoch": 4.815727220817359,
      "grad_norm": 3.9375,
      "learning_rate": 8.002605327872282e-06,
      "loss": 2.5328,
      "num_input_tokens_seen": 61664112,
      "step": 29520
    },
    {
      "epoch": 4.816542948038176,
      "grad_norm": 14.3125,
      "learning_rate": 7.995407396870862e-06,
      "loss": 3.0823,
      "num_input_tokens_seen": 61674432,
      "step": 29525
    },
    {
      "epoch": 4.817358675258993,
      "grad_norm": 18.0,
      "learning_rate": 7.988212088191307e-06,
      "loss": 2.518,
      "num_input_tokens_seen": 61684576,
      "step": 29530
    },
    {
      "epoch": 4.81817440247981,
      "grad_norm": 9.125,
      "learning_rate": 7.98101940294324e-06,
      "loss": 2.5328,
      "num_input_tokens_seen": 61696240,
      "step": 29535
    },
    {
      "epoch": 4.818990129700628,
      "grad_norm": 7.78125,
      "learning_rate": 7.973829342235847e-06,
      "loss": 2.5787,
      "num_input_tokens_seen": 61706752,
      "step": 29540
    },
    {
      "epoch": 4.8198058569214455,
      "grad_norm": 12.125,
      "learning_rate": 7.966641907177936e-06,
      "loss": 3.3015,
      "num_input_tokens_seen": 61716576,
      "step": 29545
    },
    {
      "epoch": 4.820621584142263,
      "grad_norm": 14.25,
      "learning_rate": 7.959457098877901e-06,
      "loss": 2.4609,
      "num_input_tokens_seen": 61726368,
      "step": 29550
    },
    {
      "epoch": 4.821437311363081,
      "grad_norm": 8.875,
      "learning_rate": 7.952274918443719e-06,
      "loss": 3.0672,
      "num_input_tokens_seen": 61736704,
      "step": 29555
    },
    {
      "epoch": 4.822253038583898,
      "grad_norm": 9.875,
      "learning_rate": 7.945095366982983e-06,
      "loss": 2.2774,
      "num_input_tokens_seen": 61747296,
      "step": 29560
    },
    {
      "epoch": 4.823068765804715,
      "grad_norm": 6.84375,
      "learning_rate": 7.937918445602871e-06,
      "loss": 2.2685,
      "num_input_tokens_seen": 61757616,
      "step": 29565
    },
    {
      "epoch": 4.823884493025532,
      "grad_norm": 6.21875,
      "learning_rate": 7.930744155410145e-06,
      "loss": 2.6905,
      "num_input_tokens_seen": 61767120,
      "step": 29570
    },
    {
      "epoch": 4.824700220246349,
      "grad_norm": 12.25,
      "learning_rate": 7.923572497511181e-06,
      "loss": 2.5817,
      "num_input_tokens_seen": 61777584,
      "step": 29575
    },
    {
      "epoch": 4.825515947467167,
      "grad_norm": 3.203125,
      "learning_rate": 7.916403473011927e-06,
      "loss": 1.6481,
      "num_input_tokens_seen": 61787792,
      "step": 29580
    },
    {
      "epoch": 4.826331674687984,
      "grad_norm": 7.25,
      "learning_rate": 7.909237083017953e-06,
      "loss": 1.7452,
      "num_input_tokens_seen": 61797760,
      "step": 29585
    },
    {
      "epoch": 4.827147401908801,
      "grad_norm": 12.0,
      "learning_rate": 7.902073328634389e-06,
      "loss": 5.3937,
      "num_input_tokens_seen": 61808720,
      "step": 29590
    },
    {
      "epoch": 4.8279631291296194,
      "grad_norm": 9.5625,
      "learning_rate": 7.894912210965987e-06,
      "loss": 2.2002,
      "num_input_tokens_seen": 61818624,
      "step": 29595
    },
    {
      "epoch": 4.828778856350437,
      "grad_norm": 5.71875,
      "learning_rate": 7.887753731117075e-06,
      "loss": 3.9432,
      "num_input_tokens_seen": 61829776,
      "step": 29600
    },
    {
      "epoch": 4.828778856350437,
      "eval_loss": 2.546984910964966,
      "eval_runtime": 135.0121,
      "eval_samples_per_second": 20.183,
      "eval_steps_per_second": 10.095,
      "num_input_tokens_seen": 61829776,
      "step": 29600
    },
    {
      "epoch": 4.829594583571254,
      "grad_norm": 15.125,
      "learning_rate": 7.880597890191587e-06,
      "loss": 2.3684,
      "num_input_tokens_seen": 61839296,
      "step": 29605
    },
    {
      "epoch": 4.830410310792071,
      "grad_norm": 5.46875,
      "learning_rate": 7.873444689293036e-06,
      "loss": 2.4135,
      "num_input_tokens_seen": 61850144,
      "step": 29610
    },
    {
      "epoch": 4.831226038012889,
      "grad_norm": 5.5,
      "learning_rate": 7.866294129524548e-06,
      "loss": 3.3656,
      "num_input_tokens_seen": 61859952,
      "step": 29615
    },
    {
      "epoch": 4.832041765233706,
      "grad_norm": 3.734375,
      "learning_rate": 7.859146211988811e-06,
      "loss": 2.0489,
      "num_input_tokens_seen": 61870560,
      "step": 29620
    },
    {
      "epoch": 4.832857492454523,
      "grad_norm": 5.96875,
      "learning_rate": 7.852000937788134e-06,
      "loss": 2.7118,
      "num_input_tokens_seen": 61882064,
      "step": 29625
    },
    {
      "epoch": 4.83367321967534,
      "grad_norm": 6.21875,
      "learning_rate": 7.844858308024416e-06,
      "loss": 2.4155,
      "num_input_tokens_seen": 61892912,
      "step": 29630
    },
    {
      "epoch": 4.834488946896158,
      "grad_norm": 6.96875,
      "learning_rate": 7.837718323799122e-06,
      "loss": 2.2361,
      "num_input_tokens_seen": 61902944,
      "step": 29635
    },
    {
      "epoch": 4.835304674116975,
      "grad_norm": 6.8125,
      "learning_rate": 7.83058098621334e-06,
      "loss": 2.5438,
      "num_input_tokens_seen": 61912416,
      "step": 29640
    },
    {
      "epoch": 4.8361204013377925,
      "grad_norm": 9.8125,
      "learning_rate": 7.823446296367739e-06,
      "loss": 2.3656,
      "num_input_tokens_seen": 61923968,
      "step": 29645
    },
    {
      "epoch": 4.83693612855861,
      "grad_norm": 5.21875,
      "learning_rate": 7.81631425536257e-06,
      "loss": 1.8065,
      "num_input_tokens_seen": 61935376,
      "step": 29650
    },
    {
      "epoch": 4.837751855779428,
      "grad_norm": 2.265625,
      "learning_rate": 7.809184864297689e-06,
      "loss": 3.0252,
      "num_input_tokens_seen": 61946224,
      "step": 29655
    },
    {
      "epoch": 4.838567583000245,
      "grad_norm": 4.78125,
      "learning_rate": 7.802058124272532e-06,
      "loss": 1.5386,
      "num_input_tokens_seen": 61955136,
      "step": 29660
    },
    {
      "epoch": 4.839383310221062,
      "grad_norm": 6.09375,
      "learning_rate": 7.79493403638614e-06,
      "loss": 1.2349,
      "num_input_tokens_seen": 61966608,
      "step": 29665
    },
    {
      "epoch": 4.840199037441879,
      "grad_norm": 5.9375,
      "learning_rate": 7.787812601737132e-06,
      "loss": 2.4944,
      "num_input_tokens_seen": 61977280,
      "step": 29670
    },
    {
      "epoch": 4.841014764662697,
      "grad_norm": 9.25,
      "learning_rate": 7.780693821423715e-06,
      "loss": 2.0226,
      "num_input_tokens_seen": 61988224,
      "step": 29675
    },
    {
      "epoch": 4.841830491883514,
      "grad_norm": 10.0625,
      "learning_rate": 7.773577696543705e-06,
      "loss": 3.0113,
      "num_input_tokens_seen": 62000544,
      "step": 29680
    },
    {
      "epoch": 4.842646219104331,
      "grad_norm": 7.6875,
      "learning_rate": 7.7664642281945e-06,
      "loss": 2.6332,
      "num_input_tokens_seen": 62011632,
      "step": 29685
    },
    {
      "epoch": 4.8434619463251485,
      "grad_norm": 4.40625,
      "learning_rate": 7.759353417473072e-06,
      "loss": 1.5736,
      "num_input_tokens_seen": 62022848,
      "step": 29690
    },
    {
      "epoch": 4.8442776735459665,
      "grad_norm": 8.0,
      "learning_rate": 7.752245265476016e-06,
      "loss": 2.2513,
      "num_input_tokens_seen": 62034448,
      "step": 29695
    },
    {
      "epoch": 4.845093400766784,
      "grad_norm": 5.71875,
      "learning_rate": 7.745139773299481e-06,
      "loss": 1.4916,
      "num_input_tokens_seen": 62045136,
      "step": 29700
    },
    {
      "epoch": 4.845909127987601,
      "grad_norm": 4.21875,
      "learning_rate": 7.738036942039232e-06,
      "loss": 2.4376,
      "num_input_tokens_seen": 62055600,
      "step": 29705
    },
    {
      "epoch": 4.846724855208418,
      "grad_norm": 2.234375,
      "learning_rate": 7.73093677279062e-06,
      "loss": 3.9861,
      "num_input_tokens_seen": 62065824,
      "step": 29710
    },
    {
      "epoch": 4.847540582429236,
      "grad_norm": 6.8125,
      "learning_rate": 7.72383926664857e-06,
      "loss": 1.534,
      "num_input_tokens_seen": 62076368,
      "step": 29715
    },
    {
      "epoch": 4.848356309650053,
      "grad_norm": 5.5,
      "learning_rate": 7.716744424707606e-06,
      "loss": 1.4386,
      "num_input_tokens_seen": 62086064,
      "step": 29720
    },
    {
      "epoch": 4.84917203687087,
      "grad_norm": 10.3125,
      "learning_rate": 7.709652248061858e-06,
      "loss": 2.4453,
      "num_input_tokens_seen": 62095968,
      "step": 29725
    },
    {
      "epoch": 4.849987764091688,
      "grad_norm": 8.125,
      "learning_rate": 7.702562737805017e-06,
      "loss": 1.915,
      "num_input_tokens_seen": 62107440,
      "step": 29730
    },
    {
      "epoch": 4.850803491312505,
      "grad_norm": 1.9765625,
      "learning_rate": 7.695475895030365e-06,
      "loss": 2.3427,
      "num_input_tokens_seen": 62116608,
      "step": 29735
    },
    {
      "epoch": 4.8516192185333225,
      "grad_norm": 1.453125,
      "learning_rate": 7.6883917208308e-06,
      "loss": 1.6399,
      "num_input_tokens_seen": 62126384,
      "step": 29740
    },
    {
      "epoch": 4.85243494575414,
      "grad_norm": 0.173828125,
      "learning_rate": 7.681310216298778e-06,
      "loss": 1.3435,
      "num_input_tokens_seen": 62137760,
      "step": 29745
    },
    {
      "epoch": 4.853250672974957,
      "grad_norm": 6.4375,
      "learning_rate": 7.674231382526367e-06,
      "loss": 2.5647,
      "num_input_tokens_seen": 62148448,
      "step": 29750
    },
    {
      "epoch": 4.854066400195775,
      "grad_norm": 9.0625,
      "learning_rate": 7.667155220605198e-06,
      "loss": 2.3171,
      "num_input_tokens_seen": 62158992,
      "step": 29755
    },
    {
      "epoch": 4.854882127416592,
      "grad_norm": 11.125,
      "learning_rate": 7.660081731626515e-06,
      "loss": 2.8463,
      "num_input_tokens_seen": 62170272,
      "step": 29760
    },
    {
      "epoch": 4.855697854637409,
      "grad_norm": 6.1875,
      "learning_rate": 7.653010916681141e-06,
      "loss": 2.5586,
      "num_input_tokens_seen": 62180288,
      "step": 29765
    },
    {
      "epoch": 4.856513581858227,
      "grad_norm": 11.6875,
      "learning_rate": 7.645942776859472e-06,
      "loss": 2.8107,
      "num_input_tokens_seen": 62190176,
      "step": 29770
    },
    {
      "epoch": 4.857329309079044,
      "grad_norm": 12.625,
      "learning_rate": 7.63887731325152e-06,
      "loss": 2.0999,
      "num_input_tokens_seen": 62199616,
      "step": 29775
    },
    {
      "epoch": 4.858145036299861,
      "grad_norm": 9.25,
      "learning_rate": 7.63181452694685e-06,
      "loss": 2.6412,
      "num_input_tokens_seen": 62209456,
      "step": 29780
    },
    {
      "epoch": 4.858960763520678,
      "grad_norm": 5.65625,
      "learning_rate": 7.624754419034644e-06,
      "loss": 1.8011,
      "num_input_tokens_seen": 62219408,
      "step": 29785
    },
    {
      "epoch": 4.859776490741496,
      "grad_norm": 9.4375,
      "learning_rate": 7.6176969906036645e-06,
      "loss": 3.2227,
      "num_input_tokens_seen": 62229968,
      "step": 29790
    },
    {
      "epoch": 4.8605922179623136,
      "grad_norm": 8.25,
      "learning_rate": 7.610642242742242e-06,
      "loss": 3.4279,
      "num_input_tokens_seen": 62239616,
      "step": 29795
    },
    {
      "epoch": 4.861407945183131,
      "grad_norm": 3.375,
      "learning_rate": 7.603590176538322e-06,
      "loss": 2.0286,
      "num_input_tokens_seen": 62250704,
      "step": 29800
    },
    {
      "epoch": 4.861407945183131,
      "eval_loss": 2.546522855758667,
      "eval_runtime": 134.8504,
      "eval_samples_per_second": 20.208,
      "eval_steps_per_second": 10.107,
      "num_input_tokens_seen": 62250704,
      "step": 29800
    },
    {
      "epoch": 4.862223672403948,
      "grad_norm": 7.78125,
      "learning_rate": 7.596540793079404e-06,
      "loss": 3.2345,
      "num_input_tokens_seen": 62262048,
      "step": 29805
    },
    {
      "epoch": 4.863039399624766,
      "grad_norm": 10.8125,
      "learning_rate": 7.5894940934526125e-06,
      "loss": 2.3583,
      "num_input_tokens_seen": 62271216,
      "step": 29810
    },
    {
      "epoch": 4.863855126845583,
      "grad_norm": 8.125,
      "learning_rate": 7.582450078744621e-06,
      "loss": 2.3915,
      "num_input_tokens_seen": 62282272,
      "step": 29815
    },
    {
      "epoch": 4.8646708540664,
      "grad_norm": 7.15625,
      "learning_rate": 7.575408750041707e-06,
      "loss": 1.2879,
      "num_input_tokens_seen": 62293152,
      "step": 29820
    },
    {
      "epoch": 4.865486581287217,
      "grad_norm": 2.59375,
      "learning_rate": 7.568370108429732e-06,
      "loss": 2.0493,
      "num_input_tokens_seen": 62303536,
      "step": 29825
    },
    {
      "epoch": 4.866302308508035,
      "grad_norm": 3.640625,
      "learning_rate": 7.561334154994154e-06,
      "loss": 2.7441,
      "num_input_tokens_seen": 62315040,
      "step": 29830
    },
    {
      "epoch": 4.867118035728852,
      "grad_norm": 4.8125,
      "learning_rate": 7.55430089081999e-06,
      "loss": 2.2024,
      "num_input_tokens_seen": 62325584,
      "step": 29835
    },
    {
      "epoch": 4.8679337629496695,
      "grad_norm": 6.34375,
      "learning_rate": 7.547270316991864e-06,
      "loss": 3.1147,
      "num_input_tokens_seen": 62335856,
      "step": 29840
    },
    {
      "epoch": 4.868749490170487,
      "grad_norm": 6.65625,
      "learning_rate": 7.5402424345939884e-06,
      "loss": 2.425,
      "num_input_tokens_seen": 62346928,
      "step": 29845
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 3.1875,
      "learning_rate": 7.533217244710133e-06,
      "loss": 2.006,
      "num_input_tokens_seen": 62356368,
      "step": 29850
    },
    {
      "epoch": 4.870380944612122,
      "grad_norm": 6.1875,
      "learning_rate": 7.52619474842369e-06,
      "loss": 1.2246,
      "num_input_tokens_seen": 62366816,
      "step": 29855
    },
    {
      "epoch": 4.871196671832939,
      "grad_norm": 12.3125,
      "learning_rate": 7.519174946817597e-06,
      "loss": 2.5904,
      "num_input_tokens_seen": 62377952,
      "step": 29860
    },
    {
      "epoch": 4.872012399053756,
      "grad_norm": 8.0625,
      "learning_rate": 7.512157840974407e-06,
      "loss": 2.5627,
      "num_input_tokens_seen": 62388736,
      "step": 29865
    },
    {
      "epoch": 4.872828126274574,
      "grad_norm": 8.8125,
      "learning_rate": 7.5051434319762496e-06,
      "loss": 2.8589,
      "num_input_tokens_seen": 62397968,
      "step": 29870
    },
    {
      "epoch": 4.873643853495391,
      "grad_norm": 9.4375,
      "learning_rate": 7.498131720904822e-06,
      "loss": 2.708,
      "num_input_tokens_seen": 62408064,
      "step": 29875
    },
    {
      "epoch": 4.874459580716208,
      "grad_norm": 9.9375,
      "learning_rate": 7.491122708841433e-06,
      "loss": 2.5858,
      "num_input_tokens_seen": 62418992,
      "step": 29880
    },
    {
      "epoch": 4.8752753079370255,
      "grad_norm": 0.049560546875,
      "learning_rate": 7.4841163968669524e-06,
      "loss": 2.2461,
      "num_input_tokens_seen": 62429392,
      "step": 29885
    },
    {
      "epoch": 4.8760910351578435,
      "grad_norm": 6.21875,
      "learning_rate": 7.4771127860618355e-06,
      "loss": 2.5393,
      "num_input_tokens_seen": 62438288,
      "step": 29890
    },
    {
      "epoch": 4.876906762378661,
      "grad_norm": 16.875,
      "learning_rate": 7.470111877506139e-06,
      "loss": 2.4143,
      "num_input_tokens_seen": 62447328,
      "step": 29895
    },
    {
      "epoch": 4.877722489599478,
      "grad_norm": 7.75,
      "learning_rate": 7.463113672279479e-06,
      "loss": 2.3594,
      "num_input_tokens_seen": 62457248,
      "step": 29900
    },
    {
      "epoch": 4.878538216820296,
      "grad_norm": 12.75,
      "learning_rate": 7.456118171461071e-06,
      "loss": 2.324,
      "num_input_tokens_seen": 62467056,
      "step": 29905
    },
    {
      "epoch": 4.879353944041113,
      "grad_norm": 10.0625,
      "learning_rate": 7.449125376129721e-06,
      "loss": 2.4611,
      "num_input_tokens_seen": 62477072,
      "step": 29910
    },
    {
      "epoch": 4.88016967126193,
      "grad_norm": 6.25,
      "learning_rate": 7.442135287363788e-06,
      "loss": 1.9883,
      "num_input_tokens_seen": 62487728,
      "step": 29915
    },
    {
      "epoch": 4.880985398482747,
      "grad_norm": 5.03125,
      "learning_rate": 7.435147906241247e-06,
      "loss": 2.6579,
      "num_input_tokens_seen": 62498880,
      "step": 29920
    },
    {
      "epoch": 4.881801125703564,
      "grad_norm": 4.125,
      "learning_rate": 7.428163233839624e-06,
      "loss": 3.1278,
      "num_input_tokens_seen": 62508048,
      "step": 29925
    },
    {
      "epoch": 4.882616852924382,
      "grad_norm": 8.9375,
      "learning_rate": 7.4211812712360525e-06,
      "loss": 2.5224,
      "num_input_tokens_seen": 62518464,
      "step": 29930
    },
    {
      "epoch": 4.883432580145199,
      "grad_norm": 8.3125,
      "learning_rate": 7.4142020195072464e-06,
      "loss": 2.5991,
      "num_input_tokens_seen": 62528144,
      "step": 29935
    },
    {
      "epoch": 4.884248307366017,
      "grad_norm": 8.8125,
      "learning_rate": 7.407225479729479e-06,
      "loss": 2.6162,
      "num_input_tokens_seen": 62538560,
      "step": 29940
    },
    {
      "epoch": 4.885064034586835,
      "grad_norm": 6.0625,
      "learning_rate": 7.400251652978632e-06,
      "loss": 1.4784,
      "num_input_tokens_seen": 62549648,
      "step": 29945
    },
    {
      "epoch": 4.885879761807652,
      "grad_norm": 8.625,
      "learning_rate": 7.393280540330147e-06,
      "loss": 3.1086,
      "num_input_tokens_seen": 62560480,
      "step": 29950
    },
    {
      "epoch": 4.886695489028469,
      "grad_norm": 4.96875,
      "learning_rate": 7.386312142859069e-06,
      "loss": 2.4506,
      "num_input_tokens_seen": 62569552,
      "step": 29955
    },
    {
      "epoch": 4.887511216249286,
      "grad_norm": 8.625,
      "learning_rate": 7.379346461640008e-06,
      "loss": 2.3781,
      "num_input_tokens_seen": 62580816,
      "step": 29960
    },
    {
      "epoch": 4.888326943470103,
      "grad_norm": 10.3125,
      "learning_rate": 7.372383497747149e-06,
      "loss": 2.6042,
      "num_input_tokens_seen": 62590096,
      "step": 29965
    },
    {
      "epoch": 4.889142670690921,
      "grad_norm": 9.6875,
      "learning_rate": 7.3654232522542775e-06,
      "loss": 2.6264,
      "num_input_tokens_seen": 62600256,
      "step": 29970
    },
    {
      "epoch": 4.889958397911738,
      "grad_norm": 3.875,
      "learning_rate": 7.358465726234756e-06,
      "loss": 2.0949,
      "num_input_tokens_seen": 62609792,
      "step": 29975
    },
    {
      "epoch": 4.890774125132555,
      "grad_norm": 3.984375,
      "learning_rate": 7.351510920761512e-06,
      "loss": 2.3601,
      "num_input_tokens_seen": 62620736,
      "step": 29980
    },
    {
      "epoch": 4.891589852353373,
      "grad_norm": 14.0625,
      "learning_rate": 7.344558836907067e-06,
      "loss": 2.8682,
      "num_input_tokens_seen": 62631904,
      "step": 29985
    },
    {
      "epoch": 4.8924055795741905,
      "grad_norm": 3.421875,
      "learning_rate": 7.3376094757435285e-06,
      "loss": 2.6288,
      "num_input_tokens_seen": 62642752,
      "step": 29990
    },
    {
      "epoch": 4.893221306795008,
      "grad_norm": 14.5,
      "learning_rate": 7.330662838342561e-06,
      "loss": 2.2831,
      "num_input_tokens_seen": 62653248,
      "step": 29995
    },
    {
      "epoch": 4.894037034015825,
      "grad_norm": 2.25,
      "learning_rate": 7.323718925775438e-06,
      "loss": 1.1607,
      "num_input_tokens_seen": 62662656,
      "step": 30000
    },
    {
      "epoch": 4.894037034015825,
      "eval_loss": 2.540228843688965,
      "eval_runtime": 135.0277,
      "eval_samples_per_second": 20.181,
      "eval_steps_per_second": 10.094,
      "num_input_tokens_seen": 62662656,
      "step": 30000
    },
    {
      "epoch": 4.894852761236643,
      "grad_norm": 6.53125,
      "learning_rate": 7.316777739112985e-06,
      "loss": 2.4085,
      "num_input_tokens_seen": 62671264,
      "step": 30005
    },
    {
      "epoch": 4.89566848845746,
      "grad_norm": 17.125,
      "learning_rate": 7.309839279425626e-06,
      "loss": 3.0376,
      "num_input_tokens_seen": 62682720,
      "step": 30010
    },
    {
      "epoch": 4.896484215678277,
      "grad_norm": 6.9375,
      "learning_rate": 7.302903547783366e-06,
      "loss": 1.9098,
      "num_input_tokens_seen": 62694144,
      "step": 30015
    },
    {
      "epoch": 4.897299942899094,
      "grad_norm": 3.9375,
      "learning_rate": 7.2959705452557644e-06,
      "loss": 2.577,
      "num_input_tokens_seen": 62706368,
      "step": 30020
    },
    {
      "epoch": 4.898115670119912,
      "grad_norm": 12.75,
      "learning_rate": 7.289040272911996e-06,
      "loss": 1.1663,
      "num_input_tokens_seen": 62717952,
      "step": 30025
    },
    {
      "epoch": 4.898931397340729,
      "grad_norm": 8.5625,
      "learning_rate": 7.282112731820789e-06,
      "loss": 2.4607,
      "num_input_tokens_seen": 62729552,
      "step": 30030
    },
    {
      "epoch": 4.8997471245615465,
      "grad_norm": 7.03125,
      "learning_rate": 7.275187923050447e-06,
      "loss": 1.9197,
      "num_input_tokens_seen": 62740064,
      "step": 30035
    },
    {
      "epoch": 4.900562851782364,
      "grad_norm": 7.0625,
      "learning_rate": 7.268265847668879e-06,
      "loss": 2.5057,
      "num_input_tokens_seen": 62751072,
      "step": 30040
    },
    {
      "epoch": 4.901378579003182,
      "grad_norm": 7.65625,
      "learning_rate": 7.261346506743538e-06,
      "loss": 4.4008,
      "num_input_tokens_seen": 62762304,
      "step": 30045
    },
    {
      "epoch": 4.902194306223999,
      "grad_norm": 5.625,
      "learning_rate": 7.254429901341486e-06,
      "loss": 2.3132,
      "num_input_tokens_seen": 62772432,
      "step": 30050
    },
    {
      "epoch": 4.903010033444816,
      "grad_norm": 9.3125,
      "learning_rate": 7.247516032529356e-06,
      "loss": 4.2587,
      "num_input_tokens_seen": 62782064,
      "step": 30055
    },
    {
      "epoch": 4.903825760665633,
      "grad_norm": 9.8125,
      "learning_rate": 7.240604901373338e-06,
      "loss": 2.4058,
      "num_input_tokens_seen": 62791744,
      "step": 30060
    },
    {
      "epoch": 4.904641487886451,
      "grad_norm": 5.9375,
      "learning_rate": 7.233696508939223e-06,
      "loss": 2.3155,
      "num_input_tokens_seen": 62802240,
      "step": 30065
    },
    {
      "epoch": 4.905457215107268,
      "grad_norm": 7.34375,
      "learning_rate": 7.226790856292376e-06,
      "loss": 2.2223,
      "num_input_tokens_seen": 62813120,
      "step": 30070
    },
    {
      "epoch": 4.906272942328085,
      "grad_norm": 3.515625,
      "learning_rate": 7.219887944497727e-06,
      "loss": 2.52,
      "num_input_tokens_seen": 62823792,
      "step": 30075
    },
    {
      "epoch": 4.907088669548903,
      "grad_norm": 5.625,
      "learning_rate": 7.2129877746198e-06,
      "loss": 3.1014,
      "num_input_tokens_seen": 62833440,
      "step": 30080
    },
    {
      "epoch": 4.9079043967697205,
      "grad_norm": 6.0,
      "learning_rate": 7.20609034772268e-06,
      "loss": 2.46,
      "num_input_tokens_seen": 62844544,
      "step": 30085
    },
    {
      "epoch": 4.908720123990538,
      "grad_norm": 7.46875,
      "learning_rate": 7.19919566487004e-06,
      "loss": 2.5033,
      "num_input_tokens_seen": 62854832,
      "step": 30090
    },
    {
      "epoch": 4.909535851211355,
      "grad_norm": 5.8125,
      "learning_rate": 7.192303727125132e-06,
      "loss": 1.0406,
      "num_input_tokens_seen": 62865248,
      "step": 30095
    },
    {
      "epoch": 4.910351578432172,
      "grad_norm": 5.75,
      "learning_rate": 7.185414535550777e-06,
      "loss": 1.8641,
      "num_input_tokens_seen": 62876384,
      "step": 30100
    },
    {
      "epoch": 4.91116730565299,
      "grad_norm": 3.828125,
      "learning_rate": 7.178528091209363e-06,
      "loss": 2.0712,
      "num_input_tokens_seen": 62888000,
      "step": 30105
    },
    {
      "epoch": 4.911983032873807,
      "grad_norm": 5.1875,
      "learning_rate": 7.171644395162888e-06,
      "loss": 1.693,
      "num_input_tokens_seen": 62899568,
      "step": 30110
    },
    {
      "epoch": 4.912798760094624,
      "grad_norm": 9.4375,
      "learning_rate": 7.164763448472881e-06,
      "loss": 1.6014,
      "num_input_tokens_seen": 62910944,
      "step": 30115
    },
    {
      "epoch": 4.913614487315442,
      "grad_norm": 2.140625,
      "learning_rate": 7.157885252200491e-06,
      "loss": 1.1534,
      "num_input_tokens_seen": 62921744,
      "step": 30120
    },
    {
      "epoch": 4.914430214536259,
      "grad_norm": 2.09375,
      "learning_rate": 7.151009807406403e-06,
      "loss": 2.7158,
      "num_input_tokens_seen": 62931952,
      "step": 30125
    },
    {
      "epoch": 4.915245941757076,
      "grad_norm": 5.78125,
      "learning_rate": 7.144137115150909e-06,
      "loss": 1.3309,
      "num_input_tokens_seen": 62943808,
      "step": 30130
    },
    {
      "epoch": 4.9160616689778935,
      "grad_norm": 10.625,
      "learning_rate": 7.1372671764938725e-06,
      "loss": 1.5028,
      "num_input_tokens_seen": 62955200,
      "step": 30135
    },
    {
      "epoch": 4.916877396198711,
      "grad_norm": 6.28125,
      "learning_rate": 7.130399992494705e-06,
      "loss": 1.5694,
      "num_input_tokens_seen": 62964256,
      "step": 30140
    },
    {
      "epoch": 4.917693123419529,
      "grad_norm": 3.59375,
      "learning_rate": 7.123535564212419e-06,
      "loss": 2.4667,
      "num_input_tokens_seen": 62975280,
      "step": 30145
    },
    {
      "epoch": 4.918508850640346,
      "grad_norm": 8.3125,
      "learning_rate": 7.116673892705611e-06,
      "loss": 1.4133,
      "num_input_tokens_seen": 62986288,
      "step": 30150
    },
    {
      "epoch": 4.919324577861163,
      "grad_norm": 4.28125,
      "learning_rate": 7.109814979032415e-06,
      "loss": 2.5597,
      "num_input_tokens_seen": 62997504,
      "step": 30155
    },
    {
      "epoch": 4.920140305081981,
      "grad_norm": 3.59375,
      "learning_rate": 7.102958824250577e-06,
      "loss": 2.0136,
      "num_input_tokens_seen": 63007344,
      "step": 30160
    },
    {
      "epoch": 4.920956032302798,
      "grad_norm": 8.625,
      "learning_rate": 7.096105429417393e-06,
      "loss": 2.6571,
      "num_input_tokens_seen": 63017920,
      "step": 30165
    },
    {
      "epoch": 4.921771759523615,
      "grad_norm": 5.40625,
      "learning_rate": 7.0892547955897506e-06,
      "loss": 1.8355,
      "num_input_tokens_seen": 63027088,
      "step": 30170
    },
    {
      "epoch": 4.922587486744432,
      "grad_norm": 4.71875,
      "learning_rate": 7.0824069238241e-06,
      "loss": 1.1425,
      "num_input_tokens_seen": 63037840,
      "step": 30175
    },
    {
      "epoch": 4.92340321396525,
      "grad_norm": 4.34375,
      "learning_rate": 7.075561815176462e-06,
      "loss": 2.4012,
      "num_input_tokens_seen": 63047552,
      "step": 30180
    },
    {
      "epoch": 4.9242189411860675,
      "grad_norm": 2.5,
      "learning_rate": 7.068719470702445e-06,
      "loss": 3.8797,
      "num_input_tokens_seen": 63057456,
      "step": 30185
    },
    {
      "epoch": 4.925034668406885,
      "grad_norm": 0.185546875,
      "learning_rate": 7.061879891457229e-06,
      "loss": 2.5669,
      "num_input_tokens_seen": 63068464,
      "step": 30190
    },
    {
      "epoch": 4.925850395627702,
      "grad_norm": 10.75,
      "learning_rate": 7.0550430784955515e-06,
      "loss": 2.2032,
      "num_input_tokens_seen": 63078112,
      "step": 30195
    },
    {
      "epoch": 4.92666612284852,
      "grad_norm": 5.6875,
      "learning_rate": 7.048209032871752e-06,
      "loss": 1.9432,
      "num_input_tokens_seen": 63088352,
      "step": 30200
    },
    {
      "epoch": 4.92666612284852,
      "eval_loss": 2.5601696968078613,
      "eval_runtime": 134.8372,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 63088352,
      "step": 30200
    },
    {
      "epoch": 4.927481850069337,
      "grad_norm": 6.34375,
      "learning_rate": 7.0413777556397055e-06,
      "loss": 1.7634,
      "num_input_tokens_seen": 63098240,
      "step": 30205
    },
    {
      "epoch": 4.928297577290154,
      "grad_norm": 5.84375,
      "learning_rate": 7.0345492478528925e-06,
      "loss": 3.1316,
      "num_input_tokens_seen": 63108928,
      "step": 30210
    },
    {
      "epoch": 4.929113304510971,
      "grad_norm": 7.71875,
      "learning_rate": 7.02772351056436e-06,
      "loss": 2.1743,
      "num_input_tokens_seen": 63119520,
      "step": 30215
    },
    {
      "epoch": 4.929929031731789,
      "grad_norm": 4.5625,
      "learning_rate": 7.020900544826709e-06,
      "loss": 1.6402,
      "num_input_tokens_seen": 63130384,
      "step": 30220
    },
    {
      "epoch": 4.930744758952606,
      "grad_norm": 8.0,
      "learning_rate": 7.014080351692134e-06,
      "loss": 3.2698,
      "num_input_tokens_seen": 63141024,
      "step": 30225
    },
    {
      "epoch": 4.9315604861734235,
      "grad_norm": 9.125,
      "learning_rate": 7.0072629322124024e-06,
      "loss": 2.3034,
      "num_input_tokens_seen": 63151424,
      "step": 30230
    },
    {
      "epoch": 4.932376213394241,
      "grad_norm": 8.625,
      "learning_rate": 7.000448287438827e-06,
      "loss": 2.7584,
      "num_input_tokens_seen": 63161952,
      "step": 30235
    },
    {
      "epoch": 4.933191940615059,
      "grad_norm": 5.03125,
      "learning_rate": 6.993636418422331e-06,
      "loss": 2.8559,
      "num_input_tokens_seen": 63172496,
      "step": 30240
    },
    {
      "epoch": 4.934007667835876,
      "grad_norm": 1.609375,
      "learning_rate": 6.986827326213383e-06,
      "loss": 1.6351,
      "num_input_tokens_seen": 63181344,
      "step": 30245
    },
    {
      "epoch": 4.934823395056693,
      "grad_norm": 8.4375,
      "learning_rate": 6.9800210118620205e-06,
      "loss": 1.6842,
      "num_input_tokens_seen": 63190512,
      "step": 30250
    },
    {
      "epoch": 4.935639122277511,
      "grad_norm": 7.34375,
      "learning_rate": 6.973217476417876e-06,
      "loss": 2.6684,
      "num_input_tokens_seen": 63201712,
      "step": 30255
    },
    {
      "epoch": 4.936454849498328,
      "grad_norm": 6.5,
      "learning_rate": 6.96641672093013e-06,
      "loss": 3.6638,
      "num_input_tokens_seen": 63212976,
      "step": 30260
    },
    {
      "epoch": 4.937270576719145,
      "grad_norm": 0.71484375,
      "learning_rate": 6.95961874644755e-06,
      "loss": 2.0648,
      "num_input_tokens_seen": 63224096,
      "step": 30265
    },
    {
      "epoch": 4.938086303939962,
      "grad_norm": 5.9375,
      "learning_rate": 6.952823554018476e-06,
      "loss": 1.7147,
      "num_input_tokens_seen": 63234976,
      "step": 30270
    },
    {
      "epoch": 4.938902031160779,
      "grad_norm": 6.8125,
      "learning_rate": 6.946031144690798e-06,
      "loss": 2.462,
      "num_input_tokens_seen": 63244480,
      "step": 30275
    },
    {
      "epoch": 4.939717758381597,
      "grad_norm": 4.875,
      "learning_rate": 6.939241519512005e-06,
      "loss": 1.6392,
      "num_input_tokens_seen": 63255152,
      "step": 30280
    },
    {
      "epoch": 4.940533485602415,
      "grad_norm": 4.0625,
      "learning_rate": 6.932454679529129e-06,
      "loss": 2.6507,
      "num_input_tokens_seen": 63266544,
      "step": 30285
    },
    {
      "epoch": 4.941349212823232,
      "grad_norm": 4.6875,
      "learning_rate": 6.925670625788791e-06,
      "loss": 2.1968,
      "num_input_tokens_seen": 63277792,
      "step": 30290
    },
    {
      "epoch": 4.94216494004405,
      "grad_norm": 8.125,
      "learning_rate": 6.918889359337186e-06,
      "loss": 3.3984,
      "num_input_tokens_seen": 63288912,
      "step": 30295
    },
    {
      "epoch": 4.942980667264867,
      "grad_norm": 5.375,
      "learning_rate": 6.912110881220058e-06,
      "loss": 2.0162,
      "num_input_tokens_seen": 63298928,
      "step": 30300
    },
    {
      "epoch": 4.943796394485684,
      "grad_norm": 13.3125,
      "learning_rate": 6.905335192482735e-06,
      "loss": 4.4795,
      "num_input_tokens_seen": 63309328,
      "step": 30305
    },
    {
      "epoch": 4.944612121706501,
      "grad_norm": 5.59375,
      "learning_rate": 6.8985622941701275e-06,
      "loss": 2.1001,
      "num_input_tokens_seen": 63320864,
      "step": 30310
    },
    {
      "epoch": 4.945427848927318,
      "grad_norm": 4.5,
      "learning_rate": 6.89179218732669e-06,
      "loss": 2.0733,
      "num_input_tokens_seen": 63330432,
      "step": 30315
    },
    {
      "epoch": 4.946243576148136,
      "grad_norm": 2.4375,
      "learning_rate": 6.8850248729964595e-06,
      "loss": 1.5302,
      "num_input_tokens_seen": 63340848,
      "step": 30320
    },
    {
      "epoch": 4.947059303368953,
      "grad_norm": 7.8125,
      "learning_rate": 6.8782603522230314e-06,
      "loss": 3.5655,
      "num_input_tokens_seen": 63350048,
      "step": 30325
    },
    {
      "epoch": 4.9478750305897705,
      "grad_norm": 8.9375,
      "learning_rate": 6.871498626049591e-06,
      "loss": 3.1055,
      "num_input_tokens_seen": 63360608,
      "step": 30330
    },
    {
      "epoch": 4.9486907578105885,
      "grad_norm": 6.03125,
      "learning_rate": 6.8647396955188875e-06,
      "loss": 1.945,
      "num_input_tokens_seen": 63371632,
      "step": 30335
    },
    {
      "epoch": 4.949506485031406,
      "grad_norm": 7.96875,
      "learning_rate": 6.857983561673218e-06,
      "loss": 1.5745,
      "num_input_tokens_seen": 63380816,
      "step": 30340
    },
    {
      "epoch": 4.950322212252223,
      "grad_norm": 6.25,
      "learning_rate": 6.851230225554467e-06,
      "loss": 1.7087,
      "num_input_tokens_seen": 63390720,
      "step": 30345
    },
    {
      "epoch": 4.95113793947304,
      "grad_norm": 3.359375,
      "learning_rate": 6.8444796882040946e-06,
      "loss": 1.7388,
      "num_input_tokens_seen": 63400592,
      "step": 30350
    },
    {
      "epoch": 4.951953666693858,
      "grad_norm": 1.2578125,
      "learning_rate": 6.837731950663106e-06,
      "loss": 2.0631,
      "num_input_tokens_seen": 63409728,
      "step": 30355
    },
    {
      "epoch": 4.952769393914675,
      "grad_norm": 6.125,
      "learning_rate": 6.830987013972098e-06,
      "loss": 1.7889,
      "num_input_tokens_seen": 63420544,
      "step": 30360
    },
    {
      "epoch": 4.953585121135492,
      "grad_norm": 7.90625,
      "learning_rate": 6.82424487917121e-06,
      "loss": 1.7462,
      "num_input_tokens_seen": 63429984,
      "step": 30365
    },
    {
      "epoch": 4.954400848356309,
      "grad_norm": 6.625,
      "learning_rate": 6.8175055473001735e-06,
      "loss": 1.6469,
      "num_input_tokens_seen": 63440512,
      "step": 30370
    },
    {
      "epoch": 4.955216575577127,
      "grad_norm": 7.0,
      "learning_rate": 6.8107690193982855e-06,
      "loss": 1.9943,
      "num_input_tokens_seen": 63451424,
      "step": 30375
    },
    {
      "epoch": 4.9560323027979445,
      "grad_norm": 6.90625,
      "learning_rate": 6.804035296504385e-06,
      "loss": 2.5724,
      "num_input_tokens_seen": 63462256,
      "step": 30380
    },
    {
      "epoch": 4.956848030018762,
      "grad_norm": 1.25,
      "learning_rate": 6.797304379656916e-06,
      "loss": 1.2344,
      "num_input_tokens_seen": 63473568,
      "step": 30385
    },
    {
      "epoch": 4.957663757239579,
      "grad_norm": 6.71875,
      "learning_rate": 6.790576269893861e-06,
      "loss": 2.0058,
      "num_input_tokens_seen": 63485152,
      "step": 30390
    },
    {
      "epoch": 4.958479484460397,
      "grad_norm": 11.5625,
      "learning_rate": 6.783850968252772e-06,
      "loss": 2.8123,
      "num_input_tokens_seen": 63495200,
      "step": 30395
    },
    {
      "epoch": 4.959295211681214,
      "grad_norm": 9.4375,
      "learning_rate": 6.777128475770789e-06,
      "loss": 2.651,
      "num_input_tokens_seen": 63504960,
      "step": 30400
    },
    {
      "epoch": 4.959295211681214,
      "eval_loss": 2.5388436317443848,
      "eval_runtime": 135.0431,
      "eval_samples_per_second": 20.179,
      "eval_steps_per_second": 10.093,
      "num_input_tokens_seen": 63504960,
      "step": 30400
    },
    {
      "epoch": 4.960110938902031,
      "grad_norm": 10.4375,
      "learning_rate": 6.77040879348459e-06,
      "loss": 2.8872,
      "num_input_tokens_seen": 63516224,
      "step": 30405
    },
    {
      "epoch": 4.960926666122848,
      "grad_norm": 6.0,
      "learning_rate": 6.763691922430443e-06,
      "loss": 2.5476,
      "num_input_tokens_seen": 63527232,
      "step": 30410
    },
    {
      "epoch": 4.961742393343666,
      "grad_norm": 10.625,
      "learning_rate": 6.756977863644178e-06,
      "loss": 3.3086,
      "num_input_tokens_seen": 63537936,
      "step": 30415
    },
    {
      "epoch": 4.962558120564483,
      "grad_norm": 9.8125,
      "learning_rate": 6.7502666181611804e-06,
      "loss": 2.9306,
      "num_input_tokens_seen": 63547776,
      "step": 30420
    },
    {
      "epoch": 4.9633738477853,
      "grad_norm": 10.8125,
      "learning_rate": 6.743558187016405e-06,
      "loss": 3.1777,
      "num_input_tokens_seen": 63558144,
      "step": 30425
    },
    {
      "epoch": 4.964189575006118,
      "grad_norm": 7.375,
      "learning_rate": 6.7368525712443925e-06,
      "loss": 2.2969,
      "num_input_tokens_seen": 63568336,
      "step": 30430
    },
    {
      "epoch": 4.965005302226936,
      "grad_norm": 7.09375,
      "learning_rate": 6.7301497718792155e-06,
      "loss": 2.2702,
      "num_input_tokens_seen": 63578944,
      "step": 30435
    },
    {
      "epoch": 4.965821029447753,
      "grad_norm": 9.25,
      "learning_rate": 6.723449789954544e-06,
      "loss": 3.1901,
      "num_input_tokens_seen": 63589280,
      "step": 30440
    },
    {
      "epoch": 4.96663675666857,
      "grad_norm": 8.9375,
      "learning_rate": 6.716752626503586e-06,
      "loss": 3.409,
      "num_input_tokens_seen": 63598912,
      "step": 30445
    },
    {
      "epoch": 4.967452483889387,
      "grad_norm": 6.0,
      "learning_rate": 6.710058282559131e-06,
      "loss": 3.0741,
      "num_input_tokens_seen": 63611200,
      "step": 30450
    },
    {
      "epoch": 4.968268211110205,
      "grad_norm": 6.6875,
      "learning_rate": 6.703366759153545e-06,
      "loss": 2.9854,
      "num_input_tokens_seen": 63621056,
      "step": 30455
    },
    {
      "epoch": 4.969083938331022,
      "grad_norm": 2.375,
      "learning_rate": 6.6966780573187335e-06,
      "loss": 1.4058,
      "num_input_tokens_seen": 63631312,
      "step": 30460
    },
    {
      "epoch": 4.969899665551839,
      "grad_norm": 9.6875,
      "learning_rate": 6.689992178086174e-06,
      "loss": 2.234,
      "num_input_tokens_seen": 63642624,
      "step": 30465
    },
    {
      "epoch": 4.970715392772657,
      "grad_norm": 1.859375,
      "learning_rate": 6.683309122486925e-06,
      "loss": 1.7637,
      "num_input_tokens_seen": 63652912,
      "step": 30470
    },
    {
      "epoch": 4.971531119993474,
      "grad_norm": 5.59375,
      "learning_rate": 6.676628891551584e-06,
      "loss": 2.5849,
      "num_input_tokens_seen": 63663584,
      "step": 30475
    },
    {
      "epoch": 4.9723468472142915,
      "grad_norm": 9.25,
      "learning_rate": 6.6699514863103385e-06,
      "loss": 3.2895,
      "num_input_tokens_seen": 63675504,
      "step": 30480
    },
    {
      "epoch": 4.973162574435109,
      "grad_norm": 6.65625,
      "learning_rate": 6.663276907792921e-06,
      "loss": 3.3322,
      "num_input_tokens_seen": 63685776,
      "step": 30485
    },
    {
      "epoch": 4.973978301655926,
      "grad_norm": 5.09375,
      "learning_rate": 6.656605157028634e-06,
      "loss": 1.9437,
      "num_input_tokens_seen": 63695376,
      "step": 30490
    },
    {
      "epoch": 4.974794028876744,
      "grad_norm": 3.96875,
      "learning_rate": 6.649936235046358e-06,
      "loss": 2.0423,
      "num_input_tokens_seen": 63705200,
      "step": 30495
    },
    {
      "epoch": 4.975609756097561,
      "grad_norm": 10.75,
      "learning_rate": 6.643270142874508e-06,
      "loss": 2.1188,
      "num_input_tokens_seen": 63715568,
      "step": 30500
    },
    {
      "epoch": 4.976425483318378,
      "grad_norm": 10.625,
      "learning_rate": 6.636606881541094e-06,
      "loss": 2.596,
      "num_input_tokens_seen": 63724816,
      "step": 30505
    },
    {
      "epoch": 4.977241210539196,
      "grad_norm": 11.4375,
      "learning_rate": 6.629946452073662e-06,
      "loss": 2.1754,
      "num_input_tokens_seen": 63735072,
      "step": 30510
    },
    {
      "epoch": 4.978056937760013,
      "grad_norm": 1.796875,
      "learning_rate": 6.6232888554993375e-06,
      "loss": 1.9222,
      "num_input_tokens_seen": 63744944,
      "step": 30515
    },
    {
      "epoch": 4.97887266498083,
      "grad_norm": 2.90625,
      "learning_rate": 6.616634092844817e-06,
      "loss": 1.8107,
      "num_input_tokens_seen": 63755520,
      "step": 30520
    },
    {
      "epoch": 4.9796883922016475,
      "grad_norm": 5.0,
      "learning_rate": 6.609982165136331e-06,
      "loss": 2.3244,
      "num_input_tokens_seen": 63765152,
      "step": 30525
    },
    {
      "epoch": 4.9805041194224655,
      "grad_norm": 7.46875,
      "learning_rate": 6.603333073399706e-06,
      "loss": 1.8395,
      "num_input_tokens_seen": 63775360,
      "step": 30530
    },
    {
      "epoch": 4.981319846643283,
      "grad_norm": 4.0,
      "learning_rate": 6.596686818660308e-06,
      "loss": 2.5253,
      "num_input_tokens_seen": 63786800,
      "step": 30535
    },
    {
      "epoch": 4.9821355738641,
      "grad_norm": 6.6875,
      "learning_rate": 6.590043401943066e-06,
      "loss": 3.167,
      "num_input_tokens_seen": 63797600,
      "step": 30540
    },
    {
      "epoch": 4.982951301084917,
      "grad_norm": 6.6875,
      "learning_rate": 6.583402824272494e-06,
      "loss": 1.858,
      "num_input_tokens_seen": 63808816,
      "step": 30545
    },
    {
      "epoch": 4.983767028305735,
      "grad_norm": 7.4375,
      "learning_rate": 6.576765086672634e-06,
      "loss": 1.8794,
      "num_input_tokens_seen": 63819040,
      "step": 30550
    },
    {
      "epoch": 4.984582755526552,
      "grad_norm": 7.6875,
      "learning_rate": 6.57013019016712e-06,
      "loss": 2.7738,
      "num_input_tokens_seen": 63829824,
      "step": 30555
    },
    {
      "epoch": 4.985398482747369,
      "grad_norm": 5.125,
      "learning_rate": 6.563498135779142e-06,
      "loss": 2.3265,
      "num_input_tokens_seen": 63839456,
      "step": 30560
    },
    {
      "epoch": 4.986214209968186,
      "grad_norm": 7.1875,
      "learning_rate": 6.556868924531431e-06,
      "loss": 2.0422,
      "num_input_tokens_seen": 63849904,
      "step": 30565
    },
    {
      "epoch": 4.987029937189004,
      "grad_norm": 3.375,
      "learning_rate": 6.550242557446304e-06,
      "loss": 1.8995,
      "num_input_tokens_seen": 63861792,
      "step": 30570
    },
    {
      "epoch": 4.9878456644098215,
      "grad_norm": 3.90625,
      "learning_rate": 6.543619035545634e-06,
      "loss": 3.1391,
      "num_input_tokens_seen": 63872528,
      "step": 30575
    },
    {
      "epoch": 4.988661391630639,
      "grad_norm": 5.84375,
      "learning_rate": 6.53699835985084e-06,
      "loss": 2.8571,
      "num_input_tokens_seen": 63883104,
      "step": 30580
    },
    {
      "epoch": 4.989477118851456,
      "grad_norm": 8.0625,
      "learning_rate": 6.530380531382927e-06,
      "loss": 3.0143,
      "num_input_tokens_seen": 63893264,
      "step": 30585
    },
    {
      "epoch": 4.990292846072274,
      "grad_norm": 5.875,
      "learning_rate": 6.523765551162433e-06,
      "loss": 1.7848,
      "num_input_tokens_seen": 63904400,
      "step": 30590
    },
    {
      "epoch": 4.991108573293091,
      "grad_norm": 5.5625,
      "learning_rate": 6.517153420209476e-06,
      "loss": 2.6538,
      "num_input_tokens_seen": 63915328,
      "step": 30595
    },
    {
      "epoch": 4.991924300513908,
      "grad_norm": 7.15625,
      "learning_rate": 6.510544139543739e-06,
      "loss": 4.0587,
      "num_input_tokens_seen": 63926432,
      "step": 30600
    },
    {
      "epoch": 4.991924300513908,
      "eval_loss": 2.5485336780548096,
      "eval_runtime": 134.8054,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 63926432,
      "step": 30600
    },
    {
      "epoch": 4.992740027734725,
      "grad_norm": 4.46875,
      "learning_rate": 6.503937710184452e-06,
      "loss": 3.1299,
      "num_input_tokens_seen": 63936720,
      "step": 30605
    },
    {
      "epoch": 4.993555754955543,
      "grad_norm": 13.6875,
      "learning_rate": 6.4973341331503954e-06,
      "loss": 3.4473,
      "num_input_tokens_seen": 63946576,
      "step": 30610
    },
    {
      "epoch": 4.99437148217636,
      "grad_norm": 14.1875,
      "learning_rate": 6.490733409459942e-06,
      "loss": 3.0186,
      "num_input_tokens_seen": 63957312,
      "step": 30615
    },
    {
      "epoch": 4.995187209397177,
      "grad_norm": 3.6875,
      "learning_rate": 6.484135540130995e-06,
      "loss": 2.3585,
      "num_input_tokens_seen": 63967360,
      "step": 30620
    },
    {
      "epoch": 4.9960029366179945,
      "grad_norm": 6.0625,
      "learning_rate": 6.4775405261810364e-06,
      "loss": 1.3133,
      "num_input_tokens_seen": 63976832,
      "step": 30625
    },
    {
      "epoch": 4.996818663838813,
      "grad_norm": 6.46875,
      "learning_rate": 6.470948368627092e-06,
      "loss": 3.6898,
      "num_input_tokens_seen": 63988352,
      "step": 30630
    },
    {
      "epoch": 4.99763439105963,
      "grad_norm": 7.1875,
      "learning_rate": 6.464359068485756e-06,
      "loss": 2.9102,
      "num_input_tokens_seen": 63998512,
      "step": 30635
    },
    {
      "epoch": 4.998450118280447,
      "grad_norm": 1.71875,
      "learning_rate": 6.457772626773195e-06,
      "loss": 1.3832,
      "num_input_tokens_seen": 64007648,
      "step": 30640
    },
    {
      "epoch": 4.999265845501265,
      "grad_norm": 6.125,
      "learning_rate": 6.451189044505104e-06,
      "loss": 2.035,
      "num_input_tokens_seen": 64019360,
      "step": 30645
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.1875,
      "learning_rate": 6.44460832269676e-06,
      "loss": 2.0651,
      "num_input_tokens_seen": 64029024,
      "step": 30650
    },
    {
      "epoch": 5.000815727220817,
      "grad_norm": 0.267578125,
      "learning_rate": 6.438030462363001e-06,
      "loss": 1.5691,
      "num_input_tokens_seen": 64040032,
      "step": 30655
    },
    {
      "epoch": 5.001631454441635,
      "grad_norm": 12.6875,
      "learning_rate": 6.431455464518205e-06,
      "loss": 2.0846,
      "num_input_tokens_seen": 64050192,
      "step": 30660
    },
    {
      "epoch": 5.002447181662452,
      "grad_norm": 11.5,
      "learning_rate": 6.424883330176326e-06,
      "loss": 2.6294,
      "num_input_tokens_seen": 64061088,
      "step": 30665
    },
    {
      "epoch": 5.003262908883269,
      "grad_norm": 17.25,
      "learning_rate": 6.418314060350864e-06,
      "loss": 3.112,
      "num_input_tokens_seen": 64072320,
      "step": 30670
    },
    {
      "epoch": 5.0040786361040865,
      "grad_norm": 4.21875,
      "learning_rate": 6.4117476560548895e-06,
      "loss": 1.3646,
      "num_input_tokens_seen": 64083360,
      "step": 30675
    },
    {
      "epoch": 5.004894363324905,
      "grad_norm": 4.0625,
      "learning_rate": 6.405184118301016e-06,
      "loss": 1.4828,
      "num_input_tokens_seen": 64093664,
      "step": 30680
    },
    {
      "epoch": 5.005710090545722,
      "grad_norm": 9.4375,
      "learning_rate": 6.398623448101434e-06,
      "loss": 1.49,
      "num_input_tokens_seen": 64104320,
      "step": 30685
    },
    {
      "epoch": 5.006525817766539,
      "grad_norm": 4.40625,
      "learning_rate": 6.392065646467871e-06,
      "loss": 3.3381,
      "num_input_tokens_seen": 64114320,
      "step": 30690
    },
    {
      "epoch": 5.007341544987356,
      "grad_norm": 7.96875,
      "learning_rate": 6.385510714411632e-06,
      "loss": 2.523,
      "num_input_tokens_seen": 64126544,
      "step": 30695
    },
    {
      "epoch": 5.008157272208174,
      "grad_norm": 11.75,
      "learning_rate": 6.378958652943559e-06,
      "loss": 2.8549,
      "num_input_tokens_seen": 64136672,
      "step": 30700
    },
    {
      "epoch": 5.008972999428991,
      "grad_norm": 6.59375,
      "learning_rate": 6.3724094630740776e-06,
      "loss": 1.9989,
      "num_input_tokens_seen": 64147600,
      "step": 30705
    },
    {
      "epoch": 5.009788726649808,
      "grad_norm": 8.3125,
      "learning_rate": 6.365863145813136e-06,
      "loss": 3.0857,
      "num_input_tokens_seen": 64157408,
      "step": 30710
    },
    {
      "epoch": 5.010604453870625,
      "grad_norm": 7.625,
      "learning_rate": 6.359319702170269e-06,
      "loss": 2.1562,
      "num_input_tokens_seen": 64166864,
      "step": 30715
    },
    {
      "epoch": 5.011420181091443,
      "grad_norm": 7.875,
      "learning_rate": 6.352779133154566e-06,
      "loss": 2.4016,
      "num_input_tokens_seen": 64176080,
      "step": 30720
    },
    {
      "epoch": 5.0122359083122605,
      "grad_norm": 4.15625,
      "learning_rate": 6.346241439774648e-06,
      "loss": 1.6434,
      "num_input_tokens_seen": 64187056,
      "step": 30725
    },
    {
      "epoch": 5.013051635533078,
      "grad_norm": 4.84375,
      "learning_rate": 6.339706623038716e-06,
      "loss": 2.8118,
      "num_input_tokens_seen": 64197456,
      "step": 30730
    },
    {
      "epoch": 5.013867362753895,
      "grad_norm": 14.875,
      "learning_rate": 6.333174683954532e-06,
      "loss": 2.6078,
      "num_input_tokens_seen": 64208176,
      "step": 30735
    },
    {
      "epoch": 5.014683089974713,
      "grad_norm": 8.9375,
      "learning_rate": 6.326645623529387e-06,
      "loss": 2.9301,
      "num_input_tokens_seen": 64218672,
      "step": 30740
    },
    {
      "epoch": 5.01549881719553,
      "grad_norm": 4.1875,
      "learning_rate": 6.320119442770156e-06,
      "loss": 1.8898,
      "num_input_tokens_seen": 64228672,
      "step": 30745
    },
    {
      "epoch": 5.016314544416347,
      "grad_norm": 9.1875,
      "learning_rate": 6.313596142683254e-06,
      "loss": 1.4753,
      "num_input_tokens_seen": 64239488,
      "step": 30750
    },
    {
      "epoch": 5.017130271637164,
      "grad_norm": 4.34375,
      "learning_rate": 6.307075724274647e-06,
      "loss": 3.4443,
      "num_input_tokens_seen": 64250384,
      "step": 30755
    },
    {
      "epoch": 5.017945998857982,
      "grad_norm": 2.171875,
      "learning_rate": 6.300558188549882e-06,
      "loss": 1.6053,
      "num_input_tokens_seen": 64261456,
      "step": 30760
    },
    {
      "epoch": 5.018761726078799,
      "grad_norm": 8.4375,
      "learning_rate": 6.29404353651403e-06,
      "loss": 2.849,
      "num_input_tokens_seen": 64273424,
      "step": 30765
    },
    {
      "epoch": 5.0195774532996165,
      "grad_norm": 9.0,
      "learning_rate": 6.287531769171737e-06,
      "loss": 1.2974,
      "num_input_tokens_seen": 64283088,
      "step": 30770
    },
    {
      "epoch": 5.020393180520434,
      "grad_norm": 8.1875,
      "learning_rate": 6.2810228875272045e-06,
      "loss": 3.3826,
      "num_input_tokens_seen": 64294176,
      "step": 30775
    },
    {
      "epoch": 5.021208907741252,
      "grad_norm": 10.3125,
      "learning_rate": 6.274516892584179e-06,
      "loss": 2.2277,
      "num_input_tokens_seen": 64303936,
      "step": 30780
    },
    {
      "epoch": 5.022024634962069,
      "grad_norm": 7.34375,
      "learning_rate": 6.268013785345969e-06,
      "loss": 4.192,
      "num_input_tokens_seen": 64314736,
      "step": 30785
    },
    {
      "epoch": 5.022840362182886,
      "grad_norm": 6.34375,
      "learning_rate": 6.26151356681543e-06,
      "loss": 2.7223,
      "num_input_tokens_seen": 64324976,
      "step": 30790
    },
    {
      "epoch": 5.023656089403703,
      "grad_norm": 6.0,
      "learning_rate": 6.255016237994981e-06,
      "loss": 1.7371,
      "num_input_tokens_seen": 64337424,
      "step": 30795
    },
    {
      "epoch": 5.024471816624521,
      "grad_norm": 3.28125,
      "learning_rate": 6.248521799886603e-06,
      "loss": 1.5048,
      "num_input_tokens_seen": 64346032,
      "step": 30800
    },
    {
      "epoch": 5.024471816624521,
      "eval_loss": 2.5376741886138916,
      "eval_runtime": 134.9772,
      "eval_samples_per_second": 20.189,
      "eval_steps_per_second": 10.098,
      "num_input_tokens_seen": 64346032,
      "step": 30800
    },
    {
      "epoch": 5.025287543845338,
      "grad_norm": 5.34375,
      "learning_rate": 6.242030253491798e-06,
      "loss": 2.0636,
      "num_input_tokens_seen": 64357072,
      "step": 30805
    },
    {
      "epoch": 5.026103271066155,
      "grad_norm": 5.3125,
      "learning_rate": 6.235541599811656e-06,
      "loss": 2.9576,
      "num_input_tokens_seen": 64366352,
      "step": 30810
    },
    {
      "epoch": 5.026918998286972,
      "grad_norm": 6.03125,
      "learning_rate": 6.229055839846814e-06,
      "loss": 1.9813,
      "num_input_tokens_seen": 64377504,
      "step": 30815
    },
    {
      "epoch": 5.02773472550779,
      "grad_norm": 9.5,
      "learning_rate": 6.222572974597455e-06,
      "loss": 1.6147,
      "num_input_tokens_seen": 64386512,
      "step": 30820
    },
    {
      "epoch": 5.028550452728608,
      "grad_norm": 1.515625,
      "learning_rate": 6.216093005063306e-06,
      "loss": 2.4642,
      "num_input_tokens_seen": 64396400,
      "step": 30825
    },
    {
      "epoch": 5.029366179949425,
      "grad_norm": 7.125,
      "learning_rate": 6.209615932243678e-06,
      "loss": 2.3089,
      "num_input_tokens_seen": 64406864,
      "step": 30830
    },
    {
      "epoch": 5.030181907170242,
      "grad_norm": 9.1875,
      "learning_rate": 6.203141757137399e-06,
      "loss": 3.3419,
      "num_input_tokens_seen": 64415088,
      "step": 30835
    },
    {
      "epoch": 5.03099763439106,
      "grad_norm": 4.03125,
      "learning_rate": 6.196670480742886e-06,
      "loss": 2.4497,
      "num_input_tokens_seen": 64425744,
      "step": 30840
    },
    {
      "epoch": 5.031813361611877,
      "grad_norm": 4.5,
      "learning_rate": 6.190202104058074e-06,
      "loss": 1.3324,
      "num_input_tokens_seen": 64437008,
      "step": 30845
    },
    {
      "epoch": 5.032629088832694,
      "grad_norm": 4.90625,
      "learning_rate": 6.183736628080475e-06,
      "loss": 4.1345,
      "num_input_tokens_seen": 64446976,
      "step": 30850
    },
    {
      "epoch": 5.033444816053512,
      "grad_norm": 2.75,
      "learning_rate": 6.177274053807155e-06,
      "loss": 2.1393,
      "num_input_tokens_seen": 64457040,
      "step": 30855
    },
    {
      "epoch": 5.034260543274329,
      "grad_norm": 4.15625,
      "learning_rate": 6.170814382234713e-06,
      "loss": 1.7993,
      "num_input_tokens_seen": 64467440,
      "step": 30860
    },
    {
      "epoch": 5.035076270495146,
      "grad_norm": 9.875,
      "learning_rate": 6.16435761435932e-06,
      "loss": 2.9245,
      "num_input_tokens_seen": 64478704,
      "step": 30865
    },
    {
      "epoch": 5.0358919977159635,
      "grad_norm": 11.3125,
      "learning_rate": 6.157903751176681e-06,
      "loss": 2.226,
      "num_input_tokens_seen": 64488992,
      "step": 30870
    },
    {
      "epoch": 5.0367077249367815,
      "grad_norm": 10.75,
      "learning_rate": 6.151452793682066e-06,
      "loss": 3.8606,
      "num_input_tokens_seen": 64500496,
      "step": 30875
    },
    {
      "epoch": 5.037523452157599,
      "grad_norm": 7.65625,
      "learning_rate": 6.145004742870305e-06,
      "loss": 2.6339,
      "num_input_tokens_seen": 64510416,
      "step": 30880
    },
    {
      "epoch": 5.038339179378416,
      "grad_norm": 5.6875,
      "learning_rate": 6.138559599735752e-06,
      "loss": 1.9447,
      "num_input_tokens_seen": 64521760,
      "step": 30885
    },
    {
      "epoch": 5.039154906599233,
      "grad_norm": 3.109375,
      "learning_rate": 6.132117365272344e-06,
      "loss": 2.1504,
      "num_input_tokens_seen": 64531664,
      "step": 30890
    },
    {
      "epoch": 5.039970633820051,
      "grad_norm": 7.375,
      "learning_rate": 6.125678040473545e-06,
      "loss": 2.2273,
      "num_input_tokens_seen": 64541952,
      "step": 30895
    },
    {
      "epoch": 5.040786361040868,
      "grad_norm": 7.34375,
      "learning_rate": 6.1192416263323755e-06,
      "loss": 2.7108,
      "num_input_tokens_seen": 64551488,
      "step": 30900
    },
    {
      "epoch": 5.041602088261685,
      "grad_norm": 8.875,
      "learning_rate": 6.112808123841424e-06,
      "loss": 2.8598,
      "num_input_tokens_seen": 64561616,
      "step": 30905
    },
    {
      "epoch": 5.042417815482502,
      "grad_norm": 1.78125,
      "learning_rate": 6.106377533992805e-06,
      "loss": 1.9716,
      "num_input_tokens_seen": 64571952,
      "step": 30910
    },
    {
      "epoch": 5.04323354270332,
      "grad_norm": 2.84375,
      "learning_rate": 6.099949857778204e-06,
      "loss": 3.6317,
      "num_input_tokens_seen": 64582112,
      "step": 30915
    },
    {
      "epoch": 5.0440492699241375,
      "grad_norm": 7.625,
      "learning_rate": 6.093525096188852e-06,
      "loss": 1.9227,
      "num_input_tokens_seen": 64593200,
      "step": 30920
    },
    {
      "epoch": 5.044864997144955,
      "grad_norm": 7.3125,
      "learning_rate": 6.087103250215518e-06,
      "loss": 2.8798,
      "num_input_tokens_seen": 64603936,
      "step": 30925
    },
    {
      "epoch": 5.045680724365772,
      "grad_norm": 5.0,
      "learning_rate": 6.080684320848537e-06,
      "loss": 2.1224,
      "num_input_tokens_seen": 64614096,
      "step": 30930
    },
    {
      "epoch": 5.04649645158659,
      "grad_norm": 3.34375,
      "learning_rate": 6.074268309077794e-06,
      "loss": 2.3411,
      "num_input_tokens_seen": 64624928,
      "step": 30935
    },
    {
      "epoch": 5.047312178807407,
      "grad_norm": 14.4375,
      "learning_rate": 6.067855215892709e-06,
      "loss": 3.0373,
      "num_input_tokens_seen": 64635184,
      "step": 30940
    },
    {
      "epoch": 5.048127906028224,
      "grad_norm": 7.46875,
      "learning_rate": 6.061445042282271e-06,
      "loss": 1.0678,
      "num_input_tokens_seen": 64647344,
      "step": 30945
    },
    {
      "epoch": 5.048943633249041,
      "grad_norm": 5.28125,
      "learning_rate": 6.055037789234999e-06,
      "loss": 2.0144,
      "num_input_tokens_seen": 64657680,
      "step": 30950
    },
    {
      "epoch": 5.049759360469859,
      "grad_norm": 7.28125,
      "learning_rate": 6.048633457738975e-06,
      "loss": 1.5497,
      "num_input_tokens_seen": 64667856,
      "step": 30955
    },
    {
      "epoch": 5.050575087690676,
      "grad_norm": 4.6875,
      "learning_rate": 6.042232048781837e-06,
      "loss": 1.5605,
      "num_input_tokens_seen": 64677952,
      "step": 30960
    },
    {
      "epoch": 5.051390814911493,
      "grad_norm": 8.125,
      "learning_rate": 6.035833563350757e-06,
      "loss": 2.2136,
      "num_input_tokens_seen": 64690304,
      "step": 30965
    },
    {
      "epoch": 5.052206542132311,
      "grad_norm": 7.0625,
      "learning_rate": 6.0294380024324525e-06,
      "loss": 1.2953,
      "num_input_tokens_seen": 64700528,
      "step": 30970
    },
    {
      "epoch": 5.053022269353129,
      "grad_norm": 4.28125,
      "learning_rate": 6.023045367013213e-06,
      "loss": 3.1442,
      "num_input_tokens_seen": 64712144,
      "step": 30975
    },
    {
      "epoch": 5.053837996573946,
      "grad_norm": 4.3125,
      "learning_rate": 6.016655658078851e-06,
      "loss": 2.5257,
      "num_input_tokens_seen": 64722720,
      "step": 30980
    },
    {
      "epoch": 5.054653723794763,
      "grad_norm": 4.875,
      "learning_rate": 6.010268876614753e-06,
      "loss": 1.323,
      "num_input_tokens_seen": 64732896,
      "step": 30985
    },
    {
      "epoch": 5.05546945101558,
      "grad_norm": 9.6875,
      "learning_rate": 6.0038850236058266e-06,
      "loss": 1.5968,
      "num_input_tokens_seen": 64743184,
      "step": 30990
    },
    {
      "epoch": 5.056285178236398,
      "grad_norm": 3.140625,
      "learning_rate": 5.997504100036549e-06,
      "loss": 2.158,
      "num_input_tokens_seen": 64754608,
      "step": 30995
    },
    {
      "epoch": 5.057100905457215,
      "grad_norm": 10.3125,
      "learning_rate": 5.991126106890949e-06,
      "loss": 2.8357,
      "num_input_tokens_seen": 64764608,
      "step": 31000
    },
    {
      "epoch": 5.057100905457215,
      "eval_loss": 2.5535449981689453,
      "eval_runtime": 134.8211,
      "eval_samples_per_second": 20.212,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 64764608,
      "step": 31000
    },
    {
      "epoch": 5.057916632678032,
      "grad_norm": 3.34375,
      "learning_rate": 5.984751045152576e-06,
      "loss": 1.6575,
      "num_input_tokens_seen": 64774208,
      "step": 31005
    },
    {
      "epoch": 5.058732359898849,
      "grad_norm": 2.953125,
      "learning_rate": 5.978378915804553e-06,
      "loss": 2.3217,
      "num_input_tokens_seen": 64785216,
      "step": 31010
    },
    {
      "epoch": 5.059548087119667,
      "grad_norm": 8.5625,
      "learning_rate": 5.972009719829547e-06,
      "loss": 2.5175,
      "num_input_tokens_seen": 64796016,
      "step": 31015
    },
    {
      "epoch": 5.0603638143404845,
      "grad_norm": 5.71875,
      "learning_rate": 5.965643458209755e-06,
      "loss": 2.5677,
      "num_input_tokens_seen": 64806640,
      "step": 31020
    },
    {
      "epoch": 5.061179541561302,
      "grad_norm": 6.5625,
      "learning_rate": 5.95928013192695e-06,
      "loss": 2.6477,
      "num_input_tokens_seen": 64815616,
      "step": 31025
    },
    {
      "epoch": 5.06199526878212,
      "grad_norm": 7.40625,
      "learning_rate": 5.952919741962423e-06,
      "loss": 2.0405,
      "num_input_tokens_seen": 64825360,
      "step": 31030
    },
    {
      "epoch": 5.062810996002937,
      "grad_norm": 9.125,
      "learning_rate": 5.946562289297042e-06,
      "loss": 2.5021,
      "num_input_tokens_seen": 64836768,
      "step": 31035
    },
    {
      "epoch": 5.063626723223754,
      "grad_norm": 5.5625,
      "learning_rate": 5.9402077749111855e-06,
      "loss": 2.6514,
      "num_input_tokens_seen": 64845504,
      "step": 31040
    },
    {
      "epoch": 5.064442450444571,
      "grad_norm": 12.875,
      "learning_rate": 5.933856199784821e-06,
      "loss": 4.383,
      "num_input_tokens_seen": 64856016,
      "step": 31045
    },
    {
      "epoch": 5.065258177665389,
      "grad_norm": 11.3125,
      "learning_rate": 5.927507564897419e-06,
      "loss": 2.7777,
      "num_input_tokens_seen": 64866528,
      "step": 31050
    },
    {
      "epoch": 5.066073904886206,
      "grad_norm": 9.375,
      "learning_rate": 5.9211618712280395e-06,
      "loss": 2.2448,
      "num_input_tokens_seen": 64877344,
      "step": 31055
    },
    {
      "epoch": 5.066889632107023,
      "grad_norm": 6.03125,
      "learning_rate": 5.914819119755255e-06,
      "loss": 2.4817,
      "num_input_tokens_seen": 64887504,
      "step": 31060
    },
    {
      "epoch": 5.0677053593278405,
      "grad_norm": 4.9375,
      "learning_rate": 5.908479311457205e-06,
      "loss": 1.874,
      "num_input_tokens_seen": 64898288,
      "step": 31065
    },
    {
      "epoch": 5.0685210865486585,
      "grad_norm": 11.25,
      "learning_rate": 5.902142447311559e-06,
      "loss": 2.5983,
      "num_input_tokens_seen": 64908656,
      "step": 31070
    },
    {
      "epoch": 5.069336813769476,
      "grad_norm": 9.5,
      "learning_rate": 5.895808528295546e-06,
      "loss": 2.952,
      "num_input_tokens_seen": 64919536,
      "step": 31075
    },
    {
      "epoch": 5.070152540990293,
      "grad_norm": 7.8125,
      "learning_rate": 5.889477555385941e-06,
      "loss": 3.6249,
      "num_input_tokens_seen": 64930032,
      "step": 31080
    },
    {
      "epoch": 5.07096826821111,
      "grad_norm": 3.484375,
      "learning_rate": 5.883149529559051e-06,
      "loss": 2.1797,
      "num_input_tokens_seen": 64940064,
      "step": 31085
    },
    {
      "epoch": 5.071783995431928,
      "grad_norm": 7.6875,
      "learning_rate": 5.876824451790738e-06,
      "loss": 1.6177,
      "num_input_tokens_seen": 64950464,
      "step": 31090
    },
    {
      "epoch": 5.072599722652745,
      "grad_norm": 0.8671875,
      "learning_rate": 5.87050232305642e-06,
      "loss": 1.5847,
      "num_input_tokens_seen": 64961280,
      "step": 31095
    },
    {
      "epoch": 5.073415449873562,
      "grad_norm": 2.09375,
      "learning_rate": 5.864183144331034e-06,
      "loss": 2.5036,
      "num_input_tokens_seen": 64972640,
      "step": 31100
    },
    {
      "epoch": 5.074231177094379,
      "grad_norm": 9.0625,
      "learning_rate": 5.857866916589089e-06,
      "loss": 2.5875,
      "num_input_tokens_seen": 64983888,
      "step": 31105
    },
    {
      "epoch": 5.075046904315197,
      "grad_norm": 6.09375,
      "learning_rate": 5.8515536408046216e-06,
      "loss": 2.4128,
      "num_input_tokens_seen": 64994560,
      "step": 31110
    },
    {
      "epoch": 5.0758626315360145,
      "grad_norm": 8.125,
      "learning_rate": 5.845243317951208e-06,
      "loss": 1.7371,
      "num_input_tokens_seen": 65004768,
      "step": 31115
    },
    {
      "epoch": 5.076678358756832,
      "grad_norm": 7.46875,
      "learning_rate": 5.838935949001997e-06,
      "loss": 2.2875,
      "num_input_tokens_seen": 65014816,
      "step": 31120
    },
    {
      "epoch": 5.077494085977649,
      "grad_norm": 6.125,
      "learning_rate": 5.8326315349296476e-06,
      "loss": 2.5717,
      "num_input_tokens_seen": 65024832,
      "step": 31125
    },
    {
      "epoch": 5.078309813198467,
      "grad_norm": 7.96875,
      "learning_rate": 5.826330076706396e-06,
      "loss": 1.6217,
      "num_input_tokens_seen": 65034272,
      "step": 31130
    },
    {
      "epoch": 5.079125540419284,
      "grad_norm": 5.90625,
      "learning_rate": 5.820031575303988e-06,
      "loss": 2.0515,
      "num_input_tokens_seen": 65044480,
      "step": 31135
    },
    {
      "epoch": 5.079941267640101,
      "grad_norm": 8.375,
      "learning_rate": 5.813736031693745e-06,
      "loss": 2.6278,
      "num_input_tokens_seen": 65054144,
      "step": 31140
    },
    {
      "epoch": 5.080756994860918,
      "grad_norm": 11.625,
      "learning_rate": 5.807443446846522e-06,
      "loss": 2.3139,
      "num_input_tokens_seen": 65065792,
      "step": 31145
    },
    {
      "epoch": 5.081572722081736,
      "grad_norm": 11.3125,
      "learning_rate": 5.801153821732699e-06,
      "loss": 2.9428,
      "num_input_tokens_seen": 65075840,
      "step": 31150
    },
    {
      "epoch": 5.082388449302553,
      "grad_norm": 10.0,
      "learning_rate": 5.794867157322229e-06,
      "loss": 2.9868,
      "num_input_tokens_seen": 65086080,
      "step": 31155
    },
    {
      "epoch": 5.08320417652337,
      "grad_norm": 5.6875,
      "learning_rate": 5.788583454584593e-06,
      "loss": 2.35,
      "num_input_tokens_seen": 65096384,
      "step": 31160
    },
    {
      "epoch": 5.0840199037441876,
      "grad_norm": 5.59375,
      "learning_rate": 5.7823027144888075e-06,
      "loss": 2.5418,
      "num_input_tokens_seen": 65105280,
      "step": 31165
    },
    {
      "epoch": 5.084835630965006,
      "grad_norm": 11.1875,
      "learning_rate": 5.776024938003455e-06,
      "loss": 3.0747,
      "num_input_tokens_seen": 65115184,
      "step": 31170
    },
    {
      "epoch": 5.085651358185823,
      "grad_norm": 6.71875,
      "learning_rate": 5.7697501260966345e-06,
      "loss": 2.5843,
      "num_input_tokens_seen": 65127168,
      "step": 31175
    },
    {
      "epoch": 5.08646708540664,
      "grad_norm": 11.3125,
      "learning_rate": 5.7634782797360145e-06,
      "loss": 2.3403,
      "num_input_tokens_seen": 65138080,
      "step": 31180
    },
    {
      "epoch": 5.087282812627457,
      "grad_norm": 4.34375,
      "learning_rate": 5.757209399888777e-06,
      "loss": 1.2868,
      "num_input_tokens_seen": 65147600,
      "step": 31185
    },
    {
      "epoch": 5.088098539848275,
      "grad_norm": 7.6875,
      "learning_rate": 5.750943487521679e-06,
      "loss": 2.8511,
      "num_input_tokens_seen": 65158544,
      "step": 31190
    },
    {
      "epoch": 5.088914267069092,
      "grad_norm": 4.15625,
      "learning_rate": 5.744680543600986e-06,
      "loss": 2.4018,
      "num_input_tokens_seen": 65169168,
      "step": 31195
    },
    {
      "epoch": 5.089729994289909,
      "grad_norm": 6.0625,
      "learning_rate": 5.738420569092537e-06,
      "loss": 1.7275,
      "num_input_tokens_seen": 65180560,
      "step": 31200
    },
    {
      "epoch": 5.089729994289909,
      "eval_loss": 2.5577526092529297,
      "eval_runtime": 134.8369,
      "eval_samples_per_second": 20.21,
      "eval_steps_per_second": 10.109,
      "num_input_tokens_seen": 65180560,
      "step": 31200
    },
    {
      "epoch": 5.090545721510727,
      "grad_norm": 6.8125,
      "learning_rate": 5.732163564961684e-06,
      "loss": 2.4664,
      "num_input_tokens_seen": 65191952,
      "step": 31205
    },
    {
      "epoch": 5.091361448731544,
      "grad_norm": 8.6875,
      "learning_rate": 5.725909532173354e-06,
      "loss": 4.4519,
      "num_input_tokens_seen": 65202112,
      "step": 31210
    },
    {
      "epoch": 5.0921771759523615,
      "grad_norm": 5.78125,
      "learning_rate": 5.719658471691977e-06,
      "loss": 2.8088,
      "num_input_tokens_seen": 65213296,
      "step": 31215
    },
    {
      "epoch": 5.092992903173179,
      "grad_norm": 7.34375,
      "learning_rate": 5.71341038448156e-06,
      "loss": 1.8959,
      "num_input_tokens_seen": 65224384,
      "step": 31220
    },
    {
      "epoch": 5.093808630393997,
      "grad_norm": 3.484375,
      "learning_rate": 5.707165271505635e-06,
      "loss": 1.9655,
      "num_input_tokens_seen": 65235296,
      "step": 31225
    },
    {
      "epoch": 5.094624357614814,
      "grad_norm": 4.40625,
      "learning_rate": 5.700923133727271e-06,
      "loss": 1.7492,
      "num_input_tokens_seen": 65246192,
      "step": 31230
    },
    {
      "epoch": 5.095440084835631,
      "grad_norm": 3.984375,
      "learning_rate": 5.694683972109083e-06,
      "loss": 2.1711,
      "num_input_tokens_seen": 65256608,
      "step": 31235
    },
    {
      "epoch": 5.096255812056448,
      "grad_norm": 7.4375,
      "learning_rate": 5.688447787613241e-06,
      "loss": 1.4588,
      "num_input_tokens_seen": 65266624,
      "step": 31240
    },
    {
      "epoch": 5.097071539277266,
      "grad_norm": 5.46875,
      "learning_rate": 5.6822145812014285e-06,
      "loss": 2.5992,
      "num_input_tokens_seen": 65277120,
      "step": 31245
    },
    {
      "epoch": 5.097887266498083,
      "grad_norm": 5.4375,
      "learning_rate": 5.675984353834896e-06,
      "loss": 3.2496,
      "num_input_tokens_seen": 65287280,
      "step": 31250
    },
    {
      "epoch": 5.0987029937189,
      "grad_norm": 7.4375,
      "learning_rate": 5.66975710647441e-06,
      "loss": 1.9577,
      "num_input_tokens_seen": 65298608,
      "step": 31255
    },
    {
      "epoch": 5.0995187209397175,
      "grad_norm": 4.90625,
      "learning_rate": 5.663532840080304e-06,
      "loss": 2.0327,
      "num_input_tokens_seen": 65309600,
      "step": 31260
    },
    {
      "epoch": 5.1003344481605355,
      "grad_norm": 12.4375,
      "learning_rate": 5.6573115556124325e-06,
      "loss": 1.492,
      "num_input_tokens_seen": 65319648,
      "step": 31265
    },
    {
      "epoch": 5.101150175381353,
      "grad_norm": 3.84375,
      "learning_rate": 5.651093254030185e-06,
      "loss": 1.5161,
      "num_input_tokens_seen": 65329232,
      "step": 31270
    },
    {
      "epoch": 5.10196590260217,
      "grad_norm": 11.5625,
      "learning_rate": 5.644877936292514e-06,
      "loss": 4.6343,
      "num_input_tokens_seen": 65339040,
      "step": 31275
    },
    {
      "epoch": 5.102781629822987,
      "grad_norm": 9.0,
      "learning_rate": 5.638665603357901e-06,
      "loss": 1.9148,
      "num_input_tokens_seen": 65349104,
      "step": 31280
    },
    {
      "epoch": 5.103597357043805,
      "grad_norm": 1.140625,
      "learning_rate": 5.632456256184357e-06,
      "loss": 2.6505,
      "num_input_tokens_seen": 65359648,
      "step": 31285
    },
    {
      "epoch": 5.104413084264622,
      "grad_norm": 9.375,
      "learning_rate": 5.626249895729452e-06,
      "loss": 3.299,
      "num_input_tokens_seen": 65368528,
      "step": 31290
    },
    {
      "epoch": 5.105228811485439,
      "grad_norm": 12.4375,
      "learning_rate": 5.620046522950273e-06,
      "loss": 3.1876,
      "num_input_tokens_seen": 65378976,
      "step": 31295
    },
    {
      "epoch": 5.106044538706256,
      "grad_norm": 11.6875,
      "learning_rate": 5.613846138803464e-06,
      "loss": 2.3639,
      "num_input_tokens_seen": 65389056,
      "step": 31300
    },
    {
      "epoch": 5.106860265927074,
      "grad_norm": 3.765625,
      "learning_rate": 5.607648744245206e-06,
      "loss": 1.5838,
      "num_input_tokens_seen": 65399936,
      "step": 31305
    },
    {
      "epoch": 5.1076759931478914,
      "grad_norm": 4.625,
      "learning_rate": 5.601454340231207e-06,
      "loss": 3.0206,
      "num_input_tokens_seen": 65410416,
      "step": 31310
    },
    {
      "epoch": 5.108491720368709,
      "grad_norm": 6.78125,
      "learning_rate": 5.595262927716724e-06,
      "loss": 2.3451,
      "num_input_tokens_seen": 65420656,
      "step": 31315
    },
    {
      "epoch": 5.109307447589526,
      "grad_norm": 2.734375,
      "learning_rate": 5.589074507656561e-06,
      "loss": 3.1689,
      "num_input_tokens_seen": 65432000,
      "step": 31320
    },
    {
      "epoch": 5.110123174810344,
      "grad_norm": 4.9375,
      "learning_rate": 5.582889081005044e-06,
      "loss": 2.8485,
      "num_input_tokens_seen": 65443632,
      "step": 31325
    },
    {
      "epoch": 5.110938902031161,
      "grad_norm": 7.09375,
      "learning_rate": 5.5767066487160316e-06,
      "loss": 3.7526,
      "num_input_tokens_seen": 65453312,
      "step": 31330
    },
    {
      "epoch": 5.111754629251978,
      "grad_norm": 11.6875,
      "learning_rate": 5.570527211742949e-06,
      "loss": 2.1587,
      "num_input_tokens_seen": 65464128,
      "step": 31335
    },
    {
      "epoch": 5.112570356472795,
      "grad_norm": 5.90625,
      "learning_rate": 5.564350771038731e-06,
      "loss": 1.6937,
      "num_input_tokens_seen": 65474128,
      "step": 31340
    },
    {
      "epoch": 5.113386083693613,
      "grad_norm": 0.82421875,
      "learning_rate": 5.558177327555875e-06,
      "loss": 2.5633,
      "num_input_tokens_seen": 65484512,
      "step": 31345
    },
    {
      "epoch": 5.11420181091443,
      "grad_norm": 15.375,
      "learning_rate": 5.552006882246388e-06,
      "loss": 3.2128,
      "num_input_tokens_seen": 65495776,
      "step": 31350
    },
    {
      "epoch": 5.115017538135247,
      "grad_norm": 6.5625,
      "learning_rate": 5.545839436061839e-06,
      "loss": 1.8067,
      "num_input_tokens_seen": 65506848,
      "step": 31355
    },
    {
      "epoch": 5.1158332653560645,
      "grad_norm": 13.9375,
      "learning_rate": 5.539674989953331e-06,
      "loss": 2.3077,
      "num_input_tokens_seen": 65518288,
      "step": 31360
    },
    {
      "epoch": 5.1166489925768825,
      "grad_norm": 10.4375,
      "learning_rate": 5.533513544871488e-06,
      "loss": 2.3604,
      "num_input_tokens_seen": 65529024,
      "step": 31365
    },
    {
      "epoch": 5.1174647197977,
      "grad_norm": 8.25,
      "learning_rate": 5.527355101766493e-06,
      "loss": 4.0398,
      "num_input_tokens_seen": 65538688,
      "step": 31370
    },
    {
      "epoch": 5.118280447018517,
      "grad_norm": 4.59375,
      "learning_rate": 5.521199661588044e-06,
      "loss": 3.0619,
      "num_input_tokens_seen": 65547984,
      "step": 31375
    },
    {
      "epoch": 5.119096174239334,
      "grad_norm": 8.5,
      "learning_rate": 5.5150472252853944e-06,
      "loss": 1.9902,
      "num_input_tokens_seen": 65558400,
      "step": 31380
    },
    {
      "epoch": 5.119911901460152,
      "grad_norm": 5.59375,
      "learning_rate": 5.50889779380733e-06,
      "loss": 1.9461,
      "num_input_tokens_seen": 65568816,
      "step": 31385
    },
    {
      "epoch": 5.120727628680969,
      "grad_norm": 4.9375,
      "learning_rate": 5.5027513681021605e-06,
      "loss": 3.6197,
      "num_input_tokens_seen": 65579536,
      "step": 31390
    },
    {
      "epoch": 5.121543355901786,
      "grad_norm": 7.25,
      "learning_rate": 5.4966079491177545e-06,
      "loss": 1.7787,
      "num_input_tokens_seen": 65589984,
      "step": 31395
    },
    {
      "epoch": 5.122359083122603,
      "grad_norm": 1.75,
      "learning_rate": 5.490467537801491e-06,
      "loss": 0.6242,
      "num_input_tokens_seen": 65600032,
      "step": 31400
    },
    {
      "epoch": 5.122359083122603,
      "eval_loss": 2.5458450317382812,
      "eval_runtime": 134.5275,
      "eval_samples_per_second": 20.256,
      "eval_steps_per_second": 10.132,
      "num_input_tokens_seen": 65600032,
      "step": 31400
    },
    {
      "epoch": 5.123174810343421,
      "grad_norm": 3.3125,
      "learning_rate": 5.484330135100313e-06,
      "loss": 1.4811,
      "num_input_tokens_seen": 65609696,
      "step": 31405
    },
    {
      "epoch": 5.1239905375642385,
      "grad_norm": 16.375,
      "learning_rate": 5.4781957419606785e-06,
      "loss": 2.3935,
      "num_input_tokens_seen": 65620784,
      "step": 31410
    },
    {
      "epoch": 5.124806264785056,
      "grad_norm": 1.2578125,
      "learning_rate": 5.472064359328577e-06,
      "loss": 0.8026,
      "num_input_tokens_seen": 65631760,
      "step": 31415
    },
    {
      "epoch": 5.125621992005874,
      "grad_norm": 4.09375,
      "learning_rate": 5.4659359881495565e-06,
      "loss": 1.9882,
      "num_input_tokens_seen": 65642784,
      "step": 31420
    },
    {
      "epoch": 5.126437719226691,
      "grad_norm": 4.46875,
      "learning_rate": 5.4598106293686916e-06,
      "loss": 1.474,
      "num_input_tokens_seen": 65653392,
      "step": 31425
    },
    {
      "epoch": 5.127253446447508,
      "grad_norm": 13.9375,
      "learning_rate": 5.45368828393058e-06,
      "loss": 3.1723,
      "num_input_tokens_seen": 65662032,
      "step": 31430
    },
    {
      "epoch": 5.128069173668325,
      "grad_norm": 2.171875,
      "learning_rate": 5.44756895277937e-06,
      "loss": 3.1463,
      "num_input_tokens_seen": 65672000,
      "step": 31435
    },
    {
      "epoch": 5.128884900889143,
      "grad_norm": 7.8125,
      "learning_rate": 5.441452636858746e-06,
      "loss": 2.1267,
      "num_input_tokens_seen": 65681360,
      "step": 31440
    },
    {
      "epoch": 5.12970062810996,
      "grad_norm": 13.0625,
      "learning_rate": 5.435339337111905e-06,
      "loss": 1.2637,
      "num_input_tokens_seen": 65691824,
      "step": 31445
    },
    {
      "epoch": 5.130516355330777,
      "grad_norm": 8.625,
      "learning_rate": 5.42922905448161e-06,
      "loss": 2.6203,
      "num_input_tokens_seen": 65701424,
      "step": 31450
    },
    {
      "epoch": 5.1313320825515945,
      "grad_norm": 6.75,
      "learning_rate": 5.423121789910129e-06,
      "loss": 2.3989,
      "num_input_tokens_seen": 65710192,
      "step": 31455
    },
    {
      "epoch": 5.1321478097724125,
      "grad_norm": 7.65625,
      "learning_rate": 5.417017544339287e-06,
      "loss": 0.9086,
      "num_input_tokens_seen": 65718128,
      "step": 31460
    },
    {
      "epoch": 5.13296353699323,
      "grad_norm": 5.15625,
      "learning_rate": 5.410916318710443e-06,
      "loss": 2.3129,
      "num_input_tokens_seen": 65729200,
      "step": 31465
    },
    {
      "epoch": 5.133779264214047,
      "grad_norm": 3.53125,
      "learning_rate": 5.404818113964466e-06,
      "loss": 3.1282,
      "num_input_tokens_seen": 65739136,
      "step": 31470
    },
    {
      "epoch": 5.134594991434864,
      "grad_norm": 6.5625,
      "learning_rate": 5.398722931041792e-06,
      "loss": 2.3737,
      "num_input_tokens_seen": 65750576,
      "step": 31475
    },
    {
      "epoch": 5.135410718655682,
      "grad_norm": 8.6875,
      "learning_rate": 5.392630770882367e-06,
      "loss": 2.7045,
      "num_input_tokens_seen": 65762304,
      "step": 31480
    },
    {
      "epoch": 5.136226445876499,
      "grad_norm": 4.53125,
      "learning_rate": 5.3865416344256705e-06,
      "loss": 2.1857,
      "num_input_tokens_seen": 65772672,
      "step": 31485
    },
    {
      "epoch": 5.137042173097316,
      "grad_norm": 2.171875,
      "learning_rate": 5.380455522610742e-06,
      "loss": 2.0472,
      "num_input_tokens_seen": 65781568,
      "step": 31490
    },
    {
      "epoch": 5.137857900318133,
      "grad_norm": 9.25,
      "learning_rate": 5.374372436376116e-06,
      "loss": 2.1372,
      "num_input_tokens_seen": 65791568,
      "step": 31495
    },
    {
      "epoch": 5.138673627538951,
      "grad_norm": 6.5,
      "learning_rate": 5.368292376659895e-06,
      "loss": 3.6817,
      "num_input_tokens_seen": 65801920,
      "step": 31500
    },
    {
      "epoch": 5.139489354759768,
      "grad_norm": 6.59375,
      "learning_rate": 5.362215344399701e-06,
      "loss": 2.7558,
      "num_input_tokens_seen": 65811280,
      "step": 31505
    },
    {
      "epoch": 5.1403050819805856,
      "grad_norm": 4.65625,
      "learning_rate": 5.356141340532678e-06,
      "loss": 1.0185,
      "num_input_tokens_seen": 65821712,
      "step": 31510
    },
    {
      "epoch": 5.141120809201403,
      "grad_norm": 3.484375,
      "learning_rate": 5.350070365995522e-06,
      "loss": 1.1248,
      "num_input_tokens_seen": 65832400,
      "step": 31515
    },
    {
      "epoch": 5.141936536422221,
      "grad_norm": 10.6875,
      "learning_rate": 5.344002421724459e-06,
      "loss": 2.8597,
      "num_input_tokens_seen": 65843712,
      "step": 31520
    },
    {
      "epoch": 5.142752263643038,
      "grad_norm": 9.375,
      "learning_rate": 5.337937508655228e-06,
      "loss": 3.1524,
      "num_input_tokens_seen": 65853920,
      "step": 31525
    },
    {
      "epoch": 5.143567990863855,
      "grad_norm": 8.875,
      "learning_rate": 5.331875627723126e-06,
      "loss": 3.1163,
      "num_input_tokens_seen": 65863536,
      "step": 31530
    },
    {
      "epoch": 5.144383718084672,
      "grad_norm": 7.75,
      "learning_rate": 5.325816779862963e-06,
      "loss": 1.6013,
      "num_input_tokens_seen": 65874544,
      "step": 31535
    },
    {
      "epoch": 5.14519944530549,
      "grad_norm": 3.953125,
      "learning_rate": 5.319760966009102e-06,
      "loss": 2.2666,
      "num_input_tokens_seen": 65884400,
      "step": 31540
    },
    {
      "epoch": 5.146015172526307,
      "grad_norm": 7.09375,
      "learning_rate": 5.3137081870954096e-06,
      "loss": 2.6424,
      "num_input_tokens_seen": 65895632,
      "step": 31545
    },
    {
      "epoch": 5.146830899747124,
      "grad_norm": 3.90625,
      "learning_rate": 5.307658444055313e-06,
      "loss": 4.0871,
      "num_input_tokens_seen": 65906416,
      "step": 31550
    },
    {
      "epoch": 5.1476466269679415,
      "grad_norm": 6.65625,
      "learning_rate": 5.301611737821749e-06,
      "loss": 1.9586,
      "num_input_tokens_seen": 65917248,
      "step": 31555
    },
    {
      "epoch": 5.1484623541887595,
      "grad_norm": 8.1875,
      "learning_rate": 5.295568069327206e-06,
      "loss": 2.1103,
      "num_input_tokens_seen": 65926992,
      "step": 31560
    },
    {
      "epoch": 5.149278081409577,
      "grad_norm": 6.0625,
      "learning_rate": 5.289527439503683e-06,
      "loss": 2.357,
      "num_input_tokens_seen": 65937008,
      "step": 31565
    },
    {
      "epoch": 5.150093808630394,
      "grad_norm": 7.75,
      "learning_rate": 5.28348984928273e-06,
      "loss": 1.8295,
      "num_input_tokens_seen": 65946864,
      "step": 31570
    },
    {
      "epoch": 5.150909535851211,
      "grad_norm": 6.59375,
      "learning_rate": 5.27745529959541e-06,
      "loss": 2.5535,
      "num_input_tokens_seen": 65958320,
      "step": 31575
    },
    {
      "epoch": 5.151725263072029,
      "grad_norm": 7.125,
      "learning_rate": 5.271423791372335e-06,
      "loss": 2.7574,
      "num_input_tokens_seen": 65966784,
      "step": 31580
    },
    {
      "epoch": 5.152540990292846,
      "grad_norm": 9.625,
      "learning_rate": 5.26539532554364e-06,
      "loss": 1.6575,
      "num_input_tokens_seen": 65977376,
      "step": 31585
    },
    {
      "epoch": 5.153356717513663,
      "grad_norm": 14.125,
      "learning_rate": 5.25936990303898e-06,
      "loss": 1.8816,
      "num_input_tokens_seen": 65987472,
      "step": 31590
    },
    {
      "epoch": 5.154172444734481,
      "grad_norm": 4.125,
      "learning_rate": 5.253347524787555e-06,
      "loss": 2.6775,
      "num_input_tokens_seen": 65997968,
      "step": 31595
    },
    {
      "epoch": 5.154988171955298,
      "grad_norm": 9.125,
      "learning_rate": 5.2473281917181035e-06,
      "loss": 2.2156,
      "num_input_tokens_seen": 66007440,
      "step": 31600
    },
    {
      "epoch": 5.154988171955298,
      "eval_loss": 2.548614740371704,
      "eval_runtime": 134.8377,
      "eval_samples_per_second": 20.209,
      "eval_steps_per_second": 10.108,
      "num_input_tokens_seen": 66007440,
      "step": 31600
    },
    {
      "epoch": 5.1558038991761155,
      "grad_norm": 4.53125,
      "learning_rate": 5.241311904758864e-06,
      "loss": 2.1731,
      "num_input_tokens_seen": 66018320,
      "step": 31605
    },
    {
      "epoch": 5.156619626396933,
      "grad_norm": 4.78125,
      "learning_rate": 5.23529866483764e-06,
      "loss": 3.0881,
      "num_input_tokens_seen": 66027536,
      "step": 31610
    },
    {
      "epoch": 5.157435353617751,
      "grad_norm": 8.875,
      "learning_rate": 5.229288472881732e-06,
      "loss": 2.8026,
      "num_input_tokens_seen": 66036640,
      "step": 31615
    },
    {
      "epoch": 5.158251080838568,
      "grad_norm": 2.46875,
      "learning_rate": 5.2232813298180025e-06,
      "loss": 2.0484,
      "num_input_tokens_seen": 66046416,
      "step": 31620
    },
    {
      "epoch": 5.159066808059385,
      "grad_norm": 4.625,
      "learning_rate": 5.217277236572824e-06,
      "loss": 2.4164,
      "num_input_tokens_seen": 66056624,
      "step": 31625
    },
    {
      "epoch": 5.159882535280202,
      "grad_norm": 6.59375,
      "learning_rate": 5.211276194072093e-06,
      "loss": 1.9527,
      "num_input_tokens_seen": 66066896,
      "step": 31630
    },
    {
      "epoch": 5.16069826250102,
      "grad_norm": 0.921875,
      "learning_rate": 5.205278203241254e-06,
      "loss": 1.3941,
      "num_input_tokens_seen": 66077888,
      "step": 31635
    },
    {
      "epoch": 5.161513989721837,
      "grad_norm": 9.0625,
      "learning_rate": 5.199283265005278e-06,
      "loss": 1.2259,
      "num_input_tokens_seen": 66088368,
      "step": 31640
    },
    {
      "epoch": 5.162329716942654,
      "grad_norm": 4.40625,
      "learning_rate": 5.193291380288648e-06,
      "loss": 2.1741,
      "num_input_tokens_seen": 66099520,
      "step": 31645
    },
    {
      "epoch": 5.163145444163471,
      "grad_norm": 7.09375,
      "learning_rate": 5.1873025500153995e-06,
      "loss": 2.6937,
      "num_input_tokens_seen": 66110880,
      "step": 31650
    },
    {
      "epoch": 5.1639611713842895,
      "grad_norm": 8.8125,
      "learning_rate": 5.181316775109071e-06,
      "loss": 1.6862,
      "num_input_tokens_seen": 66121072,
      "step": 31655
    },
    {
      "epoch": 5.164776898605107,
      "grad_norm": 6.375,
      "learning_rate": 5.1753340564927564e-06,
      "loss": 1.5433,
      "num_input_tokens_seen": 66130272,
      "step": 31660
    },
    {
      "epoch": 5.165592625825924,
      "grad_norm": 11.9375,
      "learning_rate": 5.169354395089068e-06,
      "loss": 2.9109,
      "num_input_tokens_seen": 66140704,
      "step": 31665
    },
    {
      "epoch": 5.166408353046741,
      "grad_norm": 4.8125,
      "learning_rate": 5.1633777918201346e-06,
      "loss": 1.189,
      "num_input_tokens_seen": 66151952,
      "step": 31670
    },
    {
      "epoch": 5.167224080267559,
      "grad_norm": 12.75,
      "learning_rate": 5.157404247607625e-06,
      "loss": 2.5385,
      "num_input_tokens_seen": 66161168,
      "step": 31675
    },
    {
      "epoch": 5.168039807488376,
      "grad_norm": 8.0625,
      "learning_rate": 5.1514337633727454e-06,
      "loss": 2.6702,
      "num_input_tokens_seen": 66170128,
      "step": 31680
    },
    {
      "epoch": 5.168855534709193,
      "grad_norm": 9.75,
      "learning_rate": 5.145466340036206e-06,
      "loss": 2.979,
      "num_input_tokens_seen": 66181056,
      "step": 31685
    },
    {
      "epoch": 5.16967126193001,
      "grad_norm": 12.4375,
      "learning_rate": 5.139501978518274e-06,
      "loss": 2.5643,
      "num_input_tokens_seen": 66192928,
      "step": 31690
    },
    {
      "epoch": 5.170486989150828,
      "grad_norm": 4.5625,
      "learning_rate": 5.133540679738716e-06,
      "loss": 3.7068,
      "num_input_tokens_seen": 66201744,
      "step": 31695
    },
    {
      "epoch": 5.171302716371645,
      "grad_norm": 8.75,
      "learning_rate": 5.127582444616838e-06,
      "loss": 2.7309,
      "num_input_tokens_seen": 66211488,
      "step": 31700
    },
    {
      "epoch": 5.1721184435924625,
      "grad_norm": 6.5,
      "learning_rate": 5.121627274071486e-06,
      "loss": 3.3983,
      "num_input_tokens_seen": 66222560,
      "step": 31705
    },
    {
      "epoch": 5.17293417081328,
      "grad_norm": 9.125,
      "learning_rate": 5.115675169021009e-06,
      "loss": 1.6988,
      "num_input_tokens_seen": 66233088,
      "step": 31710
    },
    {
      "epoch": 5.173749898034098,
      "grad_norm": 0.90625,
      "learning_rate": 5.1097261303832994e-06,
      "loss": 2.181,
      "num_input_tokens_seen": 66243952,
      "step": 31715
    },
    {
      "epoch": 5.174565625254915,
      "grad_norm": 0.07861328125,
      "learning_rate": 5.103780159075788e-06,
      "loss": 1.5375,
      "num_input_tokens_seen": 66253872,
      "step": 31720
    },
    {
      "epoch": 5.175381352475732,
      "grad_norm": 8.9375,
      "learning_rate": 5.0978372560154e-06,
      "loss": 3.1914,
      "num_input_tokens_seen": 66263152,
      "step": 31725
    },
    {
      "epoch": 5.176197079696549,
      "grad_norm": 11.0625,
      "learning_rate": 5.091897422118619e-06,
      "loss": 3.4578,
      "num_input_tokens_seen": 66273792,
      "step": 31730
    },
    {
      "epoch": 5.177012806917367,
      "grad_norm": 9.625,
      "learning_rate": 5.0859606583014305e-06,
      "loss": 2.2166,
      "num_input_tokens_seen": 66284208,
      "step": 31735
    },
    {
      "epoch": 5.177828534138184,
      "grad_norm": 5.75,
      "learning_rate": 5.080026965479365e-06,
      "loss": 2.4539,
      "num_input_tokens_seen": 66294992,
      "step": 31740
    },
    {
      "epoch": 5.178644261359001,
      "grad_norm": 20.25,
      "learning_rate": 5.074096344567475e-06,
      "loss": 1.4331,
      "num_input_tokens_seen": 66304032,
      "step": 31745
    },
    {
      "epoch": 5.1794599885798185,
      "grad_norm": 19.125,
      "learning_rate": 5.0681687964803294e-06,
      "loss": 4.2854,
      "num_input_tokens_seen": 66313328,
      "step": 31750
    },
    {
      "epoch": 5.1802757158006365,
      "grad_norm": 8.5,
      "learning_rate": 5.06224432213204e-06,
      "loss": 3.1259,
      "num_input_tokens_seen": 66321696,
      "step": 31755
    },
    {
      "epoch": 5.181091443021454,
      "grad_norm": 4.1875,
      "learning_rate": 5.056322922436224e-06,
      "loss": 1.9123,
      "num_input_tokens_seen": 66331744,
      "step": 31760
    },
    {
      "epoch": 5.181907170242271,
      "grad_norm": 11.6875,
      "learning_rate": 5.0504045983060465e-06,
      "loss": 2.8251,
      "num_input_tokens_seen": 66341456,
      "step": 31765
    },
    {
      "epoch": 5.182722897463089,
      "grad_norm": 11.0625,
      "learning_rate": 5.044489350654183e-06,
      "loss": 2.0641,
      "num_input_tokens_seen": 66350864,
      "step": 31770
    },
    {
      "epoch": 5.183538624683906,
      "grad_norm": 4.25,
      "learning_rate": 5.038577180392831e-06,
      "loss": 1.2624,
      "num_input_tokens_seen": 66361136,
      "step": 31775
    },
    {
      "epoch": 5.184354351904723,
      "grad_norm": 5.96875,
      "learning_rate": 5.032668088433729e-06,
      "loss": 2.4957,
      "num_input_tokens_seen": 66371056,
      "step": 31780
    },
    {
      "epoch": 5.18517007912554,
      "grad_norm": 10.0625,
      "learning_rate": 5.02676207568814e-06,
      "loss": 3.5543,
      "num_input_tokens_seen": 66381744,
      "step": 31785
    },
    {
      "epoch": 5.185985806346358,
      "grad_norm": 12.25,
      "learning_rate": 5.02085914306683e-06,
      "loss": 1.855,
      "num_input_tokens_seen": 66393056,
      "step": 31790
    },
    {
      "epoch": 5.186801533567175,
      "grad_norm": 6.9375,
      "learning_rate": 5.014959291480123e-06,
      "loss": 2.3199,
      "num_input_tokens_seen": 66405392,
      "step": 31795
    },
    {
      "epoch": 5.1876172607879925,
      "grad_norm": 9.375,
      "learning_rate": 5.009062521837835e-06,
      "loss": 3.6034,
      "num_input_tokens_seen": 66416480,
      "step": 31800
    },
    {
      "epoch": 5.1876172607879925,
      "eval_loss": 2.537795066833496,
      "eval_runtime": 134.81,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 66416480,
      "step": 31800
    },
    {
      "epoch": 5.18843298800881,
      "grad_norm": 7.15625,
      "learning_rate": 5.003168835049324e-06,
      "loss": 2.4163,
      "num_input_tokens_seen": 66427040,
      "step": 31805
    },
    {
      "epoch": 5.189248715229628,
      "grad_norm": 6.1875,
      "learning_rate": 4.997278232023483e-06,
      "loss": 4.4261,
      "num_input_tokens_seen": 66436864,
      "step": 31810
    },
    {
      "epoch": 5.190064442450445,
      "grad_norm": 11.9375,
      "learning_rate": 4.9913907136687036e-06,
      "loss": 5.244,
      "num_input_tokens_seen": 66446400,
      "step": 31815
    },
    {
      "epoch": 5.190880169671262,
      "grad_norm": 5.84375,
      "learning_rate": 4.985506280892918e-06,
      "loss": 2.8668,
      "num_input_tokens_seen": 66456784,
      "step": 31820
    },
    {
      "epoch": 5.191695896892079,
      "grad_norm": 4.65625,
      "learning_rate": 4.979624934603589e-06,
      "loss": 2.4323,
      "num_input_tokens_seen": 66466704,
      "step": 31825
    },
    {
      "epoch": 5.192511624112897,
      "grad_norm": 4.71875,
      "learning_rate": 4.97374667570768e-06,
      "loss": 3.0871,
      "num_input_tokens_seen": 66477280,
      "step": 31830
    },
    {
      "epoch": 5.193327351333714,
      "grad_norm": 2.5625,
      "learning_rate": 4.967871505111704e-06,
      "loss": 2.7279,
      "num_input_tokens_seen": 66487744,
      "step": 31835
    },
    {
      "epoch": 5.194143078554531,
      "grad_norm": 5.15625,
      "learning_rate": 4.961999423721686e-06,
      "loss": 3.3858,
      "num_input_tokens_seen": 66498080,
      "step": 31840
    },
    {
      "epoch": 5.194958805775348,
      "grad_norm": 2.921875,
      "learning_rate": 4.956130432443159e-06,
      "loss": 2.5598,
      "num_input_tokens_seen": 66509040,
      "step": 31845
    },
    {
      "epoch": 5.195774532996166,
      "grad_norm": 8.375,
      "learning_rate": 4.950264532181215e-06,
      "loss": 2.0359,
      "num_input_tokens_seen": 66517952,
      "step": 31850
    },
    {
      "epoch": 5.196590260216984,
      "grad_norm": 5.3125,
      "learning_rate": 4.944401723840433e-06,
      "loss": 3.504,
      "num_input_tokens_seen": 66529056,
      "step": 31855
    },
    {
      "epoch": 5.197405987437801,
      "grad_norm": 6.625,
      "learning_rate": 4.938542008324942e-06,
      "loss": 2.0019,
      "num_input_tokens_seen": 66538912,
      "step": 31860
    },
    {
      "epoch": 5.198221714658618,
      "grad_norm": 8.4375,
      "learning_rate": 4.9326853865383855e-06,
      "loss": 1.951,
      "num_input_tokens_seen": 66548272,
      "step": 31865
    },
    {
      "epoch": 5.199037441879436,
      "grad_norm": 6.75,
      "learning_rate": 4.926831859383918e-06,
      "loss": 3.8909,
      "num_input_tokens_seen": 66559104,
      "step": 31870
    },
    {
      "epoch": 5.199853169100253,
      "grad_norm": 12.5,
      "learning_rate": 4.92098142776424e-06,
      "loss": 3.1511,
      "num_input_tokens_seen": 66568928,
      "step": 31875
    },
    {
      "epoch": 5.20066889632107,
      "grad_norm": 3.28125,
      "learning_rate": 4.91513409258155e-06,
      "loss": 2.3576,
      "num_input_tokens_seen": 66579952,
      "step": 31880
    },
    {
      "epoch": 5.201484623541887,
      "grad_norm": 7.15625,
      "learning_rate": 4.909289854737581e-06,
      "loss": 2.5183,
      "num_input_tokens_seen": 66589232,
      "step": 31885
    },
    {
      "epoch": 5.202300350762705,
      "grad_norm": 9.3125,
      "learning_rate": 4.903448715133602e-06,
      "loss": 2.877,
      "num_input_tokens_seen": 66599728,
      "step": 31890
    },
    {
      "epoch": 5.203116077983522,
      "grad_norm": 7.0625,
      "learning_rate": 4.897610674670372e-06,
      "loss": 2.5958,
      "num_input_tokens_seen": 66609392,
      "step": 31895
    },
    {
      "epoch": 5.2039318052043395,
      "grad_norm": 3.578125,
      "learning_rate": 4.8917757342482e-06,
      "loss": 2.2184,
      "num_input_tokens_seen": 66618816,
      "step": 31900
    },
    {
      "epoch": 5.204747532425157,
      "grad_norm": 5.0625,
      "learning_rate": 4.885943894766909e-06,
      "loss": 2.4726,
      "num_input_tokens_seen": 66629344,
      "step": 31905
    },
    {
      "epoch": 5.205563259645975,
      "grad_norm": 6.09375,
      "learning_rate": 4.880115157125842e-06,
      "loss": 1.6628,
      "num_input_tokens_seen": 66641392,
      "step": 31910
    },
    {
      "epoch": 5.206378986866792,
      "grad_norm": 7.125,
      "learning_rate": 4.874289522223857e-06,
      "loss": 2.5919,
      "num_input_tokens_seen": 66652272,
      "step": 31915
    },
    {
      "epoch": 5.207194714087609,
      "grad_norm": 6.6875,
      "learning_rate": 4.868466990959339e-06,
      "loss": 2.502,
      "num_input_tokens_seen": 66662016,
      "step": 31920
    },
    {
      "epoch": 5.208010441308426,
      "grad_norm": 7.59375,
      "learning_rate": 4.8626475642301964e-06,
      "loss": 2.9058,
      "num_input_tokens_seen": 66672032,
      "step": 31925
    },
    {
      "epoch": 5.208826168529244,
      "grad_norm": 5.65625,
      "learning_rate": 4.856831242933871e-06,
      "loss": 2.4547,
      "num_input_tokens_seen": 66684560,
      "step": 31930
    },
    {
      "epoch": 5.209641895750061,
      "grad_norm": 17.875,
      "learning_rate": 4.851018027967294e-06,
      "loss": 3.6895,
      "num_input_tokens_seen": 66694736,
      "step": 31935
    },
    {
      "epoch": 5.210457622970878,
      "grad_norm": 5.25,
      "learning_rate": 4.845207920226946e-06,
      "loss": 1.7334,
      "num_input_tokens_seen": 66705392,
      "step": 31940
    },
    {
      "epoch": 5.211273350191696,
      "grad_norm": 5.625,
      "learning_rate": 4.839400920608825e-06,
      "loss": 1.8278,
      "num_input_tokens_seen": 66716144,
      "step": 31945
    },
    {
      "epoch": 5.2120890774125135,
      "grad_norm": 4.71875,
      "learning_rate": 4.83359703000843e-06,
      "loss": 1.5545,
      "num_input_tokens_seen": 66726768,
      "step": 31950
    },
    {
      "epoch": 5.212904804633331,
      "grad_norm": 2.296875,
      "learning_rate": 4.827796249320804e-06,
      "loss": 3.9527,
      "num_input_tokens_seen": 66737952,
      "step": 31955
    },
    {
      "epoch": 5.213720531854148,
      "grad_norm": 14.4375,
      "learning_rate": 4.82199857944049e-06,
      "loss": 2.9756,
      "num_input_tokens_seen": 66746976,
      "step": 31960
    },
    {
      "epoch": 5.214536259074965,
      "grad_norm": 10.8125,
      "learning_rate": 4.8162040212615695e-06,
      "loss": 2.7158,
      "num_input_tokens_seen": 66757648,
      "step": 31965
    },
    {
      "epoch": 5.215351986295783,
      "grad_norm": 4.40625,
      "learning_rate": 4.810412575677639e-06,
      "loss": 2.2416,
      "num_input_tokens_seen": 66768464,
      "step": 31970
    },
    {
      "epoch": 5.2161677135166,
      "grad_norm": 6.375,
      "learning_rate": 4.804624243581801e-06,
      "loss": 2.1075,
      "num_input_tokens_seen": 66778624,
      "step": 31975
    },
    {
      "epoch": 5.216983440737417,
      "grad_norm": 8.25,
      "learning_rate": 4.798839025866703e-06,
      "loss": 3.6147,
      "num_input_tokens_seen": 66789776,
      "step": 31980
    },
    {
      "epoch": 5.217799167958235,
      "grad_norm": 6.65625,
      "learning_rate": 4.793056923424491e-06,
      "loss": 1.8919,
      "num_input_tokens_seen": 66798656,
      "step": 31985
    },
    {
      "epoch": 5.218614895179052,
      "grad_norm": 7.53125,
      "learning_rate": 4.78727793714683e-06,
      "loss": 2.1721,
      "num_input_tokens_seen": 66808448,
      "step": 31990
    },
    {
      "epoch": 5.219430622399869,
      "grad_norm": 7.5625,
      "learning_rate": 4.7815020679249285e-06,
      "loss": 1.8113,
      "num_input_tokens_seen": 66818384,
      "step": 31995
    },
    {
      "epoch": 5.220246349620687,
      "grad_norm": 3.5,
      "learning_rate": 4.775729316649483e-06,
      "loss": 1.1157,
      "num_input_tokens_seen": 66829712,
      "step": 32000
    },
    {
      "epoch": 5.220246349620687,
      "eval_loss": 2.5336523056030273,
      "eval_runtime": 134.9659,
      "eval_samples_per_second": 20.19,
      "eval_steps_per_second": 10.099,
      "num_input_tokens_seen": 66829712,
      "step": 32000
    },
    {
      "epoch": 5.221062076841505,
      "grad_norm": 6.125,
      "learning_rate": 4.769959684210728e-06,
      "loss": 1.2541,
      "num_input_tokens_seen": 66839920,
      "step": 32005
    },
    {
      "epoch": 5.221877804062322,
      "grad_norm": 4.71875,
      "learning_rate": 4.764193171498426e-06,
      "loss": 1.5787,
      "num_input_tokens_seen": 66851664,
      "step": 32010
    },
    {
      "epoch": 5.222693531283139,
      "grad_norm": 5.125,
      "learning_rate": 4.75842977940183e-06,
      "loss": 2.7063,
      "num_input_tokens_seen": 66861888,
      "step": 32015
    },
    {
      "epoch": 5.223509258503956,
      "grad_norm": 7.03125,
      "learning_rate": 4.752669508809729e-06,
      "loss": 2.559,
      "num_input_tokens_seen": 66873440,
      "step": 32020
    },
    {
      "epoch": 5.224324985724774,
      "grad_norm": 3.71875,
      "learning_rate": 4.746912360610445e-06,
      "loss": 1.5809,
      "num_input_tokens_seen": 66884336,
      "step": 32025
    },
    {
      "epoch": 5.225140712945591,
      "grad_norm": 5.4375,
      "learning_rate": 4.741158335691781e-06,
      "loss": 1.2973,
      "num_input_tokens_seen": 66894608,
      "step": 32030
    },
    {
      "epoch": 5.225956440166408,
      "grad_norm": 8.9375,
      "learning_rate": 4.7354074349410994e-06,
      "loss": 2.3383,
      "num_input_tokens_seen": 66904624,
      "step": 32035
    },
    {
      "epoch": 5.226772167387225,
      "grad_norm": 4.34375,
      "learning_rate": 4.729659659245245e-06,
      "loss": 3.308,
      "num_input_tokens_seen": 66916560,
      "step": 32040
    },
    {
      "epoch": 5.227587894608043,
      "grad_norm": 8.8125,
      "learning_rate": 4.723915009490601e-06,
      "loss": 3.1977,
      "num_input_tokens_seen": 66926144,
      "step": 32045
    },
    {
      "epoch": 5.2284036218288605,
      "grad_norm": 6.46875,
      "learning_rate": 4.718173486563077e-06,
      "loss": 1.9637,
      "num_input_tokens_seen": 66936800,
      "step": 32050
    },
    {
      "epoch": 5.229219349049678,
      "grad_norm": 8.1875,
      "learning_rate": 4.71243509134808e-06,
      "loss": 2.5018,
      "num_input_tokens_seen": 66947168,
      "step": 32055
    },
    {
      "epoch": 5.230035076270495,
      "grad_norm": 6.28125,
      "learning_rate": 4.706699824730532e-06,
      "loss": 2.8128,
      "num_input_tokens_seen": 66958176,
      "step": 32060
    },
    {
      "epoch": 5.230850803491313,
      "grad_norm": 8.4375,
      "learning_rate": 4.700967687594901e-06,
      "loss": 2.9674,
      "num_input_tokens_seen": 66969808,
      "step": 32065
    },
    {
      "epoch": 5.23166653071213,
      "grad_norm": 7.6875,
      "learning_rate": 4.69523868082514e-06,
      "loss": 3.2826,
      "num_input_tokens_seen": 66981632,
      "step": 32070
    },
    {
      "epoch": 5.232482257932947,
      "grad_norm": 12.6875,
      "learning_rate": 4.689512805304747e-06,
      "loss": 2.8637,
      "num_input_tokens_seen": 66992208,
      "step": 32075
    },
    {
      "epoch": 5.233297985153764,
      "grad_norm": 7.875,
      "learning_rate": 4.683790061916707e-06,
      "loss": 2.0746,
      "num_input_tokens_seen": 67002880,
      "step": 32080
    },
    {
      "epoch": 5.234113712374582,
      "grad_norm": 7.0,
      "learning_rate": 4.678070451543551e-06,
      "loss": 1.8365,
      "num_input_tokens_seen": 67014272,
      "step": 32085
    },
    {
      "epoch": 5.234929439595399,
      "grad_norm": 14.875,
      "learning_rate": 4.6723539750673204e-06,
      "loss": 2.4565,
      "num_input_tokens_seen": 67024096,
      "step": 32090
    },
    {
      "epoch": 5.2357451668162165,
      "grad_norm": 2.984375,
      "learning_rate": 4.666640633369551e-06,
      "loss": 1.5985,
      "num_input_tokens_seen": 67033792,
      "step": 32095
    },
    {
      "epoch": 5.236560894037034,
      "grad_norm": 6.59375,
      "learning_rate": 4.660930427331323e-06,
      "loss": 4.1726,
      "num_input_tokens_seen": 67044784,
      "step": 32100
    },
    {
      "epoch": 5.237376621257852,
      "grad_norm": 0.29296875,
      "learning_rate": 4.6552233578332244e-06,
      "loss": 1.3576,
      "num_input_tokens_seen": 67056784,
      "step": 32105
    },
    {
      "epoch": 5.238192348478669,
      "grad_norm": 4.28125,
      "learning_rate": 4.649519425755347e-06,
      "loss": 1.5543,
      "num_input_tokens_seen": 67067072,
      "step": 32110
    },
    {
      "epoch": 5.239008075699486,
      "grad_norm": 10.4375,
      "learning_rate": 4.64381863197732e-06,
      "loss": 2.8785,
      "num_input_tokens_seen": 67076736,
      "step": 32115
    },
    {
      "epoch": 5.239823802920303,
      "grad_norm": 3.859375,
      "learning_rate": 4.638120977378269e-06,
      "loss": 2.0424,
      "num_input_tokens_seen": 67087712,
      "step": 32120
    },
    {
      "epoch": 5.240639530141121,
      "grad_norm": 5.125,
      "learning_rate": 4.632426462836848e-06,
      "loss": 2.3282,
      "num_input_tokens_seen": 67097968,
      "step": 32125
    },
    {
      "epoch": 5.241455257361938,
      "grad_norm": 12.0,
      "learning_rate": 4.626735089231224e-06,
      "loss": 3.935,
      "num_input_tokens_seen": 67108800,
      "step": 32130
    },
    {
      "epoch": 5.242270984582755,
      "grad_norm": 0.138671875,
      "learning_rate": 4.621046857439068e-06,
      "loss": 3.3208,
      "num_input_tokens_seen": 67119088,
      "step": 32135
    },
    {
      "epoch": 5.243086711803572,
      "grad_norm": 9.3125,
      "learning_rate": 4.615361768337587e-06,
      "loss": 2.145,
      "num_input_tokens_seen": 67130176,
      "step": 32140
    },
    {
      "epoch": 5.2439024390243905,
      "grad_norm": 6.0625,
      "learning_rate": 4.6096798228034946e-06,
      "loss": 1.6552,
      "num_input_tokens_seen": 67140512,
      "step": 32145
    },
    {
      "epoch": 5.244718166245208,
      "grad_norm": 2.46875,
      "learning_rate": 4.604001021713008e-06,
      "loss": 1.3819,
      "num_input_tokens_seen": 67150560,
      "step": 32150
    },
    {
      "epoch": 5.245533893466025,
      "grad_norm": 4.75,
      "learning_rate": 4.598325365941883e-06,
      "loss": 2.4702,
      "num_input_tokens_seen": 67160880,
      "step": 32155
    },
    {
      "epoch": 5.246349620686843,
      "grad_norm": 2.53125,
      "learning_rate": 4.5926528563653645e-06,
      "loss": 2.679,
      "num_input_tokens_seen": 67171072,
      "step": 32160
    },
    {
      "epoch": 5.24716534790766,
      "grad_norm": 2.4375,
      "learning_rate": 4.5869834938582295e-06,
      "loss": 2.5635,
      "num_input_tokens_seen": 67182512,
      "step": 32165
    },
    {
      "epoch": 5.247981075128477,
      "grad_norm": 6.90625,
      "learning_rate": 4.581317279294772e-06,
      "loss": 1.6906,
      "num_input_tokens_seen": 67192528,
      "step": 32170
    },
    {
      "epoch": 5.248796802349294,
      "grad_norm": 3.828125,
      "learning_rate": 4.57565421354878e-06,
      "loss": 2.5127,
      "num_input_tokens_seen": 67201712,
      "step": 32175
    },
    {
      "epoch": 5.249612529570112,
      "grad_norm": 8.1875,
      "learning_rate": 4.569994297493579e-06,
      "loss": 2.983,
      "num_input_tokens_seen": 67211488,
      "step": 32180
    },
    {
      "epoch": 5.250428256790929,
      "grad_norm": 10.3125,
      "learning_rate": 4.564337532002002e-06,
      "loss": 2.785,
      "num_input_tokens_seen": 67223168,
      "step": 32185
    },
    {
      "epoch": 5.251243984011746,
      "grad_norm": 11.0,
      "learning_rate": 4.55868391794638e-06,
      "loss": 2.3999,
      "num_input_tokens_seen": 67233920,
      "step": 32190
    },
    {
      "epoch": 5.2520597112325635,
      "grad_norm": 5.6875,
      "learning_rate": 4.553033456198588e-06,
      "loss": 2.4823,
      "num_input_tokens_seen": 67244016,
      "step": 32195
    },
    {
      "epoch": 5.252875438453382,
      "grad_norm": 6.625,
      "learning_rate": 4.54738614762999e-06,
      "loss": 2.284,
      "num_input_tokens_seen": 67253936,
      "step": 32200
    },
    {
      "epoch": 5.252875438453382,
      "eval_loss": 2.551058292388916,
      "eval_runtime": 134.808,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 67253936,
      "step": 32200
    },
    {
      "epoch": 5.253691165674199,
      "grad_norm": 16.25,
      "learning_rate": 4.541741993111465e-06,
      "loss": 4.9573,
      "num_input_tokens_seen": 67263344,
      "step": 32205
    },
    {
      "epoch": 5.254506892895016,
      "grad_norm": 6.46875,
      "learning_rate": 4.536100993513423e-06,
      "loss": 0.8461,
      "num_input_tokens_seen": 67272144,
      "step": 32210
    },
    {
      "epoch": 5.255322620115833,
      "grad_norm": 12.9375,
      "learning_rate": 4.530463149705768e-06,
      "loss": 1.5401,
      "num_input_tokens_seen": 67282128,
      "step": 32215
    },
    {
      "epoch": 5.256138347336651,
      "grad_norm": 8.9375,
      "learning_rate": 4.524828462557934e-06,
      "loss": 2.2447,
      "num_input_tokens_seen": 67293008,
      "step": 32220
    },
    {
      "epoch": 5.256954074557468,
      "grad_norm": 4.0,
      "learning_rate": 4.5191969329388625e-06,
      "loss": 2.0911,
      "num_input_tokens_seen": 67303792,
      "step": 32225
    },
    {
      "epoch": 5.257769801778285,
      "grad_norm": 11.0625,
      "learning_rate": 4.5135685617169965e-06,
      "loss": 3.7871,
      "num_input_tokens_seen": 67315088,
      "step": 32230
    },
    {
      "epoch": 5.258585528999102,
      "grad_norm": 9.625,
      "learning_rate": 4.507943349760313e-06,
      "loss": 1.454,
      "num_input_tokens_seen": 67325984,
      "step": 32235
    },
    {
      "epoch": 5.25940125621992,
      "grad_norm": 9.3125,
      "learning_rate": 4.502321297936277e-06,
      "loss": 3.54,
      "num_input_tokens_seen": 67336736,
      "step": 32240
    },
    {
      "epoch": 5.2602169834407375,
      "grad_norm": 6.28125,
      "learning_rate": 4.496702407111888e-06,
      "loss": 2.2236,
      "num_input_tokens_seen": 67348336,
      "step": 32245
    },
    {
      "epoch": 5.261032710661555,
      "grad_norm": 7.75,
      "learning_rate": 4.491086678153653e-06,
      "loss": 1.8731,
      "num_input_tokens_seen": 67358048,
      "step": 32250
    },
    {
      "epoch": 5.261848437882372,
      "grad_norm": 12.75,
      "learning_rate": 4.485474111927579e-06,
      "loss": 3.4648,
      "num_input_tokens_seen": 67367520,
      "step": 32255
    },
    {
      "epoch": 5.26266416510319,
      "grad_norm": 7.46875,
      "learning_rate": 4.479864709299197e-06,
      "loss": 1.6093,
      "num_input_tokens_seen": 67379680,
      "step": 32260
    },
    {
      "epoch": 5.263479892324007,
      "grad_norm": 5.15625,
      "learning_rate": 4.474258471133555e-06,
      "loss": 1.1844,
      "num_input_tokens_seen": 67389632,
      "step": 32265
    },
    {
      "epoch": 5.264295619544824,
      "grad_norm": 4.65625,
      "learning_rate": 4.4686553982952014e-06,
      "loss": 2.8365,
      "num_input_tokens_seen": 67400288,
      "step": 32270
    },
    {
      "epoch": 5.265111346765641,
      "grad_norm": 6.15625,
      "learning_rate": 4.463055491648191e-06,
      "loss": 3.1175,
      "num_input_tokens_seen": 67410336,
      "step": 32275
    },
    {
      "epoch": 5.265927073986459,
      "grad_norm": 6.84375,
      "learning_rate": 4.457458752056112e-06,
      "loss": 2.277,
      "num_input_tokens_seen": 67420944,
      "step": 32280
    },
    {
      "epoch": 5.266742801207276,
      "grad_norm": 5.90625,
      "learning_rate": 4.451865180382042e-06,
      "loss": 3.3839,
      "num_input_tokens_seen": 67431184,
      "step": 32285
    },
    {
      "epoch": 5.2675585284280935,
      "grad_norm": 9.25,
      "learning_rate": 4.4462747774885936e-06,
      "loss": 1.4052,
      "num_input_tokens_seen": 67442512,
      "step": 32290
    },
    {
      "epoch": 5.268374255648911,
      "grad_norm": 11.6875,
      "learning_rate": 4.440687544237859e-06,
      "loss": 2.108,
      "num_input_tokens_seen": 67454128,
      "step": 32295
    },
    {
      "epoch": 5.269189982869729,
      "grad_norm": 3.96875,
      "learning_rate": 4.435103481491471e-06,
      "loss": 1.6494,
      "num_input_tokens_seen": 67462240,
      "step": 32300
    },
    {
      "epoch": 5.270005710090546,
      "grad_norm": 2.203125,
      "learning_rate": 4.429522590110569e-06,
      "loss": 2.6513,
      "num_input_tokens_seen": 67473200,
      "step": 32305
    },
    {
      "epoch": 5.270821437311363,
      "grad_norm": 6.21875,
      "learning_rate": 4.423944870955779e-06,
      "loss": 1.4311,
      "num_input_tokens_seen": 67484400,
      "step": 32310
    },
    {
      "epoch": 5.27163716453218,
      "grad_norm": 4.71875,
      "learning_rate": 4.418370324887272e-06,
      "loss": 1.1231,
      "num_input_tokens_seen": 67494656,
      "step": 32315
    },
    {
      "epoch": 5.272452891752998,
      "grad_norm": 6.21875,
      "learning_rate": 4.412798952764699e-06,
      "loss": 3.011,
      "num_input_tokens_seen": 67504944,
      "step": 32320
    },
    {
      "epoch": 5.273268618973815,
      "grad_norm": 8.625,
      "learning_rate": 4.407230755447245e-06,
      "loss": 2.2878,
      "num_input_tokens_seen": 67516272,
      "step": 32325
    },
    {
      "epoch": 5.274084346194632,
      "grad_norm": 7.84375,
      "learning_rate": 4.401665733793598e-06,
      "loss": 1.6345,
      "num_input_tokens_seen": 67526672,
      "step": 32330
    },
    {
      "epoch": 5.27490007341545,
      "grad_norm": 5.53125,
      "learning_rate": 4.3961038886619425e-06,
      "loss": 1.953,
      "num_input_tokens_seen": 67537008,
      "step": 32335
    },
    {
      "epoch": 5.275715800636267,
      "grad_norm": 8.0,
      "learning_rate": 4.39054522091e-06,
      "loss": 2.5599,
      "num_input_tokens_seen": 67548528,
      "step": 32340
    },
    {
      "epoch": 5.276531527857085,
      "grad_norm": 15.625,
      "learning_rate": 4.384989731394979e-06,
      "loss": 2.9885,
      "num_input_tokens_seen": 67558592,
      "step": 32345
    },
    {
      "epoch": 5.277347255077902,
      "grad_norm": 5.25,
      "learning_rate": 4.379437420973598e-06,
      "loss": 2.4236,
      "num_input_tokens_seen": 67567600,
      "step": 32350
    },
    {
      "epoch": 5.27816298229872,
      "grad_norm": 7.875,
      "learning_rate": 4.373888290502107e-06,
      "loss": 2.086,
      "num_input_tokens_seen": 67579440,
      "step": 32355
    },
    {
      "epoch": 5.278978709519537,
      "grad_norm": 6.5,
      "learning_rate": 4.36834234083624e-06,
      "loss": 2.9554,
      "num_input_tokens_seen": 67591344,
      "step": 32360
    },
    {
      "epoch": 5.279794436740354,
      "grad_norm": 5.59375,
      "learning_rate": 4.362799572831258e-06,
      "loss": 2.9039,
      "num_input_tokens_seen": 67602144,
      "step": 32365
    },
    {
      "epoch": 5.280610163961171,
      "grad_norm": 5.5,
      "learning_rate": 4.35725998734193e-06,
      "loss": 2.2129,
      "num_input_tokens_seen": 67612032,
      "step": 32370
    },
    {
      "epoch": 5.281425891181989,
      "grad_norm": 20.125,
      "learning_rate": 4.3517235852225195e-06,
      "loss": 1.6613,
      "num_input_tokens_seen": 67621120,
      "step": 32375
    },
    {
      "epoch": 5.282241618402806,
      "grad_norm": 8.5,
      "learning_rate": 4.346190367326822e-06,
      "loss": 1.8807,
      "num_input_tokens_seen": 67631680,
      "step": 32380
    },
    {
      "epoch": 5.283057345623623,
      "grad_norm": 7.28125,
      "learning_rate": 4.340660334508115e-06,
      "loss": 2.5104,
      "num_input_tokens_seen": 67641568,
      "step": 32385
    },
    {
      "epoch": 5.2838730728444405,
      "grad_norm": 5.03125,
      "learning_rate": 4.335133487619206e-06,
      "loss": 3.0586,
      "num_input_tokens_seen": 67652336,
      "step": 32390
    },
    {
      "epoch": 5.2846888000652585,
      "grad_norm": 5.40625,
      "learning_rate": 4.329609827512409e-06,
      "loss": 2.4066,
      "num_input_tokens_seen": 67662848,
      "step": 32395
    },
    {
      "epoch": 5.285504527286076,
      "grad_norm": 5.53125,
      "learning_rate": 4.324089355039531e-06,
      "loss": 1.9202,
      "num_input_tokens_seen": 67674048,
      "step": 32400
    },
    {
      "epoch": 5.285504527286076,
      "eval_loss": 2.533475637435913,
      "eval_runtime": 134.8117,
      "eval_samples_per_second": 20.213,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 67674048,
      "step": 32400
    },
    {
      "epoch": 5.286320254506893,
      "grad_norm": 5.0625,
      "learning_rate": 4.3185720710519075e-06,
      "loss": 1.4574,
      "num_input_tokens_seen": 67685024,
      "step": 32405
    },
    {
      "epoch": 5.28713598172771,
      "grad_norm": 7.6875,
      "learning_rate": 4.3130579764003724e-06,
      "loss": 1.5337,
      "num_input_tokens_seen": 67695840,
      "step": 32410
    },
    {
      "epoch": 5.287951708948528,
      "grad_norm": 12.9375,
      "learning_rate": 4.307547071935267e-06,
      "loss": 2.3591,
      "num_input_tokens_seen": 67705408,
      "step": 32415
    },
    {
      "epoch": 5.288767436169345,
      "grad_norm": 11.5,
      "learning_rate": 4.302039358506435e-06,
      "loss": 1.9742,
      "num_input_tokens_seen": 67715168,
      "step": 32420
    },
    {
      "epoch": 5.289583163390162,
      "grad_norm": 8.25,
      "learning_rate": 4.296534836963245e-06,
      "loss": 3.153,
      "num_input_tokens_seen": 67725184,
      "step": 32425
    },
    {
      "epoch": 5.290398890610979,
      "grad_norm": 6.90625,
      "learning_rate": 4.291033508154555e-06,
      "loss": 1.133,
      "num_input_tokens_seen": 67735760,
      "step": 32430
    },
    {
      "epoch": 5.291214617831797,
      "grad_norm": 2.875,
      "learning_rate": 4.285535372928748e-06,
      "loss": 1.3177,
      "num_input_tokens_seen": 67746560,
      "step": 32435
    },
    {
      "epoch": 5.2920303450526145,
      "grad_norm": 5.125,
      "learning_rate": 4.280040432133695e-06,
      "loss": 2.1957,
      "num_input_tokens_seen": 67757488,
      "step": 32440
    },
    {
      "epoch": 5.292846072273432,
      "grad_norm": 7.375,
      "learning_rate": 4.274548686616789e-06,
      "loss": 2.8725,
      "num_input_tokens_seen": 67768752,
      "step": 32445
    },
    {
      "epoch": 5.293661799494249,
      "grad_norm": 8.8125,
      "learning_rate": 4.2690601372249364e-06,
      "loss": 2.4883,
      "num_input_tokens_seen": 67778272,
      "step": 32450
    },
    {
      "epoch": 5.294477526715067,
      "grad_norm": 7.15625,
      "learning_rate": 4.263574784804525e-06,
      "loss": 2.4971,
      "num_input_tokens_seen": 67790016,
      "step": 32455
    },
    {
      "epoch": 5.295293253935884,
      "grad_norm": 1.7421875,
      "learning_rate": 4.258092630201479e-06,
      "loss": 2.5902,
      "num_input_tokens_seen": 67799584,
      "step": 32460
    },
    {
      "epoch": 5.296108981156701,
      "grad_norm": 5.5625,
      "learning_rate": 4.252613674261202e-06,
      "loss": 3.9137,
      "num_input_tokens_seen": 67809680,
      "step": 32465
    },
    {
      "epoch": 5.296924708377518,
      "grad_norm": 5.78125,
      "learning_rate": 4.2471379178286224e-06,
      "loss": 2.5223,
      "num_input_tokens_seen": 67819696,
      "step": 32470
    },
    {
      "epoch": 5.297740435598336,
      "grad_norm": 4.09375,
      "learning_rate": 4.241665361748181e-06,
      "loss": 3.1281,
      "num_input_tokens_seen": 67829616,
      "step": 32475
    },
    {
      "epoch": 5.298556162819153,
      "grad_norm": 5.09375,
      "learning_rate": 4.2361960068637994e-06,
      "loss": 1.7755,
      "num_input_tokens_seen": 67838944,
      "step": 32480
    },
    {
      "epoch": 5.2993718900399704,
      "grad_norm": 8.25,
      "learning_rate": 4.230729854018933e-06,
      "loss": 2.3426,
      "num_input_tokens_seen": 67848096,
      "step": 32485
    },
    {
      "epoch": 5.300187617260788,
      "grad_norm": 4.15625,
      "learning_rate": 4.225266904056521e-06,
      "loss": 1.8267,
      "num_input_tokens_seen": 67859296,
      "step": 32490
    },
    {
      "epoch": 5.301003344481606,
      "grad_norm": 2.875,
      "learning_rate": 4.21980715781903e-06,
      "loss": 2.4412,
      "num_input_tokens_seen": 67870448,
      "step": 32495
    },
    {
      "epoch": 5.301819071702423,
      "grad_norm": 8.3125,
      "learning_rate": 4.214350616148416e-06,
      "loss": 1.9188,
      "num_input_tokens_seen": 67882608,
      "step": 32500
    },
    {
      "epoch": 5.30263479892324,
      "grad_norm": 2.9375,
      "learning_rate": 4.20889727988614e-06,
      "loss": 2.1714,
      "num_input_tokens_seen": 67893648,
      "step": 32505
    },
    {
      "epoch": 5.303450526144058,
      "grad_norm": 4.4375,
      "learning_rate": 4.20344714987318e-06,
      "loss": 1.9919,
      "num_input_tokens_seen": 67903344,
      "step": 32510
    },
    {
      "epoch": 5.304266253364875,
      "grad_norm": 7.5625,
      "learning_rate": 4.198000226950022e-06,
      "loss": 2.2972,
      "num_input_tokens_seen": 67915104,
      "step": 32515
    },
    {
      "epoch": 5.305081980585692,
      "grad_norm": 13.375,
      "learning_rate": 4.192556511956635e-06,
      "loss": 3.1468,
      "num_input_tokens_seen": 67924016,
      "step": 32520
    },
    {
      "epoch": 5.305897707806509,
      "grad_norm": 8.9375,
      "learning_rate": 4.18711600573252e-06,
      "loss": 3.1367,
      "num_input_tokens_seen": 67935280,
      "step": 32525
    },
    {
      "epoch": 5.306713435027326,
      "grad_norm": 5.6875,
      "learning_rate": 4.181678709116671e-06,
      "loss": 2.1795,
      "num_input_tokens_seen": 67945904,
      "step": 32530
    },
    {
      "epoch": 5.307529162248144,
      "grad_norm": 10.125,
      "learning_rate": 4.1762446229475785e-06,
      "loss": 1.7417,
      "num_input_tokens_seen": 67956576,
      "step": 32535
    },
    {
      "epoch": 5.3083448894689615,
      "grad_norm": 8.3125,
      "learning_rate": 4.17081374806326e-06,
      "loss": 1.4875,
      "num_input_tokens_seen": 67966880,
      "step": 32540
    },
    {
      "epoch": 5.309160616689779,
      "grad_norm": 3.640625,
      "learning_rate": 4.165386085301212e-06,
      "loss": 2.3882,
      "num_input_tokens_seen": 67977120,
      "step": 32545
    },
    {
      "epoch": 5.309976343910597,
      "grad_norm": 2.703125,
      "learning_rate": 4.1599616354984525e-06,
      "loss": 2.4555,
      "num_input_tokens_seen": 67987424,
      "step": 32550
    },
    {
      "epoch": 5.310792071131414,
      "grad_norm": 6.3125,
      "learning_rate": 4.154540399491508e-06,
      "loss": 1.6428,
      "num_input_tokens_seen": 67997840,
      "step": 32555
    },
    {
      "epoch": 5.311607798352231,
      "grad_norm": 8.3125,
      "learning_rate": 4.149122378116394e-06,
      "loss": 2.4289,
      "num_input_tokens_seen": 68008352,
      "step": 32560
    },
    {
      "epoch": 5.312423525573048,
      "grad_norm": 3.765625,
      "learning_rate": 4.14370757220863e-06,
      "loss": 2.1248,
      "num_input_tokens_seen": 68018960,
      "step": 32565
    },
    {
      "epoch": 5.313239252793866,
      "grad_norm": 9.75,
      "learning_rate": 4.138295982603263e-06,
      "loss": 1.7286,
      "num_input_tokens_seen": 68029760,
      "step": 32570
    },
    {
      "epoch": 5.314054980014683,
      "grad_norm": 7.78125,
      "learning_rate": 4.132887610134814e-06,
      "loss": 2.0278,
      "num_input_tokens_seen": 68041344,
      "step": 32575
    },
    {
      "epoch": 5.3148707072355,
      "grad_norm": 3.8125,
      "learning_rate": 4.127482455637335e-06,
      "loss": 2.4586,
      "num_input_tokens_seen": 68052192,
      "step": 32580
    },
    {
      "epoch": 5.3156864344563175,
      "grad_norm": 0.103515625,
      "learning_rate": 4.1220805199443545e-06,
      "loss": 1.3991,
      "num_input_tokens_seen": 68063104,
      "step": 32585
    },
    {
      "epoch": 5.3165021616771355,
      "grad_norm": 8.125,
      "learning_rate": 4.116681803888925e-06,
      "loss": 1.8607,
      "num_input_tokens_seen": 68074704,
      "step": 32590
    },
    {
      "epoch": 5.317317888897953,
      "grad_norm": 6.25,
      "learning_rate": 4.111286308303605e-06,
      "loss": 1.6929,
      "num_input_tokens_seen": 68085408,
      "step": 32595
    },
    {
      "epoch": 5.31813361611877,
      "grad_norm": 8.75,
      "learning_rate": 4.105894034020433e-06,
      "loss": 2.2426,
      "num_input_tokens_seen": 68096656,
      "step": 32600
    },
    {
      "epoch": 5.31813361611877,
      "eval_loss": 2.5445759296417236,
      "eval_runtime": 134.7521,
      "eval_samples_per_second": 20.222,
      "eval_steps_per_second": 10.115,
      "num_input_tokens_seen": 68096656,
      "step": 32600
    },
    {
      "epoch": 5.318949343339587,
      "grad_norm": 8.4375,
      "learning_rate": 4.100504981870975e-06,
      "loss": 2.0147,
      "num_input_tokens_seen": 68107840,
      "step": 32605
    },
    {
      "epoch": 5.319765070560405,
      "grad_norm": 6.90625,
      "learning_rate": 4.0951191526862915e-06,
      "loss": 3.1333,
      "num_input_tokens_seen": 68119552,
      "step": 32610
    },
    {
      "epoch": 5.320580797781222,
      "grad_norm": 1.65625,
      "learning_rate": 4.089736547296938e-06,
      "loss": 2.0765,
      "num_input_tokens_seen": 68131072,
      "step": 32615
    },
    {
      "epoch": 5.321396525002039,
      "grad_norm": 4.40625,
      "learning_rate": 4.08435716653299e-06,
      "loss": 3.0563,
      "num_input_tokens_seen": 68141904,
      "step": 32620
    },
    {
      "epoch": 5.322212252222856,
      "grad_norm": 8.5,
      "learning_rate": 4.0789810112240005e-06,
      "loss": 1.7536,
      "num_input_tokens_seen": 68153152,
      "step": 32625
    },
    {
      "epoch": 5.323027979443674,
      "grad_norm": 2.75,
      "learning_rate": 4.073608082199057e-06,
      "loss": 2.3633,
      "num_input_tokens_seen": 68162144,
      "step": 32630
    },
    {
      "epoch": 5.3238437066644915,
      "grad_norm": 8.75,
      "learning_rate": 4.068238380286718e-06,
      "loss": 2.5521,
      "num_input_tokens_seen": 68173824,
      "step": 32635
    },
    {
      "epoch": 5.324659433885309,
      "grad_norm": 12.9375,
      "learning_rate": 4.062871906315072e-06,
      "loss": 4.4844,
      "num_input_tokens_seen": 68185104,
      "step": 32640
    },
    {
      "epoch": 5.325475161106126,
      "grad_norm": 12.625,
      "learning_rate": 4.057508661111686e-06,
      "loss": 1.91,
      "num_input_tokens_seen": 68195792,
      "step": 32645
    },
    {
      "epoch": 5.326290888326944,
      "grad_norm": 3.703125,
      "learning_rate": 4.052148645503648e-06,
      "loss": 2.0108,
      "num_input_tokens_seen": 68206432,
      "step": 32650
    },
    {
      "epoch": 5.327106615547761,
      "grad_norm": 6.65625,
      "learning_rate": 4.046791860317531e-06,
      "loss": 1.552,
      "num_input_tokens_seen": 68216560,
      "step": 32655
    },
    {
      "epoch": 5.327922342768578,
      "grad_norm": 5.90625,
      "learning_rate": 4.041438306379431e-06,
      "loss": 2.4758,
      "num_input_tokens_seen": 68226208,
      "step": 32660
    },
    {
      "epoch": 5.328738069989395,
      "grad_norm": 2.265625,
      "learning_rate": 4.036087984514916e-06,
      "loss": 1.5428,
      "num_input_tokens_seen": 68237168,
      "step": 32665
    },
    {
      "epoch": 5.329553797210213,
      "grad_norm": 9.6875,
      "learning_rate": 4.030740895549084e-06,
      "loss": 2.2219,
      "num_input_tokens_seen": 68246800,
      "step": 32670
    },
    {
      "epoch": 5.33036952443103,
      "grad_norm": 7.125,
      "learning_rate": 4.025397040306531e-06,
      "loss": 1.5325,
      "num_input_tokens_seen": 68256720,
      "step": 32675
    },
    {
      "epoch": 5.331185251651847,
      "grad_norm": 2.265625,
      "learning_rate": 4.0200564196113285e-06,
      "loss": 1.7349,
      "num_input_tokens_seen": 68268000,
      "step": 32680
    },
    {
      "epoch": 5.332000978872665,
      "grad_norm": 7.0625,
      "learning_rate": 4.014719034287079e-06,
      "loss": 3.0456,
      "num_input_tokens_seen": 68278608,
      "step": 32685
    },
    {
      "epoch": 5.332816706093483,
      "grad_norm": 12.0625,
      "learning_rate": 4.0093848851568775e-06,
      "loss": 2.742,
      "num_input_tokens_seen": 68288064,
      "step": 32690
    },
    {
      "epoch": 5.3336324333143,
      "grad_norm": 7.96875,
      "learning_rate": 4.004053973043304e-06,
      "loss": 1.7614,
      "num_input_tokens_seen": 68300560,
      "step": 32695
    },
    {
      "epoch": 5.334448160535117,
      "grad_norm": 4.6875,
      "learning_rate": 3.998726298768465e-06,
      "loss": 1.8683,
      "num_input_tokens_seen": 68312400,
      "step": 32700
    },
    {
      "epoch": 5.335263887755934,
      "grad_norm": 7.59375,
      "learning_rate": 3.99340186315395e-06,
      "loss": 2.6379,
      "num_input_tokens_seen": 68323984,
      "step": 32705
    },
    {
      "epoch": 5.336079614976752,
      "grad_norm": 7.0,
      "learning_rate": 3.988080667020849e-06,
      "loss": 3.6827,
      "num_input_tokens_seen": 68334384,
      "step": 32710
    },
    {
      "epoch": 5.336895342197569,
      "grad_norm": 11.75,
      "learning_rate": 3.982762711189766e-06,
      "loss": 2.6002,
      "num_input_tokens_seen": 68345152,
      "step": 32715
    },
    {
      "epoch": 5.337711069418386,
      "grad_norm": 3.734375,
      "learning_rate": 3.977447996480785e-06,
      "loss": 2.3311,
      "num_input_tokens_seen": 68355280,
      "step": 32720
    },
    {
      "epoch": 5.338526796639204,
      "grad_norm": 5.90625,
      "learning_rate": 3.97213652371351e-06,
      "loss": 3.7039,
      "num_input_tokens_seen": 68364992,
      "step": 32725
    },
    {
      "epoch": 5.339342523860021,
      "grad_norm": 7.75,
      "learning_rate": 3.966828293707042e-06,
      "loss": 3.4688,
      "num_input_tokens_seen": 68375824,
      "step": 32730
    },
    {
      "epoch": 5.3401582510808385,
      "grad_norm": 10.5625,
      "learning_rate": 3.961523307279963e-06,
      "loss": 2.4766,
      "num_input_tokens_seen": 68384864,
      "step": 32735
    },
    {
      "epoch": 5.340973978301656,
      "grad_norm": 2.984375,
      "learning_rate": 3.956221565250382e-06,
      "loss": 2.6872,
      "num_input_tokens_seen": 68395744,
      "step": 32740
    },
    {
      "epoch": 5.341789705522474,
      "grad_norm": 7.34375,
      "learning_rate": 3.950923068435883e-06,
      "loss": 2.4929,
      "num_input_tokens_seen": 68405536,
      "step": 32745
    },
    {
      "epoch": 5.342605432743291,
      "grad_norm": 8.5625,
      "learning_rate": 3.945627817653566e-06,
      "loss": 2.6151,
      "num_input_tokens_seen": 68415440,
      "step": 32750
    },
    {
      "epoch": 5.343421159964108,
      "grad_norm": 11.6875,
      "learning_rate": 3.9403358137200335e-06,
      "loss": 3.74,
      "num_input_tokens_seen": 68426080,
      "step": 32755
    },
    {
      "epoch": 5.344236887184925,
      "grad_norm": 5.1875,
      "learning_rate": 3.9350470574513605e-06,
      "loss": 1.9137,
      "num_input_tokens_seen": 68437056,
      "step": 32760
    },
    {
      "epoch": 5.345052614405743,
      "grad_norm": 9.5,
      "learning_rate": 3.9297615496631525e-06,
      "loss": 1.9923,
      "num_input_tokens_seen": 68447056,
      "step": 32765
    },
    {
      "epoch": 5.34586834162656,
      "grad_norm": 6.09375,
      "learning_rate": 3.924479291170505e-06,
      "loss": 1.5642,
      "num_input_tokens_seen": 68457936,
      "step": 32770
    },
    {
      "epoch": 5.346684068847377,
      "grad_norm": 9.1875,
      "learning_rate": 3.919200282788002e-06,
      "loss": 2.637,
      "num_input_tokens_seen": 68468528,
      "step": 32775
    },
    {
      "epoch": 5.3474997960681945,
      "grad_norm": 7.53125,
      "learning_rate": 3.913924525329726e-06,
      "loss": 1.9295,
      "num_input_tokens_seen": 68480448,
      "step": 32780
    },
    {
      "epoch": 5.3483155232890125,
      "grad_norm": 7.96875,
      "learning_rate": 3.908652019609279e-06,
      "loss": 2.6697,
      "num_input_tokens_seen": 68492144,
      "step": 32785
    },
    {
      "epoch": 5.34913125050983,
      "grad_norm": 10.625,
      "learning_rate": 3.9033827664397364e-06,
      "loss": 2.2259,
      "num_input_tokens_seen": 68501632,
      "step": 32790
    },
    {
      "epoch": 5.349946977730647,
      "grad_norm": 9.8125,
      "learning_rate": 3.898116766633694e-06,
      "loss": 2.8508,
      "num_input_tokens_seen": 68512016,
      "step": 32795
    },
    {
      "epoch": 5.350762704951464,
      "grad_norm": 8.1875,
      "learning_rate": 3.8928540210032225e-06,
      "loss": 2.4398,
      "num_input_tokens_seen": 68521600,
      "step": 32800
    },
    {
      "epoch": 5.350762704951464,
      "eval_loss": 2.5286903381347656,
      "eval_runtime": 134.6307,
      "eval_samples_per_second": 20.241,
      "eval_steps_per_second": 10.124,
      "num_input_tokens_seen": 68521600,
      "step": 32800
    },
    {
      "epoch": 5.351578432172282,
      "grad_norm": 13.6875,
      "learning_rate": 3.887594530359909e-06,
      "loss": 3.2388,
      "num_input_tokens_seen": 68532432,
      "step": 32805
    },
    {
      "epoch": 5.352394159393099,
      "grad_norm": 11.125,
      "learning_rate": 3.88233829551484e-06,
      "loss": 1.8996,
      "num_input_tokens_seen": 68542336,
      "step": 32810
    },
    {
      "epoch": 5.353209886613916,
      "grad_norm": 6.625,
      "learning_rate": 3.877085317278581e-06,
      "loss": 1.8685,
      "num_input_tokens_seen": 68553520,
      "step": 32815
    },
    {
      "epoch": 5.354025613834733,
      "grad_norm": 8.3125,
      "learning_rate": 3.87183559646122e-06,
      "loss": 2.354,
      "num_input_tokens_seen": 68564048,
      "step": 32820
    },
    {
      "epoch": 5.354841341055551,
      "grad_norm": 5.875,
      "learning_rate": 3.866589133872317e-06,
      "loss": 2.4411,
      "num_input_tokens_seen": 68574256,
      "step": 32825
    },
    {
      "epoch": 5.3556570682763684,
      "grad_norm": 5.59375,
      "learning_rate": 3.861345930320948e-06,
      "loss": 1.5555,
      "num_input_tokens_seen": 68585120,
      "step": 32830
    },
    {
      "epoch": 5.356472795497186,
      "grad_norm": 4.9375,
      "learning_rate": 3.856105986615688e-06,
      "loss": 1.8829,
      "num_input_tokens_seen": 68594976,
      "step": 32835
    },
    {
      "epoch": 5.357288522718003,
      "grad_norm": 3.625,
      "learning_rate": 3.850869303564589e-06,
      "loss": 2.2718,
      "num_input_tokens_seen": 68605920,
      "step": 32840
    },
    {
      "epoch": 5.358104249938821,
      "grad_norm": 10.5,
      "learning_rate": 3.845635881975226e-06,
      "loss": 1.4601,
      "num_input_tokens_seen": 68615888,
      "step": 32845
    },
    {
      "epoch": 5.358919977159638,
      "grad_norm": 12.9375,
      "learning_rate": 3.840405722654647e-06,
      "loss": 3.2964,
      "num_input_tokens_seen": 68626112,
      "step": 32850
    },
    {
      "epoch": 5.359735704380455,
      "grad_norm": 5.8125,
      "learning_rate": 3.835178826409419e-06,
      "loss": 2.7309,
      "num_input_tokens_seen": 68636736,
      "step": 32855
    },
    {
      "epoch": 5.360551431601272,
      "grad_norm": 11.875,
      "learning_rate": 3.8299551940455895e-06,
      "loss": 2.652,
      "num_input_tokens_seen": 68647936,
      "step": 32860
    },
    {
      "epoch": 5.36136715882209,
      "grad_norm": 5.78125,
      "learning_rate": 3.824734826368703e-06,
      "loss": 1.6112,
      "num_input_tokens_seen": 68658336,
      "step": 32865
    },
    {
      "epoch": 5.362182886042907,
      "grad_norm": 4.96875,
      "learning_rate": 3.819517724183813e-06,
      "loss": 2.8202,
      "num_input_tokens_seen": 68669232,
      "step": 32870
    },
    {
      "epoch": 5.362998613263724,
      "grad_norm": 6.25,
      "learning_rate": 3.8143038882954648e-06,
      "loss": 1.8867,
      "num_input_tokens_seen": 68681264,
      "step": 32875
    },
    {
      "epoch": 5.3638143404845415,
      "grad_norm": 5.6875,
      "learning_rate": 3.8090933195076867e-06,
      "loss": 2.3306,
      "num_input_tokens_seen": 68690752,
      "step": 32880
    },
    {
      "epoch": 5.3646300677053596,
      "grad_norm": 7.1875,
      "learning_rate": 3.8038860186240198e-06,
      "loss": 2.892,
      "num_input_tokens_seen": 68701584,
      "step": 32885
    },
    {
      "epoch": 5.365445794926177,
      "grad_norm": 7.59375,
      "learning_rate": 3.7986819864475026e-06,
      "loss": 2.5,
      "num_input_tokens_seen": 68711952,
      "step": 32890
    },
    {
      "epoch": 5.366261522146994,
      "grad_norm": 9.125,
      "learning_rate": 3.793481223780651e-06,
      "loss": 2.4748,
      "num_input_tokens_seen": 68721808,
      "step": 32895
    },
    {
      "epoch": 5.367077249367812,
      "grad_norm": 8.375,
      "learning_rate": 3.788283731425496e-06,
      "loss": 2.6711,
      "num_input_tokens_seen": 68732656,
      "step": 32900
    },
    {
      "epoch": 5.367892976588629,
      "grad_norm": 8.8125,
      "learning_rate": 3.7830895101835488e-06,
      "loss": 2.3863,
      "num_input_tokens_seen": 68744704,
      "step": 32905
    },
    {
      "epoch": 5.368708703809446,
      "grad_norm": 10.9375,
      "learning_rate": 3.7778985608558274e-06,
      "loss": 2.5039,
      "num_input_tokens_seen": 68755264,
      "step": 32910
    },
    {
      "epoch": 5.369524431030263,
      "grad_norm": 9.9375,
      "learning_rate": 3.7727108842428443e-06,
      "loss": 1.7379,
      "num_input_tokens_seen": 68765280,
      "step": 32915
    },
    {
      "epoch": 5.370340158251081,
      "grad_norm": 9.125,
      "learning_rate": 3.7675264811446065e-06,
      "loss": 3.6476,
      "num_input_tokens_seen": 68775232,
      "step": 32920
    },
    {
      "epoch": 5.371155885471898,
      "grad_norm": 8.375,
      "learning_rate": 3.7623453523605994e-06,
      "loss": 2.0667,
      "num_input_tokens_seen": 68785760,
      "step": 32925
    },
    {
      "epoch": 5.3719716126927155,
      "grad_norm": 1.2734375,
      "learning_rate": 3.757167498689834e-06,
      "loss": 1.6156,
      "num_input_tokens_seen": 68795424,
      "step": 32930
    },
    {
      "epoch": 5.372787339913533,
      "grad_norm": 4.375,
      "learning_rate": 3.7519929209307914e-06,
      "loss": 1.691,
      "num_input_tokens_seen": 68805504,
      "step": 32935
    },
    {
      "epoch": 5.373603067134351,
      "grad_norm": 11.1875,
      "learning_rate": 3.746821619881463e-06,
      "loss": 2.1756,
      "num_input_tokens_seen": 68816416,
      "step": 32940
    },
    {
      "epoch": 5.374418794355168,
      "grad_norm": 7.78125,
      "learning_rate": 3.74165359633932e-06,
      "loss": 3.088,
      "num_input_tokens_seen": 68825600,
      "step": 32945
    },
    {
      "epoch": 5.375234521575985,
      "grad_norm": 4.25,
      "learning_rate": 3.736488851101341e-06,
      "loss": 2.5653,
      "num_input_tokens_seen": 68836592,
      "step": 32950
    },
    {
      "epoch": 5.376050248796802,
      "grad_norm": 14.1875,
      "learning_rate": 3.7313273849640035e-06,
      "loss": 3.7658,
      "num_input_tokens_seen": 68846784,
      "step": 32955
    },
    {
      "epoch": 5.37686597601762,
      "grad_norm": 6.125,
      "learning_rate": 3.7261691987232533e-06,
      "loss": 1.0972,
      "num_input_tokens_seen": 68859424,
      "step": 32960
    },
    {
      "epoch": 5.377681703238437,
      "grad_norm": 4.9375,
      "learning_rate": 3.7210142931745575e-06,
      "loss": 1.5178,
      "num_input_tokens_seen": 68869312,
      "step": 32965
    },
    {
      "epoch": 5.378497430459254,
      "grad_norm": 2.703125,
      "learning_rate": 3.7158626691128712e-06,
      "loss": 2.2884,
      "num_input_tokens_seen": 68880224,
      "step": 32970
    },
    {
      "epoch": 5.3793131576800715,
      "grad_norm": 6.0,
      "learning_rate": 3.710714327332629e-06,
      "loss": 2.0863,
      "num_input_tokens_seen": 68892112,
      "step": 32975
    },
    {
      "epoch": 5.3801288849008895,
      "grad_norm": 2.453125,
      "learning_rate": 3.7055692686277815e-06,
      "loss": 2.1197,
      "num_input_tokens_seen": 68904640,
      "step": 32980
    },
    {
      "epoch": 5.380944612121707,
      "grad_norm": 7.8125,
      "learning_rate": 3.70042749379175e-06,
      "loss": 2.9294,
      "num_input_tokens_seen": 68915632,
      "step": 32985
    },
    {
      "epoch": 5.381760339342524,
      "grad_norm": 7.59375,
      "learning_rate": 3.6952890036174693e-06,
      "loss": 2.8629,
      "num_input_tokens_seen": 68926288,
      "step": 32990
    },
    {
      "epoch": 5.382576066563341,
      "grad_norm": 6.21875,
      "learning_rate": 3.690153798897353e-06,
      "loss": 1.9659,
      "num_input_tokens_seen": 68937328,
      "step": 32995
    },
    {
      "epoch": 5.383391793784159,
      "grad_norm": 10.625,
      "learning_rate": 3.6850218804233225e-06,
      "loss": 3.0732,
      "num_input_tokens_seen": 68948064,
      "step": 33000
    },
    {
      "epoch": 5.383391793784159,
      "eval_loss": 2.550351858139038,
      "eval_runtime": 134.7768,
      "eval_samples_per_second": 20.219,
      "eval_steps_per_second": 10.113,
      "num_input_tokens_seen": 68948064,
      "step": 33000
    },
    {
      "epoch": 5.384207521004976,
      "grad_norm": 8.3125,
      "learning_rate": 3.679893248986779e-06,
      "loss": 2.0239,
      "num_input_tokens_seen": 68958288,
      "step": 33005
    },
    {
      "epoch": 5.385023248225793,
      "grad_norm": 5.1875,
      "learning_rate": 3.6747679053786147e-06,
      "loss": 2.2177,
      "num_input_tokens_seen": 68968576,
      "step": 33010
    },
    {
      "epoch": 5.38583897544661,
      "grad_norm": 6.09375,
      "learning_rate": 3.669645850389228e-06,
      "loss": 1.8081,
      "num_input_tokens_seen": 68978176,
      "step": 33015
    },
    {
      "epoch": 5.386654702667428,
      "grad_norm": 12.25,
      "learning_rate": 3.664527084808514e-06,
      "loss": 4.6706,
      "num_input_tokens_seen": 68988560,
      "step": 33020
    },
    {
      "epoch": 5.387470429888245,
      "grad_norm": 7.15625,
      "learning_rate": 3.6594116094258337e-06,
      "loss": 2.5089,
      "num_input_tokens_seen": 68998704,
      "step": 33025
    },
    {
      "epoch": 5.388286157109063,
      "grad_norm": 12.1875,
      "learning_rate": 3.6542994250300665e-06,
      "loss": 2.9712,
      "num_input_tokens_seen": 69009760,
      "step": 33030
    },
    {
      "epoch": 5.38910188432988,
      "grad_norm": 8.1875,
      "learning_rate": 3.6491905324095825e-06,
      "loss": 1.837,
      "num_input_tokens_seen": 69020944,
      "step": 33035
    },
    {
      "epoch": 5.389917611550698,
      "grad_norm": 5.6875,
      "learning_rate": 3.644084932352221e-06,
      "loss": 4.4004,
      "num_input_tokens_seen": 69031296,
      "step": 33040
    },
    {
      "epoch": 5.390733338771515,
      "grad_norm": 3.265625,
      "learning_rate": 3.6389826256453457e-06,
      "loss": 2.9087,
      "num_input_tokens_seen": 69042720,
      "step": 33045
    },
    {
      "epoch": 5.391549065992332,
      "grad_norm": 4.4375,
      "learning_rate": 3.633883613075781e-06,
      "loss": 2.2119,
      "num_input_tokens_seen": 69052672,
      "step": 33050
    },
    {
      "epoch": 5.392364793213149,
      "grad_norm": 11.6875,
      "learning_rate": 3.6287878954298693e-06,
      "loss": 1.8129,
      "num_input_tokens_seen": 69063568,
      "step": 33055
    },
    {
      "epoch": 5.393180520433967,
      "grad_norm": 3.6875,
      "learning_rate": 3.6236954734934354e-06,
      "loss": 1.7636,
      "num_input_tokens_seen": 69073744,
      "step": 33060
    },
    {
      "epoch": 5.393996247654784,
      "grad_norm": 4.9375,
      "learning_rate": 3.618606348051784e-06,
      "loss": 0.8543,
      "num_input_tokens_seen": 69082464,
      "step": 33065
    },
    {
      "epoch": 5.394811974875601,
      "grad_norm": 2.78125,
      "learning_rate": 3.6135205198897376e-06,
      "loss": 1.3975,
      "num_input_tokens_seen": 69093296,
      "step": 33070
    },
    {
      "epoch": 5.395627702096419,
      "grad_norm": 10.375,
      "learning_rate": 3.6084379897915854e-06,
      "loss": 3.1128,
      "num_input_tokens_seen": 69104224,
      "step": 33075
    },
    {
      "epoch": 5.3964434293172365,
      "grad_norm": 12.0625,
      "learning_rate": 3.6033587585411115e-06,
      "loss": 1.6582,
      "num_input_tokens_seen": 69113376,
      "step": 33080
    },
    {
      "epoch": 5.397259156538054,
      "grad_norm": 5.8125,
      "learning_rate": 3.5982828269216117e-06,
      "loss": 2.8507,
      "num_input_tokens_seen": 69124336,
      "step": 33085
    },
    {
      "epoch": 5.398074883758871,
      "grad_norm": 9.25,
      "learning_rate": 3.593210195715843e-06,
      "loss": 3.9888,
      "num_input_tokens_seen": 69135296,
      "step": 33090
    },
    {
      "epoch": 5.398890610979688,
      "grad_norm": 9.3125,
      "learning_rate": 3.5881408657060773e-06,
      "loss": 2.3739,
      "num_input_tokens_seen": 69146496,
      "step": 33095
    },
    {
      "epoch": 5.399706338200506,
      "grad_norm": 3.625,
      "learning_rate": 3.583074837674075e-06,
      "loss": 1.8028,
      "num_input_tokens_seen": 69155952,
      "step": 33100
    },
    {
      "epoch": 5.400522065421323,
      "grad_norm": 6.9375,
      "learning_rate": 3.578012112401069e-06,
      "loss": 1.9399,
      "num_input_tokens_seen": 69165904,
      "step": 33105
    },
    {
      "epoch": 5.40133779264214,
      "grad_norm": 8.625,
      "learning_rate": 3.5729526906677996e-06,
      "loss": 3.6244,
      "num_input_tokens_seen": 69176560,
      "step": 33110
    },
    {
      "epoch": 5.402153519862958,
      "grad_norm": 3.453125,
      "learning_rate": 3.5678965732545007e-06,
      "loss": 1.523,
      "num_input_tokens_seen": 69186592,
      "step": 33115
    },
    {
      "epoch": 5.402969247083775,
      "grad_norm": 10.75,
      "learning_rate": 3.562843760940876e-06,
      "loss": 2.187,
      "num_input_tokens_seen": 69195792,
      "step": 33120
    },
    {
      "epoch": 5.4037849743045925,
      "grad_norm": 5.71875,
      "learning_rate": 3.5577942545061473e-06,
      "loss": 1.3583,
      "num_input_tokens_seen": 69205552,
      "step": 33125
    },
    {
      "epoch": 5.40460070152541,
      "grad_norm": 9.5,
      "learning_rate": 3.5527480547289967e-06,
      "loss": 3.0487,
      "num_input_tokens_seen": 69216192,
      "step": 33130
    },
    {
      "epoch": 5.405416428746228,
      "grad_norm": 7.25,
      "learning_rate": 3.547705162387624e-06,
      "loss": 2.1176,
      "num_input_tokens_seen": 69226304,
      "step": 33135
    },
    {
      "epoch": 5.406232155967045,
      "grad_norm": 5.65625,
      "learning_rate": 3.542665578259699e-06,
      "loss": 3.3311,
      "num_input_tokens_seen": 69236736,
      "step": 33140
    },
    {
      "epoch": 5.407047883187862,
      "grad_norm": 4.3125,
      "learning_rate": 3.5376293031223945e-06,
      "loss": 2.5978,
      "num_input_tokens_seen": 69247776,
      "step": 33145
    },
    {
      "epoch": 5.407863610408679,
      "grad_norm": 5.53125,
      "learning_rate": 3.5325963377523614e-06,
      "loss": 1.7964,
      "num_input_tokens_seen": 69255792,
      "step": 33150
    },
    {
      "epoch": 5.408679337629497,
      "grad_norm": 8.5,
      "learning_rate": 3.5275666829257536e-06,
      "loss": 2.332,
      "num_input_tokens_seen": 69265888,
      "step": 33155
    },
    {
      "epoch": 5.409495064850314,
      "grad_norm": 7.84375,
      "learning_rate": 3.5225403394181955e-06,
      "loss": 2.1464,
      "num_input_tokens_seen": 69275840,
      "step": 33160
    },
    {
      "epoch": 5.410310792071131,
      "grad_norm": 1.0546875,
      "learning_rate": 3.517517308004828e-06,
      "loss": 2.1067,
      "num_input_tokens_seen": 69285936,
      "step": 33165
    },
    {
      "epoch": 5.411126519291948,
      "grad_norm": 8.8125,
      "learning_rate": 3.512497589460251e-06,
      "loss": 2.6663,
      "num_input_tokens_seen": 69296416,
      "step": 33170
    },
    {
      "epoch": 5.4119422465127665,
      "grad_norm": 9.6875,
      "learning_rate": 3.5074811845585727e-06,
      "loss": 2.9226,
      "num_input_tokens_seen": 69305872,
      "step": 33175
    },
    {
      "epoch": 5.412757973733584,
      "grad_norm": 8.1875,
      "learning_rate": 3.5024680940733937e-06,
      "loss": 3.3949,
      "num_input_tokens_seen": 69316816,
      "step": 33180
    },
    {
      "epoch": 5.413573700954401,
      "grad_norm": 8.875,
      "learning_rate": 3.4974583187777852e-06,
      "loss": 2.5279,
      "num_input_tokens_seen": 69326928,
      "step": 33185
    },
    {
      "epoch": 5.414389428175218,
      "grad_norm": 3.53125,
      "learning_rate": 3.4924518594443204e-06,
      "loss": 1.714,
      "num_input_tokens_seen": 69338448,
      "step": 33190
    },
    {
      "epoch": 5.415205155396036,
      "grad_norm": 4.59375,
      "learning_rate": 3.4874487168450682e-06,
      "loss": 1.8615,
      "num_input_tokens_seen": 69347312,
      "step": 33195
    },
    {
      "epoch": 5.416020882616853,
      "grad_norm": 7.78125,
      "learning_rate": 3.482448891751558e-06,
      "loss": 3.9775,
      "num_input_tokens_seen": 69357008,
      "step": 33200
    },
    {
      "epoch": 5.416020882616853,
      "eval_loss": 2.529320478439331,
      "eval_runtime": 134.8074,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 69357008,
      "step": 33200
    },
    {
      "epoch": 5.41683660983767,
      "grad_norm": 5.46875,
      "learning_rate": 3.477452384934843e-06,
      "loss": 1.408,
      "num_input_tokens_seen": 69367024,
      "step": 33205
    },
    {
      "epoch": 5.417652337058487,
      "grad_norm": 2.34375,
      "learning_rate": 3.472459197165434e-06,
      "loss": 1.645,
      "num_input_tokens_seen": 69376848,
      "step": 33210
    },
    {
      "epoch": 5.418468064279305,
      "grad_norm": 12.5625,
      "learning_rate": 3.4674693292133518e-06,
      "loss": 2.1229,
      "num_input_tokens_seen": 69386992,
      "step": 33215
    },
    {
      "epoch": 5.419283791500122,
      "grad_norm": 4.84375,
      "learning_rate": 3.4624827818480977e-06,
      "loss": 3.0874,
      "num_input_tokens_seen": 69395536,
      "step": 33220
    },
    {
      "epoch": 5.4200995187209395,
      "grad_norm": 9.5,
      "learning_rate": 3.4574995558386474e-06,
      "loss": 2.843,
      "num_input_tokens_seen": 69406144,
      "step": 33225
    },
    {
      "epoch": 5.420915245941757,
      "grad_norm": 10.5,
      "learning_rate": 3.452519651953487e-06,
      "loss": 2.2977,
      "num_input_tokens_seen": 69417696,
      "step": 33230
    },
    {
      "epoch": 5.421730973162575,
      "grad_norm": 11.4375,
      "learning_rate": 3.447543070960585e-06,
      "loss": 2.8949,
      "num_input_tokens_seen": 69427536,
      "step": 33235
    },
    {
      "epoch": 5.422546700383392,
      "grad_norm": 10.1875,
      "learning_rate": 3.4425698136273778e-06,
      "loss": 2.6095,
      "num_input_tokens_seen": 69437968,
      "step": 33240
    },
    {
      "epoch": 5.423362427604209,
      "grad_norm": 5.65625,
      "learning_rate": 3.437599880720821e-06,
      "loss": 1.8944,
      "num_input_tokens_seen": 69447936,
      "step": 33245
    },
    {
      "epoch": 5.424178154825027,
      "grad_norm": 8.375,
      "learning_rate": 3.4326332730073267e-06,
      "loss": 3.2479,
      "num_input_tokens_seen": 69458528,
      "step": 33250
    },
    {
      "epoch": 5.424993882045844,
      "grad_norm": 10.4375,
      "learning_rate": 3.427669991252813e-06,
      "loss": 2.2383,
      "num_input_tokens_seen": 69470064,
      "step": 33255
    },
    {
      "epoch": 5.425809609266661,
      "grad_norm": 5.53125,
      "learning_rate": 3.42271003622269e-06,
      "loss": 3.4743,
      "num_input_tokens_seen": 69480944,
      "step": 33260
    },
    {
      "epoch": 5.426625336487478,
      "grad_norm": 10.25,
      "learning_rate": 3.4177534086818286e-06,
      "loss": 2.3272,
      "num_input_tokens_seen": 69491008,
      "step": 33265
    },
    {
      "epoch": 5.4274410637082955,
      "grad_norm": 7.15625,
      "learning_rate": 3.412800109394612e-06,
      "loss": 2.808,
      "num_input_tokens_seen": 69501648,
      "step": 33270
    },
    {
      "epoch": 5.4282567909291135,
      "grad_norm": 5.59375,
      "learning_rate": 3.4078501391249044e-06,
      "loss": 2.886,
      "num_input_tokens_seen": 69512304,
      "step": 33275
    },
    {
      "epoch": 5.429072518149931,
      "grad_norm": 11.875,
      "learning_rate": 3.4029034986360453e-06,
      "loss": 2.659,
      "num_input_tokens_seen": 69522528,
      "step": 33280
    },
    {
      "epoch": 5.429888245370748,
      "grad_norm": 12.4375,
      "learning_rate": 3.397960188690877e-06,
      "loss": 2.4489,
      "num_input_tokens_seen": 69532704,
      "step": 33285
    },
    {
      "epoch": 5.430703972591566,
      "grad_norm": 6.15625,
      "learning_rate": 3.393020210051717e-06,
      "loss": 1.2056,
      "num_input_tokens_seen": 69542064,
      "step": 33290
    },
    {
      "epoch": 5.431519699812383,
      "grad_norm": 6.09375,
      "learning_rate": 3.3880835634803655e-06,
      "loss": 1.4297,
      "num_input_tokens_seen": 69552368,
      "step": 33295
    },
    {
      "epoch": 5.4323354270332,
      "grad_norm": 8.875,
      "learning_rate": 3.383150249738126e-06,
      "loss": 2.4705,
      "num_input_tokens_seen": 69562224,
      "step": 33300
    },
    {
      "epoch": 5.433151154254017,
      "grad_norm": 5.96875,
      "learning_rate": 3.3782202695857663e-06,
      "loss": 3.046,
      "num_input_tokens_seen": 69572112,
      "step": 33305
    },
    {
      "epoch": 5.433966881474835,
      "grad_norm": 8.4375,
      "learning_rate": 3.373293623783558e-06,
      "loss": 2.3007,
      "num_input_tokens_seen": 69584000,
      "step": 33310
    },
    {
      "epoch": 5.434782608695652,
      "grad_norm": 5.46875,
      "learning_rate": 3.368370313091257e-06,
      "loss": 1.2776,
      "num_input_tokens_seen": 69594048,
      "step": 33315
    },
    {
      "epoch": 5.4355983359164695,
      "grad_norm": 6.34375,
      "learning_rate": 3.363450338268087e-06,
      "loss": 1.4056,
      "num_input_tokens_seen": 69604560,
      "step": 33320
    },
    {
      "epoch": 5.436414063137287,
      "grad_norm": 8.6875,
      "learning_rate": 3.358533700072783e-06,
      "loss": 3.4181,
      "num_input_tokens_seen": 69614336,
      "step": 33325
    },
    {
      "epoch": 5.437229790358105,
      "grad_norm": 5.65625,
      "learning_rate": 3.3536203992635377e-06,
      "loss": 3.6387,
      "num_input_tokens_seen": 69625216,
      "step": 33330
    },
    {
      "epoch": 5.438045517578922,
      "grad_norm": 10.0625,
      "learning_rate": 3.348710436598057e-06,
      "loss": 2.6194,
      "num_input_tokens_seen": 69635632,
      "step": 33335
    },
    {
      "epoch": 5.438861244799739,
      "grad_norm": 6.21875,
      "learning_rate": 3.3438038128335155e-06,
      "loss": 2.3802,
      "num_input_tokens_seen": 69645936,
      "step": 33340
    },
    {
      "epoch": 5.439676972020556,
      "grad_norm": 14.625,
      "learning_rate": 3.338900528726571e-06,
      "loss": 3.2674,
      "num_input_tokens_seen": 69656928,
      "step": 33345
    },
    {
      "epoch": 5.440492699241374,
      "grad_norm": 4.8125,
      "learning_rate": 3.3340005850333812e-06,
      "loss": 2.105,
      "num_input_tokens_seen": 69668128,
      "step": 33350
    },
    {
      "epoch": 5.441308426462191,
      "grad_norm": 3.96875,
      "learning_rate": 3.329103982509568e-06,
      "loss": 3.1512,
      "num_input_tokens_seen": 69678864,
      "step": 33355
    },
    {
      "epoch": 5.442124153683008,
      "grad_norm": 8.5625,
      "learning_rate": 3.324210721910259e-06,
      "loss": 2.7581,
      "num_input_tokens_seen": 69689216,
      "step": 33360
    },
    {
      "epoch": 5.442939880903825,
      "grad_norm": 3.265625,
      "learning_rate": 3.319320803990053e-06,
      "loss": 2.2164,
      "num_input_tokens_seen": 69699936,
      "step": 33365
    },
    {
      "epoch": 5.443755608124643,
      "grad_norm": 4.71875,
      "learning_rate": 3.3144342295030274e-06,
      "loss": 1.7097,
      "num_input_tokens_seen": 69710640,
      "step": 33370
    },
    {
      "epoch": 5.444571335345461,
      "grad_norm": 6.90625,
      "learning_rate": 3.309550999202765e-06,
      "loss": 1.6203,
      "num_input_tokens_seen": 69721824,
      "step": 33375
    },
    {
      "epoch": 5.445387062566278,
      "grad_norm": 7.46875,
      "learning_rate": 3.3046711138423197e-06,
      "loss": 2.0463,
      "num_input_tokens_seen": 69730608,
      "step": 33380
    },
    {
      "epoch": 5.446202789787095,
      "grad_norm": 10.875,
      "learning_rate": 3.2997945741742255e-06,
      "loss": 2.4403,
      "num_input_tokens_seen": 69742128,
      "step": 33385
    },
    {
      "epoch": 5.447018517007913,
      "grad_norm": 8.4375,
      "learning_rate": 3.2949213809505082e-06,
      "loss": 2.7011,
      "num_input_tokens_seen": 69751952,
      "step": 33390
    },
    {
      "epoch": 5.44783424422873,
      "grad_norm": 9.25,
      "learning_rate": 3.2900515349226834e-06,
      "loss": 2.5441,
      "num_input_tokens_seen": 69762464,
      "step": 33395
    },
    {
      "epoch": 5.448649971449547,
      "grad_norm": 7.90625,
      "learning_rate": 3.285185036841731e-06,
      "loss": 2.9763,
      "num_input_tokens_seen": 69771824,
      "step": 33400
    },
    {
      "epoch": 5.448649971449547,
      "eval_loss": 2.5489401817321777,
      "eval_runtime": 134.7877,
      "eval_samples_per_second": 20.217,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 69771824,
      "step": 33400
    },
    {
      "epoch": 5.449465698670364,
      "grad_norm": 7.1875,
      "learning_rate": 3.2803218874581377e-06,
      "loss": 2.7255,
      "num_input_tokens_seen": 69781600,
      "step": 33405
    },
    {
      "epoch": 5.450281425891182,
      "grad_norm": 11.1875,
      "learning_rate": 3.2754620875218494e-06,
      "loss": 4.1228,
      "num_input_tokens_seen": 69791936,
      "step": 33410
    },
    {
      "epoch": 5.451097153111999,
      "grad_norm": 1.9609375,
      "learning_rate": 3.2706056377823146e-06,
      "loss": 1.7119,
      "num_input_tokens_seen": 69801552,
      "step": 33415
    },
    {
      "epoch": 5.4519128803328165,
      "grad_norm": 8.1875,
      "learning_rate": 3.2657525389884647e-06,
      "loss": 2.1158,
      "num_input_tokens_seen": 69812304,
      "step": 33420
    },
    {
      "epoch": 5.4527286075536345,
      "grad_norm": 5.625,
      "learning_rate": 3.260902791888698e-06,
      "loss": 2.3729,
      "num_input_tokens_seen": 69822688,
      "step": 33425
    },
    {
      "epoch": 5.453544334774452,
      "grad_norm": 6.21875,
      "learning_rate": 3.2560563972309166e-06,
      "loss": 2.4317,
      "num_input_tokens_seen": 69833120,
      "step": 33430
    },
    {
      "epoch": 5.454360061995269,
      "grad_norm": 0.10595703125,
      "learning_rate": 3.251213355762489e-06,
      "loss": 2.6556,
      "num_input_tokens_seen": 69844432,
      "step": 33435
    },
    {
      "epoch": 5.455175789216086,
      "grad_norm": 3.484375,
      "learning_rate": 3.2463736682302707e-06,
      "loss": 2.3437,
      "num_input_tokens_seen": 69854976,
      "step": 33440
    },
    {
      "epoch": 5.455991516436903,
      "grad_norm": 9.0625,
      "learning_rate": 3.2415373353806124e-06,
      "loss": 2.554,
      "num_input_tokens_seen": 69864944,
      "step": 33445
    },
    {
      "epoch": 5.456807243657721,
      "grad_norm": 4.9375,
      "learning_rate": 3.236704357959322e-06,
      "loss": 1.8492,
      "num_input_tokens_seen": 69874608,
      "step": 33450
    },
    {
      "epoch": 5.457622970878538,
      "grad_norm": 5.125,
      "learning_rate": 3.2318747367117154e-06,
      "loss": 4.3637,
      "num_input_tokens_seen": 69884512,
      "step": 33455
    },
    {
      "epoch": 5.458438698099355,
      "grad_norm": 8.6875,
      "learning_rate": 3.227048472382585e-06,
      "loss": 2.8495,
      "num_input_tokens_seen": 69894128,
      "step": 33460
    },
    {
      "epoch": 5.459254425320173,
      "grad_norm": 10.25,
      "learning_rate": 3.2222255657161915e-06,
      "loss": 2.8073,
      "num_input_tokens_seen": 69903952,
      "step": 33465
    },
    {
      "epoch": 5.4600701525409905,
      "grad_norm": 6.25,
      "learning_rate": 3.2174060174562924e-06,
      "loss": 1.2857,
      "num_input_tokens_seen": 69915072,
      "step": 33470
    },
    {
      "epoch": 5.460885879761808,
      "grad_norm": 7.0625,
      "learning_rate": 3.2125898283461298e-06,
      "loss": 2.9893,
      "num_input_tokens_seen": 69926032,
      "step": 33475
    },
    {
      "epoch": 5.461701606982625,
      "grad_norm": 20.0,
      "learning_rate": 3.207776999128406e-06,
      "loss": 2.7275,
      "num_input_tokens_seen": 69936800,
      "step": 33480
    },
    {
      "epoch": 5.462517334203443,
      "grad_norm": 11.1875,
      "learning_rate": 3.202967530545331e-06,
      "loss": 2.1037,
      "num_input_tokens_seen": 69948208,
      "step": 33485
    },
    {
      "epoch": 5.46333306142426,
      "grad_norm": 6.5625,
      "learning_rate": 3.1981614233385778e-06,
      "loss": 2.4547,
      "num_input_tokens_seen": 69959552,
      "step": 33490
    },
    {
      "epoch": 5.464148788645077,
      "grad_norm": 5.1875,
      "learning_rate": 3.1933586782493115e-06,
      "loss": 1.4457,
      "num_input_tokens_seen": 69968208,
      "step": 33495
    },
    {
      "epoch": 5.464964515865894,
      "grad_norm": 8.3125,
      "learning_rate": 3.188559296018184e-06,
      "loss": 3.3946,
      "num_input_tokens_seen": 69978928,
      "step": 33500
    },
    {
      "epoch": 5.465780243086712,
      "grad_norm": 3.40625,
      "learning_rate": 3.1837632773853098e-06,
      "loss": 1.4608,
      "num_input_tokens_seen": 69988816,
      "step": 33505
    },
    {
      "epoch": 5.466595970307529,
      "grad_norm": 9.5625,
      "learning_rate": 3.178970623090294e-06,
      "loss": 2.8733,
      "num_input_tokens_seen": 69998800,
      "step": 33510
    },
    {
      "epoch": 5.467411697528346,
      "grad_norm": 4.46875,
      "learning_rate": 3.174181333872234e-06,
      "loss": 2.985,
      "num_input_tokens_seen": 70008688,
      "step": 33515
    },
    {
      "epoch": 5.468227424749164,
      "grad_norm": 0.19140625,
      "learning_rate": 3.169395410469686e-06,
      "loss": 1.2965,
      "num_input_tokens_seen": 70019232,
      "step": 33520
    },
    {
      "epoch": 5.469043151969982,
      "grad_norm": 6.1875,
      "learning_rate": 3.164612853620713e-06,
      "loss": 2.6284,
      "num_input_tokens_seen": 70030400,
      "step": 33525
    },
    {
      "epoch": 5.469858879190799,
      "grad_norm": 10.125,
      "learning_rate": 3.1598336640628333e-06,
      "loss": 2.1117,
      "num_input_tokens_seen": 70041184,
      "step": 33530
    },
    {
      "epoch": 5.470674606411616,
      "grad_norm": 13.1875,
      "learning_rate": 3.155057842533063e-06,
      "loss": 2.6915,
      "num_input_tokens_seen": 70052048,
      "step": 33535
    },
    {
      "epoch": 5.471490333632433,
      "grad_norm": 0.2734375,
      "learning_rate": 3.1502853897678984e-06,
      "loss": 2.4461,
      "num_input_tokens_seen": 70062160,
      "step": 33540
    },
    {
      "epoch": 5.472306060853251,
      "grad_norm": 11.5,
      "learning_rate": 3.1455163065033017e-06,
      "loss": 2.8573,
      "num_input_tokens_seen": 70071664,
      "step": 33545
    },
    {
      "epoch": 5.473121788074068,
      "grad_norm": 4.9375,
      "learning_rate": 3.140750593474734e-06,
      "loss": 1.3993,
      "num_input_tokens_seen": 70081184,
      "step": 33550
    },
    {
      "epoch": 5.473937515294885,
      "grad_norm": 6.125,
      "learning_rate": 3.1359882514171294e-06,
      "loss": 2.0756,
      "num_input_tokens_seen": 70092528,
      "step": 33555
    },
    {
      "epoch": 5.474753242515702,
      "grad_norm": 11.75,
      "learning_rate": 3.1312292810648903e-06,
      "loss": 2.6495,
      "num_input_tokens_seen": 70104464,
      "step": 33560
    },
    {
      "epoch": 5.47556896973652,
      "grad_norm": 5.6875,
      "learning_rate": 3.1264736831519204e-06,
      "loss": 1.9641,
      "num_input_tokens_seen": 70115648,
      "step": 33565
    },
    {
      "epoch": 5.4763846969573375,
      "grad_norm": 1.390625,
      "learning_rate": 3.1217214584115863e-06,
      "loss": 2.3778,
      "num_input_tokens_seen": 70126800,
      "step": 33570
    },
    {
      "epoch": 5.477200424178155,
      "grad_norm": 9.875,
      "learning_rate": 3.116972607576746e-06,
      "loss": 2.2255,
      "num_input_tokens_seen": 70137504,
      "step": 33575
    },
    {
      "epoch": 5.478016151398972,
      "grad_norm": 12.5,
      "learning_rate": 3.1122271313797303e-06,
      "loss": 3.8461,
      "num_input_tokens_seen": 70148528,
      "step": 33580
    },
    {
      "epoch": 5.47883187861979,
      "grad_norm": 11.375,
      "learning_rate": 3.107485030552343e-06,
      "loss": 3.4812,
      "num_input_tokens_seen": 70159392,
      "step": 33585
    },
    {
      "epoch": 5.479647605840607,
      "grad_norm": 6.75,
      "learning_rate": 3.1027463058258848e-06,
      "loss": 2.4764,
      "num_input_tokens_seen": 70169040,
      "step": 33590
    },
    {
      "epoch": 5.480463333061424,
      "grad_norm": 5.59375,
      "learning_rate": 3.0980109579311273e-06,
      "loss": 1.9224,
      "num_input_tokens_seen": 70178832,
      "step": 33595
    },
    {
      "epoch": 5.481279060282241,
      "grad_norm": 6.90625,
      "learning_rate": 3.093278987598314e-06,
      "loss": 4.1438,
      "num_input_tokens_seen": 70189824,
      "step": 33600
    },
    {
      "epoch": 5.481279060282241,
      "eval_loss": 2.542243003845215,
      "eval_runtime": 134.8092,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 70189824,
      "step": 33600
    },
    {
      "epoch": 5.482094787503059,
      "grad_norm": 8.3125,
      "learning_rate": 3.0885503955571826e-06,
      "loss": 4.2614,
      "num_input_tokens_seen": 70200560,
      "step": 33605
    },
    {
      "epoch": 5.482910514723876,
      "grad_norm": 5.125,
      "learning_rate": 3.0838251825369313e-06,
      "loss": 2.9884,
      "num_input_tokens_seen": 70211088,
      "step": 33610
    },
    {
      "epoch": 5.4837262419446935,
      "grad_norm": 4.53125,
      "learning_rate": 3.0791033492662517e-06,
      "loss": 2.5172,
      "num_input_tokens_seen": 70222144,
      "step": 33615
    },
    {
      "epoch": 5.484541969165511,
      "grad_norm": 0.2578125,
      "learning_rate": 3.0743848964733203e-06,
      "loss": 1.2972,
      "num_input_tokens_seen": 70233568,
      "step": 33620
    },
    {
      "epoch": 5.485357696386329,
      "grad_norm": 6.8125,
      "learning_rate": 3.0696698248857625e-06,
      "loss": 1.912,
      "num_input_tokens_seen": 70244336,
      "step": 33625
    },
    {
      "epoch": 5.486173423607146,
      "grad_norm": 12.8125,
      "learning_rate": 3.0649581352307192e-06,
      "loss": 2.5585,
      "num_input_tokens_seen": 70255184,
      "step": 33630
    },
    {
      "epoch": 5.486989150827963,
      "grad_norm": 1.84375,
      "learning_rate": 3.060249828234776e-06,
      "loss": 1.825,
      "num_input_tokens_seen": 70265920,
      "step": 33635
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 6.28125,
      "learning_rate": 3.055544904624025e-06,
      "loss": 2.2059,
      "num_input_tokens_seen": 70276672,
      "step": 33640
    },
    {
      "epoch": 5.488620605269598,
      "grad_norm": 10.9375,
      "learning_rate": 3.050843365124026e-06,
      "loss": 2.0991,
      "num_input_tokens_seen": 70287824,
      "step": 33645
    },
    {
      "epoch": 5.489436332490415,
      "grad_norm": 8.625,
      "learning_rate": 3.0461452104598083e-06,
      "loss": 3.462,
      "num_input_tokens_seen": 70298064,
      "step": 33650
    },
    {
      "epoch": 5.490252059711232,
      "grad_norm": 4.59375,
      "learning_rate": 3.0414504413558836e-06,
      "loss": 2.6807,
      "num_input_tokens_seen": 70308208,
      "step": 33655
    },
    {
      "epoch": 5.49106778693205,
      "grad_norm": 8.0,
      "learning_rate": 3.0367590585362564e-06,
      "loss": 3.6352,
      "num_input_tokens_seen": 70317648,
      "step": 33660
    },
    {
      "epoch": 5.4918835141528675,
      "grad_norm": 8.375,
      "learning_rate": 3.0320710627243813e-06,
      "loss": 2.053,
      "num_input_tokens_seen": 70328688,
      "step": 33665
    },
    {
      "epoch": 5.492699241373685,
      "grad_norm": 5.34375,
      "learning_rate": 3.027386454643222e-06,
      "loss": 2.041,
      "num_input_tokens_seen": 70339792,
      "step": 33670
    },
    {
      "epoch": 5.493514968594502,
      "grad_norm": 5.125,
      "learning_rate": 3.0227052350151914e-06,
      "loss": 2.0909,
      "num_input_tokens_seen": 70350032,
      "step": 33675
    },
    {
      "epoch": 5.49433069581532,
      "grad_norm": 3.34375,
      "learning_rate": 3.0180274045621957e-06,
      "loss": 2.7052,
      "num_input_tokens_seen": 70360688,
      "step": 33680
    },
    {
      "epoch": 5.495146423036137,
      "grad_norm": 8.3125,
      "learning_rate": 3.013352964005625e-06,
      "loss": 2.6262,
      "num_input_tokens_seen": 70370272,
      "step": 33685
    },
    {
      "epoch": 5.495962150256954,
      "grad_norm": 7.84375,
      "learning_rate": 3.0086819140663218e-06,
      "loss": 1.4663,
      "num_input_tokens_seen": 70380080,
      "step": 33690
    },
    {
      "epoch": 5.496777877477771,
      "grad_norm": 17.0,
      "learning_rate": 3.0040142554646265e-06,
      "loss": 2.9524,
      "num_input_tokens_seen": 70390080,
      "step": 33695
    },
    {
      "epoch": 5.497593604698589,
      "grad_norm": 5.59375,
      "learning_rate": 2.999349988920361e-06,
      "loss": 1.8148,
      "num_input_tokens_seen": 70400720,
      "step": 33700
    },
    {
      "epoch": 5.498409331919406,
      "grad_norm": 11.1875,
      "learning_rate": 2.994689115152796e-06,
      "loss": 3.4553,
      "num_input_tokens_seen": 70410736,
      "step": 33705
    },
    {
      "epoch": 5.499225059140223,
      "grad_norm": 4.28125,
      "learning_rate": 2.9900316348807105e-06,
      "loss": 1.3946,
      "num_input_tokens_seen": 70420976,
      "step": 33710
    },
    {
      "epoch": 5.5000407863610405,
      "grad_norm": 16.5,
      "learning_rate": 2.985377548822338e-06,
      "loss": 2.8065,
      "num_input_tokens_seen": 70431280,
      "step": 33715
    },
    {
      "epoch": 5.500856513581859,
      "grad_norm": 8.25,
      "learning_rate": 2.980726857695404e-06,
      "loss": 2.0615,
      "num_input_tokens_seen": 70441936,
      "step": 33720
    },
    {
      "epoch": 5.501672240802676,
      "grad_norm": 12.0,
      "learning_rate": 2.9760795622171017e-06,
      "loss": 2.7577,
      "num_input_tokens_seen": 70450240,
      "step": 33725
    },
    {
      "epoch": 5.502487968023493,
      "grad_norm": 4.96875,
      "learning_rate": 2.971435663104094e-06,
      "loss": 1.4691,
      "num_input_tokens_seen": 70458880,
      "step": 33730
    },
    {
      "epoch": 5.50330369524431,
      "grad_norm": 7.84375,
      "learning_rate": 2.9667951610725385e-06,
      "loss": 2.7232,
      "num_input_tokens_seen": 70468288,
      "step": 33735
    },
    {
      "epoch": 5.504119422465128,
      "grad_norm": 12.4375,
      "learning_rate": 2.9621580568380575e-06,
      "loss": 2.0621,
      "num_input_tokens_seen": 70478384,
      "step": 33740
    },
    {
      "epoch": 5.504935149685945,
      "grad_norm": 6.5,
      "learning_rate": 2.9575243511157453e-06,
      "loss": 1.9384,
      "num_input_tokens_seen": 70489344,
      "step": 33745
    },
    {
      "epoch": 5.505750876906762,
      "grad_norm": 13.625,
      "learning_rate": 2.952894044620186e-06,
      "loss": 1.8743,
      "num_input_tokens_seen": 70500352,
      "step": 33750
    },
    {
      "epoch": 5.506566604127579,
      "grad_norm": 3.234375,
      "learning_rate": 2.948267138065419e-06,
      "loss": 2.2559,
      "num_input_tokens_seen": 70510560,
      "step": 33755
    },
    {
      "epoch": 5.507382331348397,
      "grad_norm": 0.69140625,
      "learning_rate": 2.943643632164983e-06,
      "loss": 1.3274,
      "num_input_tokens_seen": 70519920,
      "step": 33760
    },
    {
      "epoch": 5.5081980585692145,
      "grad_norm": 3.984375,
      "learning_rate": 2.939023527631879e-06,
      "loss": 3.6455,
      "num_input_tokens_seen": 70530768,
      "step": 33765
    },
    {
      "epoch": 5.509013785790032,
      "grad_norm": 9.0,
      "learning_rate": 2.934406825178576e-06,
      "loss": 2.5683,
      "num_input_tokens_seen": 70541920,
      "step": 33770
    },
    {
      "epoch": 5.50982951301085,
      "grad_norm": 6.875,
      "learning_rate": 2.9297935255170357e-06,
      "loss": 2.6147,
      "num_input_tokens_seen": 70551472,
      "step": 33775
    },
    {
      "epoch": 5.510645240231667,
      "grad_norm": 8.3125,
      "learning_rate": 2.925183629358691e-06,
      "loss": 2.1595,
      "num_input_tokens_seen": 70562112,
      "step": 33780
    },
    {
      "epoch": 5.511460967452484,
      "grad_norm": 7.65625,
      "learning_rate": 2.9205771374144346e-06,
      "loss": 1.6327,
      "num_input_tokens_seen": 70571520,
      "step": 33785
    },
    {
      "epoch": 5.512276694673301,
      "grad_norm": 7.375,
      "learning_rate": 2.915974050394657e-06,
      "loss": 2.1792,
      "num_input_tokens_seen": 70581776,
      "step": 33790
    },
    {
      "epoch": 5.513092421894118,
      "grad_norm": 11.0,
      "learning_rate": 2.9113743690092067e-06,
      "loss": 2.4173,
      "num_input_tokens_seen": 70592048,
      "step": 33795
    },
    {
      "epoch": 5.513908149114936,
      "grad_norm": 3.265625,
      "learning_rate": 2.906778093967402e-06,
      "loss": 1.7663,
      "num_input_tokens_seen": 70602704,
      "step": 33800
    },
    {
      "epoch": 5.513908149114936,
      "eval_loss": 2.5325217247009277,
      "eval_runtime": 134.7687,
      "eval_samples_per_second": 20.22,
      "eval_steps_per_second": 10.114,
      "num_input_tokens_seen": 70602704,
      "step": 33800
    },
    {
      "epoch": 5.514723876335753,
      "grad_norm": 4.21875,
      "learning_rate": 2.9021852259780656e-06,
      "loss": 1.7847,
      "num_input_tokens_seen": 70614112,
      "step": 33805
    },
    {
      "epoch": 5.5155396035565705,
      "grad_norm": 6.875,
      "learning_rate": 2.8975957657494583e-06,
      "loss": 2.2548,
      "num_input_tokens_seen": 70624400,
      "step": 33810
    },
    {
      "epoch": 5.5163553307773885,
      "grad_norm": 9.375,
      "learning_rate": 2.8930097139893417e-06,
      "loss": 3.6829,
      "num_input_tokens_seen": 70636048,
      "step": 33815
    },
    {
      "epoch": 5.517171057998206,
      "grad_norm": 15.25,
      "learning_rate": 2.888427071404945e-06,
      "loss": 3.1673,
      "num_input_tokens_seen": 70647168,
      "step": 33820
    },
    {
      "epoch": 5.517986785219023,
      "grad_norm": 4.34375,
      "learning_rate": 2.8838478387029606e-06,
      "loss": 2.8413,
      "num_input_tokens_seen": 70658976,
      "step": 33825
    },
    {
      "epoch": 5.51880251243984,
      "grad_norm": 9.75,
      "learning_rate": 2.8792720165895737e-06,
      "loss": 3.1115,
      "num_input_tokens_seen": 70670608,
      "step": 33830
    },
    {
      "epoch": 5.519618239660657,
      "grad_norm": 2.84375,
      "learning_rate": 2.874699605770423e-06,
      "loss": 2.2639,
      "num_input_tokens_seen": 70681008,
      "step": 33835
    },
    {
      "epoch": 5.520433966881475,
      "grad_norm": 8.9375,
      "learning_rate": 2.8701306069506383e-06,
      "loss": 1.9286,
      "num_input_tokens_seen": 70692576,
      "step": 33840
    },
    {
      "epoch": 5.521249694102292,
      "grad_norm": 8.4375,
      "learning_rate": 2.8655650208348178e-06,
      "loss": 1.8624,
      "num_input_tokens_seen": 70703616,
      "step": 33845
    },
    {
      "epoch": 5.522065421323109,
      "grad_norm": 2.578125,
      "learning_rate": 2.8610028481270257e-06,
      "loss": 3.7743,
      "num_input_tokens_seen": 70714848,
      "step": 33850
    },
    {
      "epoch": 5.522881148543927,
      "grad_norm": 4.3125,
      "learning_rate": 2.856444089530813e-06,
      "loss": 1.7127,
      "num_input_tokens_seen": 70726000,
      "step": 33855
    },
    {
      "epoch": 5.523696875764744,
      "grad_norm": 6.875,
      "learning_rate": 2.8518887457491955e-06,
      "loss": 2.2915,
      "num_input_tokens_seen": 70736960,
      "step": 33860
    },
    {
      "epoch": 5.524512602985562,
      "grad_norm": 8.8125,
      "learning_rate": 2.8473368174846666e-06,
      "loss": 2.4495,
      "num_input_tokens_seen": 70746688,
      "step": 33865
    },
    {
      "epoch": 5.525328330206379,
      "grad_norm": 1.578125,
      "learning_rate": 2.842788305439184e-06,
      "loss": 2.8257,
      "num_input_tokens_seen": 70758304,
      "step": 33870
    },
    {
      "epoch": 5.526144057427197,
      "grad_norm": 10.6875,
      "learning_rate": 2.8382432103141925e-06,
      "loss": 3.0169,
      "num_input_tokens_seen": 70768304,
      "step": 33875
    },
    {
      "epoch": 5.526959784648014,
      "grad_norm": 5.1875,
      "learning_rate": 2.833701532810598e-06,
      "loss": 2.1918,
      "num_input_tokens_seen": 70778944,
      "step": 33880
    },
    {
      "epoch": 5.527775511868831,
      "grad_norm": 5.0,
      "learning_rate": 2.8291632736287877e-06,
      "loss": 1.6675,
      "num_input_tokens_seen": 70790960,
      "step": 33885
    },
    {
      "epoch": 5.528591239089648,
      "grad_norm": 8.3125,
      "learning_rate": 2.824628433468615e-06,
      "loss": 2.6015,
      "num_input_tokens_seen": 70801744,
      "step": 33890
    },
    {
      "epoch": 5.529406966310466,
      "grad_norm": 9.8125,
      "learning_rate": 2.8200970130294073e-06,
      "loss": 3.1672,
      "num_input_tokens_seen": 70811344,
      "step": 33895
    },
    {
      "epoch": 5.530222693531283,
      "grad_norm": 5.375,
      "learning_rate": 2.8155690130099775e-06,
      "loss": 1.8632,
      "num_input_tokens_seen": 70820608,
      "step": 33900
    },
    {
      "epoch": 5.5310384207521,
      "grad_norm": 10.625,
      "learning_rate": 2.8110444341085895e-06,
      "loss": 3.0267,
      "num_input_tokens_seen": 70832208,
      "step": 33905
    },
    {
      "epoch": 5.5318541479729175,
      "grad_norm": 6.5,
      "learning_rate": 2.806523277022996e-06,
      "loss": 1.6474,
      "num_input_tokens_seen": 70842064,
      "step": 33910
    },
    {
      "epoch": 5.5326698751937355,
      "grad_norm": 7.40625,
      "learning_rate": 2.802005542450409e-06,
      "loss": 2.4329,
      "num_input_tokens_seen": 70851280,
      "step": 33915
    },
    {
      "epoch": 5.533485602414553,
      "grad_norm": 10.875,
      "learning_rate": 2.797491231087526e-06,
      "loss": 3.2764,
      "num_input_tokens_seen": 70860832,
      "step": 33920
    },
    {
      "epoch": 5.53430132963537,
      "grad_norm": 5.53125,
      "learning_rate": 2.7929803436305137e-06,
      "loss": 2.0809,
      "num_input_tokens_seen": 70871840,
      "step": 33925
    },
    {
      "epoch": 5.535117056856187,
      "grad_norm": 8.125,
      "learning_rate": 2.788472880774998e-06,
      "loss": 2.6532,
      "num_input_tokens_seen": 70881504,
      "step": 33930
    },
    {
      "epoch": 5.535932784077005,
      "grad_norm": 7.75,
      "learning_rate": 2.7839688432160977e-06,
      "loss": 2.5554,
      "num_input_tokens_seen": 70892464,
      "step": 33935
    },
    {
      "epoch": 5.536748511297822,
      "grad_norm": 10.625,
      "learning_rate": 2.779468231648383e-06,
      "loss": 3.6217,
      "num_input_tokens_seen": 70903888,
      "step": 33940
    },
    {
      "epoch": 5.537564238518639,
      "grad_norm": 6.59375,
      "learning_rate": 2.774971046765906e-06,
      "loss": 1.8997,
      "num_input_tokens_seen": 70914432,
      "step": 33945
    },
    {
      "epoch": 5.538379965739456,
      "grad_norm": 5.0,
      "learning_rate": 2.770477289262194e-06,
      "loss": 2.8753,
      "num_input_tokens_seen": 70926496,
      "step": 33950
    },
    {
      "epoch": 5.539195692960274,
      "grad_norm": 7.34375,
      "learning_rate": 2.765986959830233e-06,
      "loss": 3.0174,
      "num_input_tokens_seen": 70935280,
      "step": 33955
    },
    {
      "epoch": 5.5400114201810915,
      "grad_norm": 2.0,
      "learning_rate": 2.761500059162492e-06,
      "loss": 1.9065,
      "num_input_tokens_seen": 70943968,
      "step": 33960
    },
    {
      "epoch": 5.540827147401909,
      "grad_norm": 4.34375,
      "learning_rate": 2.757016587950914e-06,
      "loss": 2.7882,
      "num_input_tokens_seen": 70955440,
      "step": 33965
    },
    {
      "epoch": 5.541642874622726,
      "grad_norm": 12.0,
      "learning_rate": 2.752536546886897e-06,
      "loss": 2.7829,
      "num_input_tokens_seen": 70965936,
      "step": 33970
    },
    {
      "epoch": 5.542458601843544,
      "grad_norm": 3.453125,
      "learning_rate": 2.7480599366613234e-06,
      "loss": 2.091,
      "num_input_tokens_seen": 70976880,
      "step": 33975
    },
    {
      "epoch": 5.543274329064361,
      "grad_norm": 6.375,
      "learning_rate": 2.7435867579645473e-06,
      "loss": 1.7876,
      "num_input_tokens_seen": 70988896,
      "step": 33980
    },
    {
      "epoch": 5.544090056285178,
      "grad_norm": 2.671875,
      "learning_rate": 2.739117011486378e-06,
      "loss": 2.527,
      "num_input_tokens_seen": 70999376,
      "step": 33985
    },
    {
      "epoch": 5.544905783505996,
      "grad_norm": 5.125,
      "learning_rate": 2.7346506979161216e-06,
      "loss": 1.8347,
      "num_input_tokens_seen": 71011360,
      "step": 33990
    },
    {
      "epoch": 5.545721510726813,
      "grad_norm": 8.25,
      "learning_rate": 2.7301878179425227e-06,
      "loss": 1.6923,
      "num_input_tokens_seen": 71022528,
      "step": 33995
    },
    {
      "epoch": 5.54653723794763,
      "grad_norm": 4.09375,
      "learning_rate": 2.7257283722538244e-06,
      "loss": 2.9164,
      "num_input_tokens_seen": 71032768,
      "step": 34000
    },
    {
      "epoch": 5.54653723794763,
      "eval_loss": 2.547989845275879,
      "eval_runtime": 134.6502,
      "eval_samples_per_second": 20.238,
      "eval_steps_per_second": 10.123,
      "num_input_tokens_seen": 71032768,
      "step": 34000
    },
    {
      "epoch": 5.5473529651684474,
      "grad_norm": 14.625,
      "learning_rate": 2.7212723615377326e-06,
      "loss": 2.8876,
      "num_input_tokens_seen": 71043328,
      "step": 34005
    },
    {
      "epoch": 5.548168692389265,
      "grad_norm": 6.625,
      "learning_rate": 2.7168197864814145e-06,
      "loss": 3.3343,
      "num_input_tokens_seen": 71054096,
      "step": 34010
    },
    {
      "epoch": 5.548984419610083,
      "grad_norm": 6.5625,
      "learning_rate": 2.712370647771509e-06,
      "loss": 2.4487,
      "num_input_tokens_seen": 71064384,
      "step": 34015
    },
    {
      "epoch": 5.5498001468309,
      "grad_norm": 6.8125,
      "learning_rate": 2.707924946094137e-06,
      "loss": 2.6793,
      "num_input_tokens_seen": 71074272,
      "step": 34020
    },
    {
      "epoch": 5.550615874051717,
      "grad_norm": 14.125,
      "learning_rate": 2.7034826821348723e-06,
      "loss": 2.8552,
      "num_input_tokens_seen": 71083584,
      "step": 34025
    },
    {
      "epoch": 5.551431601272535,
      "grad_norm": 5.875,
      "learning_rate": 2.6990438565787786e-06,
      "loss": 1.2556,
      "num_input_tokens_seen": 71092560,
      "step": 34030
    },
    {
      "epoch": 5.552247328493352,
      "grad_norm": 9.625,
      "learning_rate": 2.6946084701103714e-06,
      "loss": 2.0561,
      "num_input_tokens_seen": 71103216,
      "step": 34035
    },
    {
      "epoch": 5.553063055714169,
      "grad_norm": 3.1875,
      "learning_rate": 2.6901765234136428e-06,
      "loss": 3.5946,
      "num_input_tokens_seen": 71113568,
      "step": 34040
    },
    {
      "epoch": 5.553878782934986,
      "grad_norm": 6.5625,
      "learning_rate": 2.685748017172063e-06,
      "loss": 1.2238,
      "num_input_tokens_seen": 71123872,
      "step": 34045
    },
    {
      "epoch": 5.554694510155803,
      "grad_norm": 9.375,
      "learning_rate": 2.681322952068549e-06,
      "loss": 3.0264,
      "num_input_tokens_seen": 71133792,
      "step": 34050
    },
    {
      "epoch": 5.555510237376621,
      "grad_norm": 9.4375,
      "learning_rate": 2.6769013287855137e-06,
      "loss": 2.726,
      "num_input_tokens_seen": 71143472,
      "step": 34055
    },
    {
      "epoch": 5.5563259645974385,
      "grad_norm": 12.0625,
      "learning_rate": 2.6724831480048286e-06,
      "loss": 2.5001,
      "num_input_tokens_seen": 71154528,
      "step": 34060
    },
    {
      "epoch": 5.557141691818256,
      "grad_norm": 11.0625,
      "learning_rate": 2.66806841040782e-06,
      "loss": 3.4923,
      "num_input_tokens_seen": 71166400,
      "step": 34065
    },
    {
      "epoch": 5.557957419039074,
      "grad_norm": 5.0625,
      "learning_rate": 2.6636571166753083e-06,
      "loss": 2.7179,
      "num_input_tokens_seen": 71178000,
      "step": 34070
    },
    {
      "epoch": 5.558773146259891,
      "grad_norm": 5.53125,
      "learning_rate": 2.6592492674875598e-06,
      "loss": 2.2232,
      "num_input_tokens_seen": 71187856,
      "step": 34075
    },
    {
      "epoch": 5.559588873480708,
      "grad_norm": 2.765625,
      "learning_rate": 2.6548448635243305e-06,
      "loss": 1.4215,
      "num_input_tokens_seen": 71198256,
      "step": 34080
    },
    {
      "epoch": 5.560404600701525,
      "grad_norm": 9.125,
      "learning_rate": 2.650443905464828e-06,
      "loss": 2.8933,
      "num_input_tokens_seen": 71208928,
      "step": 34085
    },
    {
      "epoch": 5.561220327922343,
      "grad_norm": 5.6875,
      "learning_rate": 2.646046393987739e-06,
      "loss": 1.9546,
      "num_input_tokens_seen": 71219168,
      "step": 34090
    },
    {
      "epoch": 5.56203605514316,
      "grad_norm": 10.3125,
      "learning_rate": 2.64165232977121e-06,
      "loss": 2.9019,
      "num_input_tokens_seen": 71229696,
      "step": 34095
    },
    {
      "epoch": 5.562851782363977,
      "grad_norm": 3.25,
      "learning_rate": 2.6372617134928695e-06,
      "loss": 1.7256,
      "num_input_tokens_seen": 71240736,
      "step": 34100
    },
    {
      "epoch": 5.5636675095847945,
      "grad_norm": 3.671875,
      "learning_rate": 2.6328745458297943e-06,
      "loss": 1.7602,
      "num_input_tokens_seen": 71249376,
      "step": 34105
    },
    {
      "epoch": 5.5644832368056125,
      "grad_norm": 14.25,
      "learning_rate": 2.6284908274585546e-06,
      "loss": 2.2534,
      "num_input_tokens_seen": 71258848,
      "step": 34110
    },
    {
      "epoch": 5.56529896402643,
      "grad_norm": 2.671875,
      "learning_rate": 2.6241105590551595e-06,
      "loss": 2.0429,
      "num_input_tokens_seen": 71267808,
      "step": 34115
    },
    {
      "epoch": 5.566114691247247,
      "grad_norm": 9.1875,
      "learning_rate": 2.6197337412951105e-06,
      "loss": 1.5822,
      "num_input_tokens_seen": 71279072,
      "step": 34120
    },
    {
      "epoch": 5.566930418468064,
      "grad_norm": 5.65625,
      "learning_rate": 2.6153603748533705e-06,
      "loss": 2.4592,
      "num_input_tokens_seen": 71289504,
      "step": 34125
    },
    {
      "epoch": 5.567746145688882,
      "grad_norm": 5.1875,
      "learning_rate": 2.6109904604043585e-06,
      "loss": 3.075,
      "num_input_tokens_seen": 71299392,
      "step": 34130
    },
    {
      "epoch": 5.568561872909699,
      "grad_norm": 3.484375,
      "learning_rate": 2.6066239986219765e-06,
      "loss": 1.5505,
      "num_input_tokens_seen": 71309664,
      "step": 34135
    },
    {
      "epoch": 5.569377600130516,
      "grad_norm": 7.53125,
      "learning_rate": 2.602260990179592e-06,
      "loss": 2.2735,
      "num_input_tokens_seen": 71320400,
      "step": 34140
    },
    {
      "epoch": 5.570193327351333,
      "grad_norm": 4.625,
      "learning_rate": 2.5979014357500248e-06,
      "loss": 2.2617,
      "num_input_tokens_seen": 71331072,
      "step": 34145
    },
    {
      "epoch": 5.571009054572151,
      "grad_norm": 3.734375,
      "learning_rate": 2.5935453360055844e-06,
      "loss": 2.1597,
      "num_input_tokens_seen": 71341824,
      "step": 34150
    },
    {
      "epoch": 5.5718247817929685,
      "grad_norm": 8.1875,
      "learning_rate": 2.5891926916180283e-06,
      "loss": 4.0901,
      "num_input_tokens_seen": 71352608,
      "step": 34155
    },
    {
      "epoch": 5.572640509013786,
      "grad_norm": 4.84375,
      "learning_rate": 2.5848435032585883e-06,
      "loss": 0.9527,
      "num_input_tokens_seen": 71362656,
      "step": 34160
    },
    {
      "epoch": 5.573456236234604,
      "grad_norm": 9.8125,
      "learning_rate": 2.58049777159797e-06,
      "loss": 2.4373,
      "num_input_tokens_seen": 71372496,
      "step": 34165
    },
    {
      "epoch": 5.574271963455421,
      "grad_norm": 0.18359375,
      "learning_rate": 2.576155497306332e-06,
      "loss": 1.6881,
      "num_input_tokens_seen": 71382528,
      "step": 34170
    },
    {
      "epoch": 5.575087690676238,
      "grad_norm": 7.28125,
      "learning_rate": 2.57181668105331e-06,
      "loss": 1.7478,
      "num_input_tokens_seen": 71393696,
      "step": 34175
    },
    {
      "epoch": 5.575903417897055,
      "grad_norm": 7.6875,
      "learning_rate": 2.567481323508014e-06,
      "loss": 3.3361,
      "num_input_tokens_seen": 71403216,
      "step": 34180
    },
    {
      "epoch": 5.576719145117872,
      "grad_norm": 9.125,
      "learning_rate": 2.5631494253389954e-06,
      "loss": 2.4501,
      "num_input_tokens_seen": 71414112,
      "step": 34185
    },
    {
      "epoch": 5.57753487233869,
      "grad_norm": 5.5,
      "learning_rate": 2.5588209872142997e-06,
      "loss": 3.7851,
      "num_input_tokens_seen": 71425344,
      "step": 34190
    },
    {
      "epoch": 5.578350599559507,
      "grad_norm": 3.21875,
      "learning_rate": 2.5544960098014186e-06,
      "loss": 1.1406,
      "num_input_tokens_seen": 71435504,
      "step": 34195
    },
    {
      "epoch": 5.579166326780324,
      "grad_norm": 11.6875,
      "learning_rate": 2.550174493767318e-06,
      "loss": 3.1158,
      "num_input_tokens_seen": 71445488,
      "step": 34200
    },
    {
      "epoch": 5.579166326780324,
      "eval_loss": 2.547051191329956,
      "eval_runtime": 134.6809,
      "eval_samples_per_second": 20.233,
      "eval_steps_per_second": 10.12,
      "num_input_tokens_seen": 71445488,
      "step": 34200
    },
    {
      "epoch": 5.5799820540011424,
      "grad_norm": 11.9375,
      "learning_rate": 2.545856439778438e-06,
      "loss": 1.919,
      "num_input_tokens_seen": 71455584,
      "step": 34205
    },
    {
      "epoch": 5.58079778122196,
      "grad_norm": 4.25,
      "learning_rate": 2.541541848500667e-06,
      "loss": 2.3945,
      "num_input_tokens_seen": 71466432,
      "step": 34210
    },
    {
      "epoch": 5.581613508442777,
      "grad_norm": 2.890625,
      "learning_rate": 2.5372307205993733e-06,
      "loss": 2.1849,
      "num_input_tokens_seen": 71476048,
      "step": 34215
    },
    {
      "epoch": 5.582429235663594,
      "grad_norm": 6.96875,
      "learning_rate": 2.5329230567393917e-06,
      "loss": 2.5195,
      "num_input_tokens_seen": 71486128,
      "step": 34220
    },
    {
      "epoch": 5.583244962884411,
      "grad_norm": 8.375,
      "learning_rate": 2.5286188575850164e-06,
      "loss": 1.8473,
      "num_input_tokens_seen": 71497328,
      "step": 34225
    },
    {
      "epoch": 5.584060690105229,
      "grad_norm": 12.4375,
      "learning_rate": 2.5243181237999984e-06,
      "loss": 2.645,
      "num_input_tokens_seen": 71507792,
      "step": 34230
    },
    {
      "epoch": 5.584876417326046,
      "grad_norm": 8.1875,
      "learning_rate": 2.520020856047578e-06,
      "loss": 1.9142,
      "num_input_tokens_seen": 71517600,
      "step": 34235
    },
    {
      "epoch": 5.585692144546863,
      "grad_norm": 6.96875,
      "learning_rate": 2.515727054990438e-06,
      "loss": 2.8674,
      "num_input_tokens_seen": 71529728,
      "step": 34240
    },
    {
      "epoch": 5.586507871767681,
      "grad_norm": 9.5625,
      "learning_rate": 2.511436721290747e-06,
      "loss": 1.9403,
      "num_input_tokens_seen": 71540784,
      "step": 34245
    },
    {
      "epoch": 5.587323598988498,
      "grad_norm": 5.8125,
      "learning_rate": 2.5071498556101164e-06,
      "loss": 3.4682,
      "num_input_tokens_seen": 71550576,
      "step": 34250
    },
    {
      "epoch": 5.5881393262093155,
      "grad_norm": 5.875,
      "learning_rate": 2.5028664586096485e-06,
      "loss": 2.4534,
      "num_input_tokens_seen": 71560656,
      "step": 34255
    },
    {
      "epoch": 5.588955053430133,
      "grad_norm": 7.9375,
      "learning_rate": 2.498586530949881e-06,
      "loss": 1.8271,
      "num_input_tokens_seen": 71571792,
      "step": 34260
    },
    {
      "epoch": 5.589770780650951,
      "grad_norm": 12.125,
      "learning_rate": 2.4943100732908427e-06,
      "loss": 2.5432,
      "num_input_tokens_seen": 71579712,
      "step": 34265
    },
    {
      "epoch": 5.590586507871768,
      "grad_norm": 1.5625,
      "learning_rate": 2.4900370862920188e-06,
      "loss": 1.4368,
      "num_input_tokens_seen": 71590704,
      "step": 34270
    },
    {
      "epoch": 5.591402235092585,
      "grad_norm": 1.8359375,
      "learning_rate": 2.4857675706123518e-06,
      "loss": 2.5995,
      "num_input_tokens_seen": 71600912,
      "step": 34275
    },
    {
      "epoch": 5.592217962313402,
      "grad_norm": 9.875,
      "learning_rate": 2.4815015269102543e-06,
      "loss": 1.6792,
      "num_input_tokens_seen": 71610848,
      "step": 34280
    },
    {
      "epoch": 5.59303368953422,
      "grad_norm": 8.375,
      "learning_rate": 2.477238955843611e-06,
      "loss": 2.04,
      "num_input_tokens_seen": 71620688,
      "step": 34285
    },
    {
      "epoch": 5.593849416755037,
      "grad_norm": 10.4375,
      "learning_rate": 2.4729798580697573e-06,
      "loss": 2.9295,
      "num_input_tokens_seen": 71631792,
      "step": 34290
    },
    {
      "epoch": 5.594665143975854,
      "grad_norm": 7.4375,
      "learning_rate": 2.4687242342455034e-06,
      "loss": 1.723,
      "num_input_tokens_seen": 71641664,
      "step": 34295
    },
    {
      "epoch": 5.5954808711966715,
      "grad_norm": 9.5,
      "learning_rate": 2.4644720850271196e-06,
      "loss": 3.2036,
      "num_input_tokens_seen": 71652096,
      "step": 34300
    },
    {
      "epoch": 5.5962965984174895,
      "grad_norm": 10.125,
      "learning_rate": 2.4602234110703364e-06,
      "loss": 3.1829,
      "num_input_tokens_seen": 71662576,
      "step": 34305
    },
    {
      "epoch": 5.597112325638307,
      "grad_norm": 10.9375,
      "learning_rate": 2.4559782130303576e-06,
      "loss": 2.9722,
      "num_input_tokens_seen": 71672560,
      "step": 34310
    },
    {
      "epoch": 5.597928052859124,
      "grad_norm": 10.0,
      "learning_rate": 2.451736491561843e-06,
      "loss": 2.5311,
      "num_input_tokens_seen": 71682288,
      "step": 34315
    },
    {
      "epoch": 5.598743780079941,
      "grad_norm": 10.0,
      "learning_rate": 2.4474982473189163e-06,
      "loss": 2.8814,
      "num_input_tokens_seen": 71693792,
      "step": 34320
    },
    {
      "epoch": 5.599559507300759,
      "grad_norm": 3.125,
      "learning_rate": 2.4432634809551796e-06,
      "loss": 1.737,
      "num_input_tokens_seen": 71702880,
      "step": 34325
    },
    {
      "epoch": 5.600375234521576,
      "grad_norm": 0.7421875,
      "learning_rate": 2.439032193123675e-06,
      "loss": 2.1028,
      "num_input_tokens_seen": 71714128,
      "step": 34330
    },
    {
      "epoch": 5.601190961742393,
      "grad_norm": 9.8125,
      "learning_rate": 2.4348043844769297e-06,
      "loss": 3.191,
      "num_input_tokens_seen": 71724640,
      "step": 34335
    },
    {
      "epoch": 5.602006688963211,
      "grad_norm": 3.5625,
      "learning_rate": 2.4305800556669146e-06,
      "loss": 1.7909,
      "num_input_tokens_seen": 71733648,
      "step": 34340
    },
    {
      "epoch": 5.602822416184028,
      "grad_norm": 9.1875,
      "learning_rate": 2.426359207345083e-06,
      "loss": 1.2112,
      "num_input_tokens_seen": 71744624,
      "step": 34345
    },
    {
      "epoch": 5.6036381434048455,
      "grad_norm": 9.1875,
      "learning_rate": 2.4221418401623396e-06,
      "loss": 2.5706,
      "num_input_tokens_seen": 71754624,
      "step": 34350
    },
    {
      "epoch": 5.604453870625663,
      "grad_norm": 12.4375,
      "learning_rate": 2.4179279547690557e-06,
      "loss": 2.6951,
      "num_input_tokens_seen": 71764720,
      "step": 34355
    },
    {
      "epoch": 5.60526959784648,
      "grad_norm": 3.90625,
      "learning_rate": 2.413717551815062e-06,
      "loss": 2.5574,
      "num_input_tokens_seen": 71774736,
      "step": 34360
    },
    {
      "epoch": 5.606085325067298,
      "grad_norm": 11.0,
      "learning_rate": 2.409510631949666e-06,
      "loss": 2.4681,
      "num_input_tokens_seen": 71785712,
      "step": 34365
    },
    {
      "epoch": 5.606901052288115,
      "grad_norm": 10.4375,
      "learning_rate": 2.405307195821618e-06,
      "loss": 3.8565,
      "num_input_tokens_seen": 71796608,
      "step": 34370
    },
    {
      "epoch": 5.607716779508932,
      "grad_norm": 6.15625,
      "learning_rate": 2.4011072440791372e-06,
      "loss": 2.0812,
      "num_input_tokens_seen": 71806048,
      "step": 34375
    },
    {
      "epoch": 5.60853250672975,
      "grad_norm": 9.75,
      "learning_rate": 2.3969107773699233e-06,
      "loss": 2.6423,
      "num_input_tokens_seen": 71816384,
      "step": 34380
    },
    {
      "epoch": 5.609348233950567,
      "grad_norm": 3.578125,
      "learning_rate": 2.3927177963411096e-06,
      "loss": 3.7692,
      "num_input_tokens_seen": 71827088,
      "step": 34385
    },
    {
      "epoch": 5.610163961171384,
      "grad_norm": 12.25,
      "learning_rate": 2.3885283016393144e-06,
      "loss": 1.9385,
      "num_input_tokens_seen": 71838288,
      "step": 34390
    },
    {
      "epoch": 5.610979688392201,
      "grad_norm": 5.625,
      "learning_rate": 2.3843422939106076e-06,
      "loss": 2.0691,
      "num_input_tokens_seen": 71848496,
      "step": 34395
    },
    {
      "epoch": 5.6117954156130185,
      "grad_norm": 2.8125,
      "learning_rate": 2.380159773800525e-06,
      "loss": 1.7717,
      "num_input_tokens_seen": 71858096,
      "step": 34400
    },
    {
      "epoch": 5.6117954156130185,
      "eval_loss": 2.5390069484710693,
      "eval_runtime": 134.7353,
      "eval_samples_per_second": 20.225,
      "eval_steps_per_second": 10.116,
      "num_input_tokens_seen": 71858096,
      "step": 34400
    },
    {
      "epoch": 5.6126111428338366,
      "grad_norm": 9.5625,
      "learning_rate": 2.3759807419540675e-06,
      "loss": 1.9853,
      "num_input_tokens_seen": 71868224,
      "step": 34405
    },
    {
      "epoch": 5.613426870054654,
      "grad_norm": 6.6875,
      "learning_rate": 2.3718051990156835e-06,
      "loss": 1.1833,
      "num_input_tokens_seen": 71877264,
      "step": 34410
    },
    {
      "epoch": 5.614242597275471,
      "grad_norm": 7.125,
      "learning_rate": 2.367633145629311e-06,
      "loss": 3.5393,
      "num_input_tokens_seen": 71888592,
      "step": 34415
    },
    {
      "epoch": 5.615058324496289,
      "grad_norm": 8.25,
      "learning_rate": 2.363464582438316e-06,
      "loss": 2.373,
      "num_input_tokens_seen": 71898640,
      "step": 34420
    },
    {
      "epoch": 5.615874051717106,
      "grad_norm": 5.625,
      "learning_rate": 2.3592995100855526e-06,
      "loss": 1.7478,
      "num_input_tokens_seen": 71909040,
      "step": 34425
    },
    {
      "epoch": 5.616689778937923,
      "grad_norm": 6.96875,
      "learning_rate": 2.3551379292133273e-06,
      "loss": 1.6532,
      "num_input_tokens_seen": 71919440,
      "step": 34430
    },
    {
      "epoch": 5.61750550615874,
      "grad_norm": 11.75,
      "learning_rate": 2.3509798404634047e-06,
      "loss": 2.5449,
      "num_input_tokens_seen": 71930384,
      "step": 34435
    },
    {
      "epoch": 5.618321233379558,
      "grad_norm": 9.125,
      "learning_rate": 2.346825244477019e-06,
      "loss": 2.4725,
      "num_input_tokens_seen": 71940016,
      "step": 34440
    },
    {
      "epoch": 5.619136960600375,
      "grad_norm": 5.53125,
      "learning_rate": 2.3426741418948545e-06,
      "loss": 2.2715,
      "num_input_tokens_seen": 71951776,
      "step": 34445
    },
    {
      "epoch": 5.6199526878211925,
      "grad_norm": 11.0625,
      "learning_rate": 2.3385265333570715e-06,
      "loss": 1.6268,
      "num_input_tokens_seen": 71962688,
      "step": 34450
    },
    {
      "epoch": 5.62076841504201,
      "grad_norm": 18.375,
      "learning_rate": 2.334382419503278e-06,
      "loss": 3.108,
      "num_input_tokens_seen": 71973040,
      "step": 34455
    },
    {
      "epoch": 5.621584142262828,
      "grad_norm": 5.71875,
      "learning_rate": 2.3302418009725465e-06,
      "loss": 3.6945,
      "num_input_tokens_seen": 71982976,
      "step": 34460
    },
    {
      "epoch": 5.622399869483645,
      "grad_norm": 5.03125,
      "learning_rate": 2.326104678403415e-06,
      "loss": 3.1249,
      "num_input_tokens_seen": 71994320,
      "step": 34465
    },
    {
      "epoch": 5.623215596704462,
      "grad_norm": 5.59375,
      "learning_rate": 2.321971052433883e-06,
      "loss": 2.1336,
      "num_input_tokens_seen": 72004960,
      "step": 34470
    },
    {
      "epoch": 5.624031323925279,
      "grad_norm": 15.125,
      "learning_rate": 2.3178409237014004e-06,
      "loss": 2.2603,
      "num_input_tokens_seen": 72015216,
      "step": 34475
    },
    {
      "epoch": 5.624847051146097,
      "grad_norm": 4.6875,
      "learning_rate": 2.313714292842889e-06,
      "loss": 2.7268,
      "num_input_tokens_seen": 72027328,
      "step": 34480
    },
    {
      "epoch": 5.625662778366914,
      "grad_norm": 4.5625,
      "learning_rate": 2.309591160494734e-06,
      "loss": 1.6208,
      "num_input_tokens_seen": 72037600,
      "step": 34485
    },
    {
      "epoch": 5.626478505587731,
      "grad_norm": 6.59375,
      "learning_rate": 2.305471527292763e-06,
      "loss": 2.2829,
      "num_input_tokens_seen": 72045760,
      "step": 34490
    },
    {
      "epoch": 5.6272942328085485,
      "grad_norm": 7.40625,
      "learning_rate": 2.3013553938722817e-06,
      "loss": 2.0435,
      "num_input_tokens_seen": 72056608,
      "step": 34495
    },
    {
      "epoch": 5.6281099600293665,
      "grad_norm": 8.1875,
      "learning_rate": 2.297242760868043e-06,
      "loss": 3.1942,
      "num_input_tokens_seen": 72066976,
      "step": 34500
    },
    {
      "epoch": 5.628925687250184,
      "grad_norm": 0.7265625,
      "learning_rate": 2.2931336289142735e-06,
      "loss": 0.8745,
      "num_input_tokens_seen": 72077504,
      "step": 34505
    },
    {
      "epoch": 5.629741414471001,
      "grad_norm": 7.9375,
      "learning_rate": 2.289027998644655e-06,
      "loss": 3.1736,
      "num_input_tokens_seen": 72086000,
      "step": 34510
    },
    {
      "epoch": 5.630557141691818,
      "grad_norm": 11.4375,
      "learning_rate": 2.2849258706923228e-06,
      "loss": 3.1483,
      "num_input_tokens_seen": 72095824,
      "step": 34515
    },
    {
      "epoch": 5.631372868912636,
      "grad_norm": 4.21875,
      "learning_rate": 2.2808272456898705e-06,
      "loss": 2.6973,
      "num_input_tokens_seen": 72106224,
      "step": 34520
    },
    {
      "epoch": 5.632188596133453,
      "grad_norm": 5.5625,
      "learning_rate": 2.2767321242693707e-06,
      "loss": 2.2799,
      "num_input_tokens_seen": 72117584,
      "step": 34525
    },
    {
      "epoch": 5.63300432335427,
      "grad_norm": 6.5,
      "learning_rate": 2.272640507062329e-06,
      "loss": 1.6744,
      "num_input_tokens_seen": 72126896,
      "step": 34530
    },
    {
      "epoch": 5.633820050575087,
      "grad_norm": 9.0625,
      "learning_rate": 2.2685523946997382e-06,
      "loss": 2.5545,
      "num_input_tokens_seen": 72138912,
      "step": 34535
    },
    {
      "epoch": 5.634635777795905,
      "grad_norm": 6.09375,
      "learning_rate": 2.2644677878120245e-06,
      "loss": 2.5968,
      "num_input_tokens_seen": 72150160,
      "step": 34540
    },
    {
      "epoch": 5.635451505016722,
      "grad_norm": 14.375,
      "learning_rate": 2.2603866870290897e-06,
      "loss": 2.964,
      "num_input_tokens_seen": 72161008,
      "step": 34545
    },
    {
      "epoch": 5.63626723223754,
      "grad_norm": 7.875,
      "learning_rate": 2.256309092980294e-06,
      "loss": 1.7179,
      "num_input_tokens_seen": 72172768,
      "step": 34550
    },
    {
      "epoch": 5.637082959458358,
      "grad_norm": 5.46875,
      "learning_rate": 2.252235006294448e-06,
      "loss": 2.795,
      "num_input_tokens_seen": 72183552,
      "step": 34555
    },
    {
      "epoch": 5.637898686679175,
      "grad_norm": 7.75,
      "learning_rate": 2.2481644275998333e-06,
      "loss": 2.7397,
      "num_input_tokens_seen": 72193664,
      "step": 34560
    },
    {
      "epoch": 5.638714413899992,
      "grad_norm": 6.5,
      "learning_rate": 2.2440973575241832e-06,
      "loss": 1.921,
      "num_input_tokens_seen": 72204592,
      "step": 34565
    },
    {
      "epoch": 5.639530141120809,
      "grad_norm": 4.28125,
      "learning_rate": 2.240033796694685e-06,
      "loss": 3.181,
      "num_input_tokens_seen": 72215760,
      "step": 34570
    },
    {
      "epoch": 5.640345868341626,
      "grad_norm": 4.53125,
      "learning_rate": 2.235973745737999e-06,
      "loss": 2.3728,
      "num_input_tokens_seen": 72226368,
      "step": 34575
    },
    {
      "epoch": 5.641161595562444,
      "grad_norm": 10.75,
      "learning_rate": 2.2319172052802263e-06,
      "loss": 2.4891,
      "num_input_tokens_seen": 72237360,
      "step": 34580
    },
    {
      "epoch": 5.641977322783261,
      "grad_norm": 8.875,
      "learning_rate": 2.2278641759469477e-06,
      "loss": 3.0054,
      "num_input_tokens_seen": 72246928,
      "step": 34585
    },
    {
      "epoch": 5.642793050004078,
      "grad_norm": 7.625,
      "learning_rate": 2.2238146583631825e-06,
      "loss": 1.6162,
      "num_input_tokens_seen": 72256400,
      "step": 34590
    },
    {
      "epoch": 5.643608777224896,
      "grad_norm": 5.25,
      "learning_rate": 2.2197686531534256e-06,
      "loss": 2.5231,
      "num_input_tokens_seen": 72265744,
      "step": 34595
    },
    {
      "epoch": 5.6444245044457135,
      "grad_norm": 9.25,
      "learning_rate": 2.2157261609416087e-06,
      "loss": 2.4476,
      "num_input_tokens_seen": 72276272,
      "step": 34600
    },
    {
      "epoch": 5.6444245044457135,
      "eval_loss": 2.53377103805542,
      "eval_runtime": 134.7344,
      "eval_samples_per_second": 20.225,
      "eval_steps_per_second": 10.116,
      "num_input_tokens_seen": 72276272,
      "step": 34600
    },
    {
      "epoch": 5.645240231666531,
      "grad_norm": 6.15625,
      "learning_rate": 2.211687182351149e-06,
      "loss": 2.8777,
      "num_input_tokens_seen": 72287232,
      "step": 34605
    },
    {
      "epoch": 5.646055958887348,
      "grad_norm": 8.4375,
      "learning_rate": 2.2076517180048993e-06,
      "loss": 1.5784,
      "num_input_tokens_seen": 72298224,
      "step": 34610
    },
    {
      "epoch": 5.646871686108166,
      "grad_norm": 0.054443359375,
      "learning_rate": 2.2036197685251834e-06,
      "loss": 3.3634,
      "num_input_tokens_seen": 72308384,
      "step": 34615
    },
    {
      "epoch": 5.647687413328983,
      "grad_norm": 3.25,
      "learning_rate": 2.199591334533771e-06,
      "loss": 1.9471,
      "num_input_tokens_seen": 72318208,
      "step": 34620
    },
    {
      "epoch": 5.6485031405498,
      "grad_norm": 2.5,
      "learning_rate": 2.1955664166519036e-06,
      "loss": 1.3057,
      "num_input_tokens_seen": 72329152,
      "step": 34625
    },
    {
      "epoch": 5.649318867770617,
      "grad_norm": 6.71875,
      "learning_rate": 2.1915450155002793e-06,
      "loss": 3.5041,
      "num_input_tokens_seen": 72340016,
      "step": 34630
    },
    {
      "epoch": 5.650134594991435,
      "grad_norm": 12.875,
      "learning_rate": 2.187527131699038e-06,
      "loss": 2.7681,
      "num_input_tokens_seen": 72351392,
      "step": 34635
    },
    {
      "epoch": 5.650950322212252,
      "grad_norm": 6.5,
      "learning_rate": 2.18351276586779e-06,
      "loss": 3.3497,
      "num_input_tokens_seen": 72361888,
      "step": 34640
    },
    {
      "epoch": 5.6517660494330695,
      "grad_norm": 10.75,
      "learning_rate": 2.1795019186256092e-06,
      "loss": 3.9367,
      "num_input_tokens_seen": 72372320,
      "step": 34645
    },
    {
      "epoch": 5.652581776653887,
      "grad_norm": 11.0,
      "learning_rate": 2.1754945905910094e-06,
      "loss": 2.655,
      "num_input_tokens_seen": 72382960,
      "step": 34650
    },
    {
      "epoch": 5.653397503874705,
      "grad_norm": 3.578125,
      "learning_rate": 2.171490782381977e-06,
      "loss": 1.4699,
      "num_input_tokens_seen": 72392912,
      "step": 34655
    },
    {
      "epoch": 5.654213231095522,
      "grad_norm": 4.78125,
      "learning_rate": 2.1674904946159425e-06,
      "loss": 1.5977,
      "num_input_tokens_seen": 72402496,
      "step": 34660
    },
    {
      "epoch": 5.655028958316339,
      "grad_norm": 3.359375,
      "learning_rate": 2.16349372790981e-06,
      "loss": 1.7798,
      "num_input_tokens_seen": 72413264,
      "step": 34665
    },
    {
      "epoch": 5.655844685537156,
      "grad_norm": 4.84375,
      "learning_rate": 2.159500482879928e-06,
      "loss": 2.6119,
      "num_input_tokens_seen": 72422864,
      "step": 34670
    },
    {
      "epoch": 5.656660412757974,
      "grad_norm": 8.875,
      "learning_rate": 2.155510760142096e-06,
      "loss": 3.4643,
      "num_input_tokens_seen": 72432944,
      "step": 34675
    },
    {
      "epoch": 5.657476139978791,
      "grad_norm": 5.78125,
      "learning_rate": 2.151524560311588e-06,
      "loss": 1.5527,
      "num_input_tokens_seen": 72443168,
      "step": 34680
    },
    {
      "epoch": 5.658291867199608,
      "grad_norm": 14.4375,
      "learning_rate": 2.147541884003129e-06,
      "loss": 3.2125,
      "num_input_tokens_seen": 72452128,
      "step": 34685
    },
    {
      "epoch": 5.659107594420425,
      "grad_norm": 4.3125,
      "learning_rate": 2.1435627318308895e-06,
      "loss": 2.0491,
      "num_input_tokens_seen": 72463520,
      "step": 34690
    },
    {
      "epoch": 5.6599233216412435,
      "grad_norm": 6.3125,
      "learning_rate": 2.139587104408511e-06,
      "loss": 2.391,
      "num_input_tokens_seen": 72474160,
      "step": 34695
    },
    {
      "epoch": 5.660739048862061,
      "grad_norm": 6.25,
      "learning_rate": 2.1356150023490783e-06,
      "loss": 1.7607,
      "num_input_tokens_seen": 72483936,
      "step": 34700
    },
    {
      "epoch": 5.661554776082878,
      "grad_norm": 10.125,
      "learning_rate": 2.1316464262651464e-06,
      "loss": 2.705,
      "num_input_tokens_seen": 72496208,
      "step": 34705
    },
    {
      "epoch": 5.662370503303695,
      "grad_norm": 6.75,
      "learning_rate": 2.1276813767687224e-06,
      "loss": 1.9463,
      "num_input_tokens_seen": 72506976,
      "step": 34710
    },
    {
      "epoch": 5.663186230524513,
      "grad_norm": 5.5625,
      "learning_rate": 2.123719854471254e-06,
      "loss": 2.936,
      "num_input_tokens_seen": 72517504,
      "step": 34715
    },
    {
      "epoch": 5.66400195774533,
      "grad_norm": 6.4375,
      "learning_rate": 2.119761859983668e-06,
      "loss": 1.6663,
      "num_input_tokens_seen": 72527792,
      "step": 34720
    },
    {
      "epoch": 5.664817684966147,
      "grad_norm": 9.25,
      "learning_rate": 2.1158073939163386e-06,
      "loss": 2.1571,
      "num_input_tokens_seen": 72538128,
      "step": 34725
    },
    {
      "epoch": 5.665633412186965,
      "grad_norm": 8.6875,
      "learning_rate": 2.111856456879088e-06,
      "loss": 2.1242,
      "num_input_tokens_seen": 72548000,
      "step": 34730
    },
    {
      "epoch": 5.666449139407782,
      "grad_norm": 8.8125,
      "learning_rate": 2.1079090494811993e-06,
      "loss": 2.4869,
      "num_input_tokens_seen": 72558096,
      "step": 34735
    },
    {
      "epoch": 5.667264866628599,
      "grad_norm": 6.3125,
      "learning_rate": 2.103965172331418e-06,
      "loss": 2.6868,
      "num_input_tokens_seen": 72568112,
      "step": 34740
    },
    {
      "epoch": 5.6680805938494165,
      "grad_norm": 5.59375,
      "learning_rate": 2.100024826037933e-06,
      "loss": 3.843,
      "num_input_tokens_seen": 72578112,
      "step": 34745
    },
    {
      "epoch": 5.668896321070234,
      "grad_norm": 8.3125,
      "learning_rate": 2.0960880112084027e-06,
      "loss": 1.8033,
      "num_input_tokens_seen": 72588816,
      "step": 34750
    },
    {
      "epoch": 5.669712048291052,
      "grad_norm": 13.625,
      "learning_rate": 2.092154728449927e-06,
      "loss": 2.2841,
      "num_input_tokens_seen": 72598864,
      "step": 34755
    },
    {
      "epoch": 5.670527775511869,
      "grad_norm": 0.15234375,
      "learning_rate": 2.0882249783690687e-06,
      "loss": 1.8777,
      "num_input_tokens_seen": 72608976,
      "step": 34760
    },
    {
      "epoch": 5.671343502732686,
      "grad_norm": 8.75,
      "learning_rate": 2.084298761571851e-06,
      "loss": 2.5514,
      "num_input_tokens_seen": 72620032,
      "step": 34765
    },
    {
      "epoch": 5.672159229953504,
      "grad_norm": 3.4375,
      "learning_rate": 2.080376078663737e-06,
      "loss": 1.5709,
      "num_input_tokens_seen": 72630608,
      "step": 34770
    },
    {
      "epoch": 5.672974957174321,
      "grad_norm": 4.0625,
      "learning_rate": 2.0764569302496593e-06,
      "loss": 1.651,
      "num_input_tokens_seen": 72642416,
      "step": 34775
    },
    {
      "epoch": 5.673790684395138,
      "grad_norm": 7.40625,
      "learning_rate": 2.0725413169339957e-06,
      "loss": 1.8498,
      "num_input_tokens_seen": 72653840,
      "step": 34780
    },
    {
      "epoch": 5.674606411615955,
      "grad_norm": 11.3125,
      "learning_rate": 2.068629239320588e-06,
      "loss": 4.7419,
      "num_input_tokens_seen": 72663408,
      "step": 34785
    },
    {
      "epoch": 5.6754221388367725,
      "grad_norm": 19.125,
      "learning_rate": 2.064720698012726e-06,
      "loss": 3.3115,
      "num_input_tokens_seen": 72673824,
      "step": 34790
    },
    {
      "epoch": 5.6762378660575905,
      "grad_norm": 16.375,
      "learning_rate": 2.0608156936131522e-06,
      "loss": 2.3557,
      "num_input_tokens_seen": 72684000,
      "step": 34795
    },
    {
      "epoch": 5.677053593278408,
      "grad_norm": 6.0625,
      "learning_rate": 2.056914226724074e-06,
      "loss": 1.8602,
      "num_input_tokens_seen": 72694032,
      "step": 34800
    },
    {
      "epoch": 5.677053593278408,
      "eval_loss": 2.5323657989501953,
      "eval_runtime": 134.7521,
      "eval_samples_per_second": 20.222,
      "eval_steps_per_second": 10.115,
      "num_input_tokens_seen": 72694032,
      "step": 34800
    },
    {
      "epoch": 5.677869320499225,
      "grad_norm": 6.40625,
      "learning_rate": 2.0530162979471385e-06,
      "loss": 1.9967,
      "num_input_tokens_seen": 72704208,
      "step": 34805
    },
    {
      "epoch": 5.678685047720043,
      "grad_norm": 6.78125,
      "learning_rate": 2.0491219078834667e-06,
      "loss": 2.0054,
      "num_input_tokens_seen": 72717472,
      "step": 34810
    },
    {
      "epoch": 5.67950077494086,
      "grad_norm": 4.8125,
      "learning_rate": 2.045231057133612e-06,
      "loss": 1.7196,
      "num_input_tokens_seen": 72728048,
      "step": 34815
    },
    {
      "epoch": 5.680316502161677,
      "grad_norm": 8.75,
      "learning_rate": 2.0413437462975944e-06,
      "loss": 2.2625,
      "num_input_tokens_seen": 72737696,
      "step": 34820
    },
    {
      "epoch": 5.681132229382494,
      "grad_norm": 6.9375,
      "learning_rate": 2.0374599759748843e-06,
      "loss": 2.6584,
      "num_input_tokens_seen": 72748304,
      "step": 34825
    },
    {
      "epoch": 5.681947956603312,
      "grad_norm": 11.0625,
      "learning_rate": 2.033579746764419e-06,
      "loss": 3.2357,
      "num_input_tokens_seen": 72757024,
      "step": 34830
    },
    {
      "epoch": 5.682763683824129,
      "grad_norm": 7.78125,
      "learning_rate": 2.029703059264565e-06,
      "loss": 2.3419,
      "num_input_tokens_seen": 72768320,
      "step": 34835
    },
    {
      "epoch": 5.6835794110449465,
      "grad_norm": 11.8125,
      "learning_rate": 2.02582991407316e-06,
      "loss": 3.2673,
      "num_input_tokens_seen": 72778656,
      "step": 34840
    },
    {
      "epoch": 5.684395138265764,
      "grad_norm": 4.21875,
      "learning_rate": 2.0219603117874992e-06,
      "loss": 1.9369,
      "num_input_tokens_seen": 72788544,
      "step": 34845
    },
    {
      "epoch": 5.685210865486582,
      "grad_norm": 6.53125,
      "learning_rate": 2.0180942530043156e-06,
      "loss": 1.7424,
      "num_input_tokens_seen": 72798352,
      "step": 34850
    },
    {
      "epoch": 5.686026592707399,
      "grad_norm": 15.8125,
      "learning_rate": 2.0142317383198107e-06,
      "loss": 2.663,
      "num_input_tokens_seen": 72809360,
      "step": 34855
    },
    {
      "epoch": 5.686842319928216,
      "grad_norm": 6.21875,
      "learning_rate": 2.0103727683296243e-06,
      "loss": 1.8504,
      "num_input_tokens_seen": 72820496,
      "step": 34860
    },
    {
      "epoch": 5.687658047149033,
      "grad_norm": 4.28125,
      "learning_rate": 2.0065173436288636e-06,
      "loss": 2.5163,
      "num_input_tokens_seen": 72830000,
      "step": 34865
    },
    {
      "epoch": 5.688473774369851,
      "grad_norm": 3.6875,
      "learning_rate": 2.002665464812087e-06,
      "loss": 1.7411,
      "num_input_tokens_seen": 72841584,
      "step": 34870
    },
    {
      "epoch": 5.689289501590668,
      "grad_norm": 5.375,
      "learning_rate": 1.998817132473291e-06,
      "loss": 1.6034,
      "num_input_tokens_seen": 72853072,
      "step": 34875
    },
    {
      "epoch": 5.690105228811485,
      "grad_norm": 12.5,
      "learning_rate": 1.9949723472059507e-06,
      "loss": 2.8051,
      "num_input_tokens_seen": 72863376,
      "step": 34880
    },
    {
      "epoch": 5.690920956032302,
      "grad_norm": 6.71875,
      "learning_rate": 1.9911311096029726e-06,
      "loss": 2.5916,
      "num_input_tokens_seen": 72875280,
      "step": 34885
    },
    {
      "epoch": 5.69173668325312,
      "grad_norm": 12.0625,
      "learning_rate": 1.9872934202567224e-06,
      "loss": 2.5245,
      "num_input_tokens_seen": 72884912,
      "step": 34890
    },
    {
      "epoch": 5.692552410473938,
      "grad_norm": 8.25,
      "learning_rate": 1.9834592797590257e-06,
      "loss": 4.7388,
      "num_input_tokens_seen": 72895264,
      "step": 34895
    },
    {
      "epoch": 5.693368137694755,
      "grad_norm": 12.625,
      "learning_rate": 1.979628688701149e-06,
      "loss": 3.1135,
      "num_input_tokens_seen": 72906160,
      "step": 34900
    },
    {
      "epoch": 5.694183864915573,
      "grad_norm": 7.125,
      "learning_rate": 1.9758016476738193e-06,
      "loss": 1.3217,
      "num_input_tokens_seen": 72918880,
      "step": 34905
    },
    {
      "epoch": 5.69499959213639,
      "grad_norm": 4.59375,
      "learning_rate": 1.971978157267221e-06,
      "loss": 2.2688,
      "num_input_tokens_seen": 72930736,
      "step": 34910
    },
    {
      "epoch": 5.695815319357207,
      "grad_norm": 9.625,
      "learning_rate": 1.968158218070973e-06,
      "loss": 3.54,
      "num_input_tokens_seen": 72941216,
      "step": 34915
    },
    {
      "epoch": 5.696631046578024,
      "grad_norm": 3.90625,
      "learning_rate": 1.9643418306741682e-06,
      "loss": 1.4181,
      "num_input_tokens_seen": 72952560,
      "step": 34920
    },
    {
      "epoch": 5.697446773798841,
      "grad_norm": 3.40625,
      "learning_rate": 1.9605289956653337e-06,
      "loss": 3.2634,
      "num_input_tokens_seen": 72962800,
      "step": 34925
    },
    {
      "epoch": 5.698262501019659,
      "grad_norm": 5.375,
      "learning_rate": 1.9567197136324626e-06,
      "loss": 3.3198,
      "num_input_tokens_seen": 72973664,
      "step": 34930
    },
    {
      "epoch": 5.699078228240476,
      "grad_norm": 4.3125,
      "learning_rate": 1.9529139851629935e-06,
      "loss": 2.6721,
      "num_input_tokens_seen": 72985504,
      "step": 34935
    },
    {
      "epoch": 5.6998939554612935,
      "grad_norm": 9.5625,
      "learning_rate": 1.949111810843812e-06,
      "loss": 2.2349,
      "num_input_tokens_seen": 72995904,
      "step": 34940
    },
    {
      "epoch": 5.7007096826821115,
      "grad_norm": 9.4375,
      "learning_rate": 1.9453131912612694e-06,
      "loss": 2.43,
      "num_input_tokens_seen": 73006592,
      "step": 34945
    },
    {
      "epoch": 5.701525409902929,
      "grad_norm": 8.0,
      "learning_rate": 1.941518127001149e-06,
      "loss": 2.3815,
      "num_input_tokens_seen": 73018432,
      "step": 34950
    },
    {
      "epoch": 5.702341137123746,
      "grad_norm": 7.21875,
      "learning_rate": 1.9377266186487107e-06,
      "loss": 1.8554,
      "num_input_tokens_seen": 73028768,
      "step": 34955
    },
    {
      "epoch": 5.703156864344563,
      "grad_norm": 5.5625,
      "learning_rate": 1.9339386667886483e-06,
      "loss": 2.602,
      "num_input_tokens_seen": 73040016,
      "step": 34960
    },
    {
      "epoch": 5.70397259156538,
      "grad_norm": 10.0625,
      "learning_rate": 1.9301542720051024e-06,
      "loss": 2.1533,
      "num_input_tokens_seen": 73049792,
      "step": 34965
    },
    {
      "epoch": 5.704788318786198,
      "grad_norm": 3.078125,
      "learning_rate": 1.926373434881684e-06,
      "loss": 2.168,
      "num_input_tokens_seen": 73058704,
      "step": 34970
    },
    {
      "epoch": 5.705604046007015,
      "grad_norm": 7.84375,
      "learning_rate": 1.9225961560014468e-06,
      "loss": 2.9266,
      "num_input_tokens_seen": 73067696,
      "step": 34975
    },
    {
      "epoch": 5.706419773227832,
      "grad_norm": 12.8125,
      "learning_rate": 1.918822435946885e-06,
      "loss": 3.9945,
      "num_input_tokens_seen": 73079440,
      "step": 34980
    },
    {
      "epoch": 5.70723550044865,
      "grad_norm": 2.5625,
      "learning_rate": 1.915052275299961e-06,
      "loss": 1.6514,
      "num_input_tokens_seen": 73089152,
      "step": 34985
    },
    {
      "epoch": 5.7080512276694675,
      "grad_norm": 9.9375,
      "learning_rate": 1.9112856746420854e-06,
      "loss": 2.1754,
      "num_input_tokens_seen": 73100144,
      "step": 34990
    },
    {
      "epoch": 5.708866954890285,
      "grad_norm": 11.8125,
      "learning_rate": 1.907522634554104e-06,
      "loss": 2.5818,
      "num_input_tokens_seen": 73110544,
      "step": 34995
    },
    {
      "epoch": 5.709682682111102,
      "grad_norm": 7.625,
      "learning_rate": 1.9037631556163337e-06,
      "loss": 2.1456,
      "num_input_tokens_seen": 73119856,
      "step": 35000
    },
    {
      "epoch": 5.709682682111102,
      "eval_loss": 2.546726703643799,
      "eval_runtime": 134.5612,
      "eval_samples_per_second": 20.251,
      "eval_steps_per_second": 10.129,
      "num_input_tokens_seen": 73119856,
      "step": 35000
    },
    {
      "epoch": 5.71049840933192,
      "grad_norm": 7.875,
      "learning_rate": 1.9000072384085272e-06,
      "loss": 2.3632,
      "num_input_tokens_seen": 73129952,
      "step": 35005
    },
    {
      "epoch": 5.711314136552737,
      "grad_norm": 7.375,
      "learning_rate": 1.8962548835098987e-06,
      "loss": 2.0775,
      "num_input_tokens_seen": 73138528,
      "step": 35010
    },
    {
      "epoch": 5.712129863773554,
      "grad_norm": 7.90625,
      "learning_rate": 1.8925060914991077e-06,
      "loss": 2.9103,
      "num_input_tokens_seen": 73148288,
      "step": 35015
    },
    {
      "epoch": 5.712945590994371,
      "grad_norm": 8.5625,
      "learning_rate": 1.888760862954264e-06,
      "loss": 2.1938,
      "num_input_tokens_seen": 73158304,
      "step": 35020
    },
    {
      "epoch": 5.713761318215189,
      "grad_norm": 4.5,
      "learning_rate": 1.8850191984529309e-06,
      "loss": 1.8173,
      "num_input_tokens_seen": 73167824,
      "step": 35025
    },
    {
      "epoch": 5.714577045436006,
      "grad_norm": 12.8125,
      "learning_rate": 1.8812810985721186e-06,
      "loss": 3.4717,
      "num_input_tokens_seen": 73179744,
      "step": 35030
    },
    {
      "epoch": 5.715392772656823,
      "grad_norm": 4.84375,
      "learning_rate": 1.8775465638882856e-06,
      "loss": 3.1615,
      "num_input_tokens_seen": 73189840,
      "step": 35035
    },
    {
      "epoch": 5.716208499877641,
      "grad_norm": 7.71875,
      "learning_rate": 1.8738155949773517e-06,
      "loss": 2.5083,
      "num_input_tokens_seen": 73199648,
      "step": 35040
    },
    {
      "epoch": 5.717024227098459,
      "grad_norm": 0.70703125,
      "learning_rate": 1.8700881924146707e-06,
      "loss": 1.502,
      "num_input_tokens_seen": 73208720,
      "step": 35045
    },
    {
      "epoch": 5.717839954319276,
      "grad_norm": 3.953125,
      "learning_rate": 1.8663643567750577e-06,
      "loss": 2.6797,
      "num_input_tokens_seen": 73220592,
      "step": 35050
    },
    {
      "epoch": 5.718655681540093,
      "grad_norm": 4.21875,
      "learning_rate": 1.8626440886327813e-06,
      "loss": 2.6372,
      "num_input_tokens_seen": 73230896,
      "step": 35055
    },
    {
      "epoch": 5.71947140876091,
      "grad_norm": 7.96875,
      "learning_rate": 1.8589273885615432e-06,
      "loss": 3.7997,
      "num_input_tokens_seen": 73242704,
      "step": 35060
    },
    {
      "epoch": 5.720287135981728,
      "grad_norm": 3.390625,
      "learning_rate": 1.8552142571345133e-06,
      "loss": 2.9957,
      "num_input_tokens_seen": 73252992,
      "step": 35065
    },
    {
      "epoch": 5.721102863202545,
      "grad_norm": 8.9375,
      "learning_rate": 1.8515046949243025e-06,
      "loss": 2.4913,
      "num_input_tokens_seen": 73263776,
      "step": 35070
    },
    {
      "epoch": 5.721918590423362,
      "grad_norm": 4.4375,
      "learning_rate": 1.8477987025029674e-06,
      "loss": 1.8598,
      "num_input_tokens_seen": 73275184,
      "step": 35075
    },
    {
      "epoch": 5.72273431764418,
      "grad_norm": 5.96875,
      "learning_rate": 1.8440962804420232e-06,
      "loss": 2.9535,
      "num_input_tokens_seen": 73285600,
      "step": 35080
    },
    {
      "epoch": 5.723550044864997,
      "grad_norm": 5.03125,
      "learning_rate": 1.8403974293124265e-06,
      "loss": 1.8279,
      "num_input_tokens_seen": 73294976,
      "step": 35085
    },
    {
      "epoch": 5.7243657720858145,
      "grad_norm": 3.984375,
      "learning_rate": 1.8367021496845854e-06,
      "loss": 3.7718,
      "num_input_tokens_seen": 73306224,
      "step": 35090
    },
    {
      "epoch": 5.725181499306632,
      "grad_norm": 7.53125,
      "learning_rate": 1.8330104421283662e-06,
      "loss": 1.6353,
      "num_input_tokens_seen": 73316448,
      "step": 35095
    },
    {
      "epoch": 5.725997226527449,
      "grad_norm": 8.625,
      "learning_rate": 1.8293223072130717e-06,
      "loss": 2.9711,
      "num_input_tokens_seen": 73327552,
      "step": 35100
    },
    {
      "epoch": 5.726812953748267,
      "grad_norm": 7.34375,
      "learning_rate": 1.8256377455074525e-06,
      "loss": 2.4942,
      "num_input_tokens_seen": 73338384,
      "step": 35105
    },
    {
      "epoch": 5.727628680969084,
      "grad_norm": 4.78125,
      "learning_rate": 1.8219567575797263e-06,
      "loss": 1.894,
      "num_input_tokens_seen": 73349216,
      "step": 35110
    },
    {
      "epoch": 5.728444408189901,
      "grad_norm": 1.9453125,
      "learning_rate": 1.8182793439975365e-06,
      "loss": 1.0833,
      "num_input_tokens_seen": 73359792,
      "step": 35115
    },
    {
      "epoch": 5.729260135410719,
      "grad_norm": 5.25,
      "learning_rate": 1.8146055053279958e-06,
      "loss": 2.4312,
      "num_input_tokens_seen": 73370544,
      "step": 35120
    },
    {
      "epoch": 5.730075862631536,
      "grad_norm": 6.15625,
      "learning_rate": 1.8109352421376486e-06,
      "loss": 2.0853,
      "num_input_tokens_seen": 73379712,
      "step": 35125
    },
    {
      "epoch": 5.730891589852353,
      "grad_norm": 13.0625,
      "learning_rate": 1.8072685549924972e-06,
      "loss": 2.7545,
      "num_input_tokens_seen": 73390752,
      "step": 35130
    },
    {
      "epoch": 5.7317073170731705,
      "grad_norm": 5.625,
      "learning_rate": 1.8036054444579982e-06,
      "loss": 2.143,
      "num_input_tokens_seen": 73400832,
      "step": 35135
    },
    {
      "epoch": 5.732523044293988,
      "grad_norm": 1.7890625,
      "learning_rate": 1.7999459110990407e-06,
      "loss": 1.5524,
      "num_input_tokens_seen": 73412800,
      "step": 35140
    },
    {
      "epoch": 5.733338771514806,
      "grad_norm": 7.6875,
      "learning_rate": 1.7962899554799712e-06,
      "loss": 1.6,
      "num_input_tokens_seen": 73423920,
      "step": 35145
    },
    {
      "epoch": 5.734154498735623,
      "grad_norm": 3.59375,
      "learning_rate": 1.7926375781645937e-06,
      "loss": 1.5129,
      "num_input_tokens_seen": 73435200,
      "step": 35150
    },
    {
      "epoch": 5.73497022595644,
      "grad_norm": 6.4375,
      "learning_rate": 1.7889887797161359e-06,
      "loss": 1.3796,
      "num_input_tokens_seen": 73447296,
      "step": 35155
    },
    {
      "epoch": 5.735785953177258,
      "grad_norm": 2.546875,
      "learning_rate": 1.7853435606973028e-06,
      "loss": 2.1926,
      "num_input_tokens_seen": 73458016,
      "step": 35160
    },
    {
      "epoch": 5.736601680398075,
      "grad_norm": 7.21875,
      "learning_rate": 1.781701921670223e-06,
      "loss": 2.1694,
      "num_input_tokens_seen": 73468784,
      "step": 35165
    },
    {
      "epoch": 5.737417407618892,
      "grad_norm": 8.4375,
      "learning_rate": 1.7780638631964886e-06,
      "loss": 1.8736,
      "num_input_tokens_seen": 73479536,
      "step": 35170
    },
    {
      "epoch": 5.738233134839709,
      "grad_norm": 6.28125,
      "learning_rate": 1.7744293858371314e-06,
      "loss": 2.0445,
      "num_input_tokens_seen": 73489024,
      "step": 35175
    },
    {
      "epoch": 5.739048862060527,
      "grad_norm": 6.0625,
      "learning_rate": 1.770798490152631e-06,
      "loss": 2.3042,
      "num_input_tokens_seen": 73499216,
      "step": 35180
    },
    {
      "epoch": 5.7398645892813445,
      "grad_norm": 2.46875,
      "learning_rate": 1.767171176702917e-06,
      "loss": 1.2762,
      "num_input_tokens_seen": 73509232,
      "step": 35185
    },
    {
      "epoch": 5.740680316502162,
      "grad_norm": 6.96875,
      "learning_rate": 1.7635474460473755e-06,
      "loss": 2.439,
      "num_input_tokens_seen": 73518752,
      "step": 35190
    },
    {
      "epoch": 5.741496043722979,
      "grad_norm": 8.4375,
      "learning_rate": 1.7599272987448206e-06,
      "loss": 2.539,
      "num_input_tokens_seen": 73527936,
      "step": 35195
    },
    {
      "epoch": 5.742311770943797,
      "grad_norm": 5.96875,
      "learning_rate": 1.7563107353535362e-06,
      "loss": 2.6911,
      "num_input_tokens_seen": 73537984,
      "step": 35200
    },
    {
      "epoch": 5.742311770943797,
      "eval_loss": 2.546726703643799,
      "eval_runtime": 134.8088,
      "eval_samples_per_second": 20.214,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 73537984,
      "step": 35200
    },
    {
      "epoch": 5.743127498164614,
      "grad_norm": 6.3125,
      "learning_rate": 1.7526977564312263e-06,
      "loss": 3.0883,
      "num_input_tokens_seen": 73548880,
      "step": 35205
    },
    {
      "epoch": 5.743943225385431,
      "grad_norm": 8.625,
      "learning_rate": 1.7490883625350701e-06,
      "loss": 3.3915,
      "num_input_tokens_seen": 73560624,
      "step": 35210
    },
    {
      "epoch": 5.744758952606248,
      "grad_norm": 4.8125,
      "learning_rate": 1.7454825542216807e-06,
      "loss": 1.0819,
      "num_input_tokens_seen": 73571536,
      "step": 35215
    },
    {
      "epoch": 5.745574679827066,
      "grad_norm": 8.875,
      "learning_rate": 1.7418803320471105e-06,
      "loss": 2.7838,
      "num_input_tokens_seen": 73582240,
      "step": 35220
    },
    {
      "epoch": 5.746390407047883,
      "grad_norm": 6.375,
      "learning_rate": 1.7382816965668737e-06,
      "loss": 3.5861,
      "num_input_tokens_seen": 73593120,
      "step": 35225
    },
    {
      "epoch": 5.7472061342687,
      "grad_norm": 4.625,
      "learning_rate": 1.7346866483359285e-06,
      "loss": 3.349,
      "num_input_tokens_seen": 73603680,
      "step": 35230
    },
    {
      "epoch": 5.7480218614895175,
      "grad_norm": 6.53125,
      "learning_rate": 1.7310951879086657e-06,
      "loss": 2.4038,
      "num_input_tokens_seen": 73615024,
      "step": 35235
    },
    {
      "epoch": 5.748837588710336,
      "grad_norm": 6.53125,
      "learning_rate": 1.7275073158389471e-06,
      "loss": 1.5901,
      "num_input_tokens_seen": 73625232,
      "step": 35240
    },
    {
      "epoch": 5.749653315931153,
      "grad_norm": 11.8125,
      "learning_rate": 1.723923032680061e-06,
      "loss": 4.5173,
      "num_input_tokens_seen": 73636096,
      "step": 35245
    },
    {
      "epoch": 5.75046904315197,
      "grad_norm": 7.375,
      "learning_rate": 1.7203423389847428e-06,
      "loss": 2.5933,
      "num_input_tokens_seen": 73645248,
      "step": 35250
    },
    {
      "epoch": 5.751284770372787,
      "grad_norm": 5.75,
      "learning_rate": 1.7167652353051928e-06,
      "loss": 1.8978,
      "num_input_tokens_seen": 73656912,
      "step": 35255
    },
    {
      "epoch": 5.752100497593605,
      "grad_norm": 13.5,
      "learning_rate": 1.7131917221930333e-06,
      "loss": 2.2594,
      "num_input_tokens_seen": 73667040,
      "step": 35260
    },
    {
      "epoch": 5.752916224814422,
      "grad_norm": 6.75,
      "learning_rate": 1.7096218001993513e-06,
      "loss": 2.3742,
      "num_input_tokens_seen": 73677072,
      "step": 35265
    },
    {
      "epoch": 5.753731952035239,
      "grad_norm": 2.171875,
      "learning_rate": 1.706055469874676e-06,
      "loss": 2.1362,
      "num_input_tokens_seen": 73686224,
      "step": 35270
    },
    {
      "epoch": 5.754547679256056,
      "grad_norm": 12.0,
      "learning_rate": 1.702492731768976e-06,
      "loss": 2.7959,
      "num_input_tokens_seen": 73697168,
      "step": 35275
    },
    {
      "epoch": 5.755363406476874,
      "grad_norm": 9.625,
      "learning_rate": 1.6989335864316724e-06,
      "loss": 1.4453,
      "num_input_tokens_seen": 73707232,
      "step": 35280
    },
    {
      "epoch": 5.7561791336976915,
      "grad_norm": 6.125,
      "learning_rate": 1.6953780344116265e-06,
      "loss": 0.9594,
      "num_input_tokens_seen": 73718160,
      "step": 35285
    },
    {
      "epoch": 5.756994860918509,
      "grad_norm": 9.8125,
      "learning_rate": 1.6918260762571497e-06,
      "loss": 2.923,
      "num_input_tokens_seen": 73729648,
      "step": 35290
    },
    {
      "epoch": 5.757810588139327,
      "grad_norm": 12.1875,
      "learning_rate": 1.6882777125160093e-06,
      "loss": 3.301,
      "num_input_tokens_seen": 73739568,
      "step": 35295
    },
    {
      "epoch": 5.758626315360144,
      "grad_norm": 7.4375,
      "learning_rate": 1.6847329437353899e-06,
      "loss": 1.9258,
      "num_input_tokens_seen": 73749776,
      "step": 35300
    },
    {
      "epoch": 5.759442042580961,
      "grad_norm": 9.4375,
      "learning_rate": 1.6811917704619511e-06,
      "loss": 3.2924,
      "num_input_tokens_seen": 73759600,
      "step": 35305
    },
    {
      "epoch": 5.760257769801778,
      "grad_norm": 5.6875,
      "learning_rate": 1.67765419324179e-06,
      "loss": 1.3106,
      "num_input_tokens_seen": 73768608,
      "step": 35310
    },
    {
      "epoch": 5.761073497022595,
      "grad_norm": 4.875,
      "learning_rate": 1.6741202126204364e-06,
      "loss": 1.9535,
      "num_input_tokens_seen": 73779168,
      "step": 35315
    },
    {
      "epoch": 5.761889224243413,
      "grad_norm": 11.25,
      "learning_rate": 1.6705898291428767e-06,
      "loss": 3.5144,
      "num_input_tokens_seen": 73789184,
      "step": 35320
    },
    {
      "epoch": 5.76270495146423,
      "grad_norm": 4.5625,
      "learning_rate": 1.6670630433535395e-06,
      "loss": 1.8432,
      "num_input_tokens_seen": 73800160,
      "step": 35325
    },
    {
      "epoch": 5.7635206786850475,
      "grad_norm": 9.6875,
      "learning_rate": 1.6635398557962979e-06,
      "loss": 3.3577,
      "num_input_tokens_seen": 73809072,
      "step": 35330
    },
    {
      "epoch": 5.7643364059058655,
      "grad_norm": 7.0,
      "learning_rate": 1.660020267014481e-06,
      "loss": 2.4257,
      "num_input_tokens_seen": 73819504,
      "step": 35335
    },
    {
      "epoch": 5.765152133126683,
      "grad_norm": 1.8203125,
      "learning_rate": 1.6565042775508438e-06,
      "loss": 2.9431,
      "num_input_tokens_seen": 73829472,
      "step": 35340
    },
    {
      "epoch": 5.7659678603475,
      "grad_norm": 12.125,
      "learning_rate": 1.6529918879475997e-06,
      "loss": 2.6111,
      "num_input_tokens_seen": 73840576,
      "step": 35345
    },
    {
      "epoch": 5.766783587568317,
      "grad_norm": 11.375,
      "learning_rate": 1.6494830987464043e-06,
      "loss": 3.5459,
      "num_input_tokens_seen": 73850752,
      "step": 35350
    },
    {
      "epoch": 5.767599314789135,
      "grad_norm": 3.828125,
      "learning_rate": 1.6459779104883555e-06,
      "loss": 2.3998,
      "num_input_tokens_seen": 73859776,
      "step": 35355
    },
    {
      "epoch": 5.768415042009952,
      "grad_norm": 4.9375,
      "learning_rate": 1.6424763237140013e-06,
      "loss": 2.5301,
      "num_input_tokens_seen": 73869104,
      "step": 35360
    },
    {
      "epoch": 5.769230769230769,
      "grad_norm": 9.4375,
      "learning_rate": 1.6389783389633207e-06,
      "loss": 1.3885,
      "num_input_tokens_seen": 73880352,
      "step": 35365
    },
    {
      "epoch": 5.770046496451586,
      "grad_norm": 10.4375,
      "learning_rate": 1.6354839567757546e-06,
      "loss": 1.858,
      "num_input_tokens_seen": 73891024,
      "step": 35370
    },
    {
      "epoch": 5.770862223672404,
      "grad_norm": 5.9375,
      "learning_rate": 1.6319931776901831e-06,
      "loss": 1.7119,
      "num_input_tokens_seen": 73902336,
      "step": 35375
    },
    {
      "epoch": 5.771677950893221,
      "grad_norm": 5.1875,
      "learning_rate": 1.6285060022449229e-06,
      "loss": 3.0047,
      "num_input_tokens_seen": 73912832,
      "step": 35380
    },
    {
      "epoch": 5.772493678114039,
      "grad_norm": 2.375,
      "learning_rate": 1.6250224309777434e-06,
      "loss": 2.1407,
      "num_input_tokens_seen": 73924496,
      "step": 35385
    },
    {
      "epoch": 5.773309405334856,
      "grad_norm": 10.5625,
      "learning_rate": 1.6215424644258515e-06,
      "loss": 1.5998,
      "num_input_tokens_seen": 73933136,
      "step": 35390
    },
    {
      "epoch": 5.774125132555674,
      "grad_norm": 9.6875,
      "learning_rate": 1.6180661031259036e-06,
      "loss": 1.894,
      "num_input_tokens_seen": 73944048,
      "step": 35395
    },
    {
      "epoch": 5.774940859776491,
      "grad_norm": 2.421875,
      "learning_rate": 1.614593347613999e-06,
      "loss": 1.2754,
      "num_input_tokens_seen": 73955216,
      "step": 35400
    },
    {
      "epoch": 5.774940859776491,
      "eval_loss": 2.546428680419922,
      "eval_runtime": 134.659,
      "eval_samples_per_second": 20.236,
      "eval_steps_per_second": 10.122,
      "num_input_tokens_seen": 73955216,
      "step": 35400
    },
    {
      "epoch": 5.775756586997308,
      "grad_norm": 7.875,
      "learning_rate": 1.6111241984256758e-06,
      "loss": 1.4236,
      "num_input_tokens_seen": 73965456,
      "step": 35405
    },
    {
      "epoch": 5.776572314218125,
      "grad_norm": 5.375,
      "learning_rate": 1.6076586560959257e-06,
      "loss": 3.0319,
      "num_input_tokens_seen": 73976192,
      "step": 35410
    },
    {
      "epoch": 5.777388041438943,
      "grad_norm": 8.8125,
      "learning_rate": 1.604196721159182e-06,
      "loss": 3.9086,
      "num_input_tokens_seen": 73985856,
      "step": 35415
    },
    {
      "epoch": 5.77820376865976,
      "grad_norm": 3.1875,
      "learning_rate": 1.6007383941493092e-06,
      "loss": 2.2557,
      "num_input_tokens_seen": 73996528,
      "step": 35420
    },
    {
      "epoch": 5.779019495880577,
      "grad_norm": 11.5625,
      "learning_rate": 1.5972836755996285e-06,
      "loss": 3.0283,
      "num_input_tokens_seen": 74005680,
      "step": 35425
    },
    {
      "epoch": 5.7798352231013945,
      "grad_norm": 5.09375,
      "learning_rate": 1.5938325660429076e-06,
      "loss": 2.8715,
      "num_input_tokens_seen": 74016384,
      "step": 35430
    },
    {
      "epoch": 5.7806509503222125,
      "grad_norm": 7.8125,
      "learning_rate": 1.5903850660113378e-06,
      "loss": 2.968,
      "num_input_tokens_seen": 74026704,
      "step": 35435
    },
    {
      "epoch": 5.78146667754303,
      "grad_norm": 4.8125,
      "learning_rate": 1.5869411760365826e-06,
      "loss": 1.6205,
      "num_input_tokens_seen": 74036544,
      "step": 35440
    },
    {
      "epoch": 5.782282404763847,
      "grad_norm": 3.28125,
      "learning_rate": 1.58350089664972e-06,
      "loss": 1.7098,
      "num_input_tokens_seen": 74047072,
      "step": 35445
    },
    {
      "epoch": 5.783098131984664,
      "grad_norm": 5.3125,
      "learning_rate": 1.5800642283812865e-06,
      "loss": 1.7302,
      "num_input_tokens_seen": 74058192,
      "step": 35450
    },
    {
      "epoch": 5.783913859205482,
      "grad_norm": 6.21875,
      "learning_rate": 1.5766311717612698e-06,
      "loss": 1.9517,
      "num_input_tokens_seen": 74068992,
      "step": 35455
    },
    {
      "epoch": 5.784729586426299,
      "grad_norm": 2.84375,
      "learning_rate": 1.5732017273190818e-06,
      "loss": 3.5216,
      "num_input_tokens_seen": 74080048,
      "step": 35460
    },
    {
      "epoch": 5.785545313647116,
      "grad_norm": 8.3125,
      "learning_rate": 1.5697758955835806e-06,
      "loss": 2.0481,
      "num_input_tokens_seen": 74090960,
      "step": 35465
    },
    {
      "epoch": 5.786361040867934,
      "grad_norm": 7.25,
      "learning_rate": 1.566353677083085e-06,
      "loss": 2.799,
      "num_input_tokens_seen": 74101232,
      "step": 35470
    },
    {
      "epoch": 5.787176768088751,
      "grad_norm": 6.46875,
      "learning_rate": 1.562935072345334e-06,
      "loss": 2.5027,
      "num_input_tokens_seen": 74111344,
      "step": 35475
    },
    {
      "epoch": 5.7879924953095685,
      "grad_norm": 5.8125,
      "learning_rate": 1.5595200818975281e-06,
      "loss": 2.4777,
      "num_input_tokens_seen": 74121104,
      "step": 35480
    },
    {
      "epoch": 5.788808222530386,
      "grad_norm": 6.78125,
      "learning_rate": 1.5561087062662905e-06,
      "loss": 2.2983,
      "num_input_tokens_seen": 74131136,
      "step": 35485
    },
    {
      "epoch": 5.789623949751203,
      "grad_norm": 5.0625,
      "learning_rate": 1.5527009459777087e-06,
      "loss": 2.568,
      "num_input_tokens_seen": 74141280,
      "step": 35490
    },
    {
      "epoch": 5.790439676972021,
      "grad_norm": 5.125,
      "learning_rate": 1.5492968015572984e-06,
      "loss": 1.637,
      "num_input_tokens_seen": 74153408,
      "step": 35495
    },
    {
      "epoch": 5.791255404192838,
      "grad_norm": 8.75,
      "learning_rate": 1.5458962735300203e-06,
      "loss": 1.8708,
      "num_input_tokens_seen": 74165248,
      "step": 35500
    },
    {
      "epoch": 5.792071131413655,
      "grad_norm": 6.34375,
      "learning_rate": 1.54249936242028e-06,
      "loss": 3.4596,
      "num_input_tokens_seen": 74175312,
      "step": 35505
    },
    {
      "epoch": 5.792886858634473,
      "grad_norm": 2.921875,
      "learning_rate": 1.5391060687519222e-06,
      "loss": 1.6101,
      "num_input_tokens_seen": 74186016,
      "step": 35510
    },
    {
      "epoch": 5.79370258585529,
      "grad_norm": 8.75,
      "learning_rate": 1.5357163930482367e-06,
      "loss": 1.4274,
      "num_input_tokens_seen": 74196992,
      "step": 35515
    },
    {
      "epoch": 5.794518313076107,
      "grad_norm": 5.25,
      "learning_rate": 1.532330335831955e-06,
      "loss": 2.471,
      "num_input_tokens_seen": 74207792,
      "step": 35520
    },
    {
      "epoch": 5.7953340402969244,
      "grad_norm": 14.875,
      "learning_rate": 1.5289478976252491e-06,
      "loss": 2.7722,
      "num_input_tokens_seen": 74217008,
      "step": 35525
    },
    {
      "epoch": 5.796149767517742,
      "grad_norm": 6.28125,
      "learning_rate": 1.5255690789497345e-06,
      "loss": 2.5851,
      "num_input_tokens_seen": 74227120,
      "step": 35530
    },
    {
      "epoch": 5.79696549473856,
      "grad_norm": 6.75,
      "learning_rate": 1.5221938803264641e-06,
      "loss": 3.8273,
      "num_input_tokens_seen": 74238304,
      "step": 35535
    },
    {
      "epoch": 5.797781221959377,
      "grad_norm": 3.234375,
      "learning_rate": 1.518822302275938e-06,
      "loss": 1.6338,
      "num_input_tokens_seen": 74249136,
      "step": 35540
    },
    {
      "epoch": 5.798596949180194,
      "grad_norm": 5.59375,
      "learning_rate": 1.5154543453180958e-06,
      "loss": 2.5172,
      "num_input_tokens_seen": 74258720,
      "step": 35545
    },
    {
      "epoch": 5.799412676401012,
      "grad_norm": 6.25,
      "learning_rate": 1.5120900099723167e-06,
      "loss": 2.5485,
      "num_input_tokens_seen": 74269584,
      "step": 35550
    },
    {
      "epoch": 5.800228403621829,
      "grad_norm": 5.78125,
      "learning_rate": 1.5087292967574273e-06,
      "loss": 2.9919,
      "num_input_tokens_seen": 74279088,
      "step": 35555
    },
    {
      "epoch": 5.801044130842646,
      "grad_norm": 1.65625,
      "learning_rate": 1.5053722061916908e-06,
      "loss": 1.7972,
      "num_input_tokens_seen": 74289312,
      "step": 35560
    },
    {
      "epoch": 5.801859858063463,
      "grad_norm": 7.15625,
      "learning_rate": 1.5020187387928124e-06,
      "loss": 2.6373,
      "num_input_tokens_seen": 74298640,
      "step": 35565
    },
    {
      "epoch": 5.802675585284281,
      "grad_norm": 0.07958984375,
      "learning_rate": 1.4986688950779343e-06,
      "loss": 1.1566,
      "num_input_tokens_seen": 74308160,
      "step": 35570
    },
    {
      "epoch": 5.803491312505098,
      "grad_norm": 8.6875,
      "learning_rate": 1.495322675563654e-06,
      "loss": 4.2447,
      "num_input_tokens_seen": 74319088,
      "step": 35575
    },
    {
      "epoch": 5.8043070397259156,
      "grad_norm": 8.4375,
      "learning_rate": 1.4919800807659922e-06,
      "loss": 2.2337,
      "num_input_tokens_seen": 74329712,
      "step": 35580
    },
    {
      "epoch": 5.805122766946733,
      "grad_norm": 6.28125,
      "learning_rate": 1.4886411112004255e-06,
      "loss": 2.1182,
      "num_input_tokens_seen": 74340640,
      "step": 35585
    },
    {
      "epoch": 5.805938494167551,
      "grad_norm": 1.109375,
      "learning_rate": 1.4853057673818588e-06,
      "loss": 1.9646,
      "num_input_tokens_seen": 74350544,
      "step": 35590
    },
    {
      "epoch": 5.806754221388368,
      "grad_norm": 4.46875,
      "learning_rate": 1.481974049824647e-06,
      "loss": 1.7603,
      "num_input_tokens_seen": 74360256,
      "step": 35595
    },
    {
      "epoch": 5.807569948609185,
      "grad_norm": 7.6875,
      "learning_rate": 1.4786459590425849e-06,
      "loss": 1.6987,
      "num_input_tokens_seen": 74371040,
      "step": 35600
    },
    {
      "epoch": 5.807569948609185,
      "eval_loss": 2.53898549079895,
      "eval_runtime": 134.6951,
      "eval_samples_per_second": 20.231,
      "eval_steps_per_second": 10.119,
      "num_input_tokens_seen": 74371040,
      "step": 35600
    },
    {
      "epoch": 5.808385675830002,
      "grad_norm": 2.140625,
      "learning_rate": 1.4753214955489036e-06,
      "loss": 2.4927,
      "num_input_tokens_seen": 74381664,
      "step": 35605
    },
    {
      "epoch": 5.80920140305082,
      "grad_norm": 9.0625,
      "learning_rate": 1.4720006598562737e-06,
      "loss": 3.0611,
      "num_input_tokens_seen": 74392912,
      "step": 35610
    },
    {
      "epoch": 5.810017130271637,
      "grad_norm": 2.609375,
      "learning_rate": 1.4686834524768185e-06,
      "loss": 1.2846,
      "num_input_tokens_seen": 74404320,
      "step": 35615
    },
    {
      "epoch": 5.810832857492454,
      "grad_norm": 5.5625,
      "learning_rate": 1.4653698739220844e-06,
      "loss": 2.6955,
      "num_input_tokens_seen": 74413904,
      "step": 35620
    },
    {
      "epoch": 5.8116485847132715,
      "grad_norm": 5.25,
      "learning_rate": 1.4620599247030715e-06,
      "loss": 1.5636,
      "num_input_tokens_seen": 74425440,
      "step": 35625
    },
    {
      "epoch": 5.8124643119340895,
      "grad_norm": 13.0,
      "learning_rate": 1.4587536053302125e-06,
      "loss": 1.8263,
      "num_input_tokens_seen": 74435184,
      "step": 35630
    },
    {
      "epoch": 5.813280039154907,
      "grad_norm": 9.625,
      "learning_rate": 1.4554509163133862e-06,
      "loss": 2.4736,
      "num_input_tokens_seen": 74444448,
      "step": 35635
    },
    {
      "epoch": 5.814095766375724,
      "grad_norm": 8.5,
      "learning_rate": 1.4521518581619098e-06,
      "loss": 1.7657,
      "num_input_tokens_seen": 74454960,
      "step": 35640
    },
    {
      "epoch": 5.814911493596542,
      "grad_norm": 4.84375,
      "learning_rate": 1.4488564313845348e-06,
      "loss": 1.7279,
      "num_input_tokens_seen": 74464624,
      "step": 35645
    },
    {
      "epoch": 5.815727220817359,
      "grad_norm": 10.0625,
      "learning_rate": 1.4455646364894603e-06,
      "loss": 2.0595,
      "num_input_tokens_seen": 74474800,
      "step": 35650
    },
    {
      "epoch": 5.816542948038176,
      "grad_norm": 10.25,
      "learning_rate": 1.4422764739843247e-06,
      "loss": 3.1276,
      "num_input_tokens_seen": 74484224,
      "step": 35655
    },
    {
      "epoch": 5.817358675258993,
      "grad_norm": 7.09375,
      "learning_rate": 1.4389919443762e-06,
      "loss": 2.675,
      "num_input_tokens_seen": 74493856,
      "step": 35660
    },
    {
      "epoch": 5.81817440247981,
      "grad_norm": 8.3125,
      "learning_rate": 1.4357110481716063e-06,
      "loss": 3.3379,
      "num_input_tokens_seen": 74506032,
      "step": 35665
    },
    {
      "epoch": 5.818990129700628,
      "grad_norm": 7.21875,
      "learning_rate": 1.4324337858764941e-06,
      "loss": 2.5635,
      "num_input_tokens_seen": 74515440,
      "step": 35670
    },
    {
      "epoch": 5.8198058569214455,
      "grad_norm": 13.625,
      "learning_rate": 1.4291601579962622e-06,
      "loss": 2.9844,
      "num_input_tokens_seen": 74526096,
      "step": 35675
    },
    {
      "epoch": 5.820621584142263,
      "grad_norm": 9.75,
      "learning_rate": 1.42589016503574e-06,
      "loss": 4.0098,
      "num_input_tokens_seen": 74536512,
      "step": 35680
    },
    {
      "epoch": 5.821437311363081,
      "grad_norm": 10.5625,
      "learning_rate": 1.4226238074992099e-06,
      "loss": 2.8852,
      "num_input_tokens_seen": 74548288,
      "step": 35685
    },
    {
      "epoch": 5.822253038583898,
      "grad_norm": 2.921875,
      "learning_rate": 1.4193610858903778e-06,
      "loss": 1.5366,
      "num_input_tokens_seen": 74558192,
      "step": 35690
    },
    {
      "epoch": 5.823068765804715,
      "grad_norm": 7.75,
      "learning_rate": 1.416102000712402e-06,
      "loss": 2.2539,
      "num_input_tokens_seen": 74569312,
      "step": 35695
    },
    {
      "epoch": 5.823884493025532,
      "grad_norm": 5.09375,
      "learning_rate": 1.4128465524678668e-06,
      "loss": 1.938,
      "num_input_tokens_seen": 74576928,
      "step": 35700
    },
    {
      "epoch": 5.824700220246349,
      "grad_norm": 11.3125,
      "learning_rate": 1.4095947416588124e-06,
      "loss": 2.3439,
      "num_input_tokens_seen": 74587664,
      "step": 35705
    },
    {
      "epoch": 5.825515947467167,
      "grad_norm": 7.375,
      "learning_rate": 1.4063465687866983e-06,
      "loss": 1.7678,
      "num_input_tokens_seen": 74600128,
      "step": 35710
    },
    {
      "epoch": 5.826331674687984,
      "grad_norm": 6.625,
      "learning_rate": 1.4031020343524438e-06,
      "loss": 4.0562,
      "num_input_tokens_seen": 74611136,
      "step": 35715
    },
    {
      "epoch": 5.827147401908801,
      "grad_norm": 4.8125,
      "learning_rate": 1.3998611388563926e-06,
      "loss": 0.8902,
      "num_input_tokens_seen": 74623008,
      "step": 35720
    },
    {
      "epoch": 5.8279631291296194,
      "grad_norm": 7.53125,
      "learning_rate": 1.3966238827983314e-06,
      "loss": 2.96,
      "num_input_tokens_seen": 74634848,
      "step": 35725
    },
    {
      "epoch": 5.828778856350437,
      "grad_norm": 11.6875,
      "learning_rate": 1.393390266677483e-06,
      "loss": 2.5517,
      "num_input_tokens_seen": 74644288,
      "step": 35730
    },
    {
      "epoch": 5.829594583571254,
      "grad_norm": 12.75,
      "learning_rate": 1.3901602909925204e-06,
      "loss": 2.3526,
      "num_input_tokens_seen": 74656336,
      "step": 35735
    },
    {
      "epoch": 5.830410310792071,
      "grad_norm": 2.3125,
      "learning_rate": 1.3869339562415373e-06,
      "loss": 1.8461,
      "num_input_tokens_seen": 74667408,
      "step": 35740
    },
    {
      "epoch": 5.831226038012889,
      "grad_norm": 4.65625,
      "learning_rate": 1.38371126292208e-06,
      "loss": 2.423,
      "num_input_tokens_seen": 74678768,
      "step": 35745
    },
    {
      "epoch": 5.832041765233706,
      "grad_norm": 7.875,
      "learning_rate": 1.3804922115311286e-06,
      "loss": 3.2759,
      "num_input_tokens_seen": 74688944,
      "step": 35750
    },
    {
      "epoch": 5.832857492454523,
      "grad_norm": 8.125,
      "learning_rate": 1.3772768025650945e-06,
      "loss": 2.0144,
      "num_input_tokens_seen": 74698720,
      "step": 35755
    },
    {
      "epoch": 5.83367321967534,
      "grad_norm": 5.46875,
      "learning_rate": 1.3740650365198448e-06,
      "loss": 1.7589,
      "num_input_tokens_seen": 74709936,
      "step": 35760
    },
    {
      "epoch": 5.834488946896158,
      "grad_norm": 8.5625,
      "learning_rate": 1.3708569138906612e-06,
      "loss": 2.3214,
      "num_input_tokens_seen": 74719792,
      "step": 35765
    },
    {
      "epoch": 5.835304674116975,
      "grad_norm": 12.0625,
      "learning_rate": 1.367652435172287e-06,
      "loss": 3.238,
      "num_input_tokens_seen": 74731040,
      "step": 35770
    },
    {
      "epoch": 5.8361204013377925,
      "grad_norm": 3.25,
      "learning_rate": 1.364451600858893e-06,
      "loss": 1.9436,
      "num_input_tokens_seen": 74742256,
      "step": 35775
    },
    {
      "epoch": 5.83693612855861,
      "grad_norm": 11.375,
      "learning_rate": 1.3612544114440823e-06,
      "loss": 2.1525,
      "num_input_tokens_seen": 74754048,
      "step": 35780
    },
    {
      "epoch": 5.837751855779428,
      "grad_norm": 7.4375,
      "learning_rate": 1.3580608674209072e-06,
      "loss": 2.0106,
      "num_input_tokens_seen": 74763952,
      "step": 35785
    },
    {
      "epoch": 5.838567583000245,
      "grad_norm": 6.875,
      "learning_rate": 1.3548709692818434e-06,
      "loss": 3.0906,
      "num_input_tokens_seen": 74772224,
      "step": 35790
    },
    {
      "epoch": 5.839383310221062,
      "grad_norm": 6.34375,
      "learning_rate": 1.3516847175188223e-06,
      "loss": 2.9262,
      "num_input_tokens_seen": 74784368,
      "step": 35795
    },
    {
      "epoch": 5.840199037441879,
      "grad_norm": 13.875,
      "learning_rate": 1.348502112623204e-06,
      "loss": 1.9141,
      "num_input_tokens_seen": 74795680,
      "step": 35800
    },
    {
      "epoch": 5.840199037441879,
      "eval_loss": 2.5437536239624023,
      "eval_runtime": 134.7998,
      "eval_samples_per_second": 20.215,
      "eval_steps_per_second": 10.111,
      "num_input_tokens_seen": 74795680,
      "step": 35800
    },
    {
      "epoch": 5.841014764662697,
      "grad_norm": 8.0625,
      "learning_rate": 1.3453231550857787e-06,
      "loss": 2.7885,
      "num_input_tokens_seen": 74806624,
      "step": 35805
    },
    {
      "epoch": 5.841830491883514,
      "grad_norm": 12.625,
      "learning_rate": 1.3421478453967878e-06,
      "loss": 2.7153,
      "num_input_tokens_seen": 74817232,
      "step": 35810
    },
    {
      "epoch": 5.842646219104331,
      "grad_norm": 7.8125,
      "learning_rate": 1.3389761840459065e-06,
      "loss": 3.4311,
      "num_input_tokens_seen": 74827904,
      "step": 35815
    },
    {
      "epoch": 5.8434619463251485,
      "grad_norm": 7.65625,
      "learning_rate": 1.3358081715222376e-06,
      "loss": 1.8821,
      "num_input_tokens_seen": 74838080,
      "step": 35820
    },
    {
      "epoch": 5.8442776735459665,
      "grad_norm": 11.6875,
      "learning_rate": 1.3326438083143295e-06,
      "loss": 2.6483,
      "num_input_tokens_seen": 74847616,
      "step": 35825
    },
    {
      "epoch": 5.845093400766784,
      "grad_norm": 4.0,
      "learning_rate": 1.3294830949101723e-06,
      "loss": 1.7411,
      "num_input_tokens_seen": 74857664,
      "step": 35830
    },
    {
      "epoch": 5.845909127987601,
      "grad_norm": 3.8125,
      "learning_rate": 1.3263260317971815e-06,
      "loss": 0.8441,
      "num_input_tokens_seen": 74867824,
      "step": 35835
    },
    {
      "epoch": 5.846724855208418,
      "grad_norm": 3.125,
      "learning_rate": 1.3231726194622208e-06,
      "loss": 2.2185,
      "num_input_tokens_seen": 74876944,
      "step": 35840
    },
    {
      "epoch": 5.847540582429236,
      "grad_norm": 2.140625,
      "learning_rate": 1.3200228583915814e-06,
      "loss": 3.4526,
      "num_input_tokens_seen": 74887088,
      "step": 35845
    },
    {
      "epoch": 5.848356309650053,
      "grad_norm": 0.80078125,
      "learning_rate": 1.3168767490709971e-06,
      "loss": 1.35,
      "num_input_tokens_seen": 74896656,
      "step": 35850
    },
    {
      "epoch": 5.84917203687087,
      "grad_norm": 4.53125,
      "learning_rate": 1.3137342919856437e-06,
      "loss": 1.8672,
      "num_input_tokens_seen": 74906880,
      "step": 35855
    },
    {
      "epoch": 5.849987764091688,
      "grad_norm": 8.5625,
      "learning_rate": 1.310595487620117e-06,
      "loss": 1.8657,
      "num_input_tokens_seen": 74917232,
      "step": 35860
    },
    {
      "epoch": 5.850803491312505,
      "grad_norm": 14.0625,
      "learning_rate": 1.3074603364584715e-06,
      "loss": 3.4065,
      "num_input_tokens_seen": 74926736,
      "step": 35865
    },
    {
      "epoch": 5.8516192185333225,
      "grad_norm": 11.1875,
      "learning_rate": 1.3043288389841758e-06,
      "loss": 2.9325,
      "num_input_tokens_seen": 74937584,
      "step": 35870
    },
    {
      "epoch": 5.85243494575414,
      "grad_norm": 2.390625,
      "learning_rate": 1.3012009956801546e-06,
      "loss": 1.7972,
      "num_input_tokens_seen": 74947008,
      "step": 35875
    },
    {
      "epoch": 5.853250672974957,
      "grad_norm": 6.96875,
      "learning_rate": 1.2980768070287586e-06,
      "loss": 2.7758,
      "num_input_tokens_seen": 74958480,
      "step": 35880
    },
    {
      "epoch": 5.854066400195775,
      "grad_norm": 9.125,
      "learning_rate": 1.2949562735117716e-06,
      "loss": 2.6259,
      "num_input_tokens_seen": 74967792,
      "step": 35885
    },
    {
      "epoch": 5.854882127416592,
      "grad_norm": 8.4375,
      "learning_rate": 1.291839395610428e-06,
      "loss": 3.4418,
      "num_input_tokens_seen": 74977520,
      "step": 35890
    },
    {
      "epoch": 5.855697854637409,
      "grad_norm": 8.75,
      "learning_rate": 1.2887261738053852e-06,
      "loss": 3.1822,
      "num_input_tokens_seen": 74988288,
      "step": 35895
    },
    {
      "epoch": 5.856513581858227,
      "grad_norm": 6.84375,
      "learning_rate": 1.2856166085767396e-06,
      "loss": 2.0171,
      "num_input_tokens_seen": 74999408,
      "step": 35900
    },
    {
      "epoch": 5.857329309079044,
      "grad_norm": 5.0625,
      "learning_rate": 1.2825107004040272e-06,
      "loss": 1.9911,
      "num_input_tokens_seen": 75008240,
      "step": 35905
    },
    {
      "epoch": 5.858145036299861,
      "grad_norm": 8.3125,
      "learning_rate": 1.2794084497662146e-06,
      "loss": 3.1941,
      "num_input_tokens_seen": 75018480,
      "step": 35910
    },
    {
      "epoch": 5.858960763520678,
      "grad_norm": 18.125,
      "learning_rate": 1.276309857141711e-06,
      "loss": 2.8652,
      "num_input_tokens_seen": 75029088,
      "step": 35915
    },
    {
      "epoch": 5.859776490741496,
      "grad_norm": 5.90625,
      "learning_rate": 1.273214923008359e-06,
      "loss": 2.3479,
      "num_input_tokens_seen": 75040048,
      "step": 35920
    },
    {
      "epoch": 5.8605922179623136,
      "grad_norm": 4.46875,
      "learning_rate": 1.2701236478434352e-06,
      "loss": 2.7387,
      "num_input_tokens_seen": 75048928,
      "step": 35925
    },
    {
      "epoch": 5.861407945183131,
      "grad_norm": 7.125,
      "learning_rate": 1.2670360321236502e-06,
      "loss": 1.8982,
      "num_input_tokens_seen": 75060000,
      "step": 35930
    },
    {
      "epoch": 5.862223672403948,
      "grad_norm": 6.15625,
      "learning_rate": 1.2639520763251617e-06,
      "loss": 2.7139,
      "num_input_tokens_seen": 75071168,
      "step": 35935
    },
    {
      "epoch": 5.863039399624766,
      "grad_norm": 11.4375,
      "learning_rate": 1.2608717809235448e-06,
      "loss": 1.9292,
      "num_input_tokens_seen": 75082272,
      "step": 35940
    },
    {
      "epoch": 5.863855126845583,
      "grad_norm": 0.08544921875,
      "learning_rate": 1.2577951463938282e-06,
      "loss": 0.8078,
      "num_input_tokens_seen": 75094512,
      "step": 35945
    },
    {
      "epoch": 5.8646708540664,
      "grad_norm": 3.078125,
      "learning_rate": 1.2547221732104569e-06,
      "loss": 2.3352,
      "num_input_tokens_seen": 75105952,
      "step": 35950
    },
    {
      "epoch": 5.865486581287217,
      "grad_norm": 3.890625,
      "learning_rate": 1.25165286184733e-06,
      "loss": 1.6971,
      "num_input_tokens_seen": 75115920,
      "step": 35955
    },
    {
      "epoch": 5.866302308508035,
      "grad_norm": 3.21875,
      "learning_rate": 1.248587212777777e-06,
      "loss": 1.4573,
      "num_input_tokens_seen": 75126304,
      "step": 35960
    },
    {
      "epoch": 5.867118035728852,
      "grad_norm": 6.6875,
      "learning_rate": 1.2455252264745532e-06,
      "loss": 2.2793,
      "num_input_tokens_seen": 75136544,
      "step": 35965
    },
    {
      "epoch": 5.8679337629496695,
      "grad_norm": 7.53125,
      "learning_rate": 1.2424669034098528e-06,
      "loss": 2.7574,
      "num_input_tokens_seen": 75147392,
      "step": 35970
    },
    {
      "epoch": 5.868749490170487,
      "grad_norm": 9.5,
      "learning_rate": 1.2394122440553185e-06,
      "loss": 1.9751,
      "num_input_tokens_seen": 75157632,
      "step": 35975
    },
    {
      "epoch": 5.869565217391305,
      "grad_norm": 5.8125,
      "learning_rate": 1.2363612488820037e-06,
      "loss": 1.771,
      "num_input_tokens_seen": 75168016,
      "step": 35980
    },
    {
      "epoch": 5.870380944612122,
      "grad_norm": 8.4375,
      "learning_rate": 1.2333139183604208e-06,
      "loss": 2.336,
      "num_input_tokens_seen": 75178096,
      "step": 35985
    },
    {
      "epoch": 5.871196671832939,
      "grad_norm": 10.125,
      "learning_rate": 1.2302702529604998e-06,
      "loss": 2.5398,
      "num_input_tokens_seen": 75188704,
      "step": 35990
    },
    {
      "epoch": 5.872012399053756,
      "grad_norm": 5.5625,
      "learning_rate": 1.227230253151615e-06,
      "loss": 2.8719,
      "num_input_tokens_seen": 75200320,
      "step": 35995
    },
    {
      "epoch": 5.872828126274574,
      "grad_norm": 11.0625,
      "learning_rate": 1.2241939194025748e-06,
      "loss": 2.1558,
      "num_input_tokens_seen": 75209824,
      "step": 36000
    },
    {
      "epoch": 5.872828126274574,
      "eval_loss": 2.543876886367798,
      "eval_runtime": 134.69,
      "eval_samples_per_second": 20.232,
      "eval_steps_per_second": 10.12,
      "num_input_tokens_seen": 75209824,
      "step": 36000
    },
    {
      "epoch": 5.873643853495391,
      "grad_norm": 4.9375,
      "learning_rate": 1.2211612521816156e-06,
      "loss": 3.0455,
      "num_input_tokens_seen": 75220640,
      "step": 36005
    },
    {
      "epoch": 5.874459580716208,
      "grad_norm": 12.0625,
      "learning_rate": 1.2181322519564137e-06,
      "loss": 3.4559,
      "num_input_tokens_seen": 75231136,
      "step": 36010
    },
    {
      "epoch": 5.8752753079370255,
      "grad_norm": 6.4375,
      "learning_rate": 1.2151069191940839e-06,
      "loss": 2.4734,
      "num_input_tokens_seen": 75243488,
      "step": 36015
    },
    {
      "epoch": 5.8760910351578435,
      "grad_norm": 8.5625,
      "learning_rate": 1.2120852543611644e-06,
      "loss": 1.5399,
      "num_input_tokens_seen": 75255216,
      "step": 36020
    },
    {
      "epoch": 5.876906762378661,
      "grad_norm": 3.953125,
      "learning_rate": 1.2090672579236379e-06,
      "loss": 1.9293,
      "num_input_tokens_seen": 75265920,
      "step": 36025
    },
    {
      "epoch": 5.877722489599478,
      "grad_norm": 3.0625,
      "learning_rate": 1.2060529303469126e-06,
      "loss": 2.5407,
      "num_input_tokens_seen": 75275520,
      "step": 36030
    },
    {
      "epoch": 5.878538216820296,
      "grad_norm": 9.0625,
      "learning_rate": 1.2030422720958445e-06,
      "loss": 2.7294,
      "num_input_tokens_seen": 75284864,
      "step": 36035
    },
    {
      "epoch": 5.879353944041113,
      "grad_norm": 14.5,
      "learning_rate": 1.200035283634704e-06,
      "loss": 2.2513,
      "num_input_tokens_seen": 75296912,
      "step": 36040
    },
    {
      "epoch": 5.88016967126193,
      "grad_norm": 6.3125,
      "learning_rate": 1.1970319654272144e-06,
      "loss": 1.4545,
      "num_input_tokens_seen": 75306672,
      "step": 36045
    },
    {
      "epoch": 5.880985398482747,
      "grad_norm": 2.5,
      "learning_rate": 1.1940323179365192e-06,
      "loss": 1.7513,
      "num_input_tokens_seen": 75316064,
      "step": 36050
    },
    {
      "epoch": 5.881801125703564,
      "grad_norm": 6.1875,
      "learning_rate": 1.1910363416252095e-06,
      "loss": 3.5684,
      "num_input_tokens_seen": 75325568,
      "step": 36055
    },
    {
      "epoch": 5.882616852924382,
      "grad_norm": 0.05224609375,
      "learning_rate": 1.1880440369552964e-06,
      "loss": 3.5902,
      "num_input_tokens_seen": 75336288,
      "step": 36060
    },
    {
      "epoch": 5.883432580145199,
      "grad_norm": 5.5625,
      "learning_rate": 1.1850554043882328e-06,
      "loss": 2.3769,
      "num_input_tokens_seen": 75347280,
      "step": 36065
    },
    {
      "epoch": 5.884248307366017,
      "grad_norm": 6.09375,
      "learning_rate": 1.1820704443849028e-06,
      "loss": 2.9168,
      "num_input_tokens_seen": 75358304,
      "step": 36070
    },
    {
      "epoch": 5.885064034586835,
      "grad_norm": 8.5,
      "learning_rate": 1.1790891574056219e-06,
      "loss": 3.3791,
      "num_input_tokens_seen": 75369184,
      "step": 36075
    },
    {
      "epoch": 5.885879761807652,
      "grad_norm": 5.4375,
      "learning_rate": 1.1761115439101523e-06,
      "loss": 1.8295,
      "num_input_tokens_seen": 75379824,
      "step": 36080
    },
    {
      "epoch": 5.886695489028469,
      "grad_norm": 6.34375,
      "learning_rate": 1.1731376043576659e-06,
      "loss": 1.8702,
      "num_input_tokens_seen": 75390688,
      "step": 36085
    },
    {
      "epoch": 5.887511216249286,
      "grad_norm": 4.03125,
      "learning_rate": 1.1701673392067875e-06,
      "loss": 1.1793,
      "num_input_tokens_seen": 75400640,
      "step": 36090
    },
    {
      "epoch": 5.888326943470103,
      "grad_norm": 6.5625,
      "learning_rate": 1.1672007489155757e-06,
      "loss": 1.9874,
      "num_input_tokens_seen": 75411088,
      "step": 36095
    },
    {
      "epoch": 5.889142670690921,
      "grad_norm": 11.375,
      "learning_rate": 1.164237833941506e-06,
      "loss": 3.1804,
      "num_input_tokens_seen": 75422512,
      "step": 36100
    },
    {
      "epoch": 5.889958397911738,
      "grad_norm": 7.5,
      "learning_rate": 1.1612785947415022e-06,
      "loss": 2.7894,
      "num_input_tokens_seen": 75433360,
      "step": 36105
    },
    {
      "epoch": 5.890774125132555,
      "grad_norm": 5.6875,
      "learning_rate": 1.1583230317719185e-06,
      "loss": 2.1666,
      "num_input_tokens_seen": 75444400,
      "step": 36110
    },
    {
      "epoch": 5.891589852353373,
      "grad_norm": 4.5625,
      "learning_rate": 1.1553711454885318e-06,
      "loss": 2.551,
      "num_input_tokens_seen": 75454544,
      "step": 36115
    },
    {
      "epoch": 5.8924055795741905,
      "grad_norm": 3.0,
      "learning_rate": 1.152422936346567e-06,
      "loss": 0.809,
      "num_input_tokens_seen": 75464624,
      "step": 36120
    },
    {
      "epoch": 5.893221306795008,
      "grad_norm": 1.4453125,
      "learning_rate": 1.1494784048006718e-06,
      "loss": 1.795,
      "num_input_tokens_seen": 75475104,
      "step": 36125
    },
    {
      "epoch": 5.894037034015825,
      "grad_norm": 5.25,
      "learning_rate": 1.1465375513049326e-06,
      "loss": 1.7194,
      "num_input_tokens_seen": 75485472,
      "step": 36130
    },
    {
      "epoch": 5.894852761236643,
      "grad_norm": 11.625,
      "learning_rate": 1.1436003763128616e-06,
      "loss": 3.6648,
      "num_input_tokens_seen": 75496656,
      "step": 36135
    },
    {
      "epoch": 5.89566848845746,
      "grad_norm": 3.828125,
      "learning_rate": 1.1406668802774106e-06,
      "loss": 1.8796,
      "num_input_tokens_seen": 75507232,
      "step": 36140
    },
    {
      "epoch": 5.896484215678277,
      "grad_norm": 7.96875,
      "learning_rate": 1.137737063650965e-06,
      "loss": 3.2753,
      "num_input_tokens_seen": 75516800,
      "step": 36145
    },
    {
      "epoch": 5.897299942899094,
      "grad_norm": 0.1611328125,
      "learning_rate": 1.1348109268853323e-06,
      "loss": 1.5365,
      "num_input_tokens_seen": 75528800,
      "step": 36150
    },
    {
      "epoch": 5.898115670119912,
      "grad_norm": 2.84375,
      "learning_rate": 1.1318884704317634e-06,
      "loss": 2.0436,
      "num_input_tokens_seen": 75540256,
      "step": 36155
    },
    {
      "epoch": 5.898931397340729,
      "grad_norm": 6.40625,
      "learning_rate": 1.1289696947409417e-06,
      "loss": 2.0902,
      "num_input_tokens_seen": 75552112,
      "step": 36160
    },
    {
      "epoch": 5.8997471245615465,
      "grad_norm": 9.375,
      "learning_rate": 1.126054600262974e-06,
      "loss": 3.0942,
      "num_input_tokens_seen": 75561888,
      "step": 36165
    },
    {
      "epoch": 5.900562851782364,
      "grad_norm": 6.0,
      "learning_rate": 1.1231431874474064e-06,
      "loss": 1.3848,
      "num_input_tokens_seen": 75571888,
      "step": 36170
    },
    {
      "epoch": 5.901378579003182,
      "grad_norm": 4.8125,
      "learning_rate": 1.12023545674321e-06,
      "loss": 2.1814,
      "num_input_tokens_seen": 75581872,
      "step": 36175
    },
    {
      "epoch": 5.902194306223999,
      "grad_norm": 9.375,
      "learning_rate": 1.117331408598804e-06,
      "loss": 2.7819,
      "num_input_tokens_seen": 75591600,
      "step": 36180
    },
    {
      "epoch": 5.903010033444816,
      "grad_norm": 9.5625,
      "learning_rate": 1.1144310434620191e-06,
      "loss": 2.5498,
      "num_input_tokens_seen": 75602912,
      "step": 36185
    },
    {
      "epoch": 5.903825760665633,
      "grad_norm": 7.09375,
      "learning_rate": 1.1115343617801365e-06,
      "loss": 1.4844,
      "num_input_tokens_seen": 75613792,
      "step": 36190
    },
    {
      "epoch": 5.904641487886451,
      "grad_norm": 2.71875,
      "learning_rate": 1.1086413639998515e-06,
      "loss": 2.5284,
      "num_input_tokens_seen": 75624720,
      "step": 36195
    },
    {
      "epoch": 5.905457215107268,
      "grad_norm": 11.875,
      "learning_rate": 1.1057520505673103e-06,
      "loss": 2.556,
      "num_input_tokens_seen": 75634096,
      "step": 36200
    },
    {
      "epoch": 5.905457215107268,
      "eval_loss": 2.5377860069274902,
      "eval_runtime": 134.8125,
      "eval_samples_per_second": 20.213,
      "eval_steps_per_second": 10.11,
      "num_input_tokens_seen": 75634096,
      "step": 36200
    },
    {
      "epoch": 5.906272942328085,
      "grad_norm": 5.5625,
      "learning_rate": 1.1028664219280727e-06,
      "loss": 2.7059,
      "num_input_tokens_seen": 75644848,
      "step": 36205
    },
    {
      "epoch": 5.907088669548903,
      "grad_norm": 2.953125,
      "learning_rate": 1.0999844785271468e-06,
      "loss": 2.6724,
      "num_input_tokens_seen": 75654928,
      "step": 36210
    },
    {
      "epoch": 5.9079043967697205,
      "grad_norm": 9.3125,
      "learning_rate": 1.097106220808955e-06,
      "loss": 3.5023,
      "num_input_tokens_seen": 75663952,
      "step": 36215
    },
    {
      "epoch": 5.908720123990538,
      "grad_norm": 11.3125,
      "learning_rate": 1.0942316492173698e-06,
      "loss": 2.9748,
      "num_input_tokens_seen": 75674032,
      "step": 36220
    },
    {
      "epoch": 5.909535851211355,
      "grad_norm": 8.375,
      "learning_rate": 1.0913607641956841e-06,
      "loss": 2.7445,
      "num_input_tokens_seen": 75684448,
      "step": 36225
    },
    {
      "epoch": 5.910351578432172,
      "grad_norm": 2.203125,
      "learning_rate": 1.0884935661866213e-06,
      "loss": 1.8918,
      "num_input_tokens_seen": 75694512,
      "step": 36230
    },
    {
      "epoch": 5.91116730565299,
      "grad_norm": 2.296875,
      "learning_rate": 1.0856300556323418e-06,
      "loss": 1.2326,
      "num_input_tokens_seen": 75705120,
      "step": 36235
    },
    {
      "epoch": 5.911983032873807,
      "grad_norm": 0.18359375,
      "learning_rate": 1.0827702329744365e-06,
      "loss": 2.3322,
      "num_input_tokens_seen": 75714496,
      "step": 36240
    },
    {
      "epoch": 5.912798760094624,
      "grad_norm": 14.25,
      "learning_rate": 1.0799140986539197e-06,
      "loss": 2.5611,
      "num_input_tokens_seen": 75726192,
      "step": 36245
    },
    {
      "epoch": 5.913614487315442,
      "grad_norm": 5.0625,
      "learning_rate": 1.0770616531112526e-06,
      "loss": 2.3584,
      "num_input_tokens_seen": 75736304,
      "step": 36250
    },
    {
      "epoch": 5.914430214536259,
      "grad_norm": 5.9375,
      "learning_rate": 1.0742128967863085e-06,
      "loss": 2.3596,
      "num_input_tokens_seen": 75745488,
      "step": 36255
    },
    {
      "epoch": 5.915245941757076,
      "grad_norm": 5.65625,
      "learning_rate": 1.071367830118411e-06,
      "loss": 1.9459,
      "num_input_tokens_seen": 75756176,
      "step": 36260
    },
    {
      "epoch": 5.9160616689778935,
      "grad_norm": 8.125,
      "learning_rate": 1.068526453546298e-06,
      "loss": 1.7194,
      "num_input_tokens_seen": 75767296,
      "step": 36265
    },
    {
      "epoch": 5.916877396198711,
      "grad_norm": 8.625,
      "learning_rate": 1.0656887675081467e-06,
      "loss": 2.5359,
      "num_input_tokens_seen": 75778384,
      "step": 36270
    },
    {
      "epoch": 5.917693123419529,
      "grad_norm": 7.8125,
      "learning_rate": 1.0628547724415628e-06,
      "loss": 1.9387,
      "num_input_tokens_seen": 75787296,
      "step": 36275
    },
    {
      "epoch": 5.918508850640346,
      "grad_norm": 2.6875,
      "learning_rate": 1.0600244687835881e-06,
      "loss": 1.1133,
      "num_input_tokens_seen": 75797280,
      "step": 36280
    },
    {
      "epoch": 5.919324577861163,
      "grad_norm": 4.875,
      "learning_rate": 1.0571978569706876e-06,
      "loss": 1.4102,
      "num_input_tokens_seen": 75807408,
      "step": 36285
    },
    {
      "epoch": 5.920140305081981,
      "grad_norm": 8.8125,
      "learning_rate": 1.0543749374387652e-06,
      "loss": 1.9541,
      "num_input_tokens_seen": 75816800,
      "step": 36290
    },
    {
      "epoch": 5.920956032302798,
      "grad_norm": 10.125,
      "learning_rate": 1.051555710623142e-06,
      "loss": 2.1324,
      "num_input_tokens_seen": 75827808,
      "step": 36295
    },
    {
      "epoch": 5.921771759523615,
      "grad_norm": 13.125,
      "learning_rate": 1.0487401769585847e-06,
      "loss": 4.4146,
      "num_input_tokens_seen": 75839584,
      "step": 36300
    },
    {
      "epoch": 5.922587486744432,
      "grad_norm": 5.84375,
      "learning_rate": 1.0459283368792845e-06,
      "loss": 1.9754,
      "num_input_tokens_seen": 75849248,
      "step": 36305
    },
    {
      "epoch": 5.92340321396525,
      "grad_norm": 4.875,
      "learning_rate": 1.043120190818858e-06,
      "loss": 2.6273,
      "num_input_tokens_seen": 75860112,
      "step": 36310
    },
    {
      "epoch": 5.9242189411860675,
      "grad_norm": 5.875,
      "learning_rate": 1.0403157392103596e-06,
      "loss": 2.5857,
      "num_input_tokens_seen": 75871200,
      "step": 36315
    },
    {
      "epoch": 5.925034668406885,
      "grad_norm": 7.03125,
      "learning_rate": 1.0375149824862735e-06,
      "loss": 2.8168,
      "num_input_tokens_seen": 75882256,
      "step": 36320
    },
    {
      "epoch": 5.925850395627702,
      "grad_norm": 2.828125,
      "learning_rate": 1.034717921078507e-06,
      "loss": 2.4316,
      "num_input_tokens_seen": 75893184,
      "step": 36325
    },
    {
      "epoch": 5.92666612284852,
      "grad_norm": 8.375,
      "learning_rate": 1.0319245554184009e-06,
      "loss": 3.6432,
      "num_input_tokens_seen": 75902352,
      "step": 36330
    },
    {
      "epoch": 5.927481850069337,
      "grad_norm": 7.3125,
      "learning_rate": 1.0291348859367361e-06,
      "loss": 1.7434,
      "num_input_tokens_seen": 75913216,
      "step": 36335
    },
    {
      "epoch": 5.928297577290154,
      "grad_norm": 6.9375,
      "learning_rate": 1.0263489130637016e-06,
      "loss": 1.9111,
      "num_input_tokens_seen": 75924304,
      "step": 36340
    },
    {
      "epoch": 5.929113304510971,
      "grad_norm": 10.875,
      "learning_rate": 1.0235666372289427e-06,
      "loss": 1.6741,
      "num_input_tokens_seen": 75934400,
      "step": 36345
    },
    {
      "epoch": 5.929929031731789,
      "grad_norm": 7.625,
      "learning_rate": 1.0207880588615076e-06,
      "loss": 2.5774,
      "num_input_tokens_seen": 75943616,
      "step": 36350
    },
    {
      "epoch": 5.930744758952606,
      "grad_norm": 10.0,
      "learning_rate": 1.0180131783898984e-06,
      "loss": 1.9461,
      "num_input_tokens_seen": 75954736,
      "step": 36355
    },
    {
      "epoch": 5.9315604861734235,
      "grad_norm": 6.125,
      "learning_rate": 1.0152419962420362e-06,
      "loss": 1.9901,
      "num_input_tokens_seen": 75965424,
      "step": 36360
    },
    {
      "epoch": 5.932376213394241,
      "grad_norm": 8.6875,
      "learning_rate": 1.0124745128452685e-06,
      "loss": 1.7863,
      "num_input_tokens_seen": 75975680,
      "step": 36365
    },
    {
      "epoch": 5.933191940615059,
      "grad_norm": 3.0,
      "learning_rate": 1.0097107286263758e-06,
      "loss": 1.6795,
      "num_input_tokens_seen": 75985360,
      "step": 36370
    },
    {
      "epoch": 5.934007667835876,
      "grad_norm": 0.77734375,
      "learning_rate": 1.00695064401157e-06,
      "loss": 2.202,
      "num_input_tokens_seen": 75994256,
      "step": 36375
    },
    {
      "epoch": 5.934823395056693,
      "grad_norm": 15.375,
      "learning_rate": 1.0041942594264886e-06,
      "loss": 2.1379,
      "num_input_tokens_seen": 76004416,
      "step": 36380
    },
    {
      "epoch": 5.935639122277511,
      "grad_norm": 3.515625,
      "learning_rate": 1.001441575296208e-06,
      "loss": 2.2854,
      "num_input_tokens_seen": 76013728,
      "step": 36385
    },
    {
      "epoch": 5.936454849498328,
      "grad_norm": 7.96875,
      "learning_rate": 9.986925920452139e-07,
      "loss": 1.5516,
      "num_input_tokens_seen": 76025168,
      "step": 36390
    },
    {
      "epoch": 5.937270576719145,
      "grad_norm": 5.625,
      "learning_rate": 9.959473100974475e-07,
      "loss": 2.4595,
      "num_input_tokens_seen": 76035776,
      "step": 36395
    },
    {
      "epoch": 5.938086303939962,
      "grad_norm": 8.1875,
      "learning_rate": 9.932057298762564e-07,
      "loss": 1.1027,
      "num_input_tokens_seen": 76046144,
      "step": 36400
    },
    {
      "epoch": 5.938086303939962,
      "eval_loss": 2.538872003555298,
      "eval_runtime": 134.7728,
      "eval_samples_per_second": 20.219,
      "eval_steps_per_second": 10.113,
      "num_input_tokens_seen": 76046144,
      "step": 36400
    },
    {
      "epoch": 5.938902031160779,
      "grad_norm": 4.34375,
      "learning_rate": 9.90467851804433e-07,
      "loss": 3.117,
      "num_input_tokens_seen": 76057040,
      "step": 36405
    },
    {
      "epoch": 5.939717758381597,
      "grad_norm": 10.25,
      "learning_rate": 9.877336763041895e-07,
      "loss": 4.3023,
      "num_input_tokens_seen": 76066784,
      "step": 36410
    },
    {
      "epoch": 5.940533485602415,
      "grad_norm": 5.5625,
      "learning_rate": 9.850032037971662e-07,
      "loss": 1.2294,
      "num_input_tokens_seen": 76076880,
      "step": 36415
    },
    {
      "epoch": 5.941349212823232,
      "grad_norm": 6.4375,
      "learning_rate": 9.822764347044406e-07,
      "loss": 2.5526,
      "num_input_tokens_seen": 76086960,
      "step": 36420
    },
    {
      "epoch": 5.94216494004405,
      "grad_norm": 10.8125,
      "learning_rate": 9.795533694465175e-07,
      "loss": 1.8407,
      "num_input_tokens_seen": 76098240,
      "step": 36425
    },
    {
      "epoch": 5.942980667264867,
      "grad_norm": 9.75,
      "learning_rate": 9.768340084433197e-07,
      "loss": 3.0187,
      "num_input_tokens_seen": 76105760,
      "step": 36430
    },
    {
      "epoch": 5.943796394485684,
      "grad_norm": 10.6875,
      "learning_rate": 9.741183521142143e-07,
      "loss": 1.9054,
      "num_input_tokens_seen": 76114832,
      "step": 36435
    },
    {
      "epoch": 5.944612121706501,
      "grad_norm": 2.546875,
      "learning_rate": 9.714064008779889e-07,
      "loss": 1.9391,
      "num_input_tokens_seen": 76124800,
      "step": 36440
    },
    {
      "epoch": 5.945427848927318,
      "grad_norm": 9.125,
      "learning_rate": 9.686981551528584e-07,
      "loss": 2.9405,
      "num_input_tokens_seen": 76136336,
      "step": 36445
    },
    {
      "epoch": 5.946243576148136,
      "grad_norm": 9.75,
      "learning_rate": 9.65993615356467e-07,
      "loss": 2.5187,
      "num_input_tokens_seen": 76147744,
      "step": 36450
    },
    {
      "epoch": 5.947059303368953,
      "grad_norm": 4.71875,
      "learning_rate": 9.632927819058917e-07,
      "loss": 1.441,
      "num_input_tokens_seen": 76157040,
      "step": 36455
    },
    {
      "epoch": 5.9478750305897705,
      "grad_norm": 3.875,
      "learning_rate": 9.605956552176305e-07,
      "loss": 2.0194,
      "num_input_tokens_seen": 76167424,
      "step": 36460
    },
    {
      "epoch": 5.9486907578105885,
      "grad_norm": 4.9375,
      "learning_rate": 9.579022357076223e-07,
      "loss": 1.8622,
      "num_input_tokens_seen": 76177216,
      "step": 36465
    },
    {
      "epoch": 5.949506485031406,
      "grad_norm": 6.71875,
      "learning_rate": 9.552125237912158e-07,
      "loss": 2.2342,
      "num_input_tokens_seen": 76186768,
      "step": 36470
    },
    {
      "epoch": 5.950322212252223,
      "grad_norm": 6.46875,
      "learning_rate": 9.525265198832096e-07,
      "loss": 1.7613,
      "num_input_tokens_seen": 76197264,
      "step": 36475
    },
    {
      "epoch": 5.95113793947304,
      "grad_norm": 3.46875,
      "learning_rate": 9.498442243978112e-07,
      "loss": 2.5866,
      "num_input_tokens_seen": 76207232,
      "step": 36480
    },
    {
      "epoch": 5.951953666693858,
      "grad_norm": 6.5,
      "learning_rate": 9.471656377486649e-07,
      "loss": 1.9161,
      "num_input_tokens_seen": 76217920,
      "step": 36485
    },
    {
      "epoch": 5.952769393914675,
      "grad_norm": 4.53125,
      "learning_rate": 9.444907603488456e-07,
      "loss": 2.4496,
      "num_input_tokens_seen": 76228608,
      "step": 36490
    },
    {
      "epoch": 5.953585121135492,
      "grad_norm": 5.0625,
      "learning_rate": 9.418195926108514e-07,
      "loss": 3.4008,
      "num_input_tokens_seen": 76238656,
      "step": 36495
    },
    {
      "epoch": 5.954400848356309,
      "grad_norm": 5.1875,
      "learning_rate": 9.391521349466053e-07,
      "loss": 2.6301,
      "num_input_tokens_seen": 76247696,
      "step": 36500
    },
    {
      "epoch": 5.955216575577127,
      "grad_norm": 5.15625,
      "learning_rate": 9.364883877674758e-07,
      "loss": 2.1221,
      "num_input_tokens_seen": 76258256,
      "step": 36505
    },
    {
      "epoch": 5.9560323027979445,
      "grad_norm": 10.5,
      "learning_rate": 9.33828351484231e-07,
      "loss": 2.3932,
      "num_input_tokens_seen": 76268544,
      "step": 36510
    },
    {
      "epoch": 5.956848030018762,
      "grad_norm": 7.84375,
      "learning_rate": 9.311720265070906e-07,
      "loss": 2.5697,
      "num_input_tokens_seen": 76277440,
      "step": 36515
    },
    {
      "epoch": 5.957663757239579,
      "grad_norm": 6.78125,
      "learning_rate": 9.285194132456931e-07,
      "loss": 3.8001,
      "num_input_tokens_seen": 76288512,
      "step": 36520
    },
    {
      "epoch": 5.958479484460397,
      "grad_norm": 6.21875,
      "learning_rate": 9.258705121091032e-07,
      "loss": 2.5367,
      "num_input_tokens_seen": 76297424,
      "step": 36525
    },
    {
      "epoch": 5.959295211681214,
      "grad_norm": 9.9375,
      "learning_rate": 9.232253235058136e-07,
      "loss": 1.6011,
      "num_input_tokens_seen": 76308160,
      "step": 36530
    },
    {
      "epoch": 5.960110938902031,
      "grad_norm": 7.0625,
      "learning_rate": 9.205838478437478e-07,
      "loss": 1.4005,
      "num_input_tokens_seen": 76320032,
      "step": 36535
    },
    {
      "epoch": 5.960926666122848,
      "grad_norm": 4.78125,
      "learning_rate": 9.179460855302524e-07,
      "loss": 1.4494,
      "num_input_tokens_seen": 76330368,
      "step": 36540
    },
    {
      "epoch": 5.961742393343666,
      "grad_norm": 6.4375,
      "learning_rate": 9.153120369721046e-07,
      "loss": 2.7868,
      "num_input_tokens_seen": 76340896,
      "step": 36545
    },
    {
      "epoch": 5.962558120564483,
      "grad_norm": 11.3125,
      "learning_rate": 9.126817025755103e-07,
      "loss": 2.6861,
      "num_input_tokens_seen": 76350032,
      "step": 36550
    },
    {
      "epoch": 5.9633738477853,
      "grad_norm": 8.625,
      "learning_rate": 9.100550827460947e-07,
      "loss": 1.9918,
      "num_input_tokens_seen": 76361312,
      "step": 36555
    },
    {
      "epoch": 5.964189575006118,
      "grad_norm": 8.4375,
      "learning_rate": 9.0743217788892e-07,
      "loss": 2.1494,
      "num_input_tokens_seen": 76370128,
      "step": 36560
    },
    {
      "epoch": 5.965005302226936,
      "grad_norm": 8.9375,
      "learning_rate": 9.048129884084683e-07,
      "loss": 1.7838,
      "num_input_tokens_seen": 76379952,
      "step": 36565
    },
    {
      "epoch": 5.965821029447753,
      "grad_norm": 8.25,
      "learning_rate": 9.021975147086553e-07,
      "loss": 3.7507,
      "num_input_tokens_seen": 76390992,
      "step": 36570
    },
    {
      "epoch": 5.96663675666857,
      "grad_norm": 4.875,
      "learning_rate": 8.995857571928141e-07,
      "loss": 1.2415,
      "num_input_tokens_seen": 76402128,
      "step": 36575
    },
    {
      "epoch": 5.967452483889387,
      "grad_norm": 3.46875,
      "learning_rate": 8.969777162637139e-07,
      "loss": 3.3688,
      "num_input_tokens_seen": 76412672,
      "step": 36580
    },
    {
      "epoch": 5.968268211110205,
      "grad_norm": 10.8125,
      "learning_rate": 8.943733923235525e-07,
      "loss": 2.6664,
      "num_input_tokens_seen": 76422704,
      "step": 36585
    },
    {
      "epoch": 5.969083938331022,
      "grad_norm": 5.4375,
      "learning_rate": 8.917727857739394e-07,
      "loss": 2.5651,
      "num_input_tokens_seen": 76432464,
      "step": 36590
    },
    {
      "epoch": 5.969899665551839,
      "grad_norm": 7.59375,
      "learning_rate": 8.891758970159258e-07,
      "loss": 1.8628,
      "num_input_tokens_seen": 76443008,
      "step": 36595
    },
    {
      "epoch": 5.970715392772657,
      "grad_norm": 4.375,
      "learning_rate": 8.86582726449986e-07,
      "loss": 2.3357,
      "num_input_tokens_seen": 76453936,
      "step": 36600
    },
    {
      "epoch": 5.970715392772657,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7745,
      "eval_samples_per_second": 20.219,
      "eval_steps_per_second": 10.113,
      "num_input_tokens_seen": 76453936,
      "step": 36600
    },
    {
      "epoch": 5.971531119993474,
      "grad_norm": 6.71875,
      "learning_rate": 8.839932744760165e-07,
      "loss": 2.2442,
      "num_input_tokens_seen": 76464688,
      "step": 36605
    },
    {
      "epoch": 5.9723468472142915,
      "grad_norm": 7.3125,
      "learning_rate": 8.814075414933482e-07,
      "loss": 1.996,
      "num_input_tokens_seen": 76474640,
      "step": 36610
    },
    {
      "epoch": 5.973162574435109,
      "grad_norm": 8.125,
      "learning_rate": 8.788255279007257e-07,
      "loss": 3.8936,
      "num_input_tokens_seen": 76484944,
      "step": 36615
    },
    {
      "epoch": 5.973978301655926,
      "grad_norm": 5.75,
      "learning_rate": 8.762472340963362e-07,
      "loss": 1.64,
      "num_input_tokens_seen": 76495760,
      "step": 36620
    },
    {
      "epoch": 5.974794028876744,
      "grad_norm": 5.28125,
      "learning_rate": 8.736726604777811e-07,
      "loss": 2.0511,
      "num_input_tokens_seen": 76505008,
      "step": 36625
    },
    {
      "epoch": 5.975609756097561,
      "grad_norm": 8.5,
      "learning_rate": 8.711018074420901e-07,
      "loss": 2.9678,
      "num_input_tokens_seen": 76516160,
      "step": 36630
    },
    {
      "epoch": 5.976425483318378,
      "grad_norm": 11.75,
      "learning_rate": 8.685346753857209e-07,
      "loss": 3.3453,
      "num_input_tokens_seen": 76527360,
      "step": 36635
    },
    {
      "epoch": 5.977241210539196,
      "grad_norm": 2.15625,
      "learning_rate": 8.659712647045654e-07,
      "loss": 2.418,
      "num_input_tokens_seen": 76537792,
      "step": 36640
    },
    {
      "epoch": 5.978056937760013,
      "grad_norm": 6.90625,
      "learning_rate": 8.634115757939209e-07,
      "loss": 3.0223,
      "num_input_tokens_seen": 76548192,
      "step": 36645
    },
    {
      "epoch": 5.97887266498083,
      "grad_norm": 7.34375,
      "learning_rate": 8.608556090485387e-07,
      "loss": 2.4447,
      "num_input_tokens_seen": 76558736,
      "step": 36650
    },
    {
      "epoch": 5.9796883922016475,
      "grad_norm": 9.5625,
      "learning_rate": 8.583033648625671e-07,
      "loss": 2.1453,
      "num_input_tokens_seen": 76569808,
      "step": 36655
    },
    {
      "epoch": 5.9805041194224655,
      "grad_norm": 11.6875,
      "learning_rate": 8.557548436295998e-07,
      "loss": 3.1722,
      "num_input_tokens_seen": 76580848,
      "step": 36660
    },
    {
      "epoch": 5.981319846643283,
      "grad_norm": 6.375,
      "learning_rate": 8.532100457426556e-07,
      "loss": 2.6172,
      "num_input_tokens_seen": 76591184,
      "step": 36665
    },
    {
      "epoch": 5.9821355738641,
      "grad_norm": 19.75,
      "learning_rate": 8.506689715941679e-07,
      "loss": 3.0932,
      "num_input_tokens_seen": 76600016,
      "step": 36670
    },
    {
      "epoch": 5.982951301084917,
      "grad_norm": 6.96875,
      "learning_rate": 8.481316215760011e-07,
      "loss": 2.2336,
      "num_input_tokens_seen": 76611648,
      "step": 36675
    },
    {
      "epoch": 5.983767028305735,
      "grad_norm": 4.9375,
      "learning_rate": 8.455979960794558e-07,
      "loss": 2.5699,
      "num_input_tokens_seen": 76620976,
      "step": 36680
    },
    {
      "epoch": 5.984582755526552,
      "grad_norm": 7.09375,
      "learning_rate": 8.430680954952364e-07,
      "loss": 2.7929,
      "num_input_tokens_seen": 76631728,
      "step": 36685
    },
    {
      "epoch": 5.985398482747369,
      "grad_norm": 15.5625,
      "learning_rate": 8.405419202134974e-07,
      "loss": 1.7288,
      "num_input_tokens_seen": 76642432,
      "step": 36690
    },
    {
      "epoch": 5.986214209968186,
      "grad_norm": 7.90625,
      "learning_rate": 8.380194706237993e-07,
      "loss": 2.1495,
      "num_input_tokens_seen": 76653600,
      "step": 36695
    },
    {
      "epoch": 5.987029937189004,
      "grad_norm": 6.46875,
      "learning_rate": 8.355007471151366e-07,
      "loss": 2.1034,
      "num_input_tokens_seen": 76663904,
      "step": 36700
    },
    {
      "epoch": 5.9878456644098215,
      "grad_norm": 0.3203125,
      "learning_rate": 8.329857500759292e-07,
      "loss": 1.1883,
      "num_input_tokens_seen": 76674880,
      "step": 36705
    },
    {
      "epoch": 5.988661391630639,
      "grad_norm": 11.0,
      "learning_rate": 8.304744798940194e-07,
      "loss": 2.8017,
      "num_input_tokens_seen": 76685600,
      "step": 36710
    },
    {
      "epoch": 5.989477118851456,
      "grad_norm": 7.15625,
      "learning_rate": 8.279669369566756e-07,
      "loss": 3.378,
      "num_input_tokens_seen": 76695392,
      "step": 36715
    },
    {
      "epoch": 5.990292846072274,
      "grad_norm": 3.453125,
      "learning_rate": 8.254631216505993e-07,
      "loss": 3.5544,
      "num_input_tokens_seen": 76704784,
      "step": 36720
    },
    {
      "epoch": 5.991108573293091,
      "grad_norm": 6.09375,
      "learning_rate": 8.229630343619038e-07,
      "loss": 1.9652,
      "num_input_tokens_seen": 76715152,
      "step": 36725
    },
    {
      "epoch": 5.991924300513908,
      "grad_norm": 7.46875,
      "learning_rate": 8.204666754761392e-07,
      "loss": 1.965,
      "num_input_tokens_seen": 76726512,
      "step": 36730
    },
    {
      "epoch": 5.992740027734725,
      "grad_norm": 7.84375,
      "learning_rate": 8.179740453782669e-07,
      "loss": 2.1505,
      "num_input_tokens_seen": 76736720,
      "step": 36735
    },
    {
      "epoch": 5.993555754955543,
      "grad_norm": 3.0,
      "learning_rate": 8.154851444526907e-07,
      "loss": 3.4087,
      "num_input_tokens_seen": 76748640,
      "step": 36740
    },
    {
      "epoch": 5.99437148217636,
      "grad_norm": 7.125,
      "learning_rate": 8.129999730832283e-07,
      "loss": 1.96,
      "num_input_tokens_seen": 76758736,
      "step": 36745
    },
    {
      "epoch": 5.995187209397177,
      "grad_norm": 4.8125,
      "learning_rate": 8.105185316531178e-07,
      "loss": 1.8015,
      "num_input_tokens_seen": 76767664,
      "step": 36750
    },
    {
      "epoch": 5.9960029366179945,
      "grad_norm": 6.78125,
      "learning_rate": 8.08040820545039e-07,
      "loss": 2.0898,
      "num_input_tokens_seen": 76778704,
      "step": 36755
    },
    {
      "epoch": 5.996818663838813,
      "grad_norm": 3.6875,
      "learning_rate": 8.055668401410782e-07,
      "loss": 2.6966,
      "num_input_tokens_seen": 76789360,
      "step": 36760
    },
    {
      "epoch": 5.99763439105963,
      "grad_norm": 1.09375,
      "learning_rate": 8.030965908227578e-07,
      "loss": 1.8653,
      "num_input_tokens_seen": 76799728,
      "step": 36765
    },
    {
      "epoch": 5.998450118280447,
      "grad_norm": 4.46875,
      "learning_rate": 8.006300729710203e-07,
      "loss": 2.5859,
      "num_input_tokens_seen": 76809392,
      "step": 36770
    },
    {
      "epoch": 5.999265845501265,
      "grad_norm": 3.390625,
      "learning_rate": 7.981672869662337e-07,
      "loss": 0.9555,
      "num_input_tokens_seen": 76818672,
      "step": 36775
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.6875,
      "learning_rate": 7.957082331881888e-07,
      "loss": 1.3854,
      "num_input_tokens_seen": 76827984,
      "step": 36780
    },
    {
      "epoch": 6.000815727220817,
      "grad_norm": 7.875,
      "learning_rate": 7.932529120161069e-07,
      "loss": 1.995,
      "num_input_tokens_seen": 76838816,
      "step": 36785
    },
    {
      "epoch": 6.001631454441635,
      "grad_norm": 6.65625,
      "learning_rate": 7.908013238286243e-07,
      "loss": 2.7211,
      "num_input_tokens_seen": 76850432,
      "step": 36790
    },
    {
      "epoch": 6.002447181662452,
      "grad_norm": 9.9375,
      "learning_rate": 7.883534690038136e-07,
      "loss": 2.9282,
      "num_input_tokens_seen": 76862288,
      "step": 36795
    },
    {
      "epoch": 6.003262908883269,
      "grad_norm": 9.6875,
      "learning_rate": 7.859093479191559e-07,
      "loss": 1.7899,
      "num_input_tokens_seen": 76873152,
      "step": 36800
    },
    {
      "epoch": 6.003262908883269,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.714,
      "eval_samples_per_second": 20.228,
      "eval_steps_per_second": 10.118,
      "num_input_tokens_seen": 76873152,
      "step": 36800
    },
    {
      "epoch": 6.0040786361040865,
      "grad_norm": 11.125,
      "learning_rate": 7.834689609515722e-07,
      "loss": 2.9852,
      "num_input_tokens_seen": 76883280,
      "step": 36805
    },
    {
      "epoch": 6.004894363324905,
      "grad_norm": 2.015625,
      "learning_rate": 7.810323084774002e-07,
      "loss": 1.0761,
      "num_input_tokens_seen": 76893440,
      "step": 36810
    },
    {
      "epoch": 6.005710090545722,
      "grad_norm": 10.125,
      "learning_rate": 7.785993908723976e-07,
      "loss": 2.0809,
      "num_input_tokens_seen": 76903248,
      "step": 36815
    },
    {
      "epoch": 6.006525817766539,
      "grad_norm": 6.8125,
      "learning_rate": 7.761702085117534e-07,
      "loss": 1.9467,
      "num_input_tokens_seen": 76912240,
      "step": 36820
    },
    {
      "epoch": 6.007341544987356,
      "grad_norm": 15.9375,
      "learning_rate": 7.737447617700844e-07,
      "loss": 2.7685,
      "num_input_tokens_seen": 76923536,
      "step": 36825
    },
    {
      "epoch": 6.008157272208174,
      "grad_norm": 4.5625,
      "learning_rate": 7.713230510214136e-07,
      "loss": 1.8432,
      "num_input_tokens_seen": 76933200,
      "step": 36830
    },
    {
      "epoch": 6.008972999428991,
      "grad_norm": 7.1875,
      "learning_rate": 7.689050766392092e-07,
      "loss": 3.5864,
      "num_input_tokens_seen": 76941984,
      "step": 36835
    },
    {
      "epoch": 6.009788726649808,
      "grad_norm": 2.578125,
      "learning_rate": 7.664908389963477e-07,
      "loss": 2.4026,
      "num_input_tokens_seen": 76952912,
      "step": 36840
    },
    {
      "epoch": 6.010604453870625,
      "grad_norm": 1.3671875,
      "learning_rate": 7.64080338465134e-07,
      "loss": 1.7915,
      "num_input_tokens_seen": 76962512,
      "step": 36845
    },
    {
      "epoch": 6.011420181091443,
      "grad_norm": 11.25,
      "learning_rate": 7.616735754173043e-07,
      "loss": 2.6586,
      "num_input_tokens_seen": 76973392,
      "step": 36850
    },
    {
      "epoch": 6.0122359083122605,
      "grad_norm": 6.78125,
      "learning_rate": 7.592705502240005e-07,
      "loss": 2.9429,
      "num_input_tokens_seen": 76984064,
      "step": 36855
    },
    {
      "epoch": 6.013051635533078,
      "grad_norm": 10.9375,
      "learning_rate": 7.568712632558095e-07,
      "loss": 2.6769,
      "num_input_tokens_seen": 76995280,
      "step": 36860
    },
    {
      "epoch": 6.013867362753895,
      "grad_norm": 7.25,
      "learning_rate": 7.544757148827297e-07,
      "loss": 1.708,
      "num_input_tokens_seen": 77006688,
      "step": 36865
    },
    {
      "epoch": 6.014683089974713,
      "grad_norm": 3.171875,
      "learning_rate": 7.520839054741797e-07,
      "loss": 1.639,
      "num_input_tokens_seen": 77018864,
      "step": 36870
    },
    {
      "epoch": 6.01549881719553,
      "grad_norm": 10.875,
      "learning_rate": 7.496958353990113e-07,
      "loss": 1.4967,
      "num_input_tokens_seen": 77030208,
      "step": 36875
    },
    {
      "epoch": 6.016314544416347,
      "grad_norm": 9.375,
      "learning_rate": 7.473115050254941e-07,
      "loss": 2.11,
      "num_input_tokens_seen": 77041664,
      "step": 36880
    },
    {
      "epoch": 6.017130271637164,
      "grad_norm": 1.5546875,
      "learning_rate": 7.449309147213173e-07,
      "loss": 3.1034,
      "num_input_tokens_seen": 77053616,
      "step": 36885
    },
    {
      "epoch": 6.017945998857982,
      "grad_norm": 6.53125,
      "learning_rate": 7.425540648536067e-07,
      "loss": 1.8416,
      "num_input_tokens_seen": 77062064,
      "step": 36890
    },
    {
      "epoch": 6.018761726078799,
      "grad_norm": 6.875,
      "learning_rate": 7.40180955788894e-07,
      "loss": 2.0576,
      "num_input_tokens_seen": 77072272,
      "step": 36895
    },
    {
      "epoch": 6.0195774532996165,
      "grad_norm": 8.25,
      "learning_rate": 7.378115878931474e-07,
      "loss": 3.5937,
      "num_input_tokens_seen": 77082320,
      "step": 36900
    },
    {
      "epoch": 6.020393180520434,
      "grad_norm": 1.953125,
      "learning_rate": 7.354459615317527e-07,
      "loss": 2.0419,
      "num_input_tokens_seen": 77091840,
      "step": 36905
    },
    {
      "epoch": 6.021208907741252,
      "grad_norm": 0.81640625,
      "learning_rate": 7.33084077069518e-07,
      "loss": 1.2786,
      "num_input_tokens_seen": 77101872,
      "step": 36910
    },
    {
      "epoch": 6.022024634962069,
      "grad_norm": 7.65625,
      "learning_rate": 7.307259348706768e-07,
      "loss": 2.5058,
      "num_input_tokens_seen": 77110336,
      "step": 36915
    },
    {
      "epoch": 6.022840362182886,
      "grad_norm": 6.84375,
      "learning_rate": 7.283715352988801e-07,
      "loss": 2.1307,
      "num_input_tokens_seen": 77120096,
      "step": 36920
    },
    {
      "epoch": 6.023656089403703,
      "grad_norm": 4.125,
      "learning_rate": 7.260208787172068e-07,
      "loss": 1.9174,
      "num_input_tokens_seen": 77131136,
      "step": 36925
    },
    {
      "epoch": 6.024471816624521,
      "grad_norm": 5.40625,
      "learning_rate": 7.23673965488167e-07,
      "loss": 1.6429,
      "num_input_tokens_seen": 77141296,
      "step": 36930
    },
    {
      "epoch": 6.025287543845338,
      "grad_norm": 6.0625,
      "learning_rate": 7.213307959736709e-07,
      "loss": 2.6221,
      "num_input_tokens_seen": 77153008,
      "step": 36935
    },
    {
      "epoch": 6.026103271066155,
      "grad_norm": 7.75,
      "learning_rate": 7.189913705350715e-07,
      "loss": 1.4906,
      "num_input_tokens_seen": 77162848,
      "step": 36940
    },
    {
      "epoch": 6.026918998286972,
      "grad_norm": 0.98046875,
      "learning_rate": 7.166556895331411e-07,
      "loss": 2.3008,
      "num_input_tokens_seen": 77173920,
      "step": 36945
    },
    {
      "epoch": 6.02773472550779,
      "grad_norm": 3.515625,
      "learning_rate": 7.143237533280639e-07,
      "loss": 2.6978,
      "num_input_tokens_seen": 77183936,
      "step": 36950
    },
    {
      "epoch": 6.028550452728608,
      "grad_norm": 4.53125,
      "learning_rate": 7.119955622794578e-07,
      "loss": 3.0205,
      "num_input_tokens_seen": 77194256,
      "step": 36955
    },
    {
      "epoch": 6.029366179949425,
      "grad_norm": 19.125,
      "learning_rate": 7.096711167463577e-07,
      "loss": 2.1577,
      "num_input_tokens_seen": 77205296,
      "step": 36960
    },
    {
      "epoch": 6.030181907170242,
      "grad_norm": 4.96875,
      "learning_rate": 7.073504170872213e-07,
      "loss": 4.1926,
      "num_input_tokens_seen": 77216608,
      "step": 36965
    },
    {
      "epoch": 6.03099763439106,
      "grad_norm": 0.73828125,
      "learning_rate": 7.05033463659932e-07,
      "loss": 2.1809,
      "num_input_tokens_seen": 77225840,
      "step": 36970
    },
    {
      "epoch": 6.031813361611877,
      "grad_norm": 4.875,
      "learning_rate": 7.027202568217928e-07,
      "loss": 2.6806,
      "num_input_tokens_seen": 77237120,
      "step": 36975
    },
    {
      "epoch": 6.032629088832694,
      "grad_norm": 11.1875,
      "learning_rate": 7.004107969295293e-07,
      "loss": 2.6365,
      "num_input_tokens_seen": 77246736,
      "step": 36980
    },
    {
      "epoch": 6.033444816053512,
      "grad_norm": 8.625,
      "learning_rate": 6.9810508433929e-07,
      "loss": 1.5145,
      "num_input_tokens_seen": 77258144,
      "step": 36985
    },
    {
      "epoch": 6.034260543274329,
      "grad_norm": 6.625,
      "learning_rate": 6.958031194066406e-07,
      "loss": 1.7922,
      "num_input_tokens_seen": 77268016,
      "step": 36990
    },
    {
      "epoch": 6.035076270495146,
      "grad_norm": 3.875,
      "learning_rate": 6.935049024865776e-07,
      "loss": 1.9024,
      "num_input_tokens_seen": 77279504,
      "step": 36995
    },
    {
      "epoch": 6.0358919977159635,
      "grad_norm": 9.6875,
      "learning_rate": 6.912104339335118e-07,
      "loss": 3.1736,
      "num_input_tokens_seen": 77290000,
      "step": 37000
    },
    {
      "epoch": 6.0358919977159635,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7763,
      "eval_samples_per_second": 20.219,
      "eval_steps_per_second": 10.113,
      "num_input_tokens_seen": 77290000,
      "step": 37000
    },
    {
      "epoch": 6.0367077249367815,
      "grad_norm": 7.40625,
      "learning_rate": 6.889197141012799e-07,
      "loss": 3.0752,
      "num_input_tokens_seen": 77298624,
      "step": 37005
    },
    {
      "epoch": 6.037523452157599,
      "grad_norm": 7.53125,
      "learning_rate": 6.866327433431435e-07,
      "loss": 2.1342,
      "num_input_tokens_seen": 77308416,
      "step": 37010
    },
    {
      "epoch": 6.038339179378416,
      "grad_norm": 9.0,
      "learning_rate": 6.843495220117735e-07,
      "loss": 1.8866,
      "num_input_tokens_seen": 77319728,
      "step": 37015
    },
    {
      "epoch": 6.039154906599233,
      "grad_norm": 6.5625,
      "learning_rate": 6.820700504592798e-07,
      "loss": 2.0237,
      "num_input_tokens_seen": 77330864,
      "step": 37020
    },
    {
      "epoch": 6.039970633820051,
      "grad_norm": 9.6875,
      "learning_rate": 6.797943290371839e-07,
      "loss": 3.3293,
      "num_input_tokens_seen": 77341616,
      "step": 37025
    },
    {
      "epoch": 6.040786361040868,
      "grad_norm": 12.5625,
      "learning_rate": 6.775223580964274e-07,
      "loss": 2.8243,
      "num_input_tokens_seen": 77354208,
      "step": 37030
    },
    {
      "epoch": 6.041602088261685,
      "grad_norm": 8.0,
      "learning_rate": 6.7525413798738e-07,
      "loss": 3.509,
      "num_input_tokens_seen": 77365424,
      "step": 37035
    },
    {
      "epoch": 6.042417815482502,
      "grad_norm": 5.6875,
      "learning_rate": 6.729896690598259e-07,
      "loss": 1.7226,
      "num_input_tokens_seen": 77374656,
      "step": 37040
    },
    {
      "epoch": 6.04323354270332,
      "grad_norm": 5.15625,
      "learning_rate": 6.707289516629772e-07,
      "loss": 2.6891,
      "num_input_tokens_seen": 77385168,
      "step": 37045
    },
    {
      "epoch": 6.0440492699241375,
      "grad_norm": 0.95703125,
      "learning_rate": 6.684719861454692e-07,
      "loss": 1.5253,
      "num_input_tokens_seen": 77396656,
      "step": 37050
    },
    {
      "epoch": 6.044864997144955,
      "grad_norm": 6.75,
      "learning_rate": 6.662187728553481e-07,
      "loss": 1.3854,
      "num_input_tokens_seen": 77406192,
      "step": 37055
    },
    {
      "epoch": 6.045680724365772,
      "grad_norm": 4.25,
      "learning_rate": 6.639693121400892e-07,
      "loss": 2.197,
      "num_input_tokens_seen": 77418224,
      "step": 37060
    },
    {
      "epoch": 6.04649645158659,
      "grad_norm": 2.0625,
      "learning_rate": 6.617236043465868e-07,
      "loss": 1.475,
      "num_input_tokens_seen": 77429408,
      "step": 37065
    },
    {
      "epoch": 6.047312178807407,
      "grad_norm": 4.875,
      "learning_rate": 6.594816498211587e-07,
      "loss": 2.0617,
      "num_input_tokens_seen": 77439424,
      "step": 37070
    },
    {
      "epoch": 6.048127906028224,
      "grad_norm": 6.375,
      "learning_rate": 6.572434489095447e-07,
      "loss": 1.9578,
      "num_input_tokens_seen": 77450400,
      "step": 37075
    },
    {
      "epoch": 6.048943633249041,
      "grad_norm": 5.21875,
      "learning_rate": 6.550090019568994e-07,
      "loss": 1.9321,
      "num_input_tokens_seen": 77461232,
      "step": 37080
    },
    {
      "epoch": 6.049759360469859,
      "grad_norm": 9.125,
      "learning_rate": 6.527783093078027e-07,
      "loss": 0.9982,
      "num_input_tokens_seen": 77472656,
      "step": 37085
    },
    {
      "epoch": 6.050575087690676,
      "grad_norm": 8.9375,
      "learning_rate": 6.5055137130626e-07,
      "loss": 3.3248,
      "num_input_tokens_seen": 77482944,
      "step": 37090
    },
    {
      "epoch": 6.051390814911493,
      "grad_norm": 13.0,
      "learning_rate": 6.483281882956854e-07,
      "loss": 1.7543,
      "num_input_tokens_seen": 77494544,
      "step": 37095
    },
    {
      "epoch": 6.052206542132311,
      "grad_norm": 4.5,
      "learning_rate": 6.461087606189298e-07,
      "loss": 1.7041,
      "num_input_tokens_seen": 77503920,
      "step": 37100
    },
    {
      "epoch": 6.053022269353129,
      "grad_norm": 5.4375,
      "learning_rate": 6.438930886182554e-07,
      "loss": 2.4937,
      "num_input_tokens_seen": 77513344,
      "step": 37105
    },
    {
      "epoch": 6.053837996573946,
      "grad_norm": 5.3125,
      "learning_rate": 6.416811726353417e-07,
      "loss": 2.0693,
      "num_input_tokens_seen": 77523312,
      "step": 37110
    },
    {
      "epoch": 6.054653723794763,
      "grad_norm": 10.0625,
      "learning_rate": 6.394730130112991e-07,
      "loss": 2.884,
      "num_input_tokens_seen": 77533088,
      "step": 37115
    },
    {
      "epoch": 6.05546945101558,
      "grad_norm": 6.125,
      "learning_rate": 6.372686100866471e-07,
      "loss": 2.1008,
      "num_input_tokens_seen": 77542928,
      "step": 37120
    },
    {
      "epoch": 6.056285178236398,
      "grad_norm": 5.3125,
      "learning_rate": 6.350679642013413e-07,
      "loss": 1.6542,
      "num_input_tokens_seen": 77553328,
      "step": 37125
    },
    {
      "epoch": 6.057100905457215,
      "grad_norm": 13.5,
      "learning_rate": 6.328710756947437e-07,
      "loss": 1.963,
      "num_input_tokens_seen": 77563808,
      "step": 37130
    },
    {
      "epoch": 6.057916632678032,
      "grad_norm": 17.25,
      "learning_rate": 6.306779449056416e-07,
      "loss": 2.9676,
      "num_input_tokens_seen": 77574160,
      "step": 37135
    },
    {
      "epoch": 6.058732359898849,
      "grad_norm": 7.46875,
      "learning_rate": 6.284885721722422e-07,
      "loss": 2.2226,
      "num_input_tokens_seen": 77585888,
      "step": 37140
    },
    {
      "epoch": 6.059548087119667,
      "grad_norm": 2.984375,
      "learning_rate": 6.26302957832181e-07,
      "loss": 2.8989,
      "num_input_tokens_seen": 77597664,
      "step": 37145
    },
    {
      "epoch": 6.0603638143404845,
      "grad_norm": 4.59375,
      "learning_rate": 6.241211022224997e-07,
      "loss": 2.5862,
      "num_input_tokens_seen": 77608592,
      "step": 37150
    },
    {
      "epoch": 6.061179541561302,
      "grad_norm": 3.25,
      "learning_rate": 6.219430056796732e-07,
      "loss": 1.7921,
      "num_input_tokens_seen": 77618736,
      "step": 37155
    },
    {
      "epoch": 6.06199526878212,
      "grad_norm": 6.6875,
      "learning_rate": 6.19768668539586e-07,
      "loss": 3.1248,
      "num_input_tokens_seen": 77629808,
      "step": 37160
    },
    {
      "epoch": 6.062810996002937,
      "grad_norm": 7.96875,
      "learning_rate": 6.175980911375528e-07,
      "loss": 1.646,
      "num_input_tokens_seen": 77640416,
      "step": 37165
    },
    {
      "epoch": 6.063626723223754,
      "grad_norm": 11.0,
      "learning_rate": 6.154312738083034e-07,
      "loss": 2.3296,
      "num_input_tokens_seen": 77651008,
      "step": 37170
    },
    {
      "epoch": 6.064442450444571,
      "grad_norm": 10.875,
      "learning_rate": 6.132682168859843e-07,
      "loss": 2.558,
      "num_input_tokens_seen": 77660224,
      "step": 37175
    },
    {
      "epoch": 6.065258177665389,
      "grad_norm": 1.15625,
      "learning_rate": 6.111089207041704e-07,
      "loss": 0.9299,
      "num_input_tokens_seen": 77669808,
      "step": 37180
    },
    {
      "epoch": 6.066073904886206,
      "grad_norm": 6.75,
      "learning_rate": 6.089533855958507e-07,
      "loss": 3.2431,
      "num_input_tokens_seen": 77679744,
      "step": 37185
    },
    {
      "epoch": 6.066889632107023,
      "grad_norm": 12.5,
      "learning_rate": 6.068016118934372e-07,
      "loss": 3.0661,
      "num_input_tokens_seen": 77688512,
      "step": 37190
    },
    {
      "epoch": 6.0677053593278405,
      "grad_norm": 8.6875,
      "learning_rate": 6.04653599928759e-07,
      "loss": 1.8537,
      "num_input_tokens_seen": 77698336,
      "step": 37195
    },
    {
      "epoch": 6.0685210865486585,
      "grad_norm": 5.25,
      "learning_rate": 6.025093500330675e-07,
      "loss": 2.7029,
      "num_input_tokens_seen": 77708416,
      "step": 37200
    },
    {
      "epoch": 6.0685210865486585,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.755,
      "eval_samples_per_second": 20.222,
      "eval_steps_per_second": 10.115,
      "num_input_tokens_seen": 77708416,
      "step": 37200
    },
    {
      "epoch": 6.069336813769476,
      "grad_norm": 8.875,
      "learning_rate": 6.003688625370291e-07,
      "loss": 1.5637,
      "num_input_tokens_seen": 77718320,
      "step": 37205
    },
    {
      "epoch": 6.070152540990293,
      "grad_norm": 8.75,
      "learning_rate": 5.982321377707406e-07,
      "loss": 1.9735,
      "num_input_tokens_seen": 77728624,
      "step": 37210
    },
    {
      "epoch": 6.07096826821111,
      "grad_norm": 6.40625,
      "learning_rate": 5.96099176063708e-07,
      "loss": 2.1515,
      "num_input_tokens_seen": 77738368,
      "step": 37215
    },
    {
      "epoch": 6.071783995431928,
      "grad_norm": 11.1875,
      "learning_rate": 5.93969977744857e-07,
      "loss": 1.6543,
      "num_input_tokens_seen": 77749792,
      "step": 37220
    },
    {
      "epoch": 6.072599722652745,
      "grad_norm": 9.125,
      "learning_rate": 5.918445431425445e-07,
      "loss": 2.1792,
      "num_input_tokens_seen": 77759264,
      "step": 37225
    },
    {
      "epoch": 6.073415449873562,
      "grad_norm": 0.1708984375,
      "learning_rate": 5.897228725845333e-07,
      "loss": 2.9578,
      "num_input_tokens_seen": 77769888,
      "step": 37230
    },
    {
      "epoch": 6.074231177094379,
      "grad_norm": 6.375,
      "learning_rate": 5.876049663980171e-07,
      "loss": 2.6253,
      "num_input_tokens_seen": 77779120,
      "step": 37235
    },
    {
      "epoch": 6.075046904315197,
      "grad_norm": 7.71875,
      "learning_rate": 5.854908249095959e-07,
      "loss": 2.9896,
      "num_input_tokens_seen": 77788288,
      "step": 37240
    },
    {
      "epoch": 6.0758626315360145,
      "grad_norm": 6.09375,
      "learning_rate": 5.833804484453031e-07,
      "loss": 2.8922,
      "num_input_tokens_seen": 77798288,
      "step": 37245
    },
    {
      "epoch": 6.076678358756832,
      "grad_norm": 7.46875,
      "learning_rate": 5.81273837330587e-07,
      "loss": 1.3703,
      "num_input_tokens_seen": 77809072,
      "step": 37250
    },
    {
      "epoch": 6.077494085977649,
      "grad_norm": 1.6015625,
      "learning_rate": 5.791709918903071e-07,
      "loss": 0.9698,
      "num_input_tokens_seen": 77819184,
      "step": 37255
    },
    {
      "epoch": 6.078309813198467,
      "grad_norm": 6.625,
      "learning_rate": 5.770719124487483e-07,
      "loss": 2.8986,
      "num_input_tokens_seen": 77830048,
      "step": 37260
    },
    {
      "epoch": 6.079125540419284,
      "grad_norm": 2.796875,
      "learning_rate": 5.749765993296241e-07,
      "loss": 2.3397,
      "num_input_tokens_seen": 77840368,
      "step": 37265
    },
    {
      "epoch": 6.079941267640101,
      "grad_norm": 9.125,
      "learning_rate": 5.728850528560509e-07,
      "loss": 3.7692,
      "num_input_tokens_seen": 77851552,
      "step": 37270
    },
    {
      "epoch": 6.080756994860918,
      "grad_norm": 9.5,
      "learning_rate": 5.707972733505707e-07,
      "loss": 2.729,
      "num_input_tokens_seen": 77862192,
      "step": 37275
    },
    {
      "epoch": 6.081572722081736,
      "grad_norm": 3.28125,
      "learning_rate": 5.687132611351509e-07,
      "loss": 1.9384,
      "num_input_tokens_seen": 77872464,
      "step": 37280
    },
    {
      "epoch": 6.082388449302553,
      "grad_norm": 14.125,
      "learning_rate": 5.666330165311651e-07,
      "loss": 3.2284,
      "num_input_tokens_seen": 77883024,
      "step": 37285
    },
    {
      "epoch": 6.08320417652337,
      "grad_norm": 7.875,
      "learning_rate": 5.645565398594204e-07,
      "loss": 1.9556,
      "num_input_tokens_seen": 77893824,
      "step": 37290
    },
    {
      "epoch": 6.0840199037441876,
      "grad_norm": 9.0625,
      "learning_rate": 5.624838314401304e-07,
      "loss": 2.8368,
      "num_input_tokens_seen": 77903056,
      "step": 37295
    },
    {
      "epoch": 6.084835630965006,
      "grad_norm": 4.1875,
      "learning_rate": 5.604148915929336e-07,
      "loss": 2.5564,
      "num_input_tokens_seen": 77914144,
      "step": 37300
    },
    {
      "epoch": 6.085651358185823,
      "grad_norm": 4.65625,
      "learning_rate": 5.583497206368887e-07,
      "loss": 3.0775,
      "num_input_tokens_seen": 77924352,
      "step": 37305
    },
    {
      "epoch": 6.08646708540664,
      "grad_norm": 14.0,
      "learning_rate": 5.562883188904688e-07,
      "loss": 2.9592,
      "num_input_tokens_seen": 77935680,
      "step": 37310
    },
    {
      "epoch": 6.087282812627457,
      "grad_norm": 12.25,
      "learning_rate": 5.542306866715724e-07,
      "loss": 2.5899,
      "num_input_tokens_seen": 77946192,
      "step": 37315
    },
    {
      "epoch": 6.088098539848275,
      "grad_norm": 4.5,
      "learning_rate": 5.52176824297504e-07,
      "loss": 1.3172,
      "num_input_tokens_seen": 77957408,
      "step": 37320
    },
    {
      "epoch": 6.088914267069092,
      "grad_norm": 5.03125,
      "learning_rate": 5.501267320850018e-07,
      "loss": 2.8001,
      "num_input_tokens_seen": 77966896,
      "step": 37325
    },
    {
      "epoch": 6.089729994289909,
      "grad_norm": 9.875,
      "learning_rate": 5.480804103502157e-07,
      "loss": 2.823,
      "num_input_tokens_seen": 77978240,
      "step": 37330
    },
    {
      "epoch": 6.090545721510727,
      "grad_norm": 4.90625,
      "learning_rate": 5.460378594087101e-07,
      "loss": 2.3668,
      "num_input_tokens_seen": 77988144,
      "step": 37335
    },
    {
      "epoch": 6.091361448731544,
      "grad_norm": 0.62890625,
      "learning_rate": 5.439990795754773e-07,
      "loss": 1.5266,
      "num_input_tokens_seen": 77997536,
      "step": 37340
    },
    {
      "epoch": 6.0921771759523615,
      "grad_norm": 14.5,
      "learning_rate": 5.419640711649188e-07,
      "loss": 2.4564,
      "num_input_tokens_seen": 78007664,
      "step": 37345
    },
    {
      "epoch": 6.092992903173179,
      "grad_norm": 4.6875,
      "learning_rate": 5.399328344908583e-07,
      "loss": 3.1869,
      "num_input_tokens_seen": 78018256,
      "step": 37350
    },
    {
      "epoch": 6.093808630393997,
      "grad_norm": 3.125,
      "learning_rate": 5.379053698665399e-07,
      "loss": 1.4886,
      "num_input_tokens_seen": 78028672,
      "step": 37355
    },
    {
      "epoch": 6.094624357614814,
      "grad_norm": 13.0,
      "learning_rate": 5.358816776046216e-07,
      "loss": 1.144,
      "num_input_tokens_seen": 78039440,
      "step": 37360
    },
    {
      "epoch": 6.095440084835631,
      "grad_norm": 9.8125,
      "learning_rate": 5.338617580171817e-07,
      "loss": 4.3136,
      "num_input_tokens_seen": 78050848,
      "step": 37365
    },
    {
      "epoch": 6.096255812056448,
      "grad_norm": 7.15625,
      "learning_rate": 5.318456114157239e-07,
      "loss": 2.7774,
      "num_input_tokens_seen": 78060528,
      "step": 37370
    },
    {
      "epoch": 6.097071539277266,
      "grad_norm": 5.40625,
      "learning_rate": 5.298332381111576e-07,
      "loss": 1.6386,
      "num_input_tokens_seen": 78071344,
      "step": 37375
    },
    {
      "epoch": 6.097887266498083,
      "grad_norm": 6.6875,
      "learning_rate": 5.27824638413818e-07,
      "loss": 2.3291,
      "num_input_tokens_seen": 78082832,
      "step": 37380
    },
    {
      "epoch": 6.0987029937189,
      "grad_norm": 4.6875,
      "learning_rate": 5.258198126334546e-07,
      "loss": 2.0565,
      "num_input_tokens_seen": 78092896,
      "step": 37385
    },
    {
      "epoch": 6.0995187209397175,
      "grad_norm": 8.625,
      "learning_rate": 5.238187610792367e-07,
      "loss": 2.3567,
      "num_input_tokens_seen": 78103872,
      "step": 37390
    },
    {
      "epoch": 6.1003344481605355,
      "grad_norm": 4.78125,
      "learning_rate": 5.218214840597563e-07,
      "loss": 2.4276,
      "num_input_tokens_seen": 78113824,
      "step": 37395
    },
    {
      "epoch": 6.101150175381353,
      "grad_norm": 6.4375,
      "learning_rate": 5.198279818830115e-07,
      "loss": 1.6779,
      "num_input_tokens_seen": 78124432,
      "step": 37400
    },
    {
      "epoch": 6.101150175381353,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7893,
      "eval_samples_per_second": 20.217,
      "eval_steps_per_second": 10.112,
      "num_input_tokens_seen": 78124432,
      "step": 37400
    },
    {
      "epoch": 6.10196590260217,
      "grad_norm": 5.9375,
      "learning_rate": 5.178382548564287e-07,
      "loss": 1.6727,
      "num_input_tokens_seen": 78133600,
      "step": 37405
    },
    {
      "epoch": 6.102781629822987,
      "grad_norm": 5.90625,
      "learning_rate": 5.15852303286854e-07,
      "loss": 2.3785,
      "num_input_tokens_seen": 78144688,
      "step": 37410
    },
    {
      "epoch": 6.103597357043805,
      "grad_norm": 12.4375,
      "learning_rate": 5.138701274805396e-07,
      "loss": 2.6792,
      "num_input_tokens_seen": 78155968,
      "step": 37415
    },
    {
      "epoch": 6.104413084264622,
      "grad_norm": 3.71875,
      "learning_rate": 5.118917277431606e-07,
      "loss": 1.6123,
      "num_input_tokens_seen": 78165616,
      "step": 37420
    },
    {
      "epoch": 6.105228811485439,
      "grad_norm": 11.75,
      "learning_rate": 5.099171043798145e-07,
      "loss": 1.8633,
      "num_input_tokens_seen": 78175168,
      "step": 37425
    },
    {
      "epoch": 6.106044538706256,
      "grad_norm": 3.359375,
      "learning_rate": 5.079462576950133e-07,
      "loss": 2.2778,
      "num_input_tokens_seen": 78184784,
      "step": 37430
    },
    {
      "epoch": 6.106860265927074,
      "grad_norm": 2.890625,
      "learning_rate": 5.059791879926862e-07,
      "loss": 2.5979,
      "num_input_tokens_seen": 78195328,
      "step": 37435
    },
    {
      "epoch": 6.1076759931478914,
      "grad_norm": 9.4375,
      "learning_rate": 5.040158955761793e-07,
      "loss": 1.1616,
      "num_input_tokens_seen": 78206128,
      "step": 37440
    },
    {
      "epoch": 6.108491720368709,
      "grad_norm": 10.5625,
      "learning_rate": 5.020563807482559e-07,
      "loss": 3.0559,
      "num_input_tokens_seen": 78215264,
      "step": 37445
    },
    {
      "epoch": 6.109307447589526,
      "grad_norm": 10.625,
      "learning_rate": 5.001006438110995e-07,
      "loss": 2.7784,
      "num_input_tokens_seen": 78224640,
      "step": 37450
    },
    {
      "epoch": 6.110123174810344,
      "grad_norm": 9.375,
      "learning_rate": 4.981486850663075e-07,
      "loss": 1.4426,
      "num_input_tokens_seen": 78235056,
      "step": 37455
    },
    {
      "epoch": 6.110938902031161,
      "grad_norm": 3.171875,
      "learning_rate": 4.962005048149005e-07,
      "loss": 1.2803,
      "num_input_tokens_seen": 78245984,
      "step": 37460
    },
    {
      "epoch": 6.111754629251978,
      "grad_norm": 14.0625,
      "learning_rate": 4.942561033573073e-07,
      "loss": 2.209,
      "num_input_tokens_seen": 78256080,
      "step": 37465
    },
    {
      "epoch": 6.112570356472795,
      "grad_norm": 5.375,
      "learning_rate": 4.923154809933827e-07,
      "loss": 2.2469,
      "num_input_tokens_seen": 78265104,
      "step": 37470
    },
    {
      "epoch": 6.113386083693613,
      "grad_norm": 5.0625,
      "learning_rate": 4.903786380223957e-07,
      "loss": 2.0929,
      "num_input_tokens_seen": 78274992,
      "step": 37475
    },
    {
      "epoch": 6.11420181091443,
      "grad_norm": 6.21875,
      "learning_rate": 4.884455747430266e-07,
      "loss": 2.3949,
      "num_input_tokens_seen": 78286320,
      "step": 37480
    },
    {
      "epoch": 6.115017538135247,
      "grad_norm": 3.9375,
      "learning_rate": 4.865162914533816e-07,
      "loss": 2.7219,
      "num_input_tokens_seen": 78298160,
      "step": 37485
    },
    {
      "epoch": 6.1158332653560645,
      "grad_norm": 17.5,
      "learning_rate": 4.845907884509809e-07,
      "loss": 2.1082,
      "num_input_tokens_seen": 78308624,
      "step": 37490
    },
    {
      "epoch": 6.1166489925768825,
      "grad_norm": 5.125,
      "learning_rate": 4.82669066032762e-07,
      "loss": 1.5608,
      "num_input_tokens_seen": 78318560,
      "step": 37495
    },
    {
      "epoch": 6.1174647197977,
      "grad_norm": 8.625,
      "learning_rate": 4.807511244950768e-07,
      "loss": 2.5177,
      "num_input_tokens_seen": 78329296,
      "step": 37500
    },
    {
      "epoch": 6.118280447018517,
      "grad_norm": 5.65625,
      "learning_rate": 4.788369641336943e-07,
      "loss": 2.8352,
      "num_input_tokens_seen": 78339648,
      "step": 37505
    },
    {
      "epoch": 6.119096174239334,
      "grad_norm": 4.71875,
      "learning_rate": 4.769265852438032e-07,
      "loss": 2.8495,
      "num_input_tokens_seen": 78350320,
      "step": 37510
    },
    {
      "epoch": 6.119911901460152,
      "grad_norm": 2.6875,
      "learning_rate": 4.750199881200124e-07,
      "loss": 1.3317,
      "num_input_tokens_seen": 78361168,
      "step": 37515
    },
    {
      "epoch": 6.120727628680969,
      "grad_norm": 7.21875,
      "learning_rate": 4.7311717305633664e-07,
      "loss": 2.4742,
      "num_input_tokens_seen": 78371328,
      "step": 37520
    },
    {
      "epoch": 6.121543355901786,
      "grad_norm": 3.46875,
      "learning_rate": 4.7121814034621623e-07,
      "loss": 2.2901,
      "num_input_tokens_seen": 78382928,
      "step": 37525
    },
    {
      "epoch": 6.122359083122603,
      "grad_norm": 6.3125,
      "learning_rate": 4.693228902825114e-07,
      "loss": 2.4296,
      "num_input_tokens_seen": 78393872,
      "step": 37530
    },
    {
      "epoch": 6.123174810343421,
      "grad_norm": 6.6875,
      "learning_rate": 4.6743142315748277e-07,
      "loss": 2.9927,
      "num_input_tokens_seen": 78405536,
      "step": 37535
    },
    {
      "epoch": 6.1239905375642385,
      "grad_norm": 8.5,
      "learning_rate": 4.655437392628276e-07,
      "loss": 2.7864,
      "num_input_tokens_seen": 78415200,
      "step": 37540
    },
    {
      "epoch": 6.124806264785056,
      "grad_norm": 9.6875,
      "learning_rate": 4.636598388896463e-07,
      "loss": 2.411,
      "num_input_tokens_seen": 78425984,
      "step": 37545
    },
    {
      "epoch": 6.125621992005874,
      "grad_norm": 9.6875,
      "learning_rate": 4.6177972232845925e-07,
      "loss": 2.0575,
      "num_input_tokens_seen": 78436480,
      "step": 37550
    },
    {
      "epoch": 6.126437719226691,
      "grad_norm": 10.5625,
      "learning_rate": 4.5990338986920953e-07,
      "loss": 2.6513,
      "num_input_tokens_seen": 78447504,
      "step": 37555
    },
    {
      "epoch": 6.127253446447508,
      "grad_norm": 10.6875,
      "learning_rate": 4.5803084180124633e-07,
      "loss": 2.9949,
      "num_input_tokens_seen": 78457632,
      "step": 37560
    },
    {
      "epoch": 6.128069173668325,
      "grad_norm": 4.0625,
      "learning_rate": 4.561620784133386e-07,
      "loss": 1.9786,
      "num_input_tokens_seen": 78469408,
      "step": 37565
    },
    {
      "epoch": 6.128884900889143,
      "grad_norm": 7.71875,
      "learning_rate": 4.5429709999367796e-07,
      "loss": 2.0313,
      "num_input_tokens_seen": 78481472,
      "step": 37570
    },
    {
      "epoch": 6.12970062810996,
      "grad_norm": 7.59375,
      "learning_rate": 4.5243590682986223e-07,
      "loss": 3.0817,
      "num_input_tokens_seen": 78490992,
      "step": 37575
    },
    {
      "epoch": 6.130516355330777,
      "grad_norm": 6.625,
      "learning_rate": 4.5057849920891735e-07,
      "loss": 1.8336,
      "num_input_tokens_seen": 78500864,
      "step": 37580
    },
    {
      "epoch": 6.1313320825515945,
      "grad_norm": 0.34375,
      "learning_rate": 4.487248774172698e-07,
      "loss": 2.036,
      "num_input_tokens_seen": 78511696,
      "step": 37585
    },
    {
      "epoch": 6.1321478097724125,
      "grad_norm": 9.5,
      "learning_rate": 4.4687504174077965e-07,
      "loss": 1.5751,
      "num_input_tokens_seen": 78522112,
      "step": 37590
    },
    {
      "epoch": 6.13296353699323,
      "grad_norm": 15.75,
      "learning_rate": 4.450289924647133e-07,
      "loss": 3.4028,
      "num_input_tokens_seen": 78532192,
      "step": 37595
    },
    {
      "epoch": 6.133779264214047,
      "grad_norm": 7.96875,
      "learning_rate": 4.431867298737513e-07,
      "loss": 3.2998,
      "num_input_tokens_seen": 78542400,
      "step": 37600
    },
    {
      "epoch": 6.133779264214047,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7096,
      "eval_samples_per_second": 20.229,
      "eval_steps_per_second": 10.118,
      "num_input_tokens_seen": 78542400,
      "step": 37600
    },
    {
      "epoch": 6.134594991434864,
      "grad_norm": 4.90625,
      "learning_rate": 4.41348254251997e-07,
      "loss": 2.2696,
      "num_input_tokens_seen": 78552080,
      "step": 37605
    },
    {
      "epoch": 6.135410718655682,
      "grad_norm": 3.03125,
      "learning_rate": 4.395135658829652e-07,
      "loss": 0.615,
      "num_input_tokens_seen": 78561856,
      "step": 37610
    },
    {
      "epoch": 6.136226445876499,
      "grad_norm": 6.15625,
      "learning_rate": 4.376826650495852e-07,
      "loss": 2.2805,
      "num_input_tokens_seen": 78573328,
      "step": 37615
    },
    {
      "epoch": 6.137042173097316,
      "grad_norm": 8.1875,
      "learning_rate": 4.358555520342117e-07,
      "loss": 2.4129,
      "num_input_tokens_seen": 78582816,
      "step": 37620
    },
    {
      "epoch": 6.137857900318133,
      "grad_norm": 6.4375,
      "learning_rate": 4.3403222711860257e-07,
      "loss": 2.962,
      "num_input_tokens_seen": 78593616,
      "step": 37625
    },
    {
      "epoch": 6.138673627538951,
      "grad_norm": 5.71875,
      "learning_rate": 4.3221269058394133e-07,
      "loss": 2.5541,
      "num_input_tokens_seen": 78605536,
      "step": 37630
    },
    {
      "epoch": 6.139489354759768,
      "grad_norm": 2.375,
      "learning_rate": 4.303969427108173e-07,
      "loss": 2.1021,
      "num_input_tokens_seen": 78617152,
      "step": 37635
    },
    {
      "epoch": 6.1403050819805856,
      "grad_norm": 5.84375,
      "learning_rate": 4.2858498377924825e-07,
      "loss": 2.5952,
      "num_input_tokens_seen": 78629088,
      "step": 37640
    },
    {
      "epoch": 6.141120809201403,
      "grad_norm": 5.9375,
      "learning_rate": 4.267768140686579e-07,
      "loss": 0.9042,
      "num_input_tokens_seen": 78639680,
      "step": 37645
    },
    {
      "epoch": 6.141936536422221,
      "grad_norm": 13.3125,
      "learning_rate": 4.2497243385788975e-07,
      "loss": 3.7055,
      "num_input_tokens_seen": 78649680,
      "step": 37650
    },
    {
      "epoch": 6.142752263643038,
      "grad_norm": 8.75,
      "learning_rate": 4.231718434251991e-07,
      "loss": 3.3527,
      "num_input_tokens_seen": 78661200,
      "step": 37655
    },
    {
      "epoch": 6.143567990863855,
      "grad_norm": 11.3125,
      "learning_rate": 4.213750430482666e-07,
      "loss": 2.4022,
      "num_input_tokens_seen": 78671248,
      "step": 37660
    },
    {
      "epoch": 6.144383718084672,
      "grad_norm": 10.9375,
      "learning_rate": 4.1958203300417054e-07,
      "loss": 2.5857,
      "num_input_tokens_seen": 78681440,
      "step": 37665
    },
    {
      "epoch": 6.14519944530549,
      "grad_norm": 11.375,
      "learning_rate": 4.177928135694259e-07,
      "loss": 3.3705,
      "num_input_tokens_seen": 78693152,
      "step": 37670
    },
    {
      "epoch": 6.146015172526307,
      "grad_norm": 9.3125,
      "learning_rate": 4.1600738501994807e-07,
      "loss": 2.4941,
      "num_input_tokens_seen": 78703936,
      "step": 37675
    },
    {
      "epoch": 6.146830899747124,
      "grad_norm": 3.34375,
      "learning_rate": 4.1422574763107237e-07,
      "loss": 2.5601,
      "num_input_tokens_seen": 78714448,
      "step": 37680
    },
    {
      "epoch": 6.1476466269679415,
      "grad_norm": 7.1875,
      "learning_rate": 4.124479016775512e-07,
      "loss": 2.2415,
      "num_input_tokens_seen": 78725568,
      "step": 37685
    },
    {
      "epoch": 6.1484623541887595,
      "grad_norm": 11.1875,
      "learning_rate": 4.106738474335514e-07,
      "loss": 3.1499,
      "num_input_tokens_seen": 78736144,
      "step": 37690
    },
    {
      "epoch": 6.149278081409577,
      "grad_norm": 8.0625,
      "learning_rate": 4.089035851726486e-07,
      "loss": 2.6113,
      "num_input_tokens_seen": 78745888,
      "step": 37695
    },
    {
      "epoch": 6.150093808630394,
      "grad_norm": 7.53125,
      "learning_rate": 4.0713711516784937e-07,
      "loss": 3.1926,
      "num_input_tokens_seen": 78757104,
      "step": 37700
    },
    {
      "epoch": 6.150909535851211,
      "grad_norm": 4.875,
      "learning_rate": 4.05374437691558e-07,
      "loss": 1.6964,
      "num_input_tokens_seen": 78767344,
      "step": 37705
    },
    {
      "epoch": 6.151725263072029,
      "grad_norm": 9.9375,
      "learning_rate": 4.036155530156044e-07,
      "loss": 2.5212,
      "num_input_tokens_seen": 78777984,
      "step": 37710
    },
    {
      "epoch": 6.152540990292846,
      "grad_norm": 16.125,
      "learning_rate": 4.018604614112298e-07,
      "loss": 3.8754,
      "num_input_tokens_seen": 78788976,
      "step": 37715
    },
    {
      "epoch": 6.153356717513663,
      "grad_norm": 6.78125,
      "learning_rate": 4.0010916314908996e-07,
      "loss": 1.253,
      "num_input_tokens_seen": 78800800,
      "step": 37720
    },
    {
      "epoch": 6.154172444734481,
      "grad_norm": 10.8125,
      "learning_rate": 3.983616584992578e-07,
      "loss": 2.4841,
      "num_input_tokens_seen": 78811568,
      "step": 37725
    },
    {
      "epoch": 6.154988171955298,
      "grad_norm": 4.5,
      "learning_rate": 3.9661794773122595e-07,
      "loss": 2.8411,
      "num_input_tokens_seen": 78821792,
      "step": 37730
    },
    {
      "epoch": 6.1558038991761155,
      "grad_norm": 7.5,
      "learning_rate": 3.9487803111388777e-07,
      "loss": 1.961,
      "num_input_tokens_seen": 78832368,
      "step": 37735
    },
    {
      "epoch": 6.156619626396933,
      "grad_norm": 12.5625,
      "learning_rate": 3.9314190891556747e-07,
      "loss": 3.5449,
      "num_input_tokens_seen": 78842160,
      "step": 37740
    },
    {
      "epoch": 6.157435353617751,
      "grad_norm": 12.4375,
      "learning_rate": 3.914095814039925e-07,
      "loss": 2.9113,
      "num_input_tokens_seen": 78852560,
      "step": 37745
    },
    {
      "epoch": 6.158251080838568,
      "grad_norm": 7.78125,
      "learning_rate": 3.896810488463104e-07,
      "loss": 2.0581,
      "num_input_tokens_seen": 78863808,
      "step": 37750
    },
    {
      "epoch": 6.159066808059385,
      "grad_norm": 4.53125,
      "learning_rate": 3.8795631150908565e-07,
      "loss": 1.3254,
      "num_input_tokens_seen": 78875072,
      "step": 37755
    },
    {
      "epoch": 6.159882535280202,
      "grad_norm": 6.1875,
      "learning_rate": 3.862353696582888e-07,
      "loss": 2.9676,
      "num_input_tokens_seen": 78884928,
      "step": 37760
    },
    {
      "epoch": 6.16069826250102,
      "grad_norm": 11.0,
      "learning_rate": 3.8451822355931313e-07,
      "loss": 2.9994,
      "num_input_tokens_seen": 78895376,
      "step": 37765
    },
    {
      "epoch": 6.161513989721837,
      "grad_norm": 14.25,
      "learning_rate": 3.82804873476969e-07,
      "loss": 2.4806,
      "num_input_tokens_seen": 78906896,
      "step": 37770
    },
    {
      "epoch": 6.162329716942654,
      "grad_norm": 5.5,
      "learning_rate": 3.810953196754702e-07,
      "loss": 2.4156,
      "num_input_tokens_seen": 78916768,
      "step": 37775
    },
    {
      "epoch": 6.163145444163471,
      "grad_norm": 10.875,
      "learning_rate": 3.793895624184529e-07,
      "loss": 3.0739,
      "num_input_tokens_seen": 78927424,
      "step": 37780
    },
    {
      "epoch": 6.1639611713842895,
      "grad_norm": 3.296875,
      "learning_rate": 3.776876019689679e-07,
      "loss": 2.1586,
      "num_input_tokens_seen": 78938368,
      "step": 37785
    },
    {
      "epoch": 6.164776898605107,
      "grad_norm": 10.1875,
      "learning_rate": 3.7598943858947743e-07,
      "loss": 3.1265,
      "num_input_tokens_seen": 78947824,
      "step": 37790
    },
    {
      "epoch": 6.165592625825924,
      "grad_norm": 10.0625,
      "learning_rate": 3.742950725418637e-07,
      "loss": 2.5158,
      "num_input_tokens_seen": 78958352,
      "step": 37795
    },
    {
      "epoch": 6.166408353046741,
      "grad_norm": 2.609375,
      "learning_rate": 3.726045040874093e-07,
      "loss": 2.5407,
      "num_input_tokens_seen": 78968368,
      "step": 37800
    },
    {
      "epoch": 6.166408353046741,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7131,
      "eval_samples_per_second": 20.228,
      "eval_steps_per_second": 10.118,
      "num_input_tokens_seen": 78968368,
      "step": 37800
    },
    {
      "epoch": 6.167224080267559,
      "grad_norm": 9.4375,
      "learning_rate": 3.709177334868308e-07,
      "loss": 1.8345,
      "num_input_tokens_seen": 78979024,
      "step": 37805
    },
    {
      "epoch": 6.168039807488376,
      "grad_norm": 8.0625,
      "learning_rate": 3.692347610002478e-07,
      "loss": 2.3452,
      "num_input_tokens_seen": 78988768,
      "step": 37810
    },
    {
      "epoch": 6.168855534709193,
      "grad_norm": 5.65625,
      "learning_rate": 3.675555868871916e-07,
      "loss": 3.0511,
      "num_input_tokens_seen": 78998128,
      "step": 37815
    },
    {
      "epoch": 6.16967126193001,
      "grad_norm": 10.375,
      "learning_rate": 3.658802114066162e-07,
      "loss": 2.3439,
      "num_input_tokens_seen": 79009536,
      "step": 37820
    },
    {
      "epoch": 6.170486989150828,
      "grad_norm": 5.0,
      "learning_rate": 3.6420863481688437e-07,
      "loss": 1.1631,
      "num_input_tokens_seen": 79020048,
      "step": 37825
    },
    {
      "epoch": 6.171302716371645,
      "grad_norm": 10.9375,
      "learning_rate": 3.625408573757705e-07,
      "loss": 1.6322,
      "num_input_tokens_seen": 79030624,
      "step": 37830
    },
    {
      "epoch": 6.1721184435924625,
      "grad_norm": 3.921875,
      "learning_rate": 3.608768793404743e-07,
      "loss": 2.5218,
      "num_input_tokens_seen": 79040512,
      "step": 37835
    },
    {
      "epoch": 6.17293417081328,
      "grad_norm": 5.21875,
      "learning_rate": 3.592167009675934e-07,
      "loss": 2.6723,
      "num_input_tokens_seen": 79050864,
      "step": 37840
    },
    {
      "epoch": 6.173749898034098,
      "grad_norm": 8.8125,
      "learning_rate": 3.575603225131563e-07,
      "loss": 2.3568,
      "num_input_tokens_seen": 79061008,
      "step": 37845
    },
    {
      "epoch": 6.174565625254915,
      "grad_norm": 9.125,
      "learning_rate": 3.55907744232592e-07,
      "loss": 2.4561,
      "num_input_tokens_seen": 79071696,
      "step": 37850
    },
    {
      "epoch": 6.175381352475732,
      "grad_norm": 5.9375,
      "learning_rate": 3.5425896638075217e-07,
      "loss": 1.9724,
      "num_input_tokens_seen": 79081568,
      "step": 37855
    },
    {
      "epoch": 6.176197079696549,
      "grad_norm": 12.25,
      "learning_rate": 3.5261398921189736e-07,
      "loss": 2.9743,
      "num_input_tokens_seen": 79090976,
      "step": 37860
    },
    {
      "epoch": 6.177012806917367,
      "grad_norm": 10.9375,
      "learning_rate": 3.509728129797024e-07,
      "loss": 2.4955,
      "num_input_tokens_seen": 79101808,
      "step": 37865
    },
    {
      "epoch": 6.177828534138184,
      "grad_norm": 4.78125,
      "learning_rate": 3.4933543793725656e-07,
      "loss": 2.5082,
      "num_input_tokens_seen": 79112272,
      "step": 37870
    },
    {
      "epoch": 6.178644261359001,
      "grad_norm": 11.0625,
      "learning_rate": 3.4770186433707163e-07,
      "loss": 3.1335,
      "num_input_tokens_seen": 79123088,
      "step": 37875
    },
    {
      "epoch": 6.1794599885798185,
      "grad_norm": 14.1875,
      "learning_rate": 3.4607209243105453e-07,
      "loss": 2.1164,
      "num_input_tokens_seen": 79133488,
      "step": 37880
    },
    {
      "epoch": 6.1802757158006365,
      "grad_norm": 5.96875,
      "learning_rate": 3.444461224705431e-07,
      "loss": 2.5156,
      "num_input_tokens_seen": 79144224,
      "step": 37885
    },
    {
      "epoch": 6.181091443021454,
      "grad_norm": 3.921875,
      "learning_rate": 3.4282395470628116e-07,
      "loss": 1.7246,
      "num_input_tokens_seen": 79153888,
      "step": 37890
    },
    {
      "epoch": 6.181907170242271,
      "grad_norm": 5.78125,
      "learning_rate": 3.4120558938842417e-07,
      "loss": 3.2096,
      "num_input_tokens_seen": 79164576,
      "step": 37895
    },
    {
      "epoch": 6.182722897463089,
      "grad_norm": 8.0625,
      "learning_rate": 3.395910267665503e-07,
      "loss": 2.5205,
      "num_input_tokens_seen": 79175104,
      "step": 37900
    },
    {
      "epoch": 6.183538624683906,
      "grad_norm": 8.9375,
      "learning_rate": 3.3798026708964094e-07,
      "loss": 2.7483,
      "num_input_tokens_seen": 79185472,
      "step": 37905
    },
    {
      "epoch": 6.184354351904723,
      "grad_norm": 7.96875,
      "learning_rate": 3.3637331060609456e-07,
      "loss": 3.2198,
      "num_input_tokens_seen": 79196848,
      "step": 37910
    },
    {
      "epoch": 6.18517007912554,
      "grad_norm": 6.75,
      "learning_rate": 3.3477015756372966e-07,
      "loss": 1.211,
      "num_input_tokens_seen": 79206096,
      "step": 37915
    },
    {
      "epoch": 6.185985806346358,
      "grad_norm": 3.875,
      "learning_rate": 3.3317080820976785e-07,
      "loss": 2.4185,
      "num_input_tokens_seen": 79215904,
      "step": 37920
    },
    {
      "epoch": 6.186801533567175,
      "grad_norm": 6.34375,
      "learning_rate": 3.315752627908508e-07,
      "loss": 3.0412,
      "num_input_tokens_seen": 79226176,
      "step": 37925
    },
    {
      "epoch": 6.1876172607879925,
      "grad_norm": 11.0,
      "learning_rate": 3.299835215530317e-07,
      "loss": 2.7692,
      "num_input_tokens_seen": 79236272,
      "step": 37930
    },
    {
      "epoch": 6.18843298800881,
      "grad_norm": 11.0625,
      "learning_rate": 3.2839558474177245e-07,
      "loss": 3.2939,
      "num_input_tokens_seen": 79246736,
      "step": 37935
    },
    {
      "epoch": 6.189248715229628,
      "grad_norm": 6.0,
      "learning_rate": 3.2681145260196056e-07,
      "loss": 2.7709,
      "num_input_tokens_seen": 79256384,
      "step": 37940
    },
    {
      "epoch": 6.190064442450445,
      "grad_norm": 3.203125,
      "learning_rate": 3.252311253778839e-07,
      "loss": 2.5421,
      "num_input_tokens_seen": 79268032,
      "step": 37945
    },
    {
      "epoch": 6.190880169671262,
      "grad_norm": 9.0,
      "learning_rate": 3.2365460331325034e-07,
      "loss": 2.6012,
      "num_input_tokens_seen": 79279328,
      "step": 37950
    },
    {
      "epoch": 6.191695896892079,
      "grad_norm": 7.125,
      "learning_rate": 3.2208188665117934e-07,
      "loss": 1.3524,
      "num_input_tokens_seen": 79289616,
      "step": 37955
    },
    {
      "epoch": 6.192511624112897,
      "grad_norm": 5.5625,
      "learning_rate": 3.205129756342018e-07,
      "loss": 2.0632,
      "num_input_tokens_seen": 79298368,
      "step": 37960
    },
    {
      "epoch": 6.193327351333714,
      "grad_norm": 8.1875,
      "learning_rate": 3.189478705042659e-07,
      "loss": 2.1626,
      "num_input_tokens_seen": 79306544,
      "step": 37965
    },
    {
      "epoch": 6.194143078554531,
      "grad_norm": 10.9375,
      "learning_rate": 3.173865715027341e-07,
      "loss": 2.0486,
      "num_input_tokens_seen": 79316144,
      "step": 37970
    },
    {
      "epoch": 6.194958805775348,
      "grad_norm": 9.4375,
      "learning_rate": 3.158290788703694e-07,
      "loss": 3.0197,
      "num_input_tokens_seen": 79325568,
      "step": 37975
    },
    {
      "epoch": 6.195774532996166,
      "grad_norm": 5.6875,
      "learning_rate": 3.1427539284736297e-07,
      "loss": 2.3251,
      "num_input_tokens_seen": 79337152,
      "step": 37980
    },
    {
      "epoch": 6.196590260216984,
      "grad_norm": 4.84375,
      "learning_rate": 3.127255136733093e-07,
      "loss": 2.7295,
      "num_input_tokens_seen": 79346832,
      "step": 37985
    },
    {
      "epoch": 6.197405987437801,
      "grad_norm": 6.5625,
      "learning_rate": 3.1117944158722544e-07,
      "loss": 1.8806,
      "num_input_tokens_seen": 79357984,
      "step": 37990
    },
    {
      "epoch": 6.198221714658618,
      "grad_norm": 9.875,
      "learning_rate": 3.0963717682752635e-07,
      "loss": 2.1924,
      "num_input_tokens_seen": 79368224,
      "step": 37995
    },
    {
      "epoch": 6.199037441879436,
      "grad_norm": 2.390625,
      "learning_rate": 3.080987196320578e-07,
      "loss": 1.2663,
      "num_input_tokens_seen": 79378528,
      "step": 38000
    },
    {
      "epoch": 6.199037441879436,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.6952,
      "eval_samples_per_second": 20.231,
      "eval_steps_per_second": 10.119,
      "num_input_tokens_seen": 79378528,
      "step": 38000
    },
    {
      "epoch": 6.199853169100253,
      "grad_norm": 0.2255859375,
      "learning_rate": 3.065640702380607e-07,
      "loss": 1.7786,
      "num_input_tokens_seen": 79389168,
      "step": 38005
    },
    {
      "epoch": 6.20066889632107,
      "grad_norm": 2.078125,
      "learning_rate": 3.050332288822011e-07,
      "loss": 1.9805,
      "num_input_tokens_seen": 79399120,
      "step": 38010
    },
    {
      "epoch": 6.201484623541887,
      "grad_norm": 3.984375,
      "learning_rate": 3.035061958005542e-07,
      "loss": 3.0661,
      "num_input_tokens_seen": 79411024,
      "step": 38015
    },
    {
      "epoch": 6.202300350762705,
      "grad_norm": 4.53125,
      "learning_rate": 3.019829712286093e-07,
      "loss": 1.5562,
      "num_input_tokens_seen": 79422688,
      "step": 38020
    },
    {
      "epoch": 6.203116077983522,
      "grad_norm": 10.5625,
      "learning_rate": 3.004635554012647e-07,
      "loss": 2.5306,
      "num_input_tokens_seen": 79432832,
      "step": 38025
    },
    {
      "epoch": 6.2039318052043395,
      "grad_norm": 9.5,
      "learning_rate": 2.9894794855283017e-07,
      "loss": 3.0771,
      "num_input_tokens_seen": 79444064,
      "step": 38030
    },
    {
      "epoch": 6.204747532425157,
      "grad_norm": 7.3125,
      "learning_rate": 2.9743615091703816e-07,
      "loss": 1.8517,
      "num_input_tokens_seen": 79453808,
      "step": 38035
    },
    {
      "epoch": 6.205563259645975,
      "grad_norm": 7.40625,
      "learning_rate": 2.959281627270216e-07,
      "loss": 2.0853,
      "num_input_tokens_seen": 79463552,
      "step": 38040
    },
    {
      "epoch": 6.206378986866792,
      "grad_norm": 0.47265625,
      "learning_rate": 2.944239842153362e-07,
      "loss": 1.8826,
      "num_input_tokens_seen": 79474240,
      "step": 38045
    },
    {
      "epoch": 6.207194714087609,
      "grad_norm": 8.6875,
      "learning_rate": 2.929236156139381e-07,
      "loss": 2.7772,
      "num_input_tokens_seen": 79484096,
      "step": 38050
    },
    {
      "epoch": 6.208010441308426,
      "grad_norm": 9.3125,
      "learning_rate": 2.9142705715420883e-07,
      "loss": 1.9524,
      "num_input_tokens_seen": 79494608,
      "step": 38055
    },
    {
      "epoch": 6.208826168529244,
      "grad_norm": 5.75,
      "learning_rate": 2.8993430906693595e-07,
      "loss": 1.8412,
      "num_input_tokens_seen": 79505360,
      "step": 38060
    },
    {
      "epoch": 6.209641895750061,
      "grad_norm": 5.65625,
      "learning_rate": 2.88445371582316e-07,
      "loss": 1.4115,
      "num_input_tokens_seen": 79515488,
      "step": 38065
    },
    {
      "epoch": 6.210457622970878,
      "grad_norm": 5.40625,
      "learning_rate": 2.8696024492996796e-07,
      "loss": 2.8746,
      "num_input_tokens_seen": 79525840,
      "step": 38070
    },
    {
      "epoch": 6.211273350191696,
      "grad_norm": 5.625,
      "learning_rate": 2.854789293389115e-07,
      "loss": 2.7681,
      "num_input_tokens_seen": 79536112,
      "step": 38075
    },
    {
      "epoch": 6.2120890774125135,
      "grad_norm": 7.84375,
      "learning_rate": 2.8400142503758606e-07,
      "loss": 1.0938,
      "num_input_tokens_seen": 79546720,
      "step": 38080
    },
    {
      "epoch": 6.212904804633331,
      "grad_norm": 4.46875,
      "learning_rate": 2.8252773225384276e-07,
      "loss": 2.302,
      "num_input_tokens_seen": 79557936,
      "step": 38085
    },
    {
      "epoch": 6.213720531854148,
      "grad_norm": 9.625,
      "learning_rate": 2.8105785121494143e-07,
      "loss": 3.8971,
      "num_input_tokens_seen": 79569584,
      "step": 38090
    },
    {
      "epoch": 6.214536259074965,
      "grad_norm": 8.6875,
      "learning_rate": 2.795917821475563e-07,
      "loss": 2.8693,
      "num_input_tokens_seen": 79579840,
      "step": 38095
    },
    {
      "epoch": 6.215351986295783,
      "grad_norm": 6.96875,
      "learning_rate": 2.78129525277776e-07,
      "loss": 1.7988,
      "num_input_tokens_seen": 79590464,
      "step": 38100
    },
    {
      "epoch": 6.2161677135166,
      "grad_norm": 5.15625,
      "learning_rate": 2.766710808310952e-07,
      "loss": 2.9476,
      "num_input_tokens_seen": 79600848,
      "step": 38105
    },
    {
      "epoch": 6.216983440737417,
      "grad_norm": 5.125,
      "learning_rate": 2.7521644903242827e-07,
      "loss": 2.1424,
      "num_input_tokens_seen": 79611408,
      "step": 38110
    },
    {
      "epoch": 6.217799167958235,
      "grad_norm": 8.375,
      "learning_rate": 2.7376563010609593e-07,
      "loss": 2.7264,
      "num_input_tokens_seen": 79621056,
      "step": 38115
    },
    {
      "epoch": 6.218614895179052,
      "grad_norm": 7.46875,
      "learning_rate": 2.72318624275833e-07,
      "loss": 1.365,
      "num_input_tokens_seen": 79631456,
      "step": 38120
    },
    {
      "epoch": 6.219430622399869,
      "grad_norm": 3.328125,
      "learning_rate": 2.7087543176478324e-07,
      "loss": 2.493,
      "num_input_tokens_seen": 79642448,
      "step": 38125
    },
    {
      "epoch": 6.220246349620687,
      "grad_norm": 4.46875,
      "learning_rate": 2.694360527955103e-07,
      "loss": 2.0805,
      "num_input_tokens_seen": 79652016,
      "step": 38130
    },
    {
      "epoch": 6.221062076841505,
      "grad_norm": 8.125,
      "learning_rate": 2.680004875899811e-07,
      "loss": 2.3353,
      "num_input_tokens_seen": 79663264,
      "step": 38135
    },
    {
      "epoch": 6.221877804062322,
      "grad_norm": 5.96875,
      "learning_rate": 2.665687363695768e-07,
      "loss": 3.5997,
      "num_input_tokens_seen": 79673168,
      "step": 38140
    },
    {
      "epoch": 6.222693531283139,
      "grad_norm": 13.0625,
      "learning_rate": 2.6514079935509584e-07,
      "loss": 3.0249,
      "num_input_tokens_seen": 79683728,
      "step": 38145
    },
    {
      "epoch": 6.223509258503956,
      "grad_norm": 7.96875,
      "learning_rate": 2.6371667676673983e-07,
      "loss": 2.2844,
      "num_input_tokens_seen": 79694672,
      "step": 38150
    },
    {
      "epoch": 6.224324985724774,
      "grad_norm": 10.625,
      "learning_rate": 2.6229636882412755e-07,
      "loss": 2.62,
      "num_input_tokens_seen": 79705216,
      "step": 38155
    },
    {
      "epoch": 6.225140712945591,
      "grad_norm": 5.78125,
      "learning_rate": 2.6087987574628935e-07,
      "loss": 1.8438,
      "num_input_tokens_seen": 79715984,
      "step": 38160
    },
    {
      "epoch": 6.225956440166408,
      "grad_norm": 1.8046875,
      "learning_rate": 2.5946719775166437e-07,
      "loss": 1.8779,
      "num_input_tokens_seen": 79725904,
      "step": 38165
    },
    {
      "epoch": 6.226772167387225,
      "grad_norm": 1.9140625,
      "learning_rate": 2.5805833505810616e-07,
      "loss": 2.4688,
      "num_input_tokens_seen": 79736944,
      "step": 38170
    },
    {
      "epoch": 6.227587894608043,
      "grad_norm": 6.4375,
      "learning_rate": 2.566532878828798e-07,
      "loss": 1.251,
      "num_input_tokens_seen": 79748304,
      "step": 38175
    },
    {
      "epoch": 6.2284036218288605,
      "grad_norm": 6.40625,
      "learning_rate": 2.552520564426619e-07,
      "loss": 1.9093,
      "num_input_tokens_seen": 79759728,
      "step": 38180
    },
    {
      "epoch": 6.229219349049678,
      "grad_norm": 10.5,
      "learning_rate": 2.5385464095353803e-07,
      "loss": 3.637,
      "num_input_tokens_seen": 79769776,
      "step": 38185
    },
    {
      "epoch": 6.230035076270495,
      "grad_norm": 6.3125,
      "learning_rate": 2.5246104163100804e-07,
      "loss": 2.6777,
      "num_input_tokens_seen": 79780208,
      "step": 38190
    },
    {
      "epoch": 6.230850803491313,
      "grad_norm": 7.375,
      "learning_rate": 2.510712586899833e-07,
      "loss": 1.783,
      "num_input_tokens_seen": 79791344,
      "step": 38195
    },
    {
      "epoch": 6.23166653071213,
      "grad_norm": 7.15625,
      "learning_rate": 2.4968529234478124e-07,
      "loss": 4.8279,
      "num_input_tokens_seen": 79802112,
      "step": 38200
    },
    {
      "epoch": 6.23166653071213,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7528,
      "eval_samples_per_second": 20.222,
      "eval_steps_per_second": 10.115,
      "num_input_tokens_seen": 79802112,
      "step": 38200
    },
    {
      "epoch": 6.232482257932947,
      "grad_norm": 6.125,
      "learning_rate": 2.483031428091448e-07,
      "loss": 1.9544,
      "num_input_tokens_seen": 79812512,
      "step": 38205
    },
    {
      "epoch": 6.233297985153764,
      "grad_norm": 13.375,
      "learning_rate": 2.469248102962091e-07,
      "loss": 2.2185,
      "num_input_tokens_seen": 79823568,
      "step": 38210
    },
    {
      "epoch": 6.234113712374582,
      "grad_norm": 7.71875,
      "learning_rate": 2.4555029501853455e-07,
      "loss": 2.3884,
      "num_input_tokens_seen": 79834336,
      "step": 38215
    },
    {
      "epoch": 6.234929439595399,
      "grad_norm": 10.25,
      "learning_rate": 2.441795971880906e-07,
      "loss": 3.7727,
      "num_input_tokens_seen": 79844960,
      "step": 38220
    },
    {
      "epoch": 6.2357451668162165,
      "grad_norm": 8.75,
      "learning_rate": 2.4281271701625255e-07,
      "loss": 1.5325,
      "num_input_tokens_seen": 79855664,
      "step": 38225
    },
    {
      "epoch": 6.236560894037034,
      "grad_norm": 8.0,
      "learning_rate": 2.4144965471381007e-07,
      "loss": 2.7582,
      "num_input_tokens_seen": 79866848,
      "step": 38230
    },
    {
      "epoch": 6.237376621257852,
      "grad_norm": 8.4375,
      "learning_rate": 2.400904104909674e-07,
      "loss": 1.6745,
      "num_input_tokens_seen": 79877536,
      "step": 38235
    },
    {
      "epoch": 6.238192348478669,
      "grad_norm": 7.59375,
      "learning_rate": 2.3873498455733725e-07,
      "loss": 2.7208,
      "num_input_tokens_seen": 79887696,
      "step": 38240
    },
    {
      "epoch": 6.239008075699486,
      "grad_norm": 2.046875,
      "learning_rate": 2.3738337712194137e-07,
      "loss": 1.8004,
      "num_input_tokens_seen": 79897888,
      "step": 38245
    },
    {
      "epoch": 6.239823802920303,
      "grad_norm": 5.59375,
      "learning_rate": 2.3603558839321305e-07,
      "loss": 2.1685,
      "num_input_tokens_seen": 79908800,
      "step": 38250
    },
    {
      "epoch": 6.240639530141121,
      "grad_norm": 13.9375,
      "learning_rate": 2.3469161857900267e-07,
      "loss": 2.2996,
      "num_input_tokens_seen": 79919232,
      "step": 38255
    },
    {
      "epoch": 6.241455257361938,
      "grad_norm": 10.625,
      "learning_rate": 2.3335146788656393e-07,
      "loss": 2.6209,
      "num_input_tokens_seen": 79930480,
      "step": 38260
    },
    {
      "epoch": 6.242270984582755,
      "grad_norm": 9.8125,
      "learning_rate": 2.3201513652256757e-07,
      "loss": 2.1849,
      "num_input_tokens_seen": 79940784,
      "step": 38265
    },
    {
      "epoch": 6.243086711803572,
      "grad_norm": 7.5625,
      "learning_rate": 2.3068262469308766e-07,
      "loss": 0.9894,
      "num_input_tokens_seen": 79950208,
      "step": 38270
    },
    {
      "epoch": 6.2439024390243905,
      "grad_norm": 2.90625,
      "learning_rate": 2.2935393260362093e-07,
      "loss": 2.6364,
      "num_input_tokens_seen": 79960944,
      "step": 38275
    },
    {
      "epoch": 6.244718166245208,
      "grad_norm": 5.21875,
      "learning_rate": 2.2802906045906458e-07,
      "loss": 1.6578,
      "num_input_tokens_seen": 79970528,
      "step": 38280
    },
    {
      "epoch": 6.245533893466025,
      "grad_norm": 0.11083984375,
      "learning_rate": 2.2670800846373018e-07,
      "loss": 1.1059,
      "num_input_tokens_seen": 79979456,
      "step": 38285
    },
    {
      "epoch": 6.246349620686843,
      "grad_norm": 8.9375,
      "learning_rate": 2.2539077682134367e-07,
      "loss": 3.0101,
      "num_input_tokens_seen": 79990864,
      "step": 38290
    },
    {
      "epoch": 6.24716534790766,
      "grad_norm": 12.125,
      "learning_rate": 2.2407736573503423e-07,
      "loss": 1.4965,
      "num_input_tokens_seen": 80001856,
      "step": 38295
    },
    {
      "epoch": 6.247981075128477,
      "grad_norm": 5.53125,
      "learning_rate": 2.2276777540735093e-07,
      "loss": 1.2453,
      "num_input_tokens_seen": 80013392,
      "step": 38300
    },
    {
      "epoch": 6.248796802349294,
      "grad_norm": 3.796875,
      "learning_rate": 2.2146200604024613e-07,
      "loss": 0.9908,
      "num_input_tokens_seen": 80024736,
      "step": 38305
    },
    {
      "epoch": 6.249612529570112,
      "grad_norm": 10.75,
      "learning_rate": 2.2016005783508375e-07,
      "loss": 2.1107,
      "num_input_tokens_seen": 80034800,
      "step": 38310
    },
    {
      "epoch": 6.250428256790929,
      "grad_norm": 3.6875,
      "learning_rate": 2.1886193099264763e-07,
      "loss": 2.7397,
      "num_input_tokens_seen": 80044960,
      "step": 38315
    },
    {
      "epoch": 6.251243984011746,
      "grad_norm": 6.25,
      "learning_rate": 2.175676257131165e-07,
      "loss": 2.3341,
      "num_input_tokens_seen": 80056800,
      "step": 38320
    },
    {
      "epoch": 6.2520597112325635,
      "grad_norm": 12.9375,
      "learning_rate": 2.162771421960974e-07,
      "loss": 2.219,
      "num_input_tokens_seen": 80067328,
      "step": 38325
    },
    {
      "epoch": 6.252875438453382,
      "grad_norm": 8.5625,
      "learning_rate": 2.1499048064059224e-07,
      "loss": 1.6496,
      "num_input_tokens_seen": 80077776,
      "step": 38330
    },
    {
      "epoch": 6.253691165674199,
      "grad_norm": 5.40625,
      "learning_rate": 2.1370764124502285e-07,
      "loss": 2.8483,
      "num_input_tokens_seen": 80086320,
      "step": 38335
    },
    {
      "epoch": 6.254506892895016,
      "grad_norm": 2.734375,
      "learning_rate": 2.1242862420721988e-07,
      "loss": 2.7111,
      "num_input_tokens_seen": 80097920,
      "step": 38340
    },
    {
      "epoch": 6.255322620115833,
      "grad_norm": 9.75,
      "learning_rate": 2.1115342972442276e-07,
      "loss": 2.112,
      "num_input_tokens_seen": 80110304,
      "step": 38345
    },
    {
      "epoch": 6.256138347336651,
      "grad_norm": 5.75,
      "learning_rate": 2.0988205799328252e-07,
      "loss": 2.0897,
      "num_input_tokens_seen": 80120256,
      "step": 38350
    },
    {
      "epoch": 6.256954074557468,
      "grad_norm": 6.0,
      "learning_rate": 2.0861450920986182e-07,
      "loss": 2.1309,
      "num_input_tokens_seen": 80131072,
      "step": 38355
    },
    {
      "epoch": 6.257769801778285,
      "grad_norm": 8.6875,
      "learning_rate": 2.07350783569632e-07,
      "loss": 1.6689,
      "num_input_tokens_seen": 80142256,
      "step": 38360
    },
    {
      "epoch": 6.258585528999102,
      "grad_norm": 3.640625,
      "learning_rate": 2.060908812674761e-07,
      "loss": 1.4719,
      "num_input_tokens_seen": 80152928,
      "step": 38365
    },
    {
      "epoch": 6.25940125621992,
      "grad_norm": 5.21875,
      "learning_rate": 2.0483480249768317e-07,
      "loss": 1.9444,
      "num_input_tokens_seen": 80163472,
      "step": 38370
    },
    {
      "epoch": 6.2602169834407375,
      "grad_norm": 7.6875,
      "learning_rate": 2.035825474539621e-07,
      "loss": 1.719,
      "num_input_tokens_seen": 80174640,
      "step": 38375
    },
    {
      "epoch": 6.261032710661555,
      "grad_norm": 9.0625,
      "learning_rate": 2.0233411632942235e-07,
      "loss": 1.7242,
      "num_input_tokens_seen": 80184272,
      "step": 38380
    },
    {
      "epoch": 6.261848437882372,
      "grad_norm": 4.90625,
      "learning_rate": 2.0108950931658764e-07,
      "loss": 1.5439,
      "num_input_tokens_seen": 80196304,
      "step": 38385
    },
    {
      "epoch": 6.26266416510319,
      "grad_norm": 8.5625,
      "learning_rate": 1.998487266073934e-07,
      "loss": 3.0629,
      "num_input_tokens_seen": 80206544,
      "step": 38390
    },
    {
      "epoch": 6.263479892324007,
      "grad_norm": 1.21875,
      "learning_rate": 1.986117683931865e-07,
      "loss": 1.6194,
      "num_input_tokens_seen": 80217744,
      "step": 38395
    },
    {
      "epoch": 6.264295619544824,
      "grad_norm": 7.46875,
      "learning_rate": 1.9737863486471442e-07,
      "loss": 3.159,
      "num_input_tokens_seen": 80229344,
      "step": 38400
    },
    {
      "epoch": 6.264295619544824,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.7266,
      "eval_samples_per_second": 20.226,
      "eval_steps_per_second": 10.117,
      "num_input_tokens_seen": 80229344,
      "step": 38400
    },
    {
      "epoch": 6.265111346765641,
      "grad_norm": 13.25,
      "learning_rate": 1.9614932621215e-07,
      "loss": 2.3409,
      "num_input_tokens_seen": 80238736,
      "step": 38405
    },
    {
      "epoch": 6.265927073986459,
      "grad_norm": 3.9375,
      "learning_rate": 1.9492384262506102e-07,
      "loss": 1.9272,
      "num_input_tokens_seen": 80248368,
      "step": 38410
    },
    {
      "epoch": 6.266742801207276,
      "grad_norm": 8.4375,
      "learning_rate": 1.9370218429243524e-07,
      "loss": 2.7534,
      "num_input_tokens_seen": 80259360,
      "step": 38415
    },
    {
      "epoch": 6.2675585284280935,
      "grad_norm": 2.9375,
      "learning_rate": 1.9248435140267197e-07,
      "loss": 3.2606,
      "num_input_tokens_seen": 80270256,
      "step": 38420
    },
    {
      "epoch": 6.268374255648911,
      "grad_norm": 15.0,
      "learning_rate": 1.9127034414356814e-07,
      "loss": 2.7334,
      "num_input_tokens_seen": 80280272,
      "step": 38425
    },
    {
      "epoch": 6.269189982869729,
      "grad_norm": 10.1875,
      "learning_rate": 1.9006016270234627e-07,
      "loss": 2.5592,
      "num_input_tokens_seen": 80290000,
      "step": 38430
    },
    {
      "epoch": 6.270005710090546,
      "grad_norm": 6.90625,
      "learning_rate": 1.888538072656293e-07,
      "loss": 3.1807,
      "num_input_tokens_seen": 80301968,
      "step": 38435
    },
    {
      "epoch": 6.270821437311363,
      "grad_norm": 5.125,
      "learning_rate": 1.8765127801944893e-07,
      "loss": 2.5819,
      "num_input_tokens_seen": 80311936,
      "step": 38440
    },
    {
      "epoch": 6.27163716453218,
      "grad_norm": 7.5625,
      "learning_rate": 1.8645257514925406e-07,
      "loss": 1.8347,
      "num_input_tokens_seen": 80322560,
      "step": 38445
    },
    {
      "epoch": 6.272452891752998,
      "grad_norm": 9.25,
      "learning_rate": 1.8525769883989685e-07,
      "loss": 2.0879,
      "num_input_tokens_seen": 80333280,
      "step": 38450
    },
    {
      "epoch": 6.273268618973815,
      "grad_norm": 3.671875,
      "learning_rate": 1.8406664927564654e-07,
      "loss": 1.992,
      "num_input_tokens_seen": 80342384,
      "step": 38455
    },
    {
      "epoch": 6.274084346194632,
      "grad_norm": 7.3125,
      "learning_rate": 1.8287942664017566e-07,
      "loss": 3.3605,
      "num_input_tokens_seen": 80352576,
      "step": 38460
    },
    {
      "epoch": 6.27490007341545,
      "grad_norm": 5.34375,
      "learning_rate": 1.8169603111656552e-07,
      "loss": 2.1352,
      "num_input_tokens_seen": 80361920,
      "step": 38465
    },
    {
      "epoch": 6.275715800636267,
      "grad_norm": 5.25,
      "learning_rate": 1.805164628873146e-07,
      "loss": 3.5429,
      "num_input_tokens_seen": 80372112,
      "step": 38470
    },
    {
      "epoch": 6.276531527857085,
      "grad_norm": 3.75,
      "learning_rate": 1.793407221343274e-07,
      "loss": 3.2005,
      "num_input_tokens_seen": 80382368,
      "step": 38475
    },
    {
      "epoch": 6.277347255077902,
      "grad_norm": 9.0,
      "learning_rate": 1.781688090389172e-07,
      "loss": 2.701,
      "num_input_tokens_seen": 80391328,
      "step": 38480
    },
    {
      "epoch": 6.27816298229872,
      "grad_norm": 18.75,
      "learning_rate": 1.770007237818061e-07,
      "loss": 2.7116,
      "num_input_tokens_seen": 80402800,
      "step": 38485
    },
    {
      "epoch": 6.278978709519537,
      "grad_norm": 10.8125,
      "learning_rate": 1.7583646654313059e-07,
      "loss": 2.6534,
      "num_input_tokens_seen": 80413136,
      "step": 38490
    },
    {
      "epoch": 6.279794436740354,
      "grad_norm": 6.84375,
      "learning_rate": 1.7467603750242757e-07,
      "loss": 2.8355,
      "num_input_tokens_seen": 80423584,
      "step": 38495
    },
    {
      "epoch": 6.280610163961171,
      "grad_norm": 5.125,
      "learning_rate": 1.7351943683865944e-07,
      "loss": 0.8452,
      "num_input_tokens_seen": 80435376,
      "step": 38500
    },
    {
      "epoch": 6.281425891181989,
      "grad_norm": 7.71875,
      "learning_rate": 1.723666647301808e-07,
      "loss": 2.313,
      "num_input_tokens_seen": 80445904,
      "step": 38505
    },
    {
      "epoch": 6.282241618402806,
      "grad_norm": 11.5625,
      "learning_rate": 1.712177213547661e-07,
      "loss": 3.6515,
      "num_input_tokens_seen": 80455344,
      "step": 38510
    },
    {
      "epoch": 6.283057345623623,
      "grad_norm": 1.609375,
      "learning_rate": 1.7007260688959581e-07,
      "loss": 2.459,
      "num_input_tokens_seen": 80466160,
      "step": 38515
    },
    {
      "epoch": 6.2838730728444405,
      "grad_norm": 8.1875,
      "learning_rate": 1.68931321511262e-07,
      "loss": 2.7,
      "num_input_tokens_seen": 80476128,
      "step": 38520
    },
    {
      "epoch": 6.2846888000652585,
      "grad_norm": 10.9375,
      "learning_rate": 1.6779386539576835e-07,
      "loss": 2.4041,
      "num_input_tokens_seen": 80485872,
      "step": 38525
    },
    {
      "epoch": 6.285504527286076,
      "grad_norm": 0.232421875,
      "learning_rate": 1.666602387185162e-07,
      "loss": 1.1816,
      "num_input_tokens_seen": 80496016,
      "step": 38530
    },
    {
      "epoch": 6.286320254506893,
      "grad_norm": 3.609375,
      "learning_rate": 1.655304416543352e-07,
      "loss": 2.8938,
      "num_input_tokens_seen": 80505536,
      "step": 38535
    },
    {
      "epoch": 6.28713598172771,
      "grad_norm": 10.6875,
      "learning_rate": 1.6440447437744698e-07,
      "loss": 4.2738,
      "num_input_tokens_seen": 80515776,
      "step": 38540
    },
    {
      "epoch": 6.287951708948528,
      "grad_norm": 11.6875,
      "learning_rate": 1.6328233706149332e-07,
      "loss": 4.796,
      "num_input_tokens_seen": 80526752,
      "step": 38545
    },
    {
      "epoch": 6.288767436169345,
      "grad_norm": 3.578125,
      "learning_rate": 1.6216402987951906e-07,
      "loss": 2.3432,
      "num_input_tokens_seen": 80536496,
      "step": 38550
    },
    {
      "epoch": 6.289583163390162,
      "grad_norm": 7.09375,
      "learning_rate": 1.6104955300398627e-07,
      "loss": 2.9521,
      "num_input_tokens_seen": 80547424,
      "step": 38555
    },
    {
      "epoch": 6.290398890610979,
      "grad_norm": 3.015625,
      "learning_rate": 1.5993890660675748e-07,
      "loss": 2.4793,
      "num_input_tokens_seen": 80557984,
      "step": 38560
    },
    {
      "epoch": 6.291214617831797,
      "grad_norm": 9.5,
      "learning_rate": 1.5883209085910678e-07,
      "loss": 3.9188,
      "num_input_tokens_seen": 80569104,
      "step": 38565
    },
    {
      "epoch": 6.2920303450526145,
      "grad_norm": 5.65625,
      "learning_rate": 1.5772910593172264e-07,
      "loss": 2.5822,
      "num_input_tokens_seen": 80579328,
      "step": 38570
    },
    {
      "epoch": 6.292846072273432,
      "grad_norm": 6.90625,
      "learning_rate": 1.5662995199469954e-07,
      "loss": 2.4876,
      "num_input_tokens_seen": 80589808,
      "step": 38575
    },
    {
      "epoch": 6.293661799494249,
      "grad_norm": 10.25,
      "learning_rate": 1.5553462921753802e-07,
      "loss": 1.3114,
      "num_input_tokens_seen": 80599456,
      "step": 38580
    },
    {
      "epoch": 6.294477526715067,
      "grad_norm": 7.40625,
      "learning_rate": 1.544431377691502e-07,
      "loss": 2.4901,
      "num_input_tokens_seen": 80609312,
      "step": 38585
    },
    {
      "epoch": 6.295293253935884,
      "grad_norm": 6.5,
      "learning_rate": 1.5335547781785975e-07,
      "loss": 1.999,
      "num_input_tokens_seen": 80619904,
      "step": 38590
    },
    {
      "epoch": 6.296108981156701,
      "grad_norm": 6.375,
      "learning_rate": 1.5227164953139917e-07,
      "loss": 3.3656,
      "num_input_tokens_seen": 80632800,
      "step": 38595
    },
    {
      "epoch": 6.296924708377518,
      "grad_norm": 14.75,
      "learning_rate": 1.511916530769042e-07,
      "loss": 1.9346,
      "num_input_tokens_seen": 80643632,
      "step": 38600
    },
    {
      "epoch": 6.296924708377518,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.6117,
      "eval_samples_per_second": 20.243,
      "eval_steps_per_second": 10.125,
      "num_input_tokens_seen": 80643632,
      "step": 38600
    },
    {
      "epoch": 6.297740435598336,
      "grad_norm": 3.875,
      "learning_rate": 1.5011548862092773e-07,
      "loss": 1.5578,
      "num_input_tokens_seen": 80653280,
      "step": 38605
    },
    {
      "epoch": 6.298556162819153,
      "grad_norm": 12.0,
      "learning_rate": 1.490431563294231e-07,
      "loss": 2.8874,
      "num_input_tokens_seen": 80664608,
      "step": 38610
    },
    {
      "epoch": 6.2993718900399704,
      "grad_norm": 9.8125,
      "learning_rate": 1.4797465636776365e-07,
      "loss": 1.9634,
      "num_input_tokens_seen": 80675408,
      "step": 38615
    },
    {
      "epoch": 6.300187617260788,
      "grad_norm": 10.5625,
      "learning_rate": 1.4690998890072027e-07,
      "loss": 1.9092,
      "num_input_tokens_seen": 80686720,
      "step": 38620
    },
    {
      "epoch": 6.301003344481606,
      "grad_norm": 5.6875,
      "learning_rate": 1.4584915409248112e-07,
      "loss": 2.5565,
      "num_input_tokens_seen": 80696592,
      "step": 38625
    },
    {
      "epoch": 6.301819071702423,
      "grad_norm": 6.375,
      "learning_rate": 1.4479215210663754e-07,
      "loss": 2.9335,
      "num_input_tokens_seen": 80707376,
      "step": 38630
    },
    {
      "epoch": 6.30263479892324,
      "grad_norm": 7.0,
      "learning_rate": 1.4373898310619528e-07,
      "loss": 4.6723,
      "num_input_tokens_seen": 80717056,
      "step": 38635
    },
    {
      "epoch": 6.303450526144058,
      "grad_norm": 10.375,
      "learning_rate": 1.4268964725356604e-07,
      "loss": 2.2538,
      "num_input_tokens_seen": 80727728,
      "step": 38640
    },
    {
      "epoch": 6.304266253364875,
      "grad_norm": 10.625,
      "learning_rate": 1.4164414471056764e-07,
      "loss": 2.6737,
      "num_input_tokens_seen": 80736944,
      "step": 38645
    },
    {
      "epoch": 6.305081980585692,
      "grad_norm": 9.375,
      "learning_rate": 1.4060247563843497e-07,
      "loss": 2.6912,
      "num_input_tokens_seen": 80747520,
      "step": 38650
    },
    {
      "epoch": 6.305897707806509,
      "grad_norm": 11.0625,
      "learning_rate": 1.3956464019780068e-07,
      "loss": 1.8926,
      "num_input_tokens_seen": 80757600,
      "step": 38655
    },
    {
      "epoch": 6.306713435027326,
      "grad_norm": 5.09375,
      "learning_rate": 1.385306385487145e-07,
      "loss": 3.3906,
      "num_input_tokens_seen": 80767072,
      "step": 38660
    },
    {
      "epoch": 6.307529162248144,
      "grad_norm": 3.40625,
      "learning_rate": 1.3750047085063222e-07,
      "loss": 1.7267,
      "num_input_tokens_seen": 80777776,
      "step": 38665
    },
    {
      "epoch": 6.3083448894689615,
      "grad_norm": 2.5,
      "learning_rate": 1.3647413726242119e-07,
      "loss": 2.2446,
      "num_input_tokens_seen": 80787712,
      "step": 38670
    },
    {
      "epoch": 6.309160616689779,
      "grad_norm": 7.3125,
      "learning_rate": 1.3545163794235205e-07,
      "loss": 3.4725,
      "num_input_tokens_seen": 80798736,
      "step": 38675
    },
    {
      "epoch": 6.309976343910597,
      "grad_norm": 4.6875,
      "learning_rate": 1.3443297304810698e-07,
      "loss": 2.3188,
      "num_input_tokens_seen": 80809040,
      "step": 38680
    },
    {
      "epoch": 6.310792071131414,
      "grad_norm": 2.953125,
      "learning_rate": 1.3341814273677977e-07,
      "loss": 1.1266,
      "num_input_tokens_seen": 80818320,
      "step": 38685
    },
    {
      "epoch": 6.311607798352231,
      "grad_norm": 7.84375,
      "learning_rate": 1.324071471648647e-07,
      "loss": 2.4951,
      "num_input_tokens_seen": 80828656,
      "step": 38690
    },
    {
      "epoch": 6.312423525573048,
      "grad_norm": 10.4375,
      "learning_rate": 1.3139998648827312e-07,
      "loss": 1.5555,
      "num_input_tokens_seen": 80839120,
      "step": 38695
    },
    {
      "epoch": 6.313239252793866,
      "grad_norm": 4.875,
      "learning_rate": 1.3039666086232526e-07,
      "loss": 2.1156,
      "num_input_tokens_seen": 80850912,
      "step": 38700
    },
    {
      "epoch": 6.314054980014683,
      "grad_norm": 6.34375,
      "learning_rate": 1.2939717044174183e-07,
      "loss": 3.3882,
      "num_input_tokens_seen": 80860240,
      "step": 38705
    },
    {
      "epoch": 6.3148707072355,
      "grad_norm": 12.0625,
      "learning_rate": 1.284015153806578e-07,
      "loss": 2.1669,
      "num_input_tokens_seen": 80868112,
      "step": 38710
    },
    {
      "epoch": 6.3156864344563175,
      "grad_norm": 6.5,
      "learning_rate": 1.274096958326171e-07,
      "loss": 2.9756,
      "num_input_tokens_seen": 80878592,
      "step": 38715
    },
    {
      "epoch": 6.3165021616771355,
      "grad_norm": 11.9375,
      "learning_rate": 1.2642171195056952e-07,
      "loss": 3.5525,
      "num_input_tokens_seen": 80887792,
      "step": 38720
    },
    {
      "epoch": 6.317317888897953,
      "grad_norm": 5.03125,
      "learning_rate": 1.2543756388687377e-07,
      "loss": 2.1238,
      "num_input_tokens_seen": 80898208,
      "step": 38725
    },
    {
      "epoch": 6.31813361611877,
      "grad_norm": 5.8125,
      "learning_rate": 1.2445725179330014e-07,
      "loss": 2.6483,
      "num_input_tokens_seen": 80908928,
      "step": 38730
    },
    {
      "epoch": 6.318949343339587,
      "grad_norm": 9.75,
      "learning_rate": 1.2348077582102212e-07,
      "loss": 2.419,
      "num_input_tokens_seen": 80919424,
      "step": 38735
    },
    {
      "epoch": 6.319765070560405,
      "grad_norm": 5.5,
      "learning_rate": 1.2250813612062762e-07,
      "loss": 1.5029,
      "num_input_tokens_seen": 80930128,
      "step": 38740
    },
    {
      "epoch": 6.320580797781222,
      "grad_norm": 8.5625,
      "learning_rate": 1.215393328421105e-07,
      "loss": 2.4536,
      "num_input_tokens_seen": 80940208,
      "step": 38745
    },
    {
      "epoch": 6.321396525002039,
      "grad_norm": 6.4375,
      "learning_rate": 1.2057436613486796e-07,
      "loss": 3.1785,
      "num_input_tokens_seen": 80949808,
      "step": 38750
    },
    {
      "epoch": 6.322212252222856,
      "grad_norm": 8.5625,
      "learning_rate": 1.1961323614771424e-07,
      "loss": 2.3832,
      "num_input_tokens_seen": 80960672,
      "step": 38755
    },
    {
      "epoch": 6.323027979443674,
      "grad_norm": 5.5625,
      "learning_rate": 1.1865594302886418e-07,
      "loss": 3.8472,
      "num_input_tokens_seen": 80971040,
      "step": 38760
    },
    {
      "epoch": 6.3238437066644915,
      "grad_norm": 8.6875,
      "learning_rate": 1.1770248692594687e-07,
      "loss": 2.1042,
      "num_input_tokens_seen": 80980080,
      "step": 38765
    },
    {
      "epoch": 6.324659433885309,
      "grad_norm": 9.625,
      "learning_rate": 1.167528679859975e-07,
      "loss": 1.7199,
      "num_input_tokens_seen": 80990080,
      "step": 38770
    },
    {
      "epoch": 6.325475161106126,
      "grad_norm": 6.40625,
      "learning_rate": 1.1580708635545446e-07,
      "loss": 2.4792,
      "num_input_tokens_seen": 81000416,
      "step": 38775
    },
    {
      "epoch": 6.326290888326944,
      "grad_norm": 2.6875,
      "learning_rate": 1.1486514218017885e-07,
      "loss": 1.9909,
      "num_input_tokens_seen": 81011376,
      "step": 38780
    },
    {
      "epoch": 6.327106615547761,
      "grad_norm": 7.03125,
      "learning_rate": 1.1392703560542117e-07,
      "loss": 2.0435,
      "num_input_tokens_seen": 81021760,
      "step": 38785
    },
    {
      "epoch": 6.327922342768578,
      "grad_norm": 7.0625,
      "learning_rate": 1.129927667758518e-07,
      "loss": 3.628,
      "num_input_tokens_seen": 81032352,
      "step": 38790
    },
    {
      "epoch": 6.328738069989395,
      "grad_norm": 7.34375,
      "learning_rate": 1.1206233583554992e-07,
      "loss": 3.9584,
      "num_input_tokens_seen": 81042832,
      "step": 38795
    },
    {
      "epoch": 6.329553797210213,
      "grad_norm": 8.375,
      "learning_rate": 1.1113574292799523e-07,
      "loss": 1.6984,
      "num_input_tokens_seen": 81051936,
      "step": 38800
    },
    {
      "epoch": 6.329553797210213,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.3488,
      "eval_samples_per_second": 20.283,
      "eval_steps_per_second": 10.145,
      "num_input_tokens_seen": 81051936,
      "step": 38800
    },
    {
      "epoch": 6.33036952443103,
      "grad_norm": 9.0,
      "learning_rate": 1.1021298819608449e-07,
      "loss": 2.267,
      "num_input_tokens_seen": 81062848,
      "step": 38805
    },
    {
      "epoch": 6.331185251651847,
      "grad_norm": 5.8125,
      "learning_rate": 1.0929407178211226e-07,
      "loss": 2.1411,
      "num_input_tokens_seen": 81072736,
      "step": 38810
    },
    {
      "epoch": 6.332000978872665,
      "grad_norm": 5.53125,
      "learning_rate": 1.0837899382779293e-07,
      "loss": 2.4041,
      "num_input_tokens_seen": 81083248,
      "step": 38815
    },
    {
      "epoch": 6.332816706093483,
      "grad_norm": 0.09228515625,
      "learning_rate": 1.0746775447423862e-07,
      "loss": 1.9642,
      "num_input_tokens_seen": 81093264,
      "step": 38820
    },
    {
      "epoch": 6.3336324333143,
      "grad_norm": 6.125,
      "learning_rate": 1.0656035386197583e-07,
      "loss": 2.0931,
      "num_input_tokens_seen": 81105168,
      "step": 38825
    },
    {
      "epoch": 6.334448160535117,
      "grad_norm": 9.3125,
      "learning_rate": 1.0565679213093982e-07,
      "loss": 2.1358,
      "num_input_tokens_seen": 81116016,
      "step": 38830
    },
    {
      "epoch": 6.335263887755934,
      "grad_norm": 5.90625,
      "learning_rate": 1.0475706942046638e-07,
      "loss": 2.0985,
      "num_input_tokens_seen": 81128192,
      "step": 38835
    },
    {
      "epoch": 6.336079614976752,
      "grad_norm": 8.1875,
      "learning_rate": 1.0386118586930282e-07,
      "loss": 2.4736,
      "num_input_tokens_seen": 81138640,
      "step": 38840
    },
    {
      "epoch": 6.336895342197569,
      "grad_norm": 11.5625,
      "learning_rate": 1.0296914161561367e-07,
      "loss": 1.9331,
      "num_input_tokens_seen": 81150960,
      "step": 38845
    },
    {
      "epoch": 6.337711069418386,
      "grad_norm": 7.59375,
      "learning_rate": 1.0208093679695552e-07,
      "loss": 1.7168,
      "num_input_tokens_seen": 81162768,
      "step": 38850
    },
    {
      "epoch": 6.338526796639204,
      "grad_norm": 10.5,
      "learning_rate": 1.0119657155030493e-07,
      "loss": 2.0389,
      "num_input_tokens_seen": 81173856,
      "step": 38855
    },
    {
      "epoch": 6.339342523860021,
      "grad_norm": 7.125,
      "learning_rate": 1.003160460120417e-07,
      "loss": 1.7666,
      "num_input_tokens_seen": 81185152,
      "step": 38860
    },
    {
      "epoch": 6.3401582510808385,
      "grad_norm": 4.125,
      "learning_rate": 9.943936031795165e-08,
      "loss": 2.2515,
      "num_input_tokens_seen": 81196720,
      "step": 38865
    },
    {
      "epoch": 6.340973978301656,
      "grad_norm": 7.96875,
      "learning_rate": 9.856651460323219e-08,
      "loss": 2.0022,
      "num_input_tokens_seen": 81207584,
      "step": 38870
    },
    {
      "epoch": 6.341789705522474,
      "grad_norm": 5.59375,
      "learning_rate": 9.769750900248953e-08,
      "loss": 2.2842,
      "num_input_tokens_seen": 81217744,
      "step": 38875
    },
    {
      "epoch": 6.342605432743291,
      "grad_norm": 14.75,
      "learning_rate": 9.683234364973038e-08,
      "loss": 1.6487,
      "num_input_tokens_seen": 81226160,
      "step": 38880
    },
    {
      "epoch": 6.343421159964108,
      "grad_norm": 8.0625,
      "learning_rate": 9.597101867837854e-08,
      "loss": 2.1504,
      "num_input_tokens_seen": 81237264,
      "step": 38885
    },
    {
      "epoch": 6.344236887184925,
      "grad_norm": 14.875,
      "learning_rate": 9.511353422125835e-08,
      "loss": 2.5346,
      "num_input_tokens_seen": 81247472,
      "step": 38890
    },
    {
      "epoch": 6.345052614405743,
      "grad_norm": 4.625,
      "learning_rate": 9.42598904106029e-08,
      "loss": 1.9476,
      "num_input_tokens_seen": 81259264,
      "step": 38895
    },
    {
      "epoch": 6.34586834162656,
      "grad_norm": 7.6875,
      "learning_rate": 9.341008737806245e-08,
      "loss": 1.8462,
      "num_input_tokens_seen": 81269632,
      "step": 38900
    },
    {
      "epoch": 6.346684068847377,
      "grad_norm": 3.21875,
      "learning_rate": 9.256412525467661e-08,
      "loss": 2.6333,
      "num_input_tokens_seen": 81280368,
      "step": 38905
    },
    {
      "epoch": 6.3474997960681945,
      "grad_norm": 4.71875,
      "learning_rate": 9.172200417091326e-08,
      "loss": 1.434,
      "num_input_tokens_seen": 81291488,
      "step": 38910
    },
    {
      "epoch": 6.3483155232890125,
      "grad_norm": 6.03125,
      "learning_rate": 9.088372425663239e-08,
      "loss": 2.1019,
      "num_input_tokens_seen": 81300912,
      "step": 38915
    },
    {
      "epoch": 6.34913125050983,
      "grad_norm": 7.21875,
      "learning_rate": 9.004928564110837e-08,
      "loss": 2.4287,
      "num_input_tokens_seen": 81310944,
      "step": 38920
    },
    {
      "epoch": 6.349946977730647,
      "grad_norm": 2.34375,
      "learning_rate": 8.92186884530244e-08,
      "loss": 2.233,
      "num_input_tokens_seen": 81321136,
      "step": 38925
    },
    {
      "epoch": 6.350762704951464,
      "grad_norm": 14.1875,
      "learning_rate": 8.83919328204641e-08,
      "loss": 2.9086,
      "num_input_tokens_seen": 81332144,
      "step": 38930
    },
    {
      "epoch": 6.351578432172282,
      "grad_norm": 6.875,
      "learning_rate": 8.756901887093105e-08,
      "loss": 2.5815,
      "num_input_tokens_seen": 81343024,
      "step": 38935
    },
    {
      "epoch": 6.352394159393099,
      "grad_norm": 7.53125,
      "learning_rate": 8.674994673132098e-08,
      "loss": 1.8414,
      "num_input_tokens_seen": 81354464,
      "step": 38940
    },
    {
      "epoch": 6.353209886613916,
      "grad_norm": 3.703125,
      "learning_rate": 8.593471652794949e-08,
      "loss": 2.4025,
      "num_input_tokens_seen": 81365216,
      "step": 38945
    },
    {
      "epoch": 6.354025613834733,
      "grad_norm": 7.84375,
      "learning_rate": 8.512332838653548e-08,
      "loss": 3.3861,
      "num_input_tokens_seen": 81374624,
      "step": 38950
    },
    {
      "epoch": 6.354841341055551,
      "grad_norm": 6.5625,
      "learning_rate": 8.431578243220106e-08,
      "loss": 3.1781,
      "num_input_tokens_seen": 81384048,
      "step": 38955
    },
    {
      "epoch": 6.3556570682763684,
      "grad_norm": 5.15625,
      "learning_rate": 8.351207878948552e-08,
      "loss": 2.7992,
      "num_input_tokens_seen": 81393536,
      "step": 38960
    },
    {
      "epoch": 6.356472795497186,
      "grad_norm": 7.0,
      "learning_rate": 8.271221758232583e-08,
      "loss": 1.6178,
      "num_input_tokens_seen": 81404128,
      "step": 38965
    },
    {
      "epoch": 6.357288522718003,
      "grad_norm": 6.125,
      "learning_rate": 8.191619893407332e-08,
      "loss": 2.1865,
      "num_input_tokens_seen": 81415680,
      "step": 38970
    },
    {
      "epoch": 6.358104249938821,
      "grad_norm": 9.6875,
      "learning_rate": 8.112402296748534e-08,
      "loss": 1.545,
      "num_input_tokens_seen": 81425568,
      "step": 38975
    },
    {
      "epoch": 6.358919977159638,
      "grad_norm": 9.1875,
      "learning_rate": 8.033568980471973e-08,
      "loss": 2.2482,
      "num_input_tokens_seen": 81434080,
      "step": 38980
    },
    {
      "epoch": 6.359735704380455,
      "grad_norm": 3.15625,
      "learning_rate": 7.955119956735146e-08,
      "loss": 1.6244,
      "num_input_tokens_seen": 81445184,
      "step": 38985
    },
    {
      "epoch": 6.360551431601272,
      "grad_norm": 4.5,
      "learning_rate": 7.877055237636155e-08,
      "loss": 1.6888,
      "num_input_tokens_seen": 81456352,
      "step": 38990
    },
    {
      "epoch": 6.36136715882209,
      "grad_norm": 10.5625,
      "learning_rate": 7.79937483521287e-08,
      "loss": 1.603,
      "num_input_tokens_seen": 81466752,
      "step": 38995
    },
    {
      "epoch": 6.362182886042907,
      "grad_norm": 7.875,
      "learning_rate": 7.722078761444873e-08,
      "loss": 3.1322,
      "num_input_tokens_seen": 81475504,
      "step": 39000
    },
    {
      "epoch": 6.362182886042907,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.4246,
      "eval_samples_per_second": 20.272,
      "eval_steps_per_second": 10.14,
      "num_input_tokens_seen": 81475504,
      "step": 39000
    },
    {
      "epoch": 6.362998613263724,
      "grad_norm": 5.59375,
      "learning_rate": 7.645167028252631e-08,
      "loss": 2.1323,
      "num_input_tokens_seen": 81486336,
      "step": 39005
    },
    {
      "epoch": 6.3638143404845415,
      "grad_norm": 6.28125,
      "learning_rate": 7.568639647496379e-08,
      "loss": 0.714,
      "num_input_tokens_seen": 81496464,
      "step": 39010
    },
    {
      "epoch": 6.3646300677053596,
      "grad_norm": 14.375,
      "learning_rate": 7.492496630977508e-08,
      "loss": 2.0721,
      "num_input_tokens_seen": 81507248,
      "step": 39015
    },
    {
      "epoch": 6.365445794926177,
      "grad_norm": 6.71875,
      "learning_rate": 7.416737990438571e-08,
      "loss": 1.6645,
      "num_input_tokens_seen": 81517984,
      "step": 39020
    },
    {
      "epoch": 6.366261522146994,
      "grad_norm": 5.875,
      "learning_rate": 7.341363737562445e-08,
      "loss": 2.4416,
      "num_input_tokens_seen": 81527440,
      "step": 39025
    },
    {
      "epoch": 6.367077249367812,
      "grad_norm": 14.875,
      "learning_rate": 7.266373883972887e-08,
      "loss": 2.7039,
      "num_input_tokens_seen": 81537648,
      "step": 39030
    },
    {
      "epoch": 6.367892976588629,
      "grad_norm": 12.3125,
      "learning_rate": 7.191768441233981e-08,
      "loss": 3.2938,
      "num_input_tokens_seen": 81548528,
      "step": 39035
    },
    {
      "epoch": 6.368708703809446,
      "grad_norm": 8.75,
      "learning_rate": 7.11754742085069e-08,
      "loss": 2.153,
      "num_input_tokens_seen": 81557632,
      "step": 39040
    },
    {
      "epoch": 6.369524431030263,
      "grad_norm": 9.625,
      "learning_rate": 7.043710834269413e-08,
      "loss": 3.7709,
      "num_input_tokens_seen": 81568448,
      "step": 39045
    },
    {
      "epoch": 6.370340158251081,
      "grad_norm": 8.8125,
      "learning_rate": 6.970258692876319e-08,
      "loss": 1.7409,
      "num_input_tokens_seen": 81577360,
      "step": 39050
    },
    {
      "epoch": 6.371155885471898,
      "grad_norm": 10.9375,
      "learning_rate": 6.897191007998738e-08,
      "loss": 3.9392,
      "num_input_tokens_seen": 81588080,
      "step": 39055
    },
    {
      "epoch": 6.3719716126927155,
      "grad_norm": 5.75,
      "learning_rate": 6.824507790904599e-08,
      "loss": 2.0936,
      "num_input_tokens_seen": 81598816,
      "step": 39060
    },
    {
      "epoch": 6.372787339913533,
      "grad_norm": 0.83203125,
      "learning_rate": 6.752209052802439e-08,
      "loss": 1.4064,
      "num_input_tokens_seen": 81609840,
      "step": 39065
    },
    {
      "epoch": 6.373603067134351,
      "grad_norm": 3.5625,
      "learning_rate": 6.680294804841946e-08,
      "loss": 1.7898,
      "num_input_tokens_seen": 81620064,
      "step": 39070
    },
    {
      "epoch": 6.374418794355168,
      "grad_norm": 7.4375,
      "learning_rate": 6.608765058112865e-08,
      "loss": 2.1981,
      "num_input_tokens_seen": 81630000,
      "step": 39075
    },
    {
      "epoch": 6.375234521575985,
      "grad_norm": 6.5,
      "learning_rate": 6.537619823646368e-08,
      "loss": 1.7153,
      "num_input_tokens_seen": 81641088,
      "step": 39080
    },
    {
      "epoch": 6.376050248796802,
      "grad_norm": 19.0,
      "learning_rate": 6.466859112413404e-08,
      "loss": 2.3992,
      "num_input_tokens_seen": 81651056,
      "step": 39085
    },
    {
      "epoch": 6.37686597601762,
      "grad_norm": 8.9375,
      "learning_rate": 6.39648293532663e-08,
      "loss": 2.9019,
      "num_input_tokens_seen": 81661632,
      "step": 39090
    },
    {
      "epoch": 6.377681703238437,
      "grad_norm": 10.4375,
      "learning_rate": 6.32649130323848e-08,
      "loss": 2.648,
      "num_input_tokens_seen": 81670640,
      "step": 39095
    },
    {
      "epoch": 6.378497430459254,
      "grad_norm": 9.1875,
      "learning_rate": 6.256884226943094e-08,
      "loss": 2.683,
      "num_input_tokens_seen": 81679648,
      "step": 39100
    },
    {
      "epoch": 6.3793131576800715,
      "grad_norm": 8.625,
      "learning_rate": 6.187661717174386e-08,
      "loss": 1.8595,
      "num_input_tokens_seen": 81690768,
      "step": 39105
    },
    {
      "epoch": 6.3801288849008895,
      "grad_norm": 6.65625,
      "learning_rate": 6.118823784607708e-08,
      "loss": 1.1874,
      "num_input_tokens_seen": 81699984,
      "step": 39110
    },
    {
      "epoch": 6.380944612121707,
      "grad_norm": 10.625,
      "learning_rate": 6.050370439858178e-08,
      "loss": 2.9432,
      "num_input_tokens_seen": 81709840,
      "step": 39115
    },
    {
      "epoch": 6.381760339342524,
      "grad_norm": 5.125,
      "learning_rate": 5.98230169348235e-08,
      "loss": 2.3406,
      "num_input_tokens_seen": 81720080,
      "step": 39120
    },
    {
      "epoch": 6.382576066563341,
      "grad_norm": 8.375,
      "learning_rate": 5.914617555977664e-08,
      "loss": 2.5976,
      "num_input_tokens_seen": 81730688,
      "step": 39125
    },
    {
      "epoch": 6.383391793784159,
      "grad_norm": 6.90625,
      "learning_rate": 5.8473180377816017e-08,
      "loss": 1.4913,
      "num_input_tokens_seen": 81741984,
      "step": 39130
    },
    {
      "epoch": 6.384207521004976,
      "grad_norm": 7.15625,
      "learning_rate": 5.780403149272251e-08,
      "loss": 3.6142,
      "num_input_tokens_seen": 81753808,
      "step": 39135
    },
    {
      "epoch": 6.385023248225793,
      "grad_norm": 0.57421875,
      "learning_rate": 5.7138729007694126e-08,
      "loss": 3.0723,
      "num_input_tokens_seen": 81764848,
      "step": 39140
    },
    {
      "epoch": 6.38583897544661,
      "grad_norm": 7.1875,
      "learning_rate": 5.64772730253238e-08,
      "loss": 3.0741,
      "num_input_tokens_seen": 81776176,
      "step": 39145
    },
    {
      "epoch": 6.386654702667428,
      "grad_norm": 11.5625,
      "learning_rate": 5.5819663647618814e-08,
      "loss": 2.1012,
      "num_input_tokens_seen": 81786448,
      "step": 39150
    },
    {
      "epoch": 6.387470429888245,
      "grad_norm": 4.1875,
      "learning_rate": 5.5165900975989723e-08,
      "loss": 2.2497,
      "num_input_tokens_seen": 81797184,
      "step": 39155
    },
    {
      "epoch": 6.388286157109063,
      "grad_norm": 2.328125,
      "learning_rate": 5.451598511125311e-08,
      "loss": 2.5918,
      "num_input_tokens_seen": 81807152,
      "step": 39160
    },
    {
      "epoch": 6.38910188432988,
      "grad_norm": 8.125,
      "learning_rate": 5.3869916153637124e-08,
      "loss": 1.9328,
      "num_input_tokens_seen": 81818192,
      "step": 39165
    },
    {
      "epoch": 6.389917611550698,
      "grad_norm": 3.265625,
      "learning_rate": 5.322769420277318e-08,
      "loss": 2.3046,
      "num_input_tokens_seen": 81828800,
      "step": 39170
    },
    {
      "epoch": 6.390733338771515,
      "grad_norm": 14.6875,
      "learning_rate": 5.258931935769873e-08,
      "loss": 3.2759,
      "num_input_tokens_seen": 81839456,
      "step": 39175
    },
    {
      "epoch": 6.391549065992332,
      "grad_norm": 7.53125,
      "learning_rate": 5.19547917168628e-08,
      "loss": 2.7797,
      "num_input_tokens_seen": 81849616,
      "step": 39180
    },
    {
      "epoch": 6.392364793213149,
      "grad_norm": 5.1875,
      "learning_rate": 5.13241113781121e-08,
      "loss": 1.6261,
      "num_input_tokens_seen": 81860432,
      "step": 39185
    },
    {
      "epoch": 6.393180520433967,
      "grad_norm": 0.294921875,
      "learning_rate": 5.0697278438707755e-08,
      "loss": 1.3949,
      "num_input_tokens_seen": 81869312,
      "step": 39190
    },
    {
      "epoch": 6.393996247654784,
      "grad_norm": 11.0,
      "learning_rate": 5.0074292995316854e-08,
      "loss": 3.0205,
      "num_input_tokens_seen": 81880000,
      "step": 39195
    },
    {
      "epoch": 6.394811974875601,
      "grad_norm": 10.1875,
      "learning_rate": 4.945515514400978e-08,
      "loss": 3.2842,
      "num_input_tokens_seen": 81889856,
      "step": 39200
    },
    {
      "epoch": 6.394811974875601,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.5786,
      "eval_samples_per_second": 20.248,
      "eval_steps_per_second": 10.128,
      "num_input_tokens_seen": 81889856,
      "step": 39200
    },
    {
      "epoch": 6.395627702096419,
      "grad_norm": 5.75,
      "learning_rate": 4.883986498026571e-08,
      "loss": 3.0712,
      "num_input_tokens_seen": 81900016,
      "step": 39205
    },
    {
      "epoch": 6.3964434293172365,
      "grad_norm": 6.21875,
      "learning_rate": 4.822842259896987e-08,
      "loss": 3.04,
      "num_input_tokens_seen": 81910944,
      "step": 39210
    },
    {
      "epoch": 6.397259156538054,
      "grad_norm": 5.1875,
      "learning_rate": 4.762082809441626e-08,
      "loss": 1.8931,
      "num_input_tokens_seen": 81921936,
      "step": 39215
    },
    {
      "epoch": 6.398074883758871,
      "grad_norm": 8.9375,
      "learning_rate": 4.7017081560302156e-08,
      "loss": 2.0019,
      "num_input_tokens_seen": 81932640,
      "step": 39220
    },
    {
      "epoch": 6.398890610979688,
      "grad_norm": 9.1875,
      "learning_rate": 4.6417183089730866e-08,
      "loss": 3.0279,
      "num_input_tokens_seen": 81941872,
      "step": 39225
    },
    {
      "epoch": 6.399706338200506,
      "grad_norm": 4.21875,
      "learning_rate": 4.5821132775217265e-08,
      "loss": 1.8014,
      "num_input_tokens_seen": 81951376,
      "step": 39230
    },
    {
      "epoch": 6.400522065421323,
      "grad_norm": 4.625,
      "learning_rate": 4.5228930708679504e-08,
      "loss": 2.6987,
      "num_input_tokens_seen": 81960416,
      "step": 39235
    },
    {
      "epoch": 6.40133779264214,
      "grad_norm": 1.5,
      "learning_rate": 4.464057698144175e-08,
      "loss": 1.5615,
      "num_input_tokens_seen": 81970560,
      "step": 39240
    },
    {
      "epoch": 6.402153519862958,
      "grad_norm": 8.4375,
      "learning_rate": 4.4056071684236974e-08,
      "loss": 2.7491,
      "num_input_tokens_seen": 81980416,
      "step": 39245
    },
    {
      "epoch": 6.402969247083775,
      "grad_norm": 11.0625,
      "learning_rate": 4.347541490719864e-08,
      "loss": 2.8378,
      "num_input_tokens_seen": 81990944,
      "step": 39250
    },
    {
      "epoch": 6.4037849743045925,
      "grad_norm": 8.125,
      "learning_rate": 4.2898606739877336e-08,
      "loss": 2.958,
      "num_input_tokens_seen": 82002032,
      "step": 39255
    },
    {
      "epoch": 6.40460070152541,
      "grad_norm": 2.921875,
      "learning_rate": 4.232564727122135e-08,
      "loss": 2.1829,
      "num_input_tokens_seen": 82012352,
      "step": 39260
    },
    {
      "epoch": 6.405416428746228,
      "grad_norm": 6.125,
      "learning_rate": 4.1756536589585004e-08,
      "loss": 3.6478,
      "num_input_tokens_seen": 82022416,
      "step": 39265
    },
    {
      "epoch": 6.406232155967045,
      "grad_norm": 9.625,
      "learning_rate": 4.119127478273976e-08,
      "loss": 2.8536,
      "num_input_tokens_seen": 82032992,
      "step": 39270
    },
    {
      "epoch": 6.407047883187862,
      "grad_norm": 10.1875,
      "learning_rate": 4.062986193784923e-08,
      "loss": 2.2831,
      "num_input_tokens_seen": 82042000,
      "step": 39275
    },
    {
      "epoch": 6.407863610408679,
      "grad_norm": 10.625,
      "learning_rate": 4.007229814149416e-08,
      "loss": 2.9864,
      "num_input_tokens_seen": 82052288,
      "step": 39280
    },
    {
      "epoch": 6.408679337629497,
      "grad_norm": 4.375,
      "learning_rate": 3.951858347965576e-08,
      "loss": 2.6389,
      "num_input_tokens_seen": 82060864,
      "step": 39285
    },
    {
      "epoch": 6.409495064850314,
      "grad_norm": 4.90625,
      "learning_rate": 3.896871803772684e-08,
      "loss": 2.2337,
      "num_input_tokens_seen": 82071424,
      "step": 39290
    },
    {
      "epoch": 6.410310792071131,
      "grad_norm": 1.515625,
      "learning_rate": 3.842270190050068e-08,
      "loss": 2.713,
      "num_input_tokens_seen": 82081456,
      "step": 39295
    },
    {
      "epoch": 6.411126519291948,
      "grad_norm": 7.75,
      "learning_rate": 3.7880535152179376e-08,
      "loss": 2.2665,
      "num_input_tokens_seen": 82091264,
      "step": 39300
    },
    {
      "epoch": 6.4119422465127665,
      "grad_norm": 12.0,
      "learning_rate": 3.734221787637382e-08,
      "loss": 2.2764,
      "num_input_tokens_seen": 82100880,
      "step": 39305
    },
    {
      "epoch": 6.412757973733584,
      "grad_norm": 5.9375,
      "learning_rate": 3.680775015609817e-08,
      "loss": 3.8364,
      "num_input_tokens_seen": 82112976,
      "step": 39310
    },
    {
      "epoch": 6.413573700954401,
      "grad_norm": 8.1875,
      "learning_rate": 3.627713207377537e-08,
      "loss": 2.8558,
      "num_input_tokens_seen": 82123040,
      "step": 39315
    },
    {
      "epoch": 6.414389428175218,
      "grad_norm": 10.875,
      "learning_rate": 3.575036371123164e-08,
      "loss": 2.046,
      "num_input_tokens_seen": 82134928,
      "step": 39320
    },
    {
      "epoch": 6.415205155396036,
      "grad_norm": 10.3125,
      "learning_rate": 3.5227445149704776e-08,
      "loss": 2.1491,
      "num_input_tokens_seen": 82145008,
      "step": 39325
    },
    {
      "epoch": 6.416020882616853,
      "grad_norm": 4.9375,
      "learning_rate": 3.470837646983027e-08,
      "loss": 3.1118,
      "num_input_tokens_seen": 82155824,
      "step": 39330
    },
    {
      "epoch": 6.41683660983767,
      "grad_norm": 11.125,
      "learning_rate": 3.419315775165799e-08,
      "loss": 0.8422,
      "num_input_tokens_seen": 82166352,
      "step": 39335
    },
    {
      "epoch": 6.417652337058487,
      "grad_norm": 8.1875,
      "learning_rate": 3.368178907464103e-08,
      "loss": 2.106,
      "num_input_tokens_seen": 82177872,
      "step": 39340
    },
    {
      "epoch": 6.418468064279305,
      "grad_norm": 11.8125,
      "learning_rate": 3.317427051763855e-08,
      "loss": 3.1297,
      "num_input_tokens_seen": 82188000,
      "step": 39345
    },
    {
      "epoch": 6.419283791500122,
      "grad_norm": 4.28125,
      "learning_rate": 3.267060215891571e-08,
      "loss": 1.352,
      "num_input_tokens_seen": 82199440,
      "step": 39350
    },
    {
      "epoch": 6.4200995187209395,
      "grad_norm": 13.3125,
      "learning_rate": 3.217078407614649e-08,
      "loss": 2.5285,
      "num_input_tokens_seen": 82208624,
      "step": 39355
    },
    {
      "epoch": 6.420915245941757,
      "grad_norm": 6.78125,
      "learning_rate": 3.1674816346405345e-08,
      "loss": 2.2726,
      "num_input_tokens_seen": 82218160,
      "step": 39360
    },
    {
      "epoch": 6.421730973162575,
      "grad_norm": 4.65625,
      "learning_rate": 3.11826990461811e-08,
      "loss": 3.6438,
      "num_input_tokens_seen": 82230352,
      "step": 39365
    },
    {
      "epoch": 6.422546700383392,
      "grad_norm": 6.84375,
      "learning_rate": 3.069443225136304e-08,
      "loss": 2.4759,
      "num_input_tokens_seen": 82241536,
      "step": 39370
    },
    {
      "epoch": 6.423362427604209,
      "grad_norm": 4.09375,
      "learning_rate": 3.021001603724372e-08,
      "loss": 1.5967,
      "num_input_tokens_seen": 82252832,
      "step": 39375
    },
    {
      "epoch": 6.424178154825027,
      "grad_norm": 4.21875,
      "learning_rate": 2.9729450478532818e-08,
      "loss": 1.9682,
      "num_input_tokens_seen": 82263392,
      "step": 39380
    },
    {
      "epoch": 6.424993882045844,
      "grad_norm": 5.59375,
      "learning_rate": 2.9252735649337726e-08,
      "loss": 1.5289,
      "num_input_tokens_seen": 82274112,
      "step": 39385
    },
    {
      "epoch": 6.425809609266661,
      "grad_norm": 9.125,
      "learning_rate": 2.8779871623171863e-08,
      "loss": 3.4401,
      "num_input_tokens_seen": 82284160,
      "step": 39390
    },
    {
      "epoch": 6.426625336487478,
      "grad_norm": 5.28125,
      "learning_rate": 2.8310858472957448e-08,
      "loss": 2.5777,
      "num_input_tokens_seen": 82294448,
      "step": 39395
    },
    {
      "epoch": 6.4274410637082955,
      "grad_norm": 7.0625,
      "learning_rate": 2.784569627101996e-08,
      "loss": 2.8503,
      "num_input_tokens_seen": 82305408,
      "step": 39400
    },
    {
      "epoch": 6.4274410637082955,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.3679,
      "eval_samples_per_second": 20.28,
      "eval_steps_per_second": 10.144,
      "num_input_tokens_seen": 82305408,
      "step": 39400
    },
    {
      "epoch": 6.4282567909291135,
      "grad_norm": 5.78125,
      "learning_rate": 2.738438508909924e-08,
      "loss": 1.3175,
      "num_input_tokens_seen": 82315808,
      "step": 39405
    },
    {
      "epoch": 6.429072518149931,
      "grad_norm": 8.9375,
      "learning_rate": 2.692692499833005e-08,
      "loss": 3.011,
      "num_input_tokens_seen": 82326832,
      "step": 39410
    },
    {
      "epoch": 6.429888245370748,
      "grad_norm": 10.9375,
      "learning_rate": 2.647331606926151e-08,
      "loss": 1.8427,
      "num_input_tokens_seen": 82337600,
      "step": 39415
    },
    {
      "epoch": 6.430703972591566,
      "grad_norm": 5.03125,
      "learning_rate": 2.6023558371843225e-08,
      "loss": 1.9736,
      "num_input_tokens_seen": 82347360,
      "step": 39420
    },
    {
      "epoch": 6.431519699812383,
      "grad_norm": 10.4375,
      "learning_rate": 2.557765197543638e-08,
      "loss": 2.5894,
      "num_input_tokens_seen": 82358544,
      "step": 39425
    },
    {
      "epoch": 6.4323354270332,
      "grad_norm": 8.0625,
      "learning_rate": 2.513559694880263e-08,
      "loss": 2.0908,
      "num_input_tokens_seen": 82367168,
      "step": 39430
    },
    {
      "epoch": 6.433151154254017,
      "grad_norm": 9.25,
      "learning_rate": 2.469739336011523e-08,
      "loss": 2.4636,
      "num_input_tokens_seen": 82375824,
      "step": 39435
    },
    {
      "epoch": 6.433966881474835,
      "grad_norm": 6.4375,
      "learning_rate": 2.4263041276947894e-08,
      "loss": 3.0135,
      "num_input_tokens_seen": 82384320,
      "step": 39440
    },
    {
      "epoch": 6.434782608695652,
      "grad_norm": 7.46875,
      "learning_rate": 2.3832540766283164e-08,
      "loss": 1.4056,
      "num_input_tokens_seen": 82394512,
      "step": 39445
    },
    {
      "epoch": 6.4355983359164695,
      "grad_norm": 6.28125,
      "learning_rate": 2.3405891894512366e-08,
      "loss": 2.2534,
      "num_input_tokens_seen": 82403376,
      "step": 39450
    },
    {
      "epoch": 6.436414063137287,
      "grad_norm": 6.375,
      "learning_rate": 2.29830947274301e-08,
      "loss": 3.3404,
      "num_input_tokens_seen": 82414784,
      "step": 39455
    },
    {
      "epoch": 6.437229790358105,
      "grad_norm": 2.078125,
      "learning_rate": 2.2564149330231432e-08,
      "loss": 1.5016,
      "num_input_tokens_seen": 82427136,
      "step": 39460
    },
    {
      "epoch": 6.438045517578922,
      "grad_norm": 1.546875,
      "learning_rate": 2.2149055767528572e-08,
      "loss": 1.9384,
      "num_input_tokens_seen": 82436800,
      "step": 39465
    },
    {
      "epoch": 6.438861244799739,
      "grad_norm": 0.8671875,
      "learning_rate": 2.1737814103334197e-08,
      "loss": 1.5498,
      "num_input_tokens_seen": 82446528,
      "step": 39470
    },
    {
      "epoch": 6.439676972020556,
      "grad_norm": 12.0625,
      "learning_rate": 2.1330424401064253e-08,
      "loss": 2.8547,
      "num_input_tokens_seen": 82455792,
      "step": 39475
    },
    {
      "epoch": 6.440492699241374,
      "grad_norm": 9.1875,
      "learning_rate": 2.092688672354348e-08,
      "loss": 2.6919,
      "num_input_tokens_seen": 82466240,
      "step": 39480
    },
    {
      "epoch": 6.441308426462191,
      "grad_norm": 4.875,
      "learning_rate": 2.0527201133005435e-08,
      "loss": 1.3794,
      "num_input_tokens_seen": 82475008,
      "step": 39485
    },
    {
      "epoch": 6.442124153683008,
      "grad_norm": 2.859375,
      "learning_rate": 2.0131367691084148e-08,
      "loss": 2.2866,
      "num_input_tokens_seen": 82484880,
      "step": 39490
    },
    {
      "epoch": 6.442939880903825,
      "grad_norm": 4.96875,
      "learning_rate": 1.9739386458819675e-08,
      "loss": 1.718,
      "num_input_tokens_seen": 82494848,
      "step": 39495
    },
    {
      "epoch": 6.443755608124643,
      "grad_norm": 8.6875,
      "learning_rate": 1.9351257496666442e-08,
      "loss": 2.0118,
      "num_input_tokens_seen": 82506272,
      "step": 39500
    },
    {
      "epoch": 6.444571335345461,
      "grad_norm": 5.1875,
      "learning_rate": 1.896698086447657e-08,
      "loss": 1.312,
      "num_input_tokens_seen": 82518704,
      "step": 39505
    },
    {
      "epoch": 6.445387062566278,
      "grad_norm": 7.6875,
      "learning_rate": 1.8586556621505436e-08,
      "loss": 2.4795,
      "num_input_tokens_seen": 82527280,
      "step": 39510
    },
    {
      "epoch": 6.446202789787095,
      "grad_norm": 6.15625,
      "learning_rate": 1.820998482642833e-08,
      "loss": 3.2399,
      "num_input_tokens_seen": 82538816,
      "step": 39515
    },
    {
      "epoch": 6.447018517007913,
      "grad_norm": 8.5625,
      "learning_rate": 1.7837265537309912e-08,
      "loss": 2.8413,
      "num_input_tokens_seen": 82548464,
      "step": 39520
    },
    {
      "epoch": 6.44783424422873,
      "grad_norm": 9.9375,
      "learning_rate": 1.7468398811629206e-08,
      "loss": 2.4373,
      "num_input_tokens_seen": 82559328,
      "step": 39525
    },
    {
      "epoch": 6.448649971449547,
      "grad_norm": 12.8125,
      "learning_rate": 1.710338470627404e-08,
      "loss": 2.7398,
      "num_input_tokens_seen": 82569888,
      "step": 39530
    },
    {
      "epoch": 6.449465698670364,
      "grad_norm": 11.1875,
      "learning_rate": 1.6742223277529945e-08,
      "loss": 1.8654,
      "num_input_tokens_seen": 82580416,
      "step": 39535
    },
    {
      "epoch": 6.450281425891182,
      "grad_norm": 7.75,
      "learning_rate": 1.6384914581094036e-08,
      "loss": 2.1097,
      "num_input_tokens_seen": 82588496,
      "step": 39540
    },
    {
      "epoch": 6.451097153111999,
      "grad_norm": 6.46875,
      "learning_rate": 1.6031458672069455e-08,
      "loss": 1.5562,
      "num_input_tokens_seen": 82598880,
      "step": 39545
    },
    {
      "epoch": 6.4519128803328165,
      "grad_norm": 5.65625,
      "learning_rate": 1.5681855604962602e-08,
      "loss": 2.7466,
      "num_input_tokens_seen": 82609360,
      "step": 39550
    },
    {
      "epoch": 6.4527286075536345,
      "grad_norm": 10.1875,
      "learning_rate": 1.5336105433683135e-08,
      "loss": 2.6548,
      "num_input_tokens_seen": 82618112,
      "step": 39555
    },
    {
      "epoch": 6.453544334774452,
      "grad_norm": 5.65625,
      "learning_rate": 1.499420821155506e-08,
      "loss": 2.6135,
      "num_input_tokens_seen": 82628160,
      "step": 39560
    },
    {
      "epoch": 6.454360061995269,
      "grad_norm": 5.78125,
      "learning_rate": 1.4656163991302874e-08,
      "loss": 2.1513,
      "num_input_tokens_seen": 82639168,
      "step": 39565
    },
    {
      "epoch": 6.455175789216086,
      "grad_norm": 7.96875,
      "learning_rate": 1.4321972825051544e-08,
      "loss": 2.153,
      "num_input_tokens_seen": 82649872,
      "step": 39570
    },
    {
      "epoch": 6.455991516436903,
      "grad_norm": 6.5625,
      "learning_rate": 1.3991634764345951e-08,
      "loss": 3.4268,
      "num_input_tokens_seen": 82661040,
      "step": 39575
    },
    {
      "epoch": 6.456807243657721,
      "grad_norm": 7.03125,
      "learning_rate": 1.3665149860120352e-08,
      "loss": 1.7663,
      "num_input_tokens_seen": 82671712,
      "step": 39580
    },
    {
      "epoch": 6.457622970878538,
      "grad_norm": 9.5625,
      "learning_rate": 1.3342518162728912e-08,
      "loss": 2.5543,
      "num_input_tokens_seen": 82681328,
      "step": 39585
    },
    {
      "epoch": 6.458438698099355,
      "grad_norm": 3.421875,
      "learning_rate": 1.30237397219235e-08,
      "loss": 3.1577,
      "num_input_tokens_seen": 82691968,
      "step": 39590
    },
    {
      "epoch": 6.459254425320173,
      "grad_norm": 6.59375,
      "learning_rate": 1.2708814586862016e-08,
      "loss": 3.3014,
      "num_input_tokens_seen": 82702032,
      "step": 39595
    },
    {
      "epoch": 6.4600701525409905,
      "grad_norm": 8.375,
      "learning_rate": 1.2397742806111168e-08,
      "loss": 2.0501,
      "num_input_tokens_seen": 82712800,
      "step": 39600
    },
    {
      "epoch": 6.4600701525409905,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.3462,
      "eval_samples_per_second": 20.283,
      "eval_steps_per_second": 10.145,
      "num_input_tokens_seen": 82712800,
      "step": 39600
    },
    {
      "epoch": 6.460885879761808,
      "grad_norm": 5.59375,
      "learning_rate": 1.209052442764369e-08,
      "loss": 1.4544,
      "num_input_tokens_seen": 82724720,
      "step": 39605
    },
    {
      "epoch": 6.461701606982625,
      "grad_norm": 4.53125,
      "learning_rate": 1.17871594988328e-08,
      "loss": 2.4737,
      "num_input_tokens_seen": 82734720,
      "step": 39610
    },
    {
      "epoch": 6.462517334203443,
      "grad_norm": 10.875,
      "learning_rate": 1.1487648066466072e-08,
      "loss": 2.4534,
      "num_input_tokens_seen": 82746320,
      "step": 39615
    },
    {
      "epoch": 6.46333306142426,
      "grad_norm": 4.28125,
      "learning_rate": 1.1191990176728784e-08,
      "loss": 3.0193,
      "num_input_tokens_seen": 82756992,
      "step": 39620
    },
    {
      "epoch": 6.464148788645077,
      "grad_norm": 3.484375,
      "learning_rate": 1.0900185875215018e-08,
      "loss": 1.8533,
      "num_input_tokens_seen": 82768560,
      "step": 39625
    },
    {
      "epoch": 6.464964515865894,
      "grad_norm": 8.1875,
      "learning_rate": 1.0612235206924891e-08,
      "loss": 1.4567,
      "num_input_tokens_seen": 82777632,
      "step": 39630
    },
    {
      "epoch": 6.465780243086712,
      "grad_norm": 9.625,
      "learning_rate": 1.0328138216264549e-08,
      "loss": 3.0455,
      "num_input_tokens_seen": 82787616,
      "step": 39635
    },
    {
      "epoch": 6.466595970307529,
      "grad_norm": 5.4375,
      "learning_rate": 1.004789494704339e-08,
      "loss": 2.3409,
      "num_input_tokens_seen": 82796992,
      "step": 39640
    },
    {
      "epoch": 6.467411697528346,
      "grad_norm": 5.40625,
      "learning_rate": 9.771505442482397e-09,
      "loss": 3.6424,
      "num_input_tokens_seen": 82807456,
      "step": 39645
    },
    {
      "epoch": 6.468227424749164,
      "grad_norm": 2.0,
      "learning_rate": 9.498969745200259e-09,
      "loss": 1.2649,
      "num_input_tokens_seen": 82818016,
      "step": 39650
    },
    {
      "epoch": 6.469043151969982,
      "grad_norm": 6.25,
      "learning_rate": 9.230287897230017e-09,
      "loss": 1.6902,
      "num_input_tokens_seen": 82827872,
      "step": 39655
    },
    {
      "epoch": 6.469858879190799,
      "grad_norm": 5.5,
      "learning_rate": 8.965459940002419e-09,
      "loss": 2.9727,
      "num_input_tokens_seen": 82837984,
      "step": 39660
    },
    {
      "epoch": 6.470674606411616,
      "grad_norm": 9.1875,
      "learning_rate": 8.704485914357019e-09,
      "loss": 2.9865,
      "num_input_tokens_seen": 82847584,
      "step": 39665
    },
    {
      "epoch": 6.471490333632433,
      "grad_norm": 6.09375,
      "learning_rate": 8.447365860539402e-09,
      "loss": 3.0881,
      "num_input_tokens_seen": 82857392,
      "step": 39670
    },
    {
      "epoch": 6.472306060853251,
      "grad_norm": 13.6875,
      "learning_rate": 8.194099818201184e-09,
      "loss": 2.4919,
      "num_input_tokens_seen": 82868144,
      "step": 39675
    },
    {
      "epoch": 6.473121788074068,
      "grad_norm": 5.75,
      "learning_rate": 7.944687826400011e-09,
      "loss": 3.0366,
      "num_input_tokens_seen": 82879424,
      "step": 39680
    },
    {
      "epoch": 6.473937515294885,
      "grad_norm": 4.71875,
      "learning_rate": 7.699129923599557e-09,
      "loss": 1.9371,
      "num_input_tokens_seen": 82888720,
      "step": 39685
    },
    {
      "epoch": 6.474753242515702,
      "grad_norm": 5.125,
      "learning_rate": 7.457426147663982e-09,
      "loss": 1.9816,
      "num_input_tokens_seen": 82899440,
      "step": 39690
    },
    {
      "epoch": 6.47556896973652,
      "grad_norm": 11.75,
      "learning_rate": 7.219576535871797e-09,
      "loss": 1.7721,
      "num_input_tokens_seen": 82908896,
      "step": 39695
    },
    {
      "epoch": 6.4763846969573375,
      "grad_norm": 6.9375,
      "learning_rate": 6.985581124896445e-09,
      "loss": 1.6336,
      "num_input_tokens_seen": 82919536,
      "step": 39700
    },
    {
      "epoch": 6.477200424178155,
      "grad_norm": 7.21875,
      "learning_rate": 6.755439950828501e-09,
      "loss": 2.6977,
      "num_input_tokens_seen": 82930080,
      "step": 39705
    },
    {
      "epoch": 6.478016151398972,
      "grad_norm": 6.0625,
      "learning_rate": 6.5291530491562444e-09,
      "loss": 2.0725,
      "num_input_tokens_seen": 82940800,
      "step": 39710
    },
    {
      "epoch": 6.47883187861979,
      "grad_norm": 9.6875,
      "learning_rate": 6.3067204547739845e-09,
      "loss": 2.6733,
      "num_input_tokens_seen": 82949920,
      "step": 39715
    },
    {
      "epoch": 6.479647605840607,
      "grad_norm": 14.0,
      "learning_rate": 6.088142201987612e-09,
      "loss": 1.7935,
      "num_input_tokens_seen": 82961248,
      "step": 39720
    },
    {
      "epoch": 6.480463333061424,
      "grad_norm": 7.53125,
      "learning_rate": 5.873418324503499e-09,
      "loss": 2.5349,
      "num_input_tokens_seen": 82972816,
      "step": 39725
    },
    {
      "epoch": 6.481279060282241,
      "grad_norm": 1.21875,
      "learning_rate": 5.6625488554340465e-09,
      "loss": 1.1489,
      "num_input_tokens_seen": 82983440,
      "step": 39730
    },
    {
      "epoch": 6.482094787503059,
      "grad_norm": 6.46875,
      "learning_rate": 5.455533827297688e-09,
      "loss": 2.0956,
      "num_input_tokens_seen": 82994512,
      "step": 39735
    },
    {
      "epoch": 6.482910514723876,
      "grad_norm": 5.0,
      "learning_rate": 5.252373272018885e-09,
      "loss": 2.0931,
      "num_input_tokens_seen": 83005104,
      "step": 39740
    },
    {
      "epoch": 6.4837262419446935,
      "grad_norm": 5.25,
      "learning_rate": 5.053067220925356e-09,
      "loss": 2.2688,
      "num_input_tokens_seen": 83017120,
      "step": 39745
    },
    {
      "epoch": 6.484541969165511,
      "grad_norm": 10.6875,
      "learning_rate": 4.857615704759177e-09,
      "loss": 2.7728,
      "num_input_tokens_seen": 83025888,
      "step": 39750
    },
    {
      "epoch": 6.485357696386329,
      "grad_norm": 9.3125,
      "learning_rate": 4.666018753654577e-09,
      "loss": 2.358,
      "num_input_tokens_seen": 83036416,
      "step": 39755
    },
    {
      "epoch": 6.486173423607146,
      "grad_norm": 6.03125,
      "learning_rate": 4.478276397162917e-09,
      "loss": 2.7612,
      "num_input_tokens_seen": 83046304,
      "step": 39760
    },
    {
      "epoch": 6.486989150827963,
      "grad_norm": 1.875,
      "learning_rate": 4.294388664233262e-09,
      "loss": 2.0276,
      "num_input_tokens_seen": 83055904,
      "step": 39765
    },
    {
      "epoch": 6.487804878048781,
      "grad_norm": 9.375,
      "learning_rate": 4.114355583223484e-09,
      "loss": 3.4098,
      "num_input_tokens_seen": 83065600,
      "step": 39770
    },
    {
      "epoch": 6.488620605269598,
      "grad_norm": 4.4375,
      "learning_rate": 3.9381771818974845e-09,
      "loss": 2.1891,
      "num_input_tokens_seen": 83076960,
      "step": 39775
    },
    {
      "epoch": 6.489436332490415,
      "grad_norm": 5.25,
      "learning_rate": 3.765853487427973e-09,
      "loss": 1.0714,
      "num_input_tokens_seen": 83088112,
      "step": 39780
    },
    {
      "epoch": 6.490252059711232,
      "grad_norm": 9.3125,
      "learning_rate": 3.5973845263825857e-09,
      "loss": 2.4803,
      "num_input_tokens_seen": 83097584,
      "step": 39785
    },
    {
      "epoch": 6.49106778693205,
      "grad_norm": 6.4375,
      "learning_rate": 3.4327703247488684e-09,
      "loss": 1.7519,
      "num_input_tokens_seen": 83106912,
      "step": 39790
    },
    {
      "epoch": 6.4918835141528675,
      "grad_norm": 8.5,
      "learning_rate": 3.2720109079037443e-09,
      "loss": 2.3986,
      "num_input_tokens_seen": 83117280,
      "step": 39795
    },
    {
      "epoch": 6.492699241373685,
      "grad_norm": 8.125,
      "learning_rate": 3.1151063006468193e-09,
      "loss": 2.6112,
      "num_input_tokens_seen": 83128704,
      "step": 39800
    },
    {
      "epoch": 6.492699241373685,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.3078,
      "eval_samples_per_second": 20.289,
      "eval_steps_per_second": 10.148,
      "num_input_tokens_seen": 83128704,
      "step": 39800
    },
    {
      "epoch": 6.493514968594502,
      "grad_norm": 5.96875,
      "learning_rate": 2.962056527169854e-09,
      "loss": 2.7234,
      "num_input_tokens_seen": 83139264,
      "step": 39805
    },
    {
      "epoch": 6.49433069581532,
      "grad_norm": 4.84375,
      "learning_rate": 2.8128616110761898e-09,
      "loss": 2.7636,
      "num_input_tokens_seen": 83149616,
      "step": 39810
    },
    {
      "epoch": 6.495146423036137,
      "grad_norm": 8.6875,
      "learning_rate": 2.6675215753724223e-09,
      "loss": 1.7141,
      "num_input_tokens_seen": 83160880,
      "step": 39815
    },
    {
      "epoch": 6.495962150256954,
      "grad_norm": 0.06494140625,
      "learning_rate": 2.5260364424739557e-09,
      "loss": 1.15,
      "num_input_tokens_seen": 83171664,
      "step": 39820
    },
    {
      "epoch": 6.496777877477771,
      "grad_norm": 0.75,
      "learning_rate": 2.3884062341994475e-09,
      "loss": 1.2836,
      "num_input_tokens_seen": 83182000,
      "step": 39825
    },
    {
      "epoch": 6.497593604698589,
      "grad_norm": 0.7109375,
      "learning_rate": 2.25463097177081e-09,
      "loss": 1.5591,
      "num_input_tokens_seen": 83190688,
      "step": 39830
    },
    {
      "epoch": 6.498409331919406,
      "grad_norm": 9.8125,
      "learning_rate": 2.1247106758215397e-09,
      "loss": 2.2348,
      "num_input_tokens_seen": 83199120,
      "step": 39835
    },
    {
      "epoch": 6.499225059140223,
      "grad_norm": 9.1875,
      "learning_rate": 1.998645366382834e-09,
      "loss": 3.8195,
      "num_input_tokens_seen": 83210832,
      "step": 39840
    },
    {
      "epoch": 6.5000407863610405,
      "grad_norm": 2.125,
      "learning_rate": 1.876435062897475e-09,
      "loss": 3.2435,
      "num_input_tokens_seen": 83222880,
      "step": 39845
    },
    {
      "epoch": 6.500856513581859,
      "grad_norm": 11.375,
      "learning_rate": 1.758079784211497e-09,
      "loss": 2.74,
      "num_input_tokens_seen": 83233184,
      "step": 39850
    },
    {
      "epoch": 6.501672240802676,
      "grad_norm": 9.75,
      "learning_rate": 1.6435795485797434e-09,
      "loss": 1.5031,
      "num_input_tokens_seen": 83243648,
      "step": 39855
    },
    {
      "epoch": 6.502487968023493,
      "grad_norm": 4.34375,
      "learning_rate": 1.5329343736547596e-09,
      "loss": 3.1175,
      "num_input_tokens_seen": 83254464,
      "step": 39860
    },
    {
      "epoch": 6.50330369524431,
      "grad_norm": 10.875,
      "learning_rate": 1.4261442765006739e-09,
      "loss": 3.1733,
      "num_input_tokens_seen": 83264800,
      "step": 39865
    },
    {
      "epoch": 6.504119422465128,
      "grad_norm": 4.40625,
      "learning_rate": 1.3232092735876445e-09,
      "loss": 1.3792,
      "num_input_tokens_seen": 83275616,
      "step": 39870
    },
    {
      "epoch": 6.504935149685945,
      "grad_norm": 6.65625,
      "learning_rate": 1.2241293807918607e-09,
      "loss": 2.2311,
      "num_input_tokens_seen": 83285712,
      "step": 39875
    },
    {
      "epoch": 6.505750876906762,
      "grad_norm": 6.78125,
      "learning_rate": 1.128904613387216e-09,
      "loss": 2.0064,
      "num_input_tokens_seen": 83297248,
      "step": 39880
    },
    {
      "epoch": 6.506566604127579,
      "grad_norm": 7.1875,
      "learning_rate": 1.0375349860591853e-09,
      "loss": 2.6212,
      "num_input_tokens_seen": 83308752,
      "step": 39885
    },
    {
      "epoch": 6.507382331348397,
      "grad_norm": 7.8125,
      "learning_rate": 9.5002051290205e-10,
      "loss": 1.6593,
      "num_input_tokens_seen": 83317312,
      "step": 39890
    },
    {
      "epoch": 6.5081980585692145,
      "grad_norm": 5.28125,
      "learning_rate": 8.663612074077954e-10,
      "loss": 3.4763,
      "num_input_tokens_seen": 83327600,
      "step": 39895
    },
    {
      "epoch": 6.509013785790032,
      "grad_norm": 17.5,
      "learning_rate": 7.865570824799884e-10,
      "loss": 2.7803,
      "num_input_tokens_seen": 83338448,
      "step": 39900
    },
    {
      "epoch": 6.50982951301085,
      "grad_norm": 9.875,
      "learning_rate": 7.106081504254514e-10,
      "loss": 3.8732,
      "num_input_tokens_seen": 83348896,
      "step": 39905
    },
    {
      "epoch": 6.510645240231667,
      "grad_norm": 6.75,
      "learning_rate": 6.385144229570372e-10,
      "loss": 1.9264,
      "num_input_tokens_seen": 83359104,
      "step": 39910
    },
    {
      "epoch": 6.511460967452484,
      "grad_norm": 4.34375,
      "learning_rate": 5.70275911190854e-10,
      "loss": 2.5507,
      "num_input_tokens_seen": 83368384,
      "step": 39915
    },
    {
      "epoch": 6.512276694673301,
      "grad_norm": 10.8125,
      "learning_rate": 5.058926256490403e-10,
      "loss": 2.3565,
      "num_input_tokens_seen": 83378352,
      "step": 39920
    },
    {
      "epoch": 6.513092421894118,
      "grad_norm": 8.3125,
      "learning_rate": 4.4536457626254134e-10,
      "loss": 3.8525,
      "num_input_tokens_seen": 83388128,
      "step": 39925
    },
    {
      "epoch": 6.513908149114936,
      "grad_norm": 10.0,
      "learning_rate": 3.88691772365557e-10,
      "loss": 4.0576,
      "num_input_tokens_seen": 83398288,
      "step": 39930
    },
    {
      "epoch": 6.514723876335753,
      "grad_norm": 4.96875,
      "learning_rate": 3.358742226955425e-10,
      "loss": 1.1232,
      "num_input_tokens_seen": 83408256,
      "step": 39935
    },
    {
      "epoch": 6.5155396035565705,
      "grad_norm": 4.28125,
      "learning_rate": 2.8691193539875925e-10,
      "loss": 2.312,
      "num_input_tokens_seen": 83418560,
      "step": 39940
    },
    {
      "epoch": 6.5163553307773885,
      "grad_norm": 11.6875,
      "learning_rate": 2.418049180274995e-10,
      "loss": 2.6746,
      "num_input_tokens_seen": 83429440,
      "step": 39945
    },
    {
      "epoch": 6.517171057998206,
      "grad_norm": 9.875,
      "learning_rate": 2.005531775373104e-10,
      "loss": 3.634,
      "num_input_tokens_seen": 83439312,
      "step": 39950
    },
    {
      "epoch": 6.517986785219023,
      "grad_norm": 7.125,
      "learning_rate": 1.6315672028699435e-10,
      "loss": 1.9588,
      "num_input_tokens_seen": 83449776,
      "step": 39955
    },
    {
      "epoch": 6.51880251243984,
      "grad_norm": 12.1875,
      "learning_rate": 1.2961555204693555e-10,
      "loss": 2.1335,
      "num_input_tokens_seen": 83460704,
      "step": 39960
    },
    {
      "epoch": 6.519618239660657,
      "grad_norm": 7.125,
      "learning_rate": 9.992967798799768e-11,
      "loss": 2.24,
      "num_input_tokens_seen": 83470784,
      "step": 39965
    },
    {
      "epoch": 6.520433966881475,
      "grad_norm": 9.3125,
      "learning_rate": 7.409910268707521e-11,
      "loss": 3.2386,
      "num_input_tokens_seen": 83481760,
      "step": 39970
    },
    {
      "epoch": 6.521249694102292,
      "grad_norm": 5.53125,
      "learning_rate": 5.212383012986877e-11,
      "loss": 1.9327,
      "num_input_tokens_seen": 83490128,
      "step": 39975
    },
    {
      "epoch": 6.522065421323109,
      "grad_norm": 0.33984375,
      "learning_rate": 3.400386370533415e-11,
      "loss": 2.0816,
      "num_input_tokens_seen": 83499984,
      "step": 39980
    },
    {
      "epoch": 6.522881148543927,
      "grad_norm": 7.84375,
      "learning_rate": 1.9739206205682258e-11,
      "loss": 3.2304,
      "num_input_tokens_seen": 83511088,
      "step": 39985
    },
    {
      "epoch": 6.523696875764744,
      "grad_norm": 9.3125,
      "learning_rate": 9.329859829154685e-12,
      "loss": 4.2414,
      "num_input_tokens_seen": 83522192,
      "step": 39990
    },
    {
      "epoch": 6.524512602985562,
      "grad_norm": 12.3125,
      "learning_rate": 2.7758261855748148e-12,
      "loss": 2.0651,
      "num_input_tokens_seen": 83532624,
      "step": 39995
    },
    {
      "epoch": 6.525328330206379,
      "grad_norm": 6.90625,
      "learning_rate": 7.710628524559838e-14,
      "loss": 2.4207,
      "num_input_tokens_seen": 83543088,
      "step": 40000
    },
    {
      "epoch": 6.525328330206379,
      "eval_loss": 2.5388035774230957,
      "eval_runtime": 134.1733,
      "eval_samples_per_second": 20.31,
      "eval_steps_per_second": 10.159,
      "num_input_tokens_seen": 83543088,
      "step": 40000
    },
    {
      "epoch": 6.525328330206379,
      "num_input_tokens_seen": 83543088,
      "step": 40000,
      "total_flos": 3.565989290886562e+18,
      "train_loss": 2.3986957787156107,
      "train_runtime": 52463.916,
      "train_samples_per_second": 3.05,
      "train_steps_per_second": 0.762
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 83543088,
  "num_train_epochs": 7,
  "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": 3.565989290886562e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}