{
  "best_global_step": 36252,
  "best_metric": 0.45250704884529114,
  "best_model_checkpoint": "saves_multiple/lntuning/llama-3-8b-instruct/train_codealpacapy_42_1760719585/checkpoint-36252",
  "epoch": 20.0,
  "eval_steps": 1908,
  "global_step": 38160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002620545073375262,
      "grad_norm": 1.343209147453308,
      "learning_rate": 5.241090146750525e-08,
      "loss": 1.1831,
      "num_input_tokens_seen": 3264,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 1.1109672784805298,
      "learning_rate": 1.1792452830188679e-07,
      "loss": 1.2437,
      "num_input_tokens_seen": 5888,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 1.108894944190979,
      "learning_rate": 1.8343815513626837e-07,
      "loss": 1.348,
      "num_input_tokens_seen": 8480,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 2.9285027980804443,
      "learning_rate": 2.489517819706499e-07,
      "loss": 1.4211,
      "num_input_tokens_seen": 11488,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 1.707275629043579,
      "learning_rate": 3.144654088050315e-07,
      "loss": 1.1751,
      "num_input_tokens_seen": 14880,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 1.2675799131393433,
      "learning_rate": 3.79979035639413e-07,
      "loss": 1.0908,
      "num_input_tokens_seen": 18592,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 1.9330321550369263,
      "learning_rate": 4.454926624737945e-07,
      "loss": 1.2527,
      "num_input_tokens_seen": 21568,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 1.7667961120605469,
      "learning_rate": 5.110062893081761e-07,
      "loss": 1.5206,
      "num_input_tokens_seen": 24160,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 3.126648426055908,
      "learning_rate": 5.765199161425577e-07,
      "loss": 1.4199,
      "num_input_tokens_seen": 27168,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 1.0990729331970215,
      "learning_rate": 6.420335429769392e-07,
      "loss": 1.1084,
      "num_input_tokens_seen": 30080,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 2.1011674404144287,
      "learning_rate": 7.075471698113208e-07,
      "loss": 1.4812,
      "num_input_tokens_seen": 32608,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 0.7242990732192993,
      "learning_rate": 7.730607966457024e-07,
      "loss": 1.1647,
      "num_input_tokens_seen": 36448,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 1.2255351543426514,
      "learning_rate": 8.38574423480084e-07,
      "loss": 1.0807,
      "num_input_tokens_seen": 38944,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 3.1643972396850586,
      "learning_rate": 9.040880503144654e-07,
      "loss": 1.4659,
      "num_input_tokens_seen": 41824,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 1.2606555223464966,
      "learning_rate": 9.69601677148847e-07,
      "loss": 1.1347,
      "num_input_tokens_seen": 44896,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 1.5462985038757324,
      "learning_rate": 1.0351153039832286e-06,
      "loss": 1.1628,
      "num_input_tokens_seen": 47616,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 1.4435269832611084,
      "learning_rate": 1.10062893081761e-06,
      "loss": 1.3055,
      "num_input_tokens_seen": 50368,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 1.5875343084335327,
      "learning_rate": 1.1661425576519918e-06,
      "loss": 1.1198,
      "num_input_tokens_seen": 54016,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 0.9061422348022461,
      "learning_rate": 1.2316561844863732e-06,
      "loss": 0.8774,
      "num_input_tokens_seen": 57888,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 1.5927069187164307,
      "learning_rate": 1.2971698113207548e-06,
      "loss": 1.3878,
      "num_input_tokens_seen": 61024,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 1.5498758554458618,
      "learning_rate": 1.3626834381551364e-06,
      "loss": 1.126,
      "num_input_tokens_seen": 64672,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 1.2413333654403687,
      "learning_rate": 1.4281970649895177e-06,
      "loss": 1.0558,
      "num_input_tokens_seen": 67616,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 0.7922526597976685,
      "learning_rate": 1.4937106918238995e-06,
      "loss": 1.0369,
      "num_input_tokens_seen": 71232,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 1.8174715042114258,
      "learning_rate": 1.559224318658281e-06,
      "loss": 0.8856,
      "num_input_tokens_seen": 74880,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 3.411499261856079,
      "learning_rate": 1.6247379454926625e-06,
      "loss": 1.2109,
      "num_input_tokens_seen": 78432,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 0.9636257290840149,
      "learning_rate": 1.690251572327044e-06,
      "loss": 1.0909,
      "num_input_tokens_seen": 81088,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 0.7775359749794006,
      "learning_rate": 1.7557651991614257e-06,
      "loss": 1.0198,
      "num_input_tokens_seen": 84896,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 0.9074620604515076,
      "learning_rate": 1.821278825995807e-06,
      "loss": 1.2627,
      "num_input_tokens_seen": 88736,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 2.5762107372283936,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 1.5872,
      "num_input_tokens_seen": 92224,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 1.743012547492981,
      "learning_rate": 1.9523060796645703e-06,
      "loss": 1.3305,
      "num_input_tokens_seen": 95168,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 1.3995047807693481,
      "learning_rate": 2.017819706498952e-06,
      "loss": 0.9558,
      "num_input_tokens_seen": 98976,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 0.7248391509056091,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.9743,
      "num_input_tokens_seen": 102176,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 1.2756186723709106,
      "learning_rate": 2.148846960167715e-06,
      "loss": 0.949,
      "num_input_tokens_seen": 105824,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 1.0563924312591553,
      "learning_rate": 2.2143605870020966e-06,
      "loss": 1.1178,
      "num_input_tokens_seen": 108512,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 1.0493009090423584,
      "learning_rate": 2.279874213836478e-06,
      "loss": 0.9909,
      "num_input_tokens_seen": 111872,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.7699776887893677,
      "learning_rate": 2.34538784067086e-06,
      "loss": 1.0426,
      "num_input_tokens_seen": 115552,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 1.45691978931427,
      "learning_rate": 2.410901467505241e-06,
      "loss": 1.239,
      "num_input_tokens_seen": 118784,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 1.730857253074646,
      "learning_rate": 2.476415094339623e-06,
      "loss": 1.0246,
      "num_input_tokens_seen": 122528,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 0.9512796998023987,
      "learning_rate": 2.5419287211740044e-06,
      "loss": 1.3984,
      "num_input_tokens_seen": 125024,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 1.230554223060608,
      "learning_rate": 2.6074423480083857e-06,
      "loss": 1.2403,
      "num_input_tokens_seen": 128096,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 1.7916316986083984,
      "learning_rate": 2.6729559748427675e-06,
      "loss": 1.5496,
      "num_input_tokens_seen": 130496,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.1201168298721313,
      "learning_rate": 2.738469601677149e-06,
      "loss": 1.2499,
      "num_input_tokens_seen": 133152,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 2.0702173709869385,
      "learning_rate": 2.8039832285115303e-06,
      "loss": 1.4203,
      "num_input_tokens_seen": 136384,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 1.3759485483169556,
      "learning_rate": 2.869496855345912e-06,
      "loss": 1.2109,
      "num_input_tokens_seen": 139456,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 4.843760013580322,
      "learning_rate": 2.935010482180294e-06,
      "loss": 1.4333,
      "num_input_tokens_seen": 142816,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 1.082629919052124,
      "learning_rate": 3.0005241090146753e-06,
      "loss": 1.2697,
      "num_input_tokens_seen": 145696,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 1.008807897567749,
      "learning_rate": 3.0660377358490567e-06,
      "loss": 1.2018,
      "num_input_tokens_seen": 149024,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 2.7648422718048096,
      "learning_rate": 3.131551362683438e-06,
      "loss": 1.3744,
      "num_input_tokens_seen": 151840,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 1.1449764966964722,
      "learning_rate": 3.1970649895178203e-06,
      "loss": 1.1831,
      "num_input_tokens_seen": 155456,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 1.542237639427185,
      "learning_rate": 3.2625786163522012e-06,
      "loss": 1.2365,
      "num_input_tokens_seen": 158528,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 1.3415334224700928,
      "learning_rate": 3.3280922431865826e-06,
      "loss": 0.9354,
      "num_input_tokens_seen": 166496,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 0.9699568748474121,
      "learning_rate": 3.393605870020965e-06,
      "loss": 1.0281,
      "num_input_tokens_seen": 169568,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.567687749862671,
      "learning_rate": 3.459119496855346e-06,
      "loss": 1.1861,
      "num_input_tokens_seen": 172704,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 0.9207450151443481,
      "learning_rate": 3.524633123689727e-06,
      "loss": 1.2347,
      "num_input_tokens_seen": 176480,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 0.7796575427055359,
      "learning_rate": 3.5901467505241094e-06,
      "loss": 1.0775,
      "num_input_tokens_seen": 181408,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 1.2570478916168213,
      "learning_rate": 3.6556603773584908e-06,
      "loss": 1.2125,
      "num_input_tokens_seen": 184736,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 1.6808645725250244,
      "learning_rate": 3.7211740041928726e-06,
      "loss": 1.1424,
      "num_input_tokens_seen": 188352,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 1.684462547302246,
      "learning_rate": 3.786687631027254e-06,
      "loss": 1.0782,
      "num_input_tokens_seen": 193408,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 1.057572603225708,
      "learning_rate": 3.852201257861635e-06,
      "loss": 1.0261,
      "num_input_tokens_seen": 196384,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 2.4915456771850586,
      "learning_rate": 3.917714884696017e-06,
      "loss": 1.2104,
      "num_input_tokens_seen": 199360,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 0.826982319355011,
      "learning_rate": 3.9832285115303985e-06,
      "loss": 1.087,
      "num_input_tokens_seen": 202848,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 1.2915829420089722,
      "learning_rate": 4.04874213836478e-06,
      "loss": 0.9271,
      "num_input_tokens_seen": 207104,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 1.8397098779678345,
      "learning_rate": 4.114255765199162e-06,
      "loss": 1.3985,
      "num_input_tokens_seen": 209600,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 1.1439168453216553,
      "learning_rate": 4.1797693920335435e-06,
      "loss": 1.2418,
      "num_input_tokens_seen": 213152,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 1.861539602279663,
      "learning_rate": 4.245283018867925e-06,
      "loss": 1.4177,
      "num_input_tokens_seen": 215616,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 1.10883629322052,
      "learning_rate": 4.310796645702306e-06,
      "loss": 1.1862,
      "num_input_tokens_seen": 218528,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 2.201871633529663,
      "learning_rate": 4.376310272536688e-06,
      "loss": 1.0566,
      "num_input_tokens_seen": 221824,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 1.0181344747543335,
      "learning_rate": 4.441823899371069e-06,
      "loss": 1.0487,
      "num_input_tokens_seen": 225280,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 1.0296322107315063,
      "learning_rate": 4.507337526205451e-06,
      "loss": 0.987,
      "num_input_tokens_seen": 230912,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 1.0923535823822021,
      "learning_rate": 4.572851153039833e-06,
      "loss": 1.146,
      "num_input_tokens_seen": 234592,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 1.1345596313476562,
      "learning_rate": 4.638364779874214e-06,
      "loss": 0.9917,
      "num_input_tokens_seen": 237600,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 1.221334457397461,
      "learning_rate": 4.703878406708595e-06,
      "loss": 1.1355,
      "num_input_tokens_seen": 241088,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 1.6890984773635864,
      "learning_rate": 4.769392033542977e-06,
      "loss": 1.2241,
      "num_input_tokens_seen": 243520,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 2.753268241882324,
      "learning_rate": 4.834905660377359e-06,
      "loss": 1.3376,
      "num_input_tokens_seen": 247456,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 0.7461223602294922,
      "learning_rate": 4.90041928721174e-06,
      "loss": 1.1097,
      "num_input_tokens_seen": 250784,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 1.3616230487823486,
      "learning_rate": 4.965932914046122e-06,
      "loss": 1.0079,
      "num_input_tokens_seen": 255040,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 1.5331827402114868,
      "learning_rate": 5.031446540880504e-06,
      "loss": 1.181,
      "num_input_tokens_seen": 258240,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 0.7402596473693848,
      "learning_rate": 5.096960167714885e-06,
      "loss": 1.016,
      "num_input_tokens_seen": 261152,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 0.989386796951294,
      "learning_rate": 5.162473794549266e-06,
      "loss": 1.346,
      "num_input_tokens_seen": 264544,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 1.8252570629119873,
      "learning_rate": 5.227987421383648e-06,
      "loss": 1.2918,
      "num_input_tokens_seen": 267136,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 1.1592501401901245,
      "learning_rate": 5.2935010482180295e-06,
      "loss": 1.2002,
      "num_input_tokens_seen": 270464,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 1.306862711906433,
      "learning_rate": 5.359014675052411e-06,
      "loss": 0.9004,
      "num_input_tokens_seen": 274528,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 1.2651081085205078,
      "learning_rate": 5.424528301886793e-06,
      "loss": 1.0571,
      "num_input_tokens_seen": 277376,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 1.424485445022583,
      "learning_rate": 5.4900419287211744e-06,
      "loss": 1.2608,
      "num_input_tokens_seen": 279616,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 3.3423519134521484,
      "learning_rate": 5.555555555555556e-06,
      "loss": 1.1058,
      "num_input_tokens_seen": 282656,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 1.6310440301895142,
      "learning_rate": 5.621069182389937e-06,
      "loss": 1.089,
      "num_input_tokens_seen": 285600,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 0.6712497472763062,
      "learning_rate": 5.686582809224319e-06,
      "loss": 0.8637,
      "num_input_tokens_seen": 289568,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 1.397294044494629,
      "learning_rate": 5.752096436058701e-06,
      "loss": 1.1407,
      "num_input_tokens_seen": 292768,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 1.8892364501953125,
      "learning_rate": 5.817610062893082e-06,
      "loss": 1.2331,
      "num_input_tokens_seen": 295456,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 1.3704931735992432,
      "learning_rate": 5.8831236897274636e-06,
      "loss": 1.1639,
      "num_input_tokens_seen": 299488,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 1.9321327209472656,
      "learning_rate": 5.948637316561846e-06,
      "loss": 1.1761,
      "num_input_tokens_seen": 302976,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 1.187753438949585,
      "learning_rate": 6.014150943396227e-06,
      "loss": 1.0252,
      "num_input_tokens_seen": 306560,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 1.4470839500427246,
      "learning_rate": 6.079664570230608e-06,
      "loss": 1.049,
      "num_input_tokens_seen": 309536,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 1.180676817893982,
      "learning_rate": 6.14517819706499e-06,
      "loss": 0.9718,
      "num_input_tokens_seen": 312576,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 1.1156386137008667,
      "learning_rate": 6.210691823899371e-06,
      "loss": 1.0296,
      "num_input_tokens_seen": 315616,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 1.016672134399414,
      "learning_rate": 6.2762054507337535e-06,
      "loss": 0.9917,
      "num_input_tokens_seen": 319136,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 1.37688148021698,
      "learning_rate": 6.341719077568134e-06,
      "loss": 1.7992,
      "num_input_tokens_seen": 321376,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 1.8812915086746216,
      "learning_rate": 6.407232704402516e-06,
      "loss": 1.0498,
      "num_input_tokens_seen": 324928,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 1.4240127801895142,
      "learning_rate": 6.472746331236898e-06,
      "loss": 1.1065,
      "num_input_tokens_seen": 329248,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 1.4426181316375732,
      "learning_rate": 6.538259958071278e-06,
      "loss": 1.1709,
      "num_input_tokens_seen": 332288,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 1.3757352828979492,
      "learning_rate": 6.60377358490566e-06,
      "loss": 1.2378,
      "num_input_tokens_seen": 334752,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 1.1979219913482666,
      "learning_rate": 6.669287211740043e-06,
      "loss": 1.1834,
      "num_input_tokens_seen": 337856,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 0.9696452617645264,
      "learning_rate": 6.734800838574423e-06,
      "loss": 1.0229,
      "num_input_tokens_seen": 341792,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 1.3187906742095947,
      "learning_rate": 6.800314465408805e-06,
      "loss": 1.246,
      "num_input_tokens_seen": 344768,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 1.3045439720153809,
      "learning_rate": 6.865828092243188e-06,
      "loss": 1.0913,
      "num_input_tokens_seen": 347360,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 1.0063735246658325,
      "learning_rate": 6.931341719077568e-06,
      "loss": 1.1348,
      "num_input_tokens_seen": 350592,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 1.1051359176635742,
      "learning_rate": 6.9968553459119495e-06,
      "loss": 1.0046,
      "num_input_tokens_seen": 353568,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 0.9048007130622864,
      "learning_rate": 7.062368972746332e-06,
      "loss": 1.2658,
      "num_input_tokens_seen": 356320,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 1.0685733556747437,
      "learning_rate": 7.127882599580712e-06,
      "loss": 0.8918,
      "num_input_tokens_seen": 359968,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 1.0294301509857178,
      "learning_rate": 7.1933962264150945e-06,
      "loss": 0.9517,
      "num_input_tokens_seen": 362752,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 1.1462289094924927,
      "learning_rate": 7.258909853249477e-06,
      "loss": 1.0871,
      "num_input_tokens_seen": 365312,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 1.0329763889312744,
      "learning_rate": 7.324423480083858e-06,
      "loss": 0.9237,
      "num_input_tokens_seen": 368896,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 1.2015678882598877,
      "learning_rate": 7.3899371069182395e-06,
      "loss": 1.0478,
      "num_input_tokens_seen": 371648,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 1.251394271850586,
      "learning_rate": 7.455450733752621e-06,
      "loss": 1.0252,
      "num_input_tokens_seen": 374848,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 2.450932741165161,
      "learning_rate": 7.520964360587003e-06,
      "loss": 1.2862,
      "num_input_tokens_seen": 377184,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 0.7118914127349854,
      "learning_rate": 7.586477987421384e-06,
      "loss": 0.9097,
      "num_input_tokens_seen": 381504,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 1.0796586275100708,
      "learning_rate": 7.651991614255765e-06,
      "loss": 0.9704,
      "num_input_tokens_seen": 384672,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 1.196104645729065,
      "learning_rate": 7.717505241090148e-06,
      "loss": 0.9229,
      "num_input_tokens_seen": 389376,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 1.0120744705200195,
      "learning_rate": 7.783018867924528e-06,
      "loss": 0.9882,
      "num_input_tokens_seen": 392512,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 1.198058009147644,
      "learning_rate": 7.848532494758911e-06,
      "loss": 0.9624,
      "num_input_tokens_seen": 396864,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 1.6415655612945557,
      "learning_rate": 7.914046121593292e-06,
      "loss": 1.0595,
      "num_input_tokens_seen": 399520,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 3.6183407306671143,
      "learning_rate": 7.979559748427674e-06,
      "loss": 1.1716,
      "num_input_tokens_seen": 402560,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 1.3114854097366333,
      "learning_rate": 8.045073375262055e-06,
      "loss": 1.0467,
      "num_input_tokens_seen": 406240,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 1.2058866024017334,
      "learning_rate": 8.110587002096436e-06,
      "loss": 1.0783,
      "num_input_tokens_seen": 408960,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 0.9372329711914062,
      "learning_rate": 8.176100628930818e-06,
      "loss": 1.1026,
      "num_input_tokens_seen": 412224,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 0.9503316879272461,
      "learning_rate": 8.241614255765199e-06,
      "loss": 0.8968,
      "num_input_tokens_seen": 415296,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 2.33463716506958,
      "learning_rate": 8.307127882599582e-06,
      "loss": 1.1176,
      "num_input_tokens_seen": 417504,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 0.9516662955284119,
      "learning_rate": 8.372641509433962e-06,
      "loss": 0.9334,
      "num_input_tokens_seen": 420288,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 1.5801048278808594,
      "learning_rate": 8.438155136268345e-06,
      "loss": 1.1693,
      "num_input_tokens_seen": 422560,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 2.3130335807800293,
      "learning_rate": 8.503668763102726e-06,
      "loss": 1.1429,
      "num_input_tokens_seen": 425472,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 1.3732738494873047,
      "learning_rate": 8.569182389937106e-06,
      "loss": 0.9116,
      "num_input_tokens_seen": 429056,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 1.312196969985962,
      "learning_rate": 8.634696016771489e-06,
      "loss": 0.8381,
      "num_input_tokens_seen": 433024,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 1.1372328996658325,
      "learning_rate": 8.70020964360587e-06,
      "loss": 0.9827,
      "num_input_tokens_seen": 435776,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 1.2353954315185547,
      "learning_rate": 8.765723270440252e-06,
      "loss": 1.0833,
      "num_input_tokens_seen": 438432,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 1.5042916536331177,
      "learning_rate": 8.831236897274633e-06,
      "loss": 0.8643,
      "num_input_tokens_seen": 442400,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 1.6977049112319946,
      "learning_rate": 8.896750524109015e-06,
      "loss": 0.998,
      "num_input_tokens_seen": 445888,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 0.5969731211662292,
      "learning_rate": 8.962264150943396e-06,
      "loss": 0.9173,
      "num_input_tokens_seen": 448864,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 1.1736764907836914,
      "learning_rate": 9.027777777777777e-06,
      "loss": 1.0871,
      "num_input_tokens_seen": 452640,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 1.214680790901184,
      "learning_rate": 9.09329140461216e-06,
      "loss": 1.0848,
      "num_input_tokens_seen": 455296,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 2.4893646240234375,
      "learning_rate": 9.15880503144654e-06,
      "loss": 1.0012,
      "num_input_tokens_seen": 457888,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 1.2184022665023804,
      "learning_rate": 9.224318658280923e-06,
      "loss": 0.888,
      "num_input_tokens_seen": 461664,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 0.9097460508346558,
      "learning_rate": 9.289832285115305e-06,
      "loss": 0.9659,
      "num_input_tokens_seen": 464352,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 1.3677425384521484,
      "learning_rate": 9.355345911949686e-06,
      "loss": 0.9356,
      "num_input_tokens_seen": 468256,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 2.634624481201172,
      "learning_rate": 9.420859538784067e-06,
      "loss": 1.0206,
      "num_input_tokens_seen": 471104,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 2.564868211746216,
      "learning_rate": 9.486373165618449e-06,
      "loss": 1.1537,
      "num_input_tokens_seen": 473696,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 1.140637993812561,
      "learning_rate": 9.551886792452832e-06,
      "loss": 0.8547,
      "num_input_tokens_seen": 476512,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 1.0203235149383545,
      "learning_rate": 9.617400419287211e-06,
      "loss": 0.8992,
      "num_input_tokens_seen": 480128,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 2.1597654819488525,
      "learning_rate": 9.682914046121595e-06,
      "loss": 0.9139,
      "num_input_tokens_seen": 483808,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 1.7101713418960571,
      "learning_rate": 9.748427672955976e-06,
      "loss": 1.1133,
      "num_input_tokens_seen": 486240,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 0.8150107860565186,
      "learning_rate": 9.813941299790357e-06,
      "loss": 1.1958,
      "num_input_tokens_seen": 488448,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 2.707306146621704,
      "learning_rate": 9.879454926624739e-06,
      "loss": 0.9657,
      "num_input_tokens_seen": 491392,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 1.2705503702163696,
      "learning_rate": 9.94496855345912e-06,
      "loss": 0.8531,
      "num_input_tokens_seen": 495232,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 1.5551810264587402,
      "learning_rate": 1.0010482180293501e-05,
      "loss": 0.8805,
      "num_input_tokens_seen": 498016,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 1.4043138027191162,
      "learning_rate": 1.0075995807127883e-05,
      "loss": 0.8561,
      "num_input_tokens_seen": 501152,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 1.0459619760513306,
      "learning_rate": 1.0141509433962266e-05,
      "loss": 0.9297,
      "num_input_tokens_seen": 503872,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 1.0693553686141968,
      "learning_rate": 1.0207023060796646e-05,
      "loss": 0.8837,
      "num_input_tokens_seen": 507328,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 0.8091657161712646,
      "learning_rate": 1.0272536687631027e-05,
      "loss": 0.7091,
      "num_input_tokens_seen": 510912,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 0.7429041862487793,
      "learning_rate": 1.033805031446541e-05,
      "loss": 0.8577,
      "num_input_tokens_seen": 514304,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 1.2163338661193848,
      "learning_rate": 1.040356394129979e-05,
      "loss": 0.874,
      "num_input_tokens_seen": 517440,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 1.5204576253890991,
      "learning_rate": 1.0469077568134173e-05,
      "loss": 1.1037,
      "num_input_tokens_seen": 520416,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 1.16669499874115,
      "learning_rate": 1.0534591194968554e-05,
      "loss": 1.0693,
      "num_input_tokens_seen": 523456,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 0.8941885828971863,
      "learning_rate": 1.0600104821802936e-05,
      "loss": 0.7801,
      "num_input_tokens_seen": 526720,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 1.18173348903656,
      "learning_rate": 1.0665618448637317e-05,
      "loss": 0.7334,
      "num_input_tokens_seen": 530272,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 1.3914932012557983,
      "learning_rate": 1.0731132075471698e-05,
      "loss": 0.9566,
      "num_input_tokens_seen": 533408,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 3.1085400581359863,
      "learning_rate": 1.079664570230608e-05,
      "loss": 1.0542,
      "num_input_tokens_seen": 536768,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 0.5314303636550903,
      "learning_rate": 1.0862159329140461e-05,
      "loss": 0.8357,
      "num_input_tokens_seen": 541184,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 1.3817431926727295,
      "learning_rate": 1.0927672955974844e-05,
      "loss": 0.7165,
      "num_input_tokens_seen": 544576,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 0.9881147146224976,
      "learning_rate": 1.0993186582809224e-05,
      "loss": 0.8549,
      "num_input_tokens_seen": 549408,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 1.355368971824646,
      "learning_rate": 1.1058700209643607e-05,
      "loss": 0.8129,
      "num_input_tokens_seen": 552032,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 1.5781805515289307,
      "learning_rate": 1.1124213836477988e-05,
      "loss": 0.7198,
      "num_input_tokens_seen": 555872,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 1.1056443452835083,
      "learning_rate": 1.118972746331237e-05,
      "loss": 0.923,
      "num_input_tokens_seen": 559136,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 0.743483304977417,
      "learning_rate": 1.1255241090146751e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 561984,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 0.790386438369751,
      "learning_rate": 1.1320754716981132e-05,
      "loss": 0.7921,
      "num_input_tokens_seen": 564448,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 1.5456799268722534,
      "learning_rate": 1.1386268343815514e-05,
      "loss": 1.0655,
      "num_input_tokens_seen": 566752,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 0.8966153860092163,
      "learning_rate": 1.1451781970649895e-05,
      "loss": 0.774,
      "num_input_tokens_seen": 569856,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 1.147105097770691,
      "learning_rate": 1.1517295597484278e-05,
      "loss": 0.7655,
      "num_input_tokens_seen": 573056,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 0.9686053395271301,
      "learning_rate": 1.158280922431866e-05,
      "loss": 0.8161,
      "num_input_tokens_seen": 575712,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 0.7557579874992371,
      "learning_rate": 1.164832285115304e-05,
      "loss": 0.7528,
      "num_input_tokens_seen": 578880,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 0.8687390089035034,
      "learning_rate": 1.1713836477987422e-05,
      "loss": 0.7216,
      "num_input_tokens_seen": 581920,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 1.029868721961975,
      "learning_rate": 1.1779350104821804e-05,
      "loss": 0.8706,
      "num_input_tokens_seen": 584672,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 0.9713388681411743,
      "learning_rate": 1.1844863731656185e-05,
      "loss": 0.7196,
      "num_input_tokens_seen": 588704,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 1.6351550817489624,
      "learning_rate": 1.1910377358490566e-05,
      "loss": 0.7647,
      "num_input_tokens_seen": 591712,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 1.0455248355865479,
      "learning_rate": 1.197589098532495e-05,
      "loss": 0.7805,
      "num_input_tokens_seen": 595104,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 0.7474347352981567,
      "learning_rate": 1.204140461215933e-05,
      "loss": 0.698,
      "num_input_tokens_seen": 599808,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 0.8501096963882446,
      "learning_rate": 1.210691823899371e-05,
      "loss": 0.6407,
      "num_input_tokens_seen": 602720,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 0.913571298122406,
      "learning_rate": 1.2172431865828094e-05,
      "loss": 0.6805,
      "num_input_tokens_seen": 606176,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 0.742790937423706,
      "learning_rate": 1.2237945492662473e-05,
      "loss": 0.718,
      "num_input_tokens_seen": 608864,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 0.7411170601844788,
      "learning_rate": 1.2303459119496856e-05,
      "loss": 0.6831,
      "num_input_tokens_seen": 611968,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 0.8818872570991516,
      "learning_rate": 1.2368972746331238e-05,
      "loss": 0.8422,
      "num_input_tokens_seen": 614848,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 0.7389304041862488,
      "learning_rate": 1.243448637316562e-05,
      "loss": 0.9827,
      "num_input_tokens_seen": 618080,
      "step": 950
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 0.8644487857818604,
      "learning_rate": 1.25e-05,
      "loss": 0.5562,
      "num_input_tokens_seen": 620736,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 1.0139617919921875,
      "learning_rate": 1.2565513626834382e-05,
      "loss": 0.6546,
      "num_input_tokens_seen": 624480,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 1.712612509727478,
      "learning_rate": 1.2631027253668765e-05,
      "loss": 0.6289,
      "num_input_tokens_seen": 628448,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 0.5471469163894653,
      "learning_rate": 1.2696540880503146e-05,
      "loss": 0.8385,
      "num_input_tokens_seen": 632000,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 0.8407309055328369,
      "learning_rate": 1.2762054507337526e-05,
      "loss": 0.8011,
      "num_input_tokens_seen": 635680,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 1.5319130420684814,
      "learning_rate": 1.2827568134171907e-05,
      "loss": 1.0264,
      "num_input_tokens_seen": 638496,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 1.2722903490066528,
      "learning_rate": 1.289308176100629e-05,
      "loss": 0.652,
      "num_input_tokens_seen": 642272,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 1.1453731060028076,
      "learning_rate": 1.2958595387840672e-05,
      "loss": 0.8193,
      "num_input_tokens_seen": 645888,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 0.7337562441825867,
      "learning_rate": 1.3024109014675053e-05,
      "loss": 0.526,
      "num_input_tokens_seen": 651808,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 1.0158809423446655,
      "learning_rate": 1.3089622641509436e-05,
      "loss": 0.7769,
      "num_input_tokens_seen": 655488,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 0.5695878863334656,
      "learning_rate": 1.3155136268343814e-05,
      "loss": 0.6392,
      "num_input_tokens_seen": 658976,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 1.034735083580017,
      "learning_rate": 1.3220649895178197e-05,
      "loss": 0.6963,
      "num_input_tokens_seen": 662048,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 0.7402485609054565,
      "learning_rate": 1.3286163522012579e-05,
      "loss": 0.7808,
      "num_input_tokens_seen": 664640,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 0.9524266719818115,
      "learning_rate": 1.3351677148846962e-05,
      "loss": 0.6198,
      "num_input_tokens_seen": 669056,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 0.9390379786491394,
      "learning_rate": 1.3417190775681343e-05,
      "loss": 0.6533,
      "num_input_tokens_seen": 672128,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 1.2434316873550415,
      "learning_rate": 1.3482704402515725e-05,
      "loss": 0.6627,
      "num_input_tokens_seen": 675136,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 1.3294882774353027,
      "learning_rate": 1.3548218029350104e-05,
      "loss": 0.7828,
      "num_input_tokens_seen": 678976,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 0.9729113578796387,
      "learning_rate": 1.3613731656184486e-05,
      "loss": 0.9475,
      "num_input_tokens_seen": 682176,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 1.6343663930892944,
      "learning_rate": 1.3679245283018869e-05,
      "loss": 0.7612,
      "num_input_tokens_seen": 684768,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 0.6969051957130432,
      "learning_rate": 1.374475890985325e-05,
      "loss": 0.7655,
      "num_input_tokens_seen": 687392,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 2.180385112762451,
      "learning_rate": 1.3810272536687632e-05,
      "loss": 0.6834,
      "num_input_tokens_seen": 690496,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.4164618253707886,
      "learning_rate": 1.3875786163522015e-05,
      "loss": 0.9208,
      "num_input_tokens_seen": 693376,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 1.2477823495864868,
      "learning_rate": 1.3941299790356394e-05,
      "loss": 0.7404,
      "num_input_tokens_seen": 695808,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 0.4741480052471161,
      "learning_rate": 1.4006813417190776e-05,
      "loss": 0.5877,
      "num_input_tokens_seen": 699424,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 0.6420347690582275,
      "learning_rate": 1.4072327044025157e-05,
      "loss": 0.6302,
      "num_input_tokens_seen": 703584,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.9677038788795471,
      "learning_rate": 1.413784067085954e-05,
      "loss": 0.7182,
      "num_input_tokens_seen": 706144,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 0.6366807818412781,
      "learning_rate": 1.4203354297693922e-05,
      "loss": 0.7308,
      "num_input_tokens_seen": 708864,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 0.5160654783248901,
      "learning_rate": 1.4268867924528303e-05,
      "loss": 0.7117,
      "num_input_tokens_seen": 712352,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 0.8714132308959961,
      "learning_rate": 1.4334381551362683e-05,
      "loss": 0.8671,
      "num_input_tokens_seen": 715296,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 0.6451776623725891,
      "learning_rate": 1.4399895178197064e-05,
      "loss": 0.6207,
      "num_input_tokens_seen": 718528,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 0.674884021282196,
      "learning_rate": 1.4465408805031447e-05,
      "loss": 0.6151,
      "num_input_tokens_seen": 722752,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 0.781731903553009,
      "learning_rate": 1.4530922431865828e-05,
      "loss": 0.689,
      "num_input_tokens_seen": 725824,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 0.6618977785110474,
      "learning_rate": 1.4596436058700211e-05,
      "loss": 0.7411,
      "num_input_tokens_seen": 730080,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 0.8655752539634705,
      "learning_rate": 1.4661949685534593e-05,
      "loss": 0.5858,
      "num_input_tokens_seen": 733216,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 0.9829463362693787,
      "learning_rate": 1.4727463312368974e-05,
      "loss": 0.7963,
      "num_input_tokens_seen": 736416,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 0.7128252387046814,
      "learning_rate": 1.4792976939203354e-05,
      "loss": 0.6192,
      "num_input_tokens_seen": 739808,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 1.0163557529449463,
      "learning_rate": 1.4858490566037735e-05,
      "loss": 0.7157,
      "num_input_tokens_seen": 742400,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 1.338437557220459,
      "learning_rate": 1.4924004192872118e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 745760,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 0.8404232859611511,
      "learning_rate": 1.49895178197065e-05,
      "loss": 0.746,
      "num_input_tokens_seen": 748736,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 0.7094617486000061,
      "learning_rate": 1.5055031446540883e-05,
      "loss": 0.6771,
      "num_input_tokens_seen": 752032,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 0.5193654894828796,
      "learning_rate": 1.5120545073375264e-05,
      "loss": 0.5946,
      "num_input_tokens_seen": 756736,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 0.5351244807243347,
      "learning_rate": 1.5186058700209644e-05,
      "loss": 0.617,
      "num_input_tokens_seen": 759680,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 0.9575039148330688,
      "learning_rate": 1.5251572327044025e-05,
      "loss": 0.783,
      "num_input_tokens_seen": 762560,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 0.5449291467666626,
      "learning_rate": 1.5317085953878407e-05,
      "loss": 0.5559,
      "num_input_tokens_seen": 765600,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 0.534072756767273,
      "learning_rate": 1.538259958071279e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 768736,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 0.8012651205062866,
      "learning_rate": 1.5448113207547173e-05,
      "loss": 0.7999,
      "num_input_tokens_seen": 771872,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 0.48578888177871704,
      "learning_rate": 1.5513626834381552e-05,
      "loss": 0.7693,
      "num_input_tokens_seen": 775008,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 0.5356956720352173,
      "learning_rate": 1.5579140461215932e-05,
      "loss": 0.5689,
      "num_input_tokens_seen": 777920,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 0.706897497177124,
      "learning_rate": 1.5644654088050315e-05,
      "loss": 0.6846,
      "num_input_tokens_seen": 781056,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 0.40814968943595886,
      "learning_rate": 1.5710167714884695e-05,
      "loss": 0.6956,
      "num_input_tokens_seen": 784128,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 0.7029784917831421,
      "learning_rate": 1.5775681341719078e-05,
      "loss": 0.7071,
      "num_input_tokens_seen": 786624,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 0.9399550557136536,
      "learning_rate": 1.584119496855346e-05,
      "loss": 0.6347,
      "num_input_tokens_seen": 789664,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 0.6366162300109863,
      "learning_rate": 1.5906708595387844e-05,
      "loss": 0.5269,
      "num_input_tokens_seen": 793824,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 0.8696964383125305,
      "learning_rate": 1.597222222222222e-05,
      "loss": 0.8113,
      "num_input_tokens_seen": 796544,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 0.4651728868484497,
      "learning_rate": 1.6037735849056604e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 799520,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 1.1752064228057861,
      "learning_rate": 1.6103249475890987e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 802464,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 0.8762246966362,
      "learning_rate": 1.6168763102725366e-05,
      "loss": 0.7404,
      "num_input_tokens_seen": 805536,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 0.48181095719337463,
      "learning_rate": 1.623427672955975e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 808448,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 0.5052818059921265,
      "learning_rate": 1.6299790356394132e-05,
      "loss": 0.5692,
      "num_input_tokens_seen": 812160,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 2.070528984069824,
      "learning_rate": 1.6365303983228512e-05,
      "loss": 0.8446,
      "num_input_tokens_seen": 814752,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 0.7502148151397705,
      "learning_rate": 1.6430817610062892e-05,
      "loss": 0.529,
      "num_input_tokens_seen": 817568,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 0.9317643642425537,
      "learning_rate": 1.6496331236897275e-05,
      "loss": 0.618,
      "num_input_tokens_seen": 820384,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 0.5271543264389038,
      "learning_rate": 1.6561844863731658e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 823040,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 0.7330589890480042,
      "learning_rate": 1.6627358490566038e-05,
      "loss": 0.7655,
      "num_input_tokens_seen": 825600,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 0.8550035953521729,
      "learning_rate": 1.669287211740042e-05,
      "loss": 0.6828,
      "num_input_tokens_seen": 828032,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 0.8806240558624268,
      "learning_rate": 1.6758385744234804e-05,
      "loss": 0.598,
      "num_input_tokens_seen": 831296,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 0.5277919173240662,
      "learning_rate": 1.6823899371069183e-05,
      "loss": 0.7169,
      "num_input_tokens_seen": 834208,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 1.0344310998916626,
      "learning_rate": 1.6889412997903563e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 838048,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 1.1174685955047607,
      "learning_rate": 1.6954926624737946e-05,
      "loss": 0.6662,
      "num_input_tokens_seen": 841536,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 0.6034958958625793,
      "learning_rate": 1.702044025157233e-05,
      "loss": 0.5336,
      "num_input_tokens_seen": 845184,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 0.4211331605911255,
      "learning_rate": 1.708595387840671e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 848000,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 1.4025675058364868,
      "learning_rate": 1.7151467505241092e-05,
      "loss": 0.6809,
      "num_input_tokens_seen": 851040,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 0.5889501571655273,
      "learning_rate": 1.7216981132075472e-05,
      "loss": 0.69,
      "num_input_tokens_seen": 853920,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 1.136643648147583,
      "learning_rate": 1.7282494758909855e-05,
      "loss": 0.6153,
      "num_input_tokens_seen": 857312,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.6726347804069519,
      "learning_rate": 1.7348008385744234e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 859904,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 0.548473060131073,
      "learning_rate": 1.7413522012578618e-05,
      "loss": 0.6601,
      "num_input_tokens_seen": 863680,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 0.6414380669593811,
      "learning_rate": 1.7479035639413e-05,
      "loss": 0.7199,
      "num_input_tokens_seen": 866880,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 0.8938717246055603,
      "learning_rate": 1.754454926624738e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 869856,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 1.0255354642868042,
      "learning_rate": 1.761006289308176e-05,
      "loss": 0.6852,
      "num_input_tokens_seen": 873472,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 0.6468727588653564,
      "learning_rate": 1.7675576519916143e-05,
      "loss": 0.5787,
      "num_input_tokens_seen": 876704,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 0.4935464560985565,
      "learning_rate": 1.7741090146750526e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 880576,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 1.0734366178512573,
      "learning_rate": 1.7806603773584906e-05,
      "loss": 0.5917,
      "num_input_tokens_seen": 883680,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 0.850071907043457,
      "learning_rate": 1.787211740041929e-05,
      "loss": 0.73,
      "num_input_tokens_seen": 886464,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 0.8696297407150269,
      "learning_rate": 1.7937631027253672e-05,
      "loss": 0.6766,
      "num_input_tokens_seen": 889536,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 0.6797934770584106,
      "learning_rate": 1.8003144654088048e-05,
      "loss": 0.6193,
      "num_input_tokens_seen": 894272,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 0.400920569896698,
      "learning_rate": 1.806865828092243e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 897120,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 0.8614768981933594,
      "learning_rate": 1.8134171907756814e-05,
      "loss": 0.5893,
      "num_input_tokens_seen": 899808,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 0.9038190245628357,
      "learning_rate": 1.8199685534591197e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 903520,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 0.5398830771446228,
      "learning_rate": 1.8265199161425577e-05,
      "loss": 0.6196,
      "num_input_tokens_seen": 907200,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 0.3598730266094208,
      "learning_rate": 1.833071278825996e-05,
      "loss": 0.5963,
      "num_input_tokens_seen": 911392,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 0.822754979133606,
      "learning_rate": 1.839622641509434e-05,
      "loss": 0.7127,
      "num_input_tokens_seen": 913440,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 0.6897557973861694,
      "learning_rate": 1.846174004192872e-05,
      "loss": 0.5527,
      "num_input_tokens_seen": 917888,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 0.45046466588974,
      "learning_rate": 1.8527253668763103e-05,
      "loss": 0.6356,
      "num_input_tokens_seen": 921376,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 0.5241090059280396,
      "learning_rate": 1.8592767295597486e-05,
      "loss": 0.6561,
      "num_input_tokens_seen": 924800,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 1.2513654232025146,
      "learning_rate": 1.865828092243187e-05,
      "loss": 0.7697,
      "num_input_tokens_seen": 927648,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 0.8891732692718506,
      "learning_rate": 1.872379454926625e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 930144,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 0.5309391617774963,
      "learning_rate": 1.878930817610063e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 933184,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.6914190053939819,
      "learning_rate": 1.885482180293501e-05,
      "loss": 0.6451,
      "num_input_tokens_seen": 937376,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 0.7349117398262024,
      "learning_rate": 1.892033542976939e-05,
      "loss": 0.6051,
      "num_input_tokens_seen": 940160,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 0.8994513154029846,
      "learning_rate": 1.8985849056603774e-05,
      "loss": 0.6754,
      "num_input_tokens_seen": 942816,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 0.6240000128746033,
      "learning_rate": 1.9051362683438157e-05,
      "loss": 0.653,
      "num_input_tokens_seen": 945952,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 1.2164403200149536,
      "learning_rate": 1.9116876310272537e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 949472,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 0.9077129364013672,
      "learning_rate": 1.918238993710692e-05,
      "loss": 0.6275,
      "num_input_tokens_seen": 952672,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 0.42716047167778015,
      "learning_rate": 1.92479035639413e-05,
      "loss": 0.7223,
      "num_input_tokens_seen": 955520,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 0.6579738259315491,
      "learning_rate": 1.9313417190775683e-05,
      "loss": 0.6358,
      "num_input_tokens_seen": 958464,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 0.9565011858940125,
      "learning_rate": 1.9378930817610062e-05,
      "loss": 0.5915,
      "num_input_tokens_seen": 960768,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 1.4946515560150146,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.8072,
      "num_input_tokens_seen": 963616,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 0.5033466815948486,
      "learning_rate": 1.950995807127883e-05,
      "loss": 0.7734,
      "num_input_tokens_seen": 967936,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 0.5553821921348572,
      "learning_rate": 1.9575471698113208e-05,
      "loss": 0.5603,
      "num_input_tokens_seen": 971744,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 0.6516197919845581,
      "learning_rate": 1.9640985324947588e-05,
      "loss": 0.6896,
      "num_input_tokens_seen": 974080,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 0.5350182056427002,
      "learning_rate": 1.970649895178197e-05,
      "loss": 0.5451,
      "num_input_tokens_seen": 976992,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 0.968264639377594,
      "learning_rate": 1.9772012578616354e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 979648,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 0.5379883050918579,
      "learning_rate": 1.9837526205450734e-05,
      "loss": 0.6765,
      "num_input_tokens_seen": 983136,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 0.547812283039093,
      "learning_rate": 1.9903039832285117e-05,
      "loss": 0.6083,
      "num_input_tokens_seen": 986976,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 0.4975142478942871,
      "learning_rate": 1.99685534591195e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 990272,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 0.4096323549747467,
      "learning_rate": 2.003406708595388e-05,
      "loss": 0.5687,
      "num_input_tokens_seen": 994368,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 0.6081419587135315,
      "learning_rate": 2.009958071278826e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 997504,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 0.5777926445007324,
      "learning_rate": 2.0165094339622642e-05,
      "loss": 0.6636,
      "num_input_tokens_seen": 1000352,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 0.4271259009838104,
      "learning_rate": 2.0230607966457025e-05,
      "loss": 0.6473,
      "num_input_tokens_seen": 1003616,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 1.6693871021270752,
      "learning_rate": 2.0296121593291405e-05,
      "loss": 0.7402,
      "num_input_tokens_seen": 1006240,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 1.2268826961517334,
      "learning_rate": 2.0361635220125788e-05,
      "loss": 0.647,
      "num_input_tokens_seen": 1009440,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 0.5996467471122742,
      "learning_rate": 2.0427148846960168e-05,
      "loss": 0.7354,
      "num_input_tokens_seen": 1011936,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 0.44003647565841675,
      "learning_rate": 2.049266247379455e-05,
      "loss": 0.5667,
      "num_input_tokens_seen": 1015040,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 0.5931268930435181,
      "learning_rate": 2.055817610062893e-05,
      "loss": 0.698,
      "num_input_tokens_seen": 1018240,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 0.47107502818107605,
      "learning_rate": 2.0623689727463314e-05,
      "loss": 0.5864,
      "num_input_tokens_seen": 1022336,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 1.0270696878433228,
      "learning_rate": 2.0689203354297697e-05,
      "loss": 0.6134,
      "num_input_tokens_seen": 1025184,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 2.0067732334136963,
      "learning_rate": 2.0754716981132076e-05,
      "loss": 0.6512,
      "num_input_tokens_seen": 1028448,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.9994735717773438,
      "learning_rate": 2.0820230607966456e-05,
      "loss": 0.6254,
      "num_input_tokens_seen": 1031072,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 0.6910018920898438,
      "learning_rate": 2.088574423480084e-05,
      "loss": 0.6073,
      "num_input_tokens_seen": 1034208,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 0.7203481793403625,
      "learning_rate": 2.0951257861635222e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 1038624,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 0.5300828218460083,
      "learning_rate": 2.1016771488469602e-05,
      "loss": 0.6905,
      "num_input_tokens_seen": 1041152,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 0.5784085988998413,
      "learning_rate": 2.1082285115303985e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 1043904,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 0.5367663502693176,
      "learning_rate": 2.1147798742138368e-05,
      "loss": 0.578,
      "num_input_tokens_seen": 1047456,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 0.5236906409263611,
      "learning_rate": 2.1213312368972748e-05,
      "loss": 0.6271,
      "num_input_tokens_seen": 1050304,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 0.6373969316482544,
      "learning_rate": 2.1278825995807127e-05,
      "loss": 0.6134,
      "num_input_tokens_seen": 1053600,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 0.5248720645904541,
      "learning_rate": 2.134433962264151e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 1056672,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 0.5489718317985535,
      "learning_rate": 2.1409853249475894e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 1059904,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 0.5621293187141418,
      "learning_rate": 2.1475366876310273e-05,
      "loss": 0.5861,
      "num_input_tokens_seen": 1063360,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 1.2503104209899902,
      "learning_rate": 2.1540880503144656e-05,
      "loss": 0.605,
      "num_input_tokens_seen": 1065728,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 0.6432293057441711,
      "learning_rate": 2.160639412997904e-05,
      "loss": 0.5751,
      "num_input_tokens_seen": 1068448,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 0.45566943287849426,
      "learning_rate": 2.1671907756813416e-05,
      "loss": 0.6013,
      "num_input_tokens_seen": 1071104,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 0.9269049167633057,
      "learning_rate": 2.17374213836478e-05,
      "loss": 0.5121,
      "num_input_tokens_seen": 1075072,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 0.549279510974884,
      "learning_rate": 2.1802935010482182e-05,
      "loss": 0.5984,
      "num_input_tokens_seen": 1077824,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 0.616353452205658,
      "learning_rate": 2.186844863731656e-05,
      "loss": 0.5406,
      "num_input_tokens_seen": 1080832,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 1.418969988822937,
      "learning_rate": 2.1933962264150945e-05,
      "loss": 0.6457,
      "num_input_tokens_seen": 1084096,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 0.4689745008945465,
      "learning_rate": 2.1999475890985328e-05,
      "loss": 0.6039,
      "num_input_tokens_seen": 1087104,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 0.5936911702156067,
      "learning_rate": 2.2064989517819707e-05,
      "loss": 0.6796,
      "num_input_tokens_seen": 1090304,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 0.686089277267456,
      "learning_rate": 2.2130503144654087e-05,
      "loss": 0.8619,
      "num_input_tokens_seen": 1094112,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 0.5676605701446533,
      "learning_rate": 2.219601677148847e-05,
      "loss": 0.6187,
      "num_input_tokens_seen": 1098496,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 0.428975373506546,
      "learning_rate": 2.2261530398322853e-05,
      "loss": 0.6041,
      "num_input_tokens_seen": 1100992,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 0.7928570508956909,
      "learning_rate": 2.2327044025157233e-05,
      "loss": 0.6141,
      "num_input_tokens_seen": 1103776,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 0.5834746360778809,
      "learning_rate": 2.2392557651991616e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 1106368,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 0.6980623006820679,
      "learning_rate": 2.2458071278825996e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 1109376,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 0.34664392471313477,
      "learning_rate": 2.252358490566038e-05,
      "loss": 0.5239,
      "num_input_tokens_seen": 1112448,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 0.30467915534973145,
      "learning_rate": 2.258909853249476e-05,
      "loss": 0.5674,
      "num_input_tokens_seen": 1115520,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 0.6384235620498657,
      "learning_rate": 2.265461215932914e-05,
      "loss": 0.7186,
      "num_input_tokens_seen": 1118784,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 2.530316114425659,
      "learning_rate": 2.2720125786163524e-05,
      "loss": 0.8266,
      "num_input_tokens_seen": 1122112,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 0.6948792338371277,
      "learning_rate": 2.2785639412997904e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 1125952,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 0.7654098868370056,
      "learning_rate": 2.2851153039832284e-05,
      "loss": 0.7551,
      "num_input_tokens_seen": 1129024,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 0.6663278341293335,
      "learning_rate": 2.2916666666666667e-05,
      "loss": 0.5506,
      "num_input_tokens_seen": 1131488,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 0.4190826117992401,
      "learning_rate": 2.298218029350105e-05,
      "loss": 0.535,
      "num_input_tokens_seen": 1134176,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 0.3732849359512329,
      "learning_rate": 2.304769392033543e-05,
      "loss": 0.6003,
      "num_input_tokens_seen": 1137888,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 0.6432092785835266,
      "learning_rate": 2.3113207547169813e-05,
      "loss": 0.6974,
      "num_input_tokens_seen": 1141184,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 1.4237207174301147,
      "learning_rate": 2.3178721174004196e-05,
      "loss": 0.6366,
      "num_input_tokens_seen": 1144064,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 0.7515240907669067,
      "learning_rate": 2.3244234800838576e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 1147584,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 0.31113767623901367,
      "learning_rate": 2.3309748427672955e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 1154784,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 1.0642861127853394,
      "learning_rate": 2.3375262054507338e-05,
      "loss": 0.6265,
      "num_input_tokens_seen": 1157440,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 0.36927223205566406,
      "learning_rate": 2.344077568134172e-05,
      "loss": 0.7302,
      "num_input_tokens_seen": 1160224,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 0.8782644271850586,
      "learning_rate": 2.35062893081761e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 1164480,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.7930328249931335,
      "learning_rate": 2.3571802935010484e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 1168704,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 0.6184717416763306,
      "learning_rate": 2.3637316561844867e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 1172384,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 1.049674391746521,
      "learning_rate": 2.3702830188679247e-05,
      "loss": 0.5985,
      "num_input_tokens_seen": 1175392,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 0.6138635873794556,
      "learning_rate": 2.3768343815513627e-05,
      "loss": 0.5284,
      "num_input_tokens_seen": 1178368,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 0.5973663330078125,
      "learning_rate": 2.383385744234801e-05,
      "loss": 0.5446,
      "num_input_tokens_seen": 1182240,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 0.4099355638027191,
      "learning_rate": 2.3899371069182393e-05,
      "loss": 0.5348,
      "num_input_tokens_seen": 1185504,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 0.908696174621582,
      "learning_rate": 2.3964884696016772e-05,
      "loss": 0.5613,
      "num_input_tokens_seen": 1188288,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 0.677991509437561,
      "learning_rate": 2.4030398322851155e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 1192064,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 0.3512945771217346,
      "learning_rate": 2.4095911949685535e-05,
      "loss": 0.5507,
      "num_input_tokens_seen": 1195104,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 0.5711775422096252,
      "learning_rate": 2.4161425576519918e-05,
      "loss": 0.7647,
      "num_input_tokens_seen": 1197920,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 0.3926745057106018,
      "learning_rate": 2.4226939203354298e-05,
      "loss": 0.5312,
      "num_input_tokens_seen": 1200896,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.5924448370933533,
      "learning_rate": 2.429245283018868e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 1204032,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 0.4702952206134796,
      "learning_rate": 2.4357966457023064e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 1207264,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 0.777739942073822,
      "learning_rate": 2.4423480083857444e-05,
      "loss": 0.6502,
      "num_input_tokens_seen": 1209984,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 0.7062153220176697,
      "learning_rate": 2.4488993710691823e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 1213856,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 0.3863323926925659,
      "learning_rate": 2.4554507337526206e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 1217408,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 0.32669752836227417,
      "learning_rate": 2.4620020964360586e-05,
      "loss": 0.5747,
      "num_input_tokens_seen": 1221024,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 0.5157504677772522,
      "learning_rate": 2.468553459119497e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 1224320,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 0.48793697357177734,
      "learning_rate": 2.4751048218029352e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 1228000,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 0.5853852033615112,
      "learning_rate": 2.4816561844863735e-05,
      "loss": 0.5499,
      "num_input_tokens_seen": 1232320,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 0.6188923716545105,
      "learning_rate": 2.488207547169811e-05,
      "loss": 0.6565,
      "num_input_tokens_seen": 1238272,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 1.5665374994277954,
      "learning_rate": 2.4947589098532495e-05,
      "loss": 0.5784,
      "num_input_tokens_seen": 1241344,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5528630614280701,
      "eval_runtime": 9.8092,
      "eval_samples_per_second": 86.45,
      "eval_steps_per_second": 21.612,
      "num_input_tokens_seen": 1243048,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 0.5315430164337158,
      "learning_rate": 2.5013102725366878e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 1244520,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 0.510593056678772,
      "learning_rate": 2.507861635220126e-05,
      "loss": 0.5776,
      "num_input_tokens_seen": 1247528,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 0.6076985597610474,
      "learning_rate": 2.514412997903564e-05,
      "loss": 0.6784,
      "num_input_tokens_seen": 1250824,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 0.5665467977523804,
      "learning_rate": 2.520964360587002e-05,
      "loss": 0.5598,
      "num_input_tokens_seen": 1254024,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 0.5951313972473145,
      "learning_rate": 2.5275157232704407e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 1260328,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 0.5544217228889465,
      "learning_rate": 2.5340670859538783e-05,
      "loss": 0.8094,
      "num_input_tokens_seen": 1263656,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 0.4806751608848572,
      "learning_rate": 2.540618448637317e-05,
      "loss": 0.6078,
      "num_input_tokens_seen": 1267528,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 0.9600875973701477,
      "learning_rate": 2.547169811320755e-05,
      "loss": 0.7629,
      "num_input_tokens_seen": 1270536,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 0.7039252519607544,
      "learning_rate": 2.553721174004193e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 1273416,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 0.6879808306694031,
      "learning_rate": 2.5602725366876312e-05,
      "loss": 0.4154,
      "num_input_tokens_seen": 1276232,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 0.3796679973602295,
      "learning_rate": 2.566823899371069e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 1279400,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 0.4857006072998047,
      "learning_rate": 2.5733752620545075e-05,
      "loss": 0.5027,
      "num_input_tokens_seen": 1283752,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 0.50080406665802,
      "learning_rate": 2.5799266247379454e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 1287048,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 0.5953927636146545,
      "learning_rate": 2.586477987421384e-05,
      "loss": 0.6104,
      "num_input_tokens_seen": 1289544,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 0.5033283829689026,
      "learning_rate": 2.593029350104822e-05,
      "loss": 0.5571,
      "num_input_tokens_seen": 1292392,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 0.7264817357063293,
      "learning_rate": 2.59958071278826e-05,
      "loss": 0.6196,
      "num_input_tokens_seen": 1295400,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 0.5026444792747498,
      "learning_rate": 2.6061320754716983e-05,
      "loss": 0.5815,
      "num_input_tokens_seen": 1298152,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 0.8700653910636902,
      "learning_rate": 2.6126834381551363e-05,
      "loss": 0.6678,
      "num_input_tokens_seen": 1301384,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 0.8970664143562317,
      "learning_rate": 2.6192348008385746e-05,
      "loss": 0.6309,
      "num_input_tokens_seen": 1304296,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 0.6467858552932739,
      "learning_rate": 2.6257861635220126e-05,
      "loss": 0.6762,
      "num_input_tokens_seen": 1307048,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 0.415096640586853,
      "learning_rate": 2.6323375262054505e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 1310792,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 0.6109135150909424,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.6869,
      "num_input_tokens_seen": 1313448,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 1.3250572681427002,
      "learning_rate": 2.645440251572327e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 1316200,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 0.48217079043388367,
      "learning_rate": 2.6519916142557655e-05,
      "loss": 0.913,
      "num_input_tokens_seen": 1319720,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 0.6280933618545532,
      "learning_rate": 2.6585429769392034e-05,
      "loss": 0.6185,
      "num_input_tokens_seen": 1323432,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 0.4328954517841339,
      "learning_rate": 2.6650943396226417e-05,
      "loss": 0.5989,
      "num_input_tokens_seen": 1326760,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 0.413892537355423,
      "learning_rate": 2.6716457023060797e-05,
      "loss": 0.6147,
      "num_input_tokens_seen": 1330120,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 0.6723844408988953,
      "learning_rate": 2.6781970649895177e-05,
      "loss": 0.6078,
      "num_input_tokens_seen": 1333032,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 1.2715914249420166,
      "learning_rate": 2.6847484276729563e-05,
      "loss": 0.6603,
      "num_input_tokens_seen": 1336296,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 0.41916462779045105,
      "learning_rate": 2.6912997903563943e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 1339592,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 0.9256888031959534,
      "learning_rate": 2.6978511530398326e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 1342664,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 0.485557496547699,
      "learning_rate": 2.7044025157232706e-05,
      "loss": 0.6841,
      "num_input_tokens_seen": 1345288,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 0.4912087619304657,
      "learning_rate": 2.710953878406709e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 1348488,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 0.47980934381484985,
      "learning_rate": 2.717505241090147e-05,
      "loss": 0.6628,
      "num_input_tokens_seen": 1352712,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 0.3808802366256714,
      "learning_rate": 2.7240566037735848e-05,
      "loss": 0.4837,
      "num_input_tokens_seen": 1355496,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 0.26491254568099976,
      "learning_rate": 2.7306079664570235e-05,
      "loss": 0.5812,
      "num_input_tokens_seen": 1360552,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 0.3598582148551941,
      "learning_rate": 2.737159329140461e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 1363432,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 0.6359404921531677,
      "learning_rate": 2.7437106918238997e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 1366824,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 1.0946526527404785,
      "learning_rate": 2.7502620545073377e-05,
      "loss": 0.6245,
      "num_input_tokens_seen": 1370312,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 0.6258524060249329,
      "learning_rate": 2.7568134171907757e-05,
      "loss": 0.7979,
      "num_input_tokens_seen": 1373160,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 0.6376269459724426,
      "learning_rate": 2.763364779874214e-05,
      "loss": 0.6182,
      "num_input_tokens_seen": 1376008,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 0.6939829587936401,
      "learning_rate": 2.769916142557652e-05,
      "loss": 0.6495,
      "num_input_tokens_seen": 1378440,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.4480949342250824,
      "learning_rate": 2.7764675052410906e-05,
      "loss": 0.5838,
      "num_input_tokens_seen": 1382152,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 0.49712562561035156,
      "learning_rate": 2.7830188679245282e-05,
      "loss": 0.7021,
      "num_input_tokens_seen": 1385416,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 0.6405726671218872,
      "learning_rate": 2.789570230607967e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 1389288,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 0.8812804222106934,
      "learning_rate": 2.796121593291405e-05,
      "loss": 0.5987,
      "num_input_tokens_seen": 1392136,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 0.36248019337654114,
      "learning_rate": 2.8026729559748428e-05,
      "loss": 0.602,
      "num_input_tokens_seen": 1395208,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 0.4272415339946747,
      "learning_rate": 2.809224318658281e-05,
      "loss": 0.4732,
      "num_input_tokens_seen": 1398024,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 0.3842897117137909,
      "learning_rate": 2.815775681341719e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 1401256,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 0.9625896215438843,
      "learning_rate": 2.8223270440251577e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 1403848,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 1.1892682313919067,
      "learning_rate": 2.8288784067085954e-05,
      "loss": 0.6888,
      "num_input_tokens_seen": 1408040,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 0.9216629266738892,
      "learning_rate": 2.8354297693920333e-05,
      "loss": 0.6682,
      "num_input_tokens_seen": 1411688,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 0.47279977798461914,
      "learning_rate": 2.841981132075472e-05,
      "loss": 0.649,
      "num_input_tokens_seen": 1415464,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 0.32775816321372986,
      "learning_rate": 2.84853249475891e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 1418664,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 0.49404430389404297,
      "learning_rate": 2.8550838574423482e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 1422088,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 1.5067708492279053,
      "learning_rate": 2.8616352201257862e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 1424872,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 0.39068540930747986,
      "learning_rate": 2.868186582809225e-05,
      "loss": 0.6702,
      "num_input_tokens_seen": 1429320,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 0.45949918031692505,
      "learning_rate": 2.8747379454926625e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 1433704,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 0.4891259968280792,
      "learning_rate": 2.8812893081761005e-05,
      "loss": 0.5766,
      "num_input_tokens_seen": 1436872,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 0.42042872309684753,
      "learning_rate": 2.887840670859539e-05,
      "loss": 0.5705,
      "num_input_tokens_seen": 1440104,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 0.48968225717544556,
      "learning_rate": 2.894392033542977e-05,
      "loss": 0.7222,
      "num_input_tokens_seen": 1443176,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 0.3787473440170288,
      "learning_rate": 2.9009433962264154e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 1446440,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 0.7802485227584839,
      "learning_rate": 2.9074947589098533e-05,
      "loss": 0.5191,
      "num_input_tokens_seen": 1450120,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 0.9139600396156311,
      "learning_rate": 2.9140461215932913e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 1453608,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 0.485975056886673,
      "learning_rate": 2.9205974842767296e-05,
      "loss": 0.6952,
      "num_input_tokens_seen": 1457832,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 0.3590909242630005,
      "learning_rate": 2.9271488469601676e-05,
      "loss": 0.5753,
      "num_input_tokens_seen": 1462280,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 0.40272244811058044,
      "learning_rate": 2.9337002096436062e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 1466024,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 0.44099730253219604,
      "learning_rate": 2.9402515723270442e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 1468872,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 1.2808579206466675,
      "learning_rate": 2.9468029350104825e-05,
      "loss": 0.5453,
      "num_input_tokens_seen": 1471336,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 0.7395802736282349,
      "learning_rate": 2.9533542976939205e-05,
      "loss": 0.6751,
      "num_input_tokens_seen": 1473928,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 0.731631338596344,
      "learning_rate": 2.9599056603773584e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 1478728,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 0.490701287984848,
      "learning_rate": 2.9664570230607968e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 1481576,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 0.6436353921890259,
      "learning_rate": 2.9730083857442347e-05,
      "loss": 0.6262,
      "num_input_tokens_seen": 1484104,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 0.5397545695304871,
      "learning_rate": 2.9795597484276734e-05,
      "loss": 0.6654,
      "num_input_tokens_seen": 1487048,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 0.43101733922958374,
      "learning_rate": 2.9861111111111113e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 1490152,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 0.4019523859024048,
      "learning_rate": 2.9926624737945496e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 1492648,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 0.5774397850036621,
      "learning_rate": 2.9992138364779876e-05,
      "loss": 0.6649,
      "num_input_tokens_seen": 1495272,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 0.5073925852775574,
      "learning_rate": 3.0057651991614256e-05,
      "loss": 0.5289,
      "num_input_tokens_seen": 1497832,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 0.41508376598358154,
      "learning_rate": 3.012316561844864e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 1501896,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 0.8944088816642761,
      "learning_rate": 3.018867924528302e-05,
      "loss": 0.5484,
      "num_input_tokens_seen": 1504328,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 0.4802808463573456,
      "learning_rate": 3.0254192872117405e-05,
      "loss": 0.5986,
      "num_input_tokens_seen": 1508264,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 0.544326663017273,
      "learning_rate": 3.0319706498951785e-05,
      "loss": 0.6587,
      "num_input_tokens_seen": 1511048,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 0.7336867451667786,
      "learning_rate": 3.038522012578616e-05,
      "loss": 0.7748,
      "num_input_tokens_seen": 1513928,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 0.39215075969696045,
      "learning_rate": 3.0450733752620547e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 1517512,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 0.5664416551589966,
      "learning_rate": 3.0516247379454927e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 1521256,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 0.6488329172134399,
      "learning_rate": 3.0581761006289314e-05,
      "loss": 0.7401,
      "num_input_tokens_seen": 1523144,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 0.7484251260757446,
      "learning_rate": 3.064727463312369e-05,
      "loss": 0.5736,
      "num_input_tokens_seen": 1527624,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 0.47259199619293213,
      "learning_rate": 3.071278825995807e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 1531560,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 0.7131131887435913,
      "learning_rate": 3.0778301886792456e-05,
      "loss": 0.5391,
      "num_input_tokens_seen": 1533832,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 0.39848592877388,
      "learning_rate": 3.084381551362683e-05,
      "loss": 0.678,
      "num_input_tokens_seen": 1536584,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 0.4182441234588623,
      "learning_rate": 3.0909329140461215e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 1539720,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 0.5100913047790527,
      "learning_rate": 3.09748427672956e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 1543112,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 0.637817919254303,
      "learning_rate": 3.104035639412998e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 1546568,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 0.8535350561141968,
      "learning_rate": 3.1105870020964365e-05,
      "loss": 0.5849,
      "num_input_tokens_seen": 1549864,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 0.3796142637729645,
      "learning_rate": 3.117138364779874e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 1552456,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.4755980372428894,
      "learning_rate": 3.1236897274633124e-05,
      "loss": 0.622,
      "num_input_tokens_seen": 1555240,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 0.6945704221725464,
      "learning_rate": 3.130241090146751e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 1558568,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 0.9565206170082092,
      "learning_rate": 3.136792452830189e-05,
      "loss": 0.5943,
      "num_input_tokens_seen": 1561288,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 1.0266082286834717,
      "learning_rate": 3.1433438155136266e-05,
      "loss": 0.6215,
      "num_input_tokens_seen": 1564200,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 0.44354161620140076,
      "learning_rate": 3.1498951781970656e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 1567208,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 0.42970752716064453,
      "learning_rate": 3.156446540880503e-05,
      "loss": 0.506,
      "num_input_tokens_seen": 1571240,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 1.0522375106811523,
      "learning_rate": 3.162997903563941e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 1574376,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 0.552370011806488,
      "learning_rate": 3.16954926624738e-05,
      "loss": 0.6266,
      "num_input_tokens_seen": 1577672,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 0.29685813188552856,
      "learning_rate": 3.1761006289308175e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 1581192,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 0.30302947759628296,
      "learning_rate": 3.182651991614256e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 1584424,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 0.694570004940033,
      "learning_rate": 3.189203354297694e-05,
      "loss": 0.5475,
      "num_input_tokens_seen": 1587912,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 0.3402726352214813,
      "learning_rate": 3.1957547169811324e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 1591624,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 0.7148197889328003,
      "learning_rate": 3.202306079664571e-05,
      "loss": 0.5938,
      "num_input_tokens_seen": 1594376,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 0.4059036374092102,
      "learning_rate": 3.2088574423480084e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 1597608,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 0.6032410264015198,
      "learning_rate": 3.215408805031447e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 1600136,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 0.7118191719055176,
      "learning_rate": 3.221960167714885e-05,
      "loss": 0.5717,
      "num_input_tokens_seen": 1603592,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 0.6439579725265503,
      "learning_rate": 3.228511530398323e-05,
      "loss": 0.6756,
      "num_input_tokens_seen": 1606504,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 0.543817937374115,
      "learning_rate": 3.235062893081761e-05,
      "loss": 0.4558,
      "num_input_tokens_seen": 1613064,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 0.38759538531303406,
      "learning_rate": 3.241614255765199e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 1616136,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 0.8490569591522217,
      "learning_rate": 3.2481656184486375e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 1618792,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 1.0777177810668945,
      "learning_rate": 3.254716981132075e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 1621096,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 0.5562697649002075,
      "learning_rate": 3.261268343815514e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 1623720,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 0.4781709611415863,
      "learning_rate": 3.267819706498952e-05,
      "loss": 0.6029,
      "num_input_tokens_seen": 1626888,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 0.5206438899040222,
      "learning_rate": 3.27437106918239e-05,
      "loss": 0.7034,
      "num_input_tokens_seen": 1630472,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 0.4388439357280731,
      "learning_rate": 3.2809224318658284e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 1633192,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 0.40257447957992554,
      "learning_rate": 3.287473794549266e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 1637128,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 0.3714028596878052,
      "learning_rate": 3.294025157232704e-05,
      "loss": 0.58,
      "num_input_tokens_seen": 1640456,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.6187935471534729,
      "learning_rate": 3.3005765199161426e-05,
      "loss": 0.6766,
      "num_input_tokens_seen": 1644008,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 0.4316597580909729,
      "learning_rate": 3.307127882599581e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 1647720,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 0.7974053025245667,
      "learning_rate": 3.313679245283019e-05,
      "loss": 0.7013,
      "num_input_tokens_seen": 1650440,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 0.42015540599823,
      "learning_rate": 3.320230607966457e-05,
      "loss": 0.6788,
      "num_input_tokens_seen": 1653352,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 0.33355313539505005,
      "learning_rate": 3.326781970649895e-05,
      "loss": 0.5331,
      "num_input_tokens_seen": 1656360,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 0.4426059126853943,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.6558,
      "num_input_tokens_seen": 1659336,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 1.0931233167648315,
      "learning_rate": 3.339884696016772e-05,
      "loss": 0.7668,
      "num_input_tokens_seen": 1661896,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 0.2790219187736511,
      "learning_rate": 3.3464360587002094e-05,
      "loss": 0.4095,
      "num_input_tokens_seen": 1665832,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 0.8114433288574219,
      "learning_rate": 3.3529874213836484e-05,
      "loss": 0.5978,
      "num_input_tokens_seen": 1668520,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 0.45077043771743774,
      "learning_rate": 3.359538784067086e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 1671912,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 0.38930144906044006,
      "learning_rate": 3.3660901467505244e-05,
      "loss": 0.5608,
      "num_input_tokens_seen": 1675080,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 0.66218101978302,
      "learning_rate": 3.3726415094339627e-05,
      "loss": 0.3943,
      "num_input_tokens_seen": 1681512,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 0.3745499849319458,
      "learning_rate": 3.3791928721174e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 1685000,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 0.7422523498535156,
      "learning_rate": 3.3857442348008386e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 1687528,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 0.5150085091590881,
      "learning_rate": 3.392295597484277e-05,
      "loss": 0.678,
      "num_input_tokens_seen": 1690280,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 0.42299750447273254,
      "learning_rate": 3.398846960167715e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 1693352,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 0.5055049061775208,
      "learning_rate": 3.4053983228511535e-05,
      "loss": 0.545,
      "num_input_tokens_seen": 1696456,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 0.3428303301334381,
      "learning_rate": 3.411949685534591e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 1700488,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 0.6357167363166809,
      "learning_rate": 3.4185010482180295e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 1703304,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 0.9292840957641602,
      "learning_rate": 3.425052410901468e-05,
      "loss": 0.6015,
      "num_input_tokens_seen": 1705864,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 0.40891826152801514,
      "learning_rate": 3.431603773584906e-05,
      "loss": 0.6626,
      "num_input_tokens_seen": 1708616,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 0.26198071241378784,
      "learning_rate": 3.438155136268344e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 1712232,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 0.960710346698761,
      "learning_rate": 3.444706498951782e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 1715144,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 0.9511292576789856,
      "learning_rate": 3.45125786163522e-05,
      "loss": 0.5039,
      "num_input_tokens_seen": 1717800,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 0.6524335145950317,
      "learning_rate": 3.457809224318658e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 1721128,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 0.3060436248779297,
      "learning_rate": 3.464360587002097e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 1724328,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.6536955833435059,
      "learning_rate": 3.4709119496855346e-05,
      "loss": 0.6352,
      "num_input_tokens_seen": 1728328,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 0.4824545979499817,
      "learning_rate": 3.477463312368973e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 1731080,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 0.4386121332645416,
      "learning_rate": 3.484014675052411e-05,
      "loss": 0.5592,
      "num_input_tokens_seen": 1736008,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 0.9528040289878845,
      "learning_rate": 3.490566037735849e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 1739368,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 0.6796497702598572,
      "learning_rate": 3.497117400419288e-05,
      "loss": 0.7577,
      "num_input_tokens_seen": 1742408,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 0.5138905048370361,
      "learning_rate": 3.5036687631027254e-05,
      "loss": 0.5743,
      "num_input_tokens_seen": 1746120,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 0.555175244808197,
      "learning_rate": 3.510220125786164e-05,
      "loss": 0.5831,
      "num_input_tokens_seen": 1749096,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 1.6514177322387695,
      "learning_rate": 3.516771488469602e-05,
      "loss": 0.5841,
      "num_input_tokens_seen": 1751624,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 0.8816514015197754,
      "learning_rate": 3.5233228511530397e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 1754216,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 0.3587028682231903,
      "learning_rate": 3.529874213836478e-05,
      "loss": 0.6022,
      "num_input_tokens_seen": 1758280,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 0.7351616024971008,
      "learning_rate": 3.536425576519916e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 1761352,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 0.3281859755516052,
      "learning_rate": 3.5429769392033546e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 1764648,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 0.6006975769996643,
      "learning_rate": 3.549528301886792e-05,
      "loss": 0.6528,
      "num_input_tokens_seen": 1767592,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 0.3159387409687042,
      "learning_rate": 3.556079664570231e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 1774600,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 0.889501690864563,
      "learning_rate": 3.562631027253669e-05,
      "loss": 0.6395,
      "num_input_tokens_seen": 1778248,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 0.4202078878879547,
      "learning_rate": 3.569182389937107e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 1781608,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 0.3617425262928009,
      "learning_rate": 3.5757337526205454e-05,
      "loss": 0.7021,
      "num_input_tokens_seen": 1784776,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 0.8963645100593567,
      "learning_rate": 3.582285115303983e-05,
      "loss": 0.6131,
      "num_input_tokens_seen": 1787688,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 0.4587342441082001,
      "learning_rate": 3.588836477987422e-05,
      "loss": 0.6515,
      "num_input_tokens_seen": 1791016,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 1.0174391269683838,
      "learning_rate": 3.59538784067086e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 1794280,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 0.6612292528152466,
      "learning_rate": 3.601939203354298e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 1797480,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 0.48800399899482727,
      "learning_rate": 3.608490566037736e-05,
      "loss": 0.6521,
      "num_input_tokens_seen": 1801160,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 0.6244049072265625,
      "learning_rate": 3.615041928721174e-05,
      "loss": 0.6512,
      "num_input_tokens_seen": 1803464,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 0.606228232383728,
      "learning_rate": 3.621593291404612e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 1806728,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 0.5249738097190857,
      "learning_rate": 3.6281446540880505e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 1810088,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 0.3758264482021332,
      "learning_rate": 3.634696016771489e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 1815144,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 0.33723676204681396,
      "learning_rate": 3.6412473794549265e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 1818792,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 0.4735422134399414,
      "learning_rate": 3.647798742138365e-05,
      "loss": 0.6442,
      "num_input_tokens_seen": 1821768,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 0.8027122616767883,
      "learning_rate": 3.654350104821803e-05,
      "loss": 0.6495,
      "num_input_tokens_seen": 1825512,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 0.5549845695495605,
      "learning_rate": 3.6609014675052414e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 1828968,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 0.5758669376373291,
      "learning_rate": 3.66745283018868e-05,
      "loss": 0.6161,
      "num_input_tokens_seen": 1831336,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 0.3780134618282318,
      "learning_rate": 3.6740041928721173e-05,
      "loss": 0.5989,
      "num_input_tokens_seen": 1834536,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 0.5243107676506042,
      "learning_rate": 3.6805555555555556e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 1837480,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 2.6274378299713135,
      "learning_rate": 3.687106918238994e-05,
      "loss": 0.6474,
      "num_input_tokens_seen": 1840808,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 0.43743327260017395,
      "learning_rate": 3.6936582809224316e-05,
      "loss": 0.6704,
      "num_input_tokens_seen": 1844392,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 0.48877644538879395,
      "learning_rate": 3.7002096436058706e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 1848008,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 0.5371912717819214,
      "learning_rate": 3.706761006289308e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 1850888,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 0.7853414416313171,
      "learning_rate": 3.7133123689727465e-05,
      "loss": 0.7311,
      "num_input_tokens_seen": 1854312,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 0.4478694200515747,
      "learning_rate": 3.719863731656185e-05,
      "loss": 0.526,
      "num_input_tokens_seen": 1858792,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 0.5025287866592407,
      "learning_rate": 3.7264150943396224e-05,
      "loss": 0.5428,
      "num_input_tokens_seen": 1861672,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 0.44468066096305847,
      "learning_rate": 3.732966457023061e-05,
      "loss": 0.5008,
      "num_input_tokens_seen": 1864360,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 0.867247462272644,
      "learning_rate": 3.739517819706499e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 1867048,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 0.39406123757362366,
      "learning_rate": 3.7460691823899374e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 1870280,
      "step": 2860
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 0.7261024713516235,
      "learning_rate": 3.752620545073376e-05,
      "loss": 0.6862,
      "num_input_tokens_seen": 1873896,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 0.33163994550704956,
      "learning_rate": 3.759171907756814e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 1876808,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 0.580349862575531,
      "learning_rate": 3.7657232704402516e-05,
      "loss": 0.5882,
      "num_input_tokens_seen": 1880168,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 1.0247117280960083,
      "learning_rate": 3.77227463312369e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 1882824,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 0.33168861269950867,
      "learning_rate": 3.778825995807128e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 1886184,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 0.37908995151519775,
      "learning_rate": 3.785377358490566e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 1888968,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 0.633793294429779,
      "learning_rate": 3.791928721174005e-05,
      "loss": 0.6421,
      "num_input_tokens_seen": 1892232,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 0.5135588049888611,
      "learning_rate": 3.7984800838574425e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 1895240,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 0.27480000257492065,
      "learning_rate": 3.805031446540881e-05,
      "loss": 0.3688,
      "num_input_tokens_seen": 1899144,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 0.40404564142227173,
      "learning_rate": 3.811582809224319e-05,
      "loss": 0.4898,
      "num_input_tokens_seen": 1901896,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.5006172060966492,
      "learning_rate": 3.818134171907757e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 1904936,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 0.41066259145736694,
      "learning_rate": 3.824685534591195e-05,
      "loss": 0.5985,
      "num_input_tokens_seen": 1908328,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 0.44623473286628723,
      "learning_rate": 3.831236897274633e-05,
      "loss": 0.676,
      "num_input_tokens_seen": 1911496,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 0.37284231185913086,
      "learning_rate": 3.8377882599580716e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 1913992,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 0.5823168754577637,
      "learning_rate": 3.844339622641509e-05,
      "loss": 0.5896,
      "num_input_tokens_seen": 1917640,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 0.46112269163131714,
      "learning_rate": 3.8508909853249476e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 1920616,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 0.5427432060241699,
      "learning_rate": 3.857442348008386e-05,
      "loss": 0.4754,
      "num_input_tokens_seen": 1923208,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 0.45294591784477234,
      "learning_rate": 3.863993710691824e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 1926952,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 0.6103368401527405,
      "learning_rate": 3.8705450733752625e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 1930696,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 0.32274118065834045,
      "learning_rate": 3.8770964360587e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 1933864,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 0.5873011350631714,
      "learning_rate": 3.883647798742139e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 1938440,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 0.3198045492172241,
      "learning_rate": 3.890199161425577e-05,
      "loss": 0.6165,
      "num_input_tokens_seen": 1942184,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 0.4906301200389862,
      "learning_rate": 3.8967505241090144e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 1945928,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 0.7319912314414978,
      "learning_rate": 3.9033018867924534e-05,
      "loss": 0.5936,
      "num_input_tokens_seen": 1949416,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 0.32603374123573303,
      "learning_rate": 3.909853249475891e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 1952008,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 0.8006625771522522,
      "learning_rate": 3.916404612159329e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 1955656,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 0.721966564655304,
      "learning_rate": 3.9229559748427676e-05,
      "loss": 0.6924,
      "num_input_tokens_seen": 1958632,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.3891414403915405,
      "learning_rate": 3.929507337526205e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 1961256,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 0.3575946092605591,
      "learning_rate": 3.9360587002096435e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 1965480,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 0.6758716106414795,
      "learning_rate": 3.942610062893082e-05,
      "loss": 0.5126,
      "num_input_tokens_seen": 1968552,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 0.32746800780296326,
      "learning_rate": 3.94916142557652e-05,
      "loss": 0.4815,
      "num_input_tokens_seen": 1971400,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 0.3012511730194092,
      "learning_rate": 3.9557127882599585e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 1974312,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 0.3521183729171753,
      "learning_rate": 3.962264150943397e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 1976904,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 0.748531699180603,
      "learning_rate": 3.9688155136268344e-05,
      "loss": 0.5869,
      "num_input_tokens_seen": 1979432,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 0.7045887112617493,
      "learning_rate": 3.975366876310273e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 1982952,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 1.0392208099365234,
      "learning_rate": 3.981918238993711e-05,
      "loss": 0.5123,
      "num_input_tokens_seen": 1986408,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 0.960878312587738,
      "learning_rate": 3.9884696016771486e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 1988936,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 0.502743124961853,
      "learning_rate": 3.9950209643605876e-05,
      "loss": 0.6923,
      "num_input_tokens_seen": 1992264,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 0.5978726744651794,
      "learning_rate": 4.001572327044025e-05,
      "loss": 0.6282,
      "num_input_tokens_seen": 1995240,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.5107831954956055,
      "learning_rate": 4.0081236897274636e-05,
      "loss": 0.5954,
      "num_input_tokens_seen": 1998408,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 0.5064271092414856,
      "learning_rate": 4.014675052410902e-05,
      "loss": 0.5268,
      "num_input_tokens_seen": 2001416,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 0.4002953767776489,
      "learning_rate": 4.0212264150943395e-05,
      "loss": 0.5263,
      "num_input_tokens_seen": 2003752,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 0.40201064944267273,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 2006728,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 0.5645060539245605,
      "learning_rate": 4.034329140461216e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 2009928,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 0.22789491713047028,
      "learning_rate": 4.0408805031446544e-05,
      "loss": 0.3863,
      "num_input_tokens_seen": 2012648,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 0.23358005285263062,
      "learning_rate": 4.047431865828093e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 2016936,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 0.6937338709831238,
      "learning_rate": 4.0539832285115304e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 2019720,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 0.6360812783241272,
      "learning_rate": 4.0605345911949687e-05,
      "loss": 0.4734,
      "num_input_tokens_seen": 2022760,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 0.6253808736801147,
      "learning_rate": 4.067085953878407e-05,
      "loss": 0.4565,
      "num_input_tokens_seen": 2025544,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 0.443312406539917,
      "learning_rate": 4.073637316561845e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 2028872,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 0.3713800311088562,
      "learning_rate": 4.080188679245283e-05,
      "loss": 0.4627,
      "num_input_tokens_seen": 2031816,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 0.48972561955451965,
      "learning_rate": 4.086740041928722e-05,
      "loss": 0.6454,
      "num_input_tokens_seen": 2034280,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 0.3429109454154968,
      "learning_rate": 4.0932914046121595e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 2038056,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 0.5275916457176208,
      "learning_rate": 4.099842767295597e-05,
      "loss": 0.5351,
      "num_input_tokens_seen": 2041192,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 0.46190714836120605,
      "learning_rate": 4.106394129979036e-05,
      "loss": 0.5383,
      "num_input_tokens_seen": 2043720,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 0.5321171283721924,
      "learning_rate": 4.112945492662474e-05,
      "loss": 0.6144,
      "num_input_tokens_seen": 2046792,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 0.3846127688884735,
      "learning_rate": 4.119496855345912e-05,
      "loss": 0.6489,
      "num_input_tokens_seen": 2049768,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 0.5265184640884399,
      "learning_rate": 4.1260482180293504e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 2052840,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 0.43091216683387756,
      "learning_rate": 4.132599580712788e-05,
      "loss": 0.6021,
      "num_input_tokens_seen": 2055432,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 0.6130406856536865,
      "learning_rate": 4.139150943396227e-05,
      "loss": 0.65,
      "num_input_tokens_seen": 2058792,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 0.3081125020980835,
      "learning_rate": 4.1457023060796646e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 2061960,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 0.43860509991645813,
      "learning_rate": 4.152253668763103e-05,
      "loss": 0.6858,
      "num_input_tokens_seen": 2064712,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 0.5450680255889893,
      "learning_rate": 4.158805031446541e-05,
      "loss": 0.457,
      "num_input_tokens_seen": 2068360,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.4129868745803833,
      "learning_rate": 4.1653563941299795e-05,
      "loss": 0.528,
      "num_input_tokens_seen": 2071304,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 0.47325676679611206,
      "learning_rate": 4.171907756813417e-05,
      "loss": 0.7248,
      "num_input_tokens_seen": 2075016,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 0.21501410007476807,
      "learning_rate": 4.1784591194968555e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 2079016,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 0.37771373987197876,
      "learning_rate": 4.185010482180294e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 2081960,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 0.6592081189155579,
      "learning_rate": 4.1915618448637314e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 2084648,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 0.48395490646362305,
      "learning_rate": 4.1981132075471704e-05,
      "loss": 0.4799,
      "num_input_tokens_seen": 2088840,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 0.8975138068199158,
      "learning_rate": 4.204664570230608e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 2091944,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 0.3844721019268036,
      "learning_rate": 4.2112159329140463e-05,
      "loss": 0.588,
      "num_input_tokens_seen": 2095144,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 0.7249630093574524,
      "learning_rate": 4.2177672955974846e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 2098984,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 0.24474464356899261,
      "learning_rate": 4.224318658280922e-05,
      "loss": 0.4094,
      "num_input_tokens_seen": 2101736,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 0.3592902421951294,
      "learning_rate": 4.2308700209643606e-05,
      "loss": 0.5744,
      "num_input_tokens_seen": 2104904,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 0.49667927622795105,
      "learning_rate": 4.237421383647799e-05,
      "loss": 0.5889,
      "num_input_tokens_seen": 2108392,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.48664242029190063,
      "learning_rate": 4.243972746331237e-05,
      "loss": 0.6416,
      "num_input_tokens_seen": 2112360,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 0.5645898580551147,
      "learning_rate": 4.2505241090146755e-05,
      "loss": 0.5619,
      "num_input_tokens_seen": 2117992,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 0.568172812461853,
      "learning_rate": 4.257075471698113e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 2121096,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 0.5583387613296509,
      "learning_rate": 4.2636268343815514e-05,
      "loss": 0.528,
      "num_input_tokens_seen": 2125064,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 0.4617266058921814,
      "learning_rate": 4.27017819706499e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 2129032,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 1.5707463026046753,
      "learning_rate": 4.276729559748428e-05,
      "loss": 0.6396,
      "num_input_tokens_seen": 2131656,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 2.1442434787750244,
      "learning_rate": 4.283280922431866e-05,
      "loss": 0.6051,
      "num_input_tokens_seen": 2134760,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 0.8837007880210876,
      "learning_rate": 4.289832285115305e-05,
      "loss": 0.6293,
      "num_input_tokens_seen": 2137960,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 0.6241505146026611,
      "learning_rate": 4.296383647798742e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 2140552,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 0.8786905407905579,
      "learning_rate": 4.3029350104821806e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 2143272,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 0.3488609194755554,
      "learning_rate": 4.309486373165619e-05,
      "loss": 0.5472,
      "num_input_tokens_seen": 2147304,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 0.518197774887085,
      "learning_rate": 4.3160377358490565e-05,
      "loss": 0.6954,
      "num_input_tokens_seen": 2151624,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 0.47903206944465637,
      "learning_rate": 4.322589098532495e-05,
      "loss": 0.6115,
      "num_input_tokens_seen": 2154248,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 0.5837568044662476,
      "learning_rate": 4.329140461215933e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 2157000,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 0.5834053754806519,
      "learning_rate": 4.335691823899371e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 2160904,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 0.375508576631546,
      "learning_rate": 4.34224318658281e-05,
      "loss": 0.5898,
      "num_input_tokens_seen": 2163720,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 0.34189823269844055,
      "learning_rate": 4.3487945492662474e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 2167912,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 0.30721285939216614,
      "learning_rate": 4.355345911949686e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 2171592,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 0.47784438729286194,
      "learning_rate": 4.361897274633124e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 2173864,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 0.7459078431129456,
      "learning_rate": 4.368448637316562e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 2176744,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 0.4961254894733429,
      "learning_rate": 4.375e-05,
      "loss": 0.7122,
      "num_input_tokens_seen": 2179528,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 0.8311951756477356,
      "learning_rate": 4.381551362683438e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 2182856,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 0.31831803917884827,
      "learning_rate": 4.3881027253668766e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 2185832,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 0.3371331989765167,
      "learning_rate": 4.394654088050314e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 2189352,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.9163993000984192,
      "learning_rate": 4.401205450733753e-05,
      "loss": 0.5215,
      "num_input_tokens_seen": 2192264,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 0.2656177580356598,
      "learning_rate": 4.407756813417191e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 2195560,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 0.3057287931442261,
      "learning_rate": 4.414308176100629e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 2199080,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 0.34862565994262695,
      "learning_rate": 4.4208595387840674e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 2202408,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 0.5817009210586548,
      "learning_rate": 4.427410901467505e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 2206056,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 0.6498614549636841,
      "learning_rate": 4.433962264150944e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 2208456,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 0.271220862865448,
      "learning_rate": 4.440513626834382e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 2211848,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 0.3262849748134613,
      "learning_rate": 4.44706498951782e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 2215112,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 0.5797364115715027,
      "learning_rate": 4.453616352201258e-05,
      "loss": 0.5557,
      "num_input_tokens_seen": 2219528,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 0.8057603240013123,
      "learning_rate": 4.460167714884696e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 2223624,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 0.4726003110408783,
      "learning_rate": 4.466719077568134e-05,
      "loss": 0.5201,
      "num_input_tokens_seen": 2227464,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 0.2382994294166565,
      "learning_rate": 4.4732704402515725e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 2231240,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.6309396028518677,
      "learning_rate": 4.479821802935011e-05,
      "loss": 0.6151,
      "num_input_tokens_seen": 2234600,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 0.38780179619789124,
      "learning_rate": 4.4863731656184485e-05,
      "loss": 0.5416,
      "num_input_tokens_seen": 2238056,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 0.38917863368988037,
      "learning_rate": 4.4929245283018875e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 2241640,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 0.4071073830127716,
      "learning_rate": 4.499475890985325e-05,
      "loss": 0.5784,
      "num_input_tokens_seen": 2245832,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 0.5710375905036926,
      "learning_rate": 4.5060272536687634e-05,
      "loss": 0.7591,
      "num_input_tokens_seen": 2248552,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.5244611501693726,
      "learning_rate": 4.512578616352202e-05,
      "loss": 0.6065,
      "num_input_tokens_seen": 2251432,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 0.8536500334739685,
      "learning_rate": 4.519129979035639e-05,
      "loss": 0.6152,
      "num_input_tokens_seen": 2254408,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 0.28498679399490356,
      "learning_rate": 4.525681341719078e-05,
      "loss": 0.7231,
      "num_input_tokens_seen": 2257032,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 0.4505796432495117,
      "learning_rate": 4.532232704402516e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 2259560,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 0.3603012263774872,
      "learning_rate": 4.5387840670859536e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 2263240,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 0.4512142837047577,
      "learning_rate": 4.5453354297693926e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 2267144,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 0.8294135332107544,
      "learning_rate": 4.55188679245283e-05,
      "loss": 0.595,
      "num_input_tokens_seen": 2270856,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.49112969636917114,
      "learning_rate": 4.5584381551362685e-05,
      "loss": 0.6156,
      "num_input_tokens_seen": 2273672,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 0.4266447424888611,
      "learning_rate": 4.564989517819707e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 2276520,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 0.402415931224823,
      "learning_rate": 4.571540880503145e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 2279368,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 0.4393315613269806,
      "learning_rate": 4.578092243186583e-05,
      "loss": 0.7671,
      "num_input_tokens_seen": 2281992,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 0.3720022737979889,
      "learning_rate": 4.584643605870021e-05,
      "loss": 0.4897,
      "num_input_tokens_seen": 2285576,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 0.8820444345474243,
      "learning_rate": 4.5911949685534594e-05,
      "loss": 0.4318,
      "num_input_tokens_seen": 2288072,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 0.5302202105522156,
      "learning_rate": 4.5977463312368977e-05,
      "loss": 0.4113,
      "num_input_tokens_seen": 2291336,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 0.5523110628128052,
      "learning_rate": 4.604297693920336e-05,
      "loss": 0.5747,
      "num_input_tokens_seen": 2294152,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 0.396359384059906,
      "learning_rate": 4.6108490566037736e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 2296872,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 0.4807191491127014,
      "learning_rate": 4.617400419287212e-05,
      "loss": 0.4836,
      "num_input_tokens_seen": 2299528,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 0.6817732453346252,
      "learning_rate": 4.62395178197065e-05,
      "loss": 0.6899,
      "num_input_tokens_seen": 2301864,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 0.30229610204696655,
      "learning_rate": 4.630503144654088e-05,
      "loss": 0.4539,
      "num_input_tokens_seen": 2306088,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 0.5086938142776489,
      "learning_rate": 4.637054507337527e-05,
      "loss": 0.5822,
      "num_input_tokens_seen": 2309416,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 0.6819168925285339,
      "learning_rate": 4.6436058700209645e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 2312328,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 0.5524068474769592,
      "learning_rate": 4.650157232704403e-05,
      "loss": 0.6029,
      "num_input_tokens_seen": 2314472,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 0.40533149242401123,
      "learning_rate": 4.656708595387841e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 2318664,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 0.5225434303283691,
      "learning_rate": 4.663259958071279e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 2323784,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 0.6387107968330383,
      "learning_rate": 4.669811320754717e-05,
      "loss": 0.5607,
      "num_input_tokens_seen": 2326536,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 0.37480664253234863,
      "learning_rate": 4.676362683438155e-05,
      "loss": 0.5098,
      "num_input_tokens_seen": 2330440,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 0.08939900994300842,
      "learning_rate": 4.6829140461215936e-05,
      "loss": 0.5987,
      "num_input_tokens_seen": 2335688,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 0.31556135416030884,
      "learning_rate": 4.689465408805032e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 2338568,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 0.5322635769844055,
      "learning_rate": 4.69601677148847e-05,
      "loss": 0.5476,
      "num_input_tokens_seen": 2342120,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 0.5392477512359619,
      "learning_rate": 4.702568134171908e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 2346184,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 0.34723249077796936,
      "learning_rate": 4.709119496855346e-05,
      "loss": 0.625,
      "num_input_tokens_seen": 2348936,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.5420917868614197,
      "learning_rate": 4.7156708595387845e-05,
      "loss": 0.5294,
      "num_input_tokens_seen": 2351976,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 0.266939640045166,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 2357192,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 0.7240489721298218,
      "learning_rate": 4.728773584905661e-05,
      "loss": 0.6847,
      "num_input_tokens_seen": 2360072,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 0.5809356570243835,
      "learning_rate": 4.735324947589099e-05,
      "loss": 0.4899,
      "num_input_tokens_seen": 2363656,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 0.6671231389045715,
      "learning_rate": 4.7418763102725364e-05,
      "loss": 0.5697,
      "num_input_tokens_seen": 2366216,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 0.5524076223373413,
      "learning_rate": 4.7484276729559753e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 2369448,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 0.4837782680988312,
      "learning_rate": 4.754979035639413e-05,
      "loss": 0.6856,
      "num_input_tokens_seen": 2372744,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 0.5065600275993347,
      "learning_rate": 4.761530398322851e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 2375080,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 0.49417760968208313,
      "learning_rate": 4.7680817610062896e-05,
      "loss": 0.5465,
      "num_input_tokens_seen": 2377864,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 0.5547209978103638,
      "learning_rate": 4.774633123689728e-05,
      "loss": 0.5268,
      "num_input_tokens_seen": 2381224,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 0.3833390772342682,
      "learning_rate": 4.7811844863731655e-05,
      "loss": 0.5485,
      "num_input_tokens_seen": 2384424,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 0.4416307210922241,
      "learning_rate": 4.787735849056604e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 2388936,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.5276903510093689,
      "learning_rate": 4.794287211740042e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 2392744,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 0.7202589511871338,
      "learning_rate": 4.8008385744234804e-05,
      "loss": 0.6272,
      "num_input_tokens_seen": 2395112,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 0.4432857930660248,
      "learning_rate": 4.807389937106919e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 2398344,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 0.3674360513687134,
      "learning_rate": 4.8139412997903564e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 2401704,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 0.676524817943573,
      "learning_rate": 4.8204926624737954e-05,
      "loss": 0.4079,
      "num_input_tokens_seen": 2404168,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 0.38717254996299744,
      "learning_rate": 4.827044025157233e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 2407688,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 0.3733403980731964,
      "learning_rate": 4.8335953878406706e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 2411528,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 0.5918490886688232,
      "learning_rate": 4.8401467505241096e-05,
      "loss": 0.5761,
      "num_input_tokens_seen": 2414440,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 0.33076900243759155,
      "learning_rate": 4.846698113207547e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 2417448,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 0.49627265334129333,
      "learning_rate": 4.8532494758909855e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 2420680,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.502655029296875,
      "learning_rate": 4.859800838574424e-05,
      "loss": 0.5087,
      "num_input_tokens_seen": 2423752,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 0.44927382469177246,
      "learning_rate": 4.8663522012578615e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 2426216,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.514447808265686,
      "learning_rate": 4.8729035639413e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 2429640,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 0.6143369078636169,
      "learning_rate": 4.879454926624738e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 2433224,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 0.5335900783538818,
      "learning_rate": 4.8860062893081764e-05,
      "loss": 0.579,
      "num_input_tokens_seen": 2437576,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 0.6349103450775146,
      "learning_rate": 4.892557651991615e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 2439976,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 0.4818267226219177,
      "learning_rate": 4.899109014675053e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 2443272,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 0.4876686632633209,
      "learning_rate": 4.9056603773584906e-05,
      "loss": 0.5768,
      "num_input_tokens_seen": 2445832,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 0.4319758117198944,
      "learning_rate": 4.912211740041929e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 2448456,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 0.7715495824813843,
      "learning_rate": 4.918763102725367e-05,
      "loss": 0.6647,
      "num_input_tokens_seen": 2451208,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 0.6240746378898621,
      "learning_rate": 4.925314465408805e-05,
      "loss": 0.4989,
      "num_input_tokens_seen": 2454152,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 1.0925219058990479,
      "learning_rate": 4.931865828092244e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 2457736,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 0.48760131001472473,
      "learning_rate": 4.9384171907756815e-05,
      "loss": 0.638,
      "num_input_tokens_seen": 2461384,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 0.5855780839920044,
      "learning_rate": 4.944968553459119e-05,
      "loss": 0.4021,
      "num_input_tokens_seen": 2464776,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.9770728349685669,
      "learning_rate": 4.951519916142558e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 2467144,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 0.4186440110206604,
      "learning_rate": 4.958071278825996e-05,
      "loss": 0.6297,
      "num_input_tokens_seen": 2469736,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 0.4647338092327118,
      "learning_rate": 4.964622641509434e-05,
      "loss": 0.3917,
      "num_input_tokens_seen": 2474536,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 0.40726739168167114,
      "learning_rate": 4.9711740041928724e-05,
      "loss": 0.5553,
      "num_input_tokens_seen": 2477704,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 0.3790849447250366,
      "learning_rate": 4.977725366876311e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 2480296,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 0.40735065937042236,
      "learning_rate": 4.984276729559749e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 2483368,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 0.3461471199989319,
      "learning_rate": 4.9908280922431866e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 2486440,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 0.38950541615486145,
      "learning_rate": 4.997379454926625e-05,
      "loss": 0.3809,
      "num_input_tokens_seen": 2489352,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.49833449721336365,
      "eval_runtime": 9.8038,
      "eval_samples_per_second": 86.497,
      "eval_steps_per_second": 21.624,
      "num_input_tokens_seen": 2489456,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 0.5281547904014587,
      "learning_rate": 4.999999905865133e-05,
      "loss": 0.5823,
      "num_input_tokens_seen": 2492112,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 0.3927661180496216,
      "learning_rate": 4.999999330596526e-05,
      "loss": 0.4117,
      "num_input_tokens_seen": 2494800,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 0.6650112271308899,
      "learning_rate": 4.9999982323565806e-05,
      "loss": 0.527,
      "num_input_tokens_seen": 2498896,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 0.45997366309165955,
      "learning_rate": 4.999996611145527e-05,
      "loss": 0.5782,
      "num_input_tokens_seen": 2502160,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.3432539701461792,
      "learning_rate": 4.9999944669637024e-05,
      "loss": 0.5311,
      "num_input_tokens_seen": 2504752,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 0.5930073261260986,
      "learning_rate": 4.999991799811557e-05,
      "loss": 0.518,
      "num_input_tokens_seen": 2507536,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 0.41388165950775146,
      "learning_rate": 4.99998860968965e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 2510352,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 0.4849447011947632,
      "learning_rate": 4.999984896598646e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 2512784,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 0.6814974546432495,
      "learning_rate": 4.999980660539323e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 2516304,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 0.6170077919960022,
      "learning_rate": 4.999975901512567e-05,
      "loss": 0.5568,
      "num_input_tokens_seen": 2519184,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 0.3396223187446594,
      "learning_rate": 4.999970619519374e-05,
      "loss": 0.5921,
      "num_input_tokens_seen": 2522448,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 0.5683161616325378,
      "learning_rate": 4.999964814560849e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 2525744,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 0.4875989854335785,
      "learning_rate": 4.999958486638205e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 2528240,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 0.5541222095489502,
      "learning_rate": 4.999951635752767e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 2530928,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 0.432606041431427,
      "learning_rate": 4.999944261905968e-05,
      "loss": 0.4985,
      "num_input_tokens_seen": 2533936,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 0.48526430130004883,
      "learning_rate": 4.99993636509935e-05,
      "loss": 0.6236,
      "num_input_tokens_seen": 2536720,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.2603623569011688,
      "learning_rate": 4.999927945334565e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 2540656,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 0.38862499594688416,
      "learning_rate": 4.9999190026133745e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 2545040,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 0.6088616847991943,
      "learning_rate": 4.9999095369376506e-05,
      "loss": 0.7432,
      "num_input_tokens_seen": 2547888,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 0.61790531873703,
      "learning_rate": 4.99989954830937e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 2550256,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 0.37450966238975525,
      "learning_rate": 4.9998890367306255e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 2553712,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 0.5666323304176331,
      "learning_rate": 4.999878002203614e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 2557392,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 0.43842971324920654,
      "learning_rate": 4.999866444730645e-05,
      "loss": 0.6943,
      "num_input_tokens_seen": 2560656,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 0.36431044340133667,
      "learning_rate": 4.999854364314136e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 2563984,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 0.421047568321228,
      "learning_rate": 4.999841760956614e-05,
      "loss": 0.6158,
      "num_input_tokens_seen": 2567088,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 0.33520326018333435,
      "learning_rate": 4.999828634660715e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 2570288,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 0.428559809923172,
      "learning_rate": 4.9998149854291856e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 2573136,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 0.5236575603485107,
      "learning_rate": 4.99980081326488e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 2576912,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.8575135469436646,
      "learning_rate": 4.999786118170763e-05,
      "loss": 0.5519,
      "num_input_tokens_seen": 2579824,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 0.6354091763496399,
      "learning_rate": 4.99977090014991e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 2583664,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 0.355674147605896,
      "learning_rate": 4.9997551592055034e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 2587344,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.44578275084495544,
      "learning_rate": 4.999738895340836e-05,
      "loss": 0.5013,
      "num_input_tokens_seen": 2590800,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 0.6696840524673462,
      "learning_rate": 4.999722108559311e-05,
      "loss": 0.6497,
      "num_input_tokens_seen": 2593872,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 0.4123932719230652,
      "learning_rate": 4.9997047988644385e-05,
      "loss": 0.5713,
      "num_input_tokens_seen": 2596976,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 0.7604090571403503,
      "learning_rate": 4.99968696625984e-05,
      "loss": 0.5131,
      "num_input_tokens_seen": 2599824,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 0.46573391556739807,
      "learning_rate": 4.999668610749246e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 2602992,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.6632283926010132,
      "learning_rate": 4.999649732336498e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 2606448,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 0.48741471767425537,
      "learning_rate": 4.999630331025542e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 2610864,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 0.23424816131591797,
      "learning_rate": 4.9996104068204385e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 2614256,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 0.3715061843395233,
      "learning_rate": 4.999589959725355e-05,
      "loss": 0.669,
      "num_input_tokens_seen": 2617072,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.7061977982521057,
      "learning_rate": 4.999568989744568e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 2619792,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 0.37580177187919617,
      "learning_rate": 4.999547496882466e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 2622448,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 0.7939099669456482,
      "learning_rate": 4.999525481143543e-05,
      "loss": 0.5537,
      "num_input_tokens_seen": 2625328,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 0.6788330674171448,
      "learning_rate": 4.999502942532406e-05,
      "loss": 0.5503,
      "num_input_tokens_seen": 2628336,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 0.5238294005393982,
      "learning_rate": 4.999479881053769e-05,
      "loss": 0.5037,
      "num_input_tokens_seen": 2634128,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 0.3134955167770386,
      "learning_rate": 4.999456296712457e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 2638672,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 0.2638826072216034,
      "learning_rate": 4.9994321895134025e-05,
      "loss": 0.5387,
      "num_input_tokens_seen": 2642768,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 0.8358477354049683,
      "learning_rate": 4.999407559461649e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 2645552,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 0.5982584357261658,
      "learning_rate": 4.999382406562349e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 2647728,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 0.3170035481452942,
      "learning_rate": 4.999356730820765e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 2650992,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 0.4449063241481781,
      "learning_rate": 4.999330532242266e-05,
      "loss": 0.5324,
      "num_input_tokens_seen": 2653744,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 0.6684060096740723,
      "learning_rate": 4.999303810832333e-05,
      "loss": 0.3152,
      "num_input_tokens_seen": 2656336,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.39154380559921265,
      "learning_rate": 4.999276566596557e-05,
      "loss": 0.7671,
      "num_input_tokens_seen": 2659216,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 0.5578533411026001,
      "learning_rate": 4.9992487995406364e-05,
      "loss": 0.568,
      "num_input_tokens_seen": 2662224,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 0.425534725189209,
      "learning_rate": 4.9992205096703806e-05,
      "loss": 0.3932,
      "num_input_tokens_seen": 2665936,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 0.3798469603061676,
      "learning_rate": 4.9991916969917054e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 2669648,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 0.8567971587181091,
      "learning_rate": 4.99916236151064e-05,
      "loss": 0.741,
      "num_input_tokens_seen": 2673552,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 0.43744510412216187,
      "learning_rate": 4.999132503233321e-05,
      "loss": 0.6466,
      "num_input_tokens_seen": 2675824,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 1.0393890142440796,
      "learning_rate": 4.999102122165994e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 2678256,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 0.3131391108036041,
      "learning_rate": 4.9990712183150135e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 2682224,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 0.47425544261932373,
      "learning_rate": 4.9990397916868456e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 2684848,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 0.3411799371242523,
      "learning_rate": 4.999007842288064e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 2688816,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 0.4673491418361664,
      "learning_rate": 4.9989753701253514e-05,
      "loss": 0.9526,
      "num_input_tokens_seen": 2691728,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 1.0026285648345947,
      "learning_rate": 4.998942375205502e-05,
      "loss": 0.6099,
      "num_input_tokens_seen": 2694608,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.383327454328537,
      "learning_rate": 4.998908857535416e-05,
      "loss": 0.5546,
      "num_input_tokens_seen": 2697872,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 0.328510582447052,
      "learning_rate": 4.9988748171221065e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 2701040,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 0.7375602126121521,
      "learning_rate": 4.998840253972695e-05,
      "loss": 0.6623,
      "num_input_tokens_seen": 2703760,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 0.29057005047798157,
      "learning_rate": 4.9988051680944094e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 2706480,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 0.3575519621372223,
      "learning_rate": 4.998769559494591e-05,
      "loss": 0.3676,
      "num_input_tokens_seen": 2709488,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 0.2398511916399002,
      "learning_rate": 4.998733428180689e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 2712880,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 0.4493565261363983,
      "learning_rate": 4.998696774160259e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 2716112,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 0.45296886563301086,
      "learning_rate": 4.9986595974409714e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 2719088,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 0.458746075630188,
      "learning_rate": 4.998621898030603e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 2722064,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 0.4648973047733307,
      "learning_rate": 4.998583675937039e-05,
      "loss": 0.3668,
      "num_input_tokens_seen": 2725072,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 0.42623743414878845,
      "learning_rate": 4.998544931168274e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 2728368,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 0.3581874370574951,
      "learning_rate": 4.9985056637324154e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 2731984,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.5758193135261536,
      "learning_rate": 4.998465873637677e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 2735632,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 0.3537597060203552,
      "learning_rate": 4.9984255608923806e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 2738928,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 0.5726908445358276,
      "learning_rate": 4.998384725504961e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 2743248,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 0.5242956280708313,
      "learning_rate": 4.9983433674839596e-05,
      "loss": 0.6058,
      "num_input_tokens_seen": 2745744,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 1.298636794090271,
      "learning_rate": 4.998301486838028e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 2748432,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 0.5056727528572083,
      "learning_rate": 4.998259083575929e-05,
      "loss": 0.5992,
      "num_input_tokens_seen": 2751280,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 0.5466222763061523,
      "learning_rate": 4.99821615770653e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 2754096,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 0.47651946544647217,
      "learning_rate": 4.998172709238813e-05,
      "loss": 0.6438,
      "num_input_tokens_seen": 2757200,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.6016761064529419,
      "learning_rate": 4.9981287381818654e-05,
      "loss": 0.547,
      "num_input_tokens_seen": 2760240,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 1.021174430847168,
      "learning_rate": 4.9980842445448856e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 2763248,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 0.4226081669330597,
      "learning_rate": 4.9980392283371815e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 2766192,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 0.33363211154937744,
      "learning_rate": 4.99799368956817e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 2770160,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 0.296336829662323,
      "learning_rate": 4.997947628247378e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 2773424,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 0.617171585559845,
      "learning_rate": 4.9979010443844395e-05,
      "loss": 0.4853,
      "num_input_tokens_seen": 2776240,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 0.6158663630485535,
      "learning_rate": 4.9978539379891e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 2779088,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 0.3329649865627289,
      "learning_rate": 4.997806309071214e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 2782032,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 0.5062671899795532,
      "learning_rate": 4.997758157640745e-05,
      "loss": 0.6582,
      "num_input_tokens_seen": 2786000,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 0.6084141731262207,
      "learning_rate": 4.997709483707765e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 2789584,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 0.4202485680580139,
      "learning_rate": 4.997660287282456e-05,
      "loss": 0.5653,
      "num_input_tokens_seen": 2792976,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 0.5464098453521729,
      "learning_rate": 4.99761056837511e-05,
      "loss": 0.5426,
      "num_input_tokens_seen": 2795856,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 0.5281027555465698,
      "learning_rate": 4.997560326996128e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 2798384,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 0.3271937668323517,
      "learning_rate": 4.997509563156018e-05,
      "loss": 0.3948,
      "num_input_tokens_seen": 2801552,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 0.6900986433029175,
      "learning_rate": 4.997458276865401e-05,
      "loss": 0.5673,
      "num_input_tokens_seen": 2804816,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 0.9722169041633606,
      "learning_rate": 4.997406468135005e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 2807536,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.28372064232826233,
      "learning_rate": 4.9973541369756686e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 2812080,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 1.6152405738830566,
      "learning_rate": 4.9973012833983365e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 2814768,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 0.6999671459197998,
      "learning_rate": 4.9972479074140684e-05,
      "loss": 0.56,
      "num_input_tokens_seen": 2817328,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 0.47730496525764465,
      "learning_rate": 4.9971940090340276e-05,
      "loss": 0.632,
      "num_input_tokens_seen": 2820624,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 0.4055004119873047,
      "learning_rate": 4.997139588269489e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 2824720,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 0.368868887424469,
      "learning_rate": 4.997084645131838e-05,
      "loss": 0.6056,
      "num_input_tokens_seen": 2828560,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 0.3762630522251129,
      "learning_rate": 4.9970291796325666e-05,
      "loss": 0.5709,
      "num_input_tokens_seen": 2832080,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 0.8151198625564575,
      "learning_rate": 4.996973191783279e-05,
      "loss": 0.6202,
      "num_input_tokens_seen": 2834640,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 0.3949766457080841,
      "learning_rate": 4.996916681595687e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 2837136,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 0.4674360752105713,
      "learning_rate": 4.996859649081611e-05,
      "loss": 0.4998,
      "num_input_tokens_seen": 2841040,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 0.7122979164123535,
      "learning_rate": 4.996802094252983e-05,
      "loss": 0.5675,
      "num_input_tokens_seen": 2843984,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 0.6229714751243591,
      "learning_rate": 4.996744017121841e-05,
      "loss": 0.6839,
      "num_input_tokens_seen": 2846832,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 0.23388148844242096,
      "learning_rate": 4.9966854177003355e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 2850512,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 0.3867023289203644,
      "learning_rate": 4.996626296000724e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 2854800,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 0.6645195484161377,
      "learning_rate": 4.996566652035374e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 2857552,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 0.6311596035957336,
      "learning_rate": 4.996506485816763e-05,
      "loss": 0.5247,
      "num_input_tokens_seen": 2860432,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 0.3345215320587158,
      "learning_rate": 4.996445797357477e-05,
      "loss": 0.5884,
      "num_input_tokens_seen": 2863568,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 0.35921889543533325,
      "learning_rate": 4.996384586670211e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 2865968,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 0.5295884013175964,
      "learning_rate": 4.99632285376777e-05,
      "loss": 0.6692,
      "num_input_tokens_seen": 2868848,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 0.45511046051979065,
      "learning_rate": 4.9962605986630674e-05,
      "loss": 0.4339,
      "num_input_tokens_seen": 2871792,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 0.32977160811424255,
      "learning_rate": 4.996197821369126e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 2874832,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 1.127699375152588,
      "learning_rate": 4.9961345218990796e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 2877648,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 0.5754979252815247,
      "learning_rate": 4.996070700266168e-05,
      "loss": 0.5282,
      "num_input_tokens_seen": 2880016,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 0.5154451727867126,
      "learning_rate": 4.9960063564837424e-05,
      "loss": 0.6297,
      "num_input_tokens_seen": 2883184,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 0.39759135246276855,
      "learning_rate": 4.995941490565262e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 2887408,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 0.25050073862075806,
      "learning_rate": 4.995876102524299e-05,
      "loss": 0.2914,
      "num_input_tokens_seen": 2892912,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 0.42155107855796814,
      "learning_rate": 4.995810192374529e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 2895504,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 0.4472399353981018,
      "learning_rate": 4.99574376012974e-05,
      "loss": 0.5931,
      "num_input_tokens_seen": 2900208,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 0.3814808428287506,
      "learning_rate": 4.995676805803829e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 2903568,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 0.5156609416007996,
      "learning_rate": 4.995609329410804e-05,
      "loss": 0.6024,
      "num_input_tokens_seen": 2906864,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 0.4570445418357849,
      "learning_rate": 4.9955413309647776e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 2910480,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 0.4262515902519226,
      "learning_rate": 4.9954728104799756e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 2913552,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 0.4844858646392822,
      "learning_rate": 4.995403767970731e-05,
      "loss": 0.5777,
      "num_input_tokens_seen": 2916368,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 0.28090155124664307,
      "learning_rate": 4.995334203451488e-05,
      "loss": 0.5548,
      "num_input_tokens_seen": 2918960,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 0.3161591589450836,
      "learning_rate": 4.995264116936798e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 2922768,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 0.4541705250740051,
      "learning_rate": 4.995193508441322e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 2925840,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 0.3434845507144928,
      "learning_rate": 4.99512237797983e-05,
      "loss": 0.4159,
      "num_input_tokens_seen": 2929008,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 0.5106178522109985,
      "learning_rate": 4.995050725567203e-05,
      "loss": 0.5212,
      "num_input_tokens_seen": 2931632,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 0.45632967352867126,
      "learning_rate": 4.994978551218429e-05,
      "loss": 0.587,
      "num_input_tokens_seen": 2935824,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 0.5586308836936951,
      "learning_rate": 4.9949058549486066e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 2938832,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 0.36611276865005493,
      "learning_rate": 4.994832636772943e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 2942096,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 0.4213748872280121,
      "learning_rate": 4.994758896706754e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 2945968,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 0.23573896288871765,
      "learning_rate": 4.994684634765465e-05,
      "loss": 0.5515,
      "num_input_tokens_seen": 2949456,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 0.49947208166122437,
      "learning_rate": 4.9946098509646125e-05,
      "loss": 0.5318,
      "num_input_tokens_seen": 2952592,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 0.3086763620376587,
      "learning_rate": 4.994534545319838e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 2955760,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 0.43818050622940063,
      "learning_rate": 4.994458717846896e-05,
      "loss": 0.3314,
      "num_input_tokens_seen": 2958768,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 1.2556806802749634,
      "learning_rate": 4.9943823685616485e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 2961552,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 0.37019798159599304,
      "learning_rate": 4.994305497480067e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 2964464,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 0.408856600522995,
      "learning_rate": 4.994228104618233e-05,
      "loss": 0.578,
      "num_input_tokens_seen": 2966832,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 0.4250966012477875,
      "learning_rate": 4.994150189992334e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 2969968,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 0.5942469835281372,
      "learning_rate": 4.99407175361867e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 2973040,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 0.3298306167125702,
      "learning_rate": 4.99399279551365e-05,
      "loss": 0.5525,
      "num_input_tokens_seen": 2977520,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 0.42414388060569763,
      "learning_rate": 4.9939133156937886e-05,
      "loss": 0.4032,
      "num_input_tokens_seen": 2980880,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 0.4367046356201172,
      "learning_rate": 4.993833314175716e-05,
      "loss": 0.6272,
      "num_input_tokens_seen": 2983632,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 0.44117268919944763,
      "learning_rate": 4.993752790976163e-05,
      "loss": 0.4645,
      "num_input_tokens_seen": 2986256,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 0.44632089138031006,
      "learning_rate": 4.993671746111977e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 2991056,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 0.5637568831443787,
      "learning_rate": 4.993590179600111e-05,
      "loss": 0.389,
      "num_input_tokens_seen": 2994320,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 0.9562558531761169,
      "learning_rate": 4.9935080914576284e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 2997104,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 0.4557703137397766,
      "learning_rate": 4.9934254817016995e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 3000112,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 0.30491042137145996,
      "learning_rate": 4.993342350349607e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 3003568,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.3650408387184143,
      "learning_rate": 4.9932586974187404e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 3007696,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 0.4906516969203949,
      "learning_rate": 4.9931745229265986e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 3011472,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 0.525799036026001,
      "learning_rate": 4.99308982689079e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 3014320,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 0.7595609426498413,
      "learning_rate": 4.993004609329033e-05,
      "loss": 0.6122,
      "num_input_tokens_seen": 3018224,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 0.523311972618103,
      "learning_rate": 4.992918870259153e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 3021072,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 0.367504358291626,
      "learning_rate": 4.992832609699086e-05,
      "loss": 0.3809,
      "num_input_tokens_seen": 3023600,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 0.5528583526611328,
      "learning_rate": 4.992745827666877e-05,
      "loss": 0.5211,
      "num_input_tokens_seen": 3026768,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 0.4033782482147217,
      "learning_rate": 4.99265852418068e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 3033424,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 0.7311928868293762,
      "learning_rate": 4.992570699258757e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 3035952,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 0.6832736134529114,
      "learning_rate": 4.99248235291948e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 3039216,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 0.4957515001296997,
      "learning_rate": 4.9923934851813314e-05,
      "loss": 0.7263,
      "num_input_tokens_seen": 3042224,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 0.3428243398666382,
      "learning_rate": 4.9923040960628996e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 3044752,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.5221303701400757,
      "learning_rate": 4.9922141855828856e-05,
      "loss": 0.5997,
      "num_input_tokens_seen": 3048336,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 0.9065271019935608,
      "learning_rate": 4.992123753760096e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 3050864,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 1.0184180736541748,
      "learning_rate": 4.9920328006134484e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 3053712,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 0.4849921464920044,
      "learning_rate": 4.991941326161971e-05,
      "loss": 0.6858,
      "num_input_tokens_seen": 3056944,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 0.4381405711174011,
      "learning_rate": 4.991849330424796e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 3060176,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 0.4994543492794037,
      "learning_rate": 4.9917568134211703e-05,
      "loss": 0.7853,
      "num_input_tokens_seen": 3062352,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 0.6190600395202637,
      "learning_rate": 4.991663775170447e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 3064848,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 0.49057286977767944,
      "learning_rate": 4.991570215692088e-05,
      "loss": 0.5748,
      "num_input_tokens_seen": 3068368,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 0.6265596747398376,
      "learning_rate": 4.991476135005665e-05,
      "loss": 0.6162,
      "num_input_tokens_seen": 3071536,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 0.47758904099464417,
      "learning_rate": 4.991381533130859e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 3074224,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 0.47745808959007263,
      "learning_rate": 4.99128641008746e-05,
      "loss": 0.5503,
      "num_input_tokens_seen": 3077744,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 0.6882988214492798,
      "learning_rate": 4.9911907658953664e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 3080656,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.1916760802268982,
      "learning_rate": 4.991094600574584e-05,
      "loss": 0.5352,
      "num_input_tokens_seen": 3084624,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 0.9615076184272766,
      "learning_rate": 4.9909979141452324e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 3087760,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 0.4086228907108307,
      "learning_rate": 4.9909007066275346e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 3091440,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 0.6849853992462158,
      "learning_rate": 4.990802978041828e-05,
      "loss": 0.498,
      "num_input_tokens_seen": 3094192,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 0.5424833297729492,
      "learning_rate": 4.990704728408555e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 3097584,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 0.4518975019454956,
      "learning_rate": 4.990605957748268e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 3101232,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.581921398639679,
      "learning_rate": 4.990506666081628e-05,
      "loss": 0.346,
      "num_input_tokens_seen": 3103888,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 0.5667752623558044,
      "learning_rate": 4.990406853429408e-05,
      "loss": 0.3818,
      "num_input_tokens_seen": 3111408,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 0.29168465733528137,
      "learning_rate": 4.990306519812485e-05,
      "loss": 0.4217,
      "num_input_tokens_seen": 3114288,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 0.56925368309021,
      "learning_rate": 4.99020566525185e-05,
      "loss": 0.5789,
      "num_input_tokens_seen": 3117296,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 0.7955440282821655,
      "learning_rate": 4.9901042897685996e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 3119792,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 0.4511343836784363,
      "learning_rate": 4.99000239338394e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 3122544,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.5842865109443665,
      "learning_rate": 4.989899976119187e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 3125648,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 0.6200059652328491,
      "learning_rate": 4.989797037995766e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 3128400,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 0.6594765782356262,
      "learning_rate": 4.989693579035209e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 3131280,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 0.3955945372581482,
      "learning_rate": 4.989589599259159e-05,
      "loss": 0.5178,
      "num_input_tokens_seen": 3134288,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 0.6964133977890015,
      "learning_rate": 4.9894850986893685e-05,
      "loss": 0.5263,
      "num_input_tokens_seen": 3136784,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 0.5633686780929565,
      "learning_rate": 4.9893800773476965e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 3139664,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 0.399397611618042,
      "learning_rate": 4.9892745352561134e-05,
      "loss": 0.7237,
      "num_input_tokens_seen": 3142896,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 0.30845990777015686,
      "learning_rate": 4.9891684724366955e-05,
      "loss": 0.5492,
      "num_input_tokens_seen": 3146224,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 0.5849590301513672,
      "learning_rate": 4.989061888911631e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 3149136,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 0.5950363874435425,
      "learning_rate": 4.988954784703217e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 3151984,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 0.3964073359966278,
      "learning_rate": 4.9888471598338574e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 3155440,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 0.23665273189544678,
      "learning_rate": 4.988739014326066e-05,
      "loss": 0.3924,
      "num_input_tokens_seen": 3159216,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.4784731864929199,
      "learning_rate": 4.9886303482024656e-05,
      "loss": 0.5995,
      "num_input_tokens_seen": 3162640,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 0.32561954855918884,
      "learning_rate": 4.988521161485789e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 3165968,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 0.5064007639884949,
      "learning_rate": 4.9884114541988747e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 3168688,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 0.44577956199645996,
      "learning_rate": 4.988301226364674e-05,
      "loss": 0.5934,
      "num_input_tokens_seen": 3172400,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 0.47541874647140503,
      "learning_rate": 4.988190478006245e-05,
      "loss": 0.623,
      "num_input_tokens_seen": 3175472,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 0.5876806378364563,
      "learning_rate": 4.988079209146754e-05,
      "loss": 0.7344,
      "num_input_tokens_seen": 3178032,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 0.41741204261779785,
      "learning_rate": 4.9879674198094794e-05,
      "loss": 0.3894,
      "num_input_tokens_seen": 3181744,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 0.42559370398521423,
      "learning_rate": 4.987855110017804e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 3184656,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 0.3630039393901825,
      "learning_rate": 4.9877422797952224e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 3187504,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 0.4576730728149414,
      "learning_rate": 4.987628929165338e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 3190608,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 0.3491947650909424,
      "learning_rate": 4.9875150581518615e-05,
      "loss": 0.5397,
      "num_input_tokens_seen": 3193584,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 0.4079686403274536,
      "learning_rate": 4.987400666778614e-05,
      "loss": 0.6812,
      "num_input_tokens_seen": 3197488,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 0.4653405547142029,
      "learning_rate": 4.987285755069524e-05,
      "loss": 0.5982,
      "num_input_tokens_seen": 3200592,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 0.5794954299926758,
      "learning_rate": 4.987170323048632e-05,
      "loss": 0.4898,
      "num_input_tokens_seen": 3204208,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 0.6039353013038635,
      "learning_rate": 4.987054370740083e-05,
      "loss": 0.6028,
      "num_input_tokens_seen": 3208048,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 0.4130508303642273,
      "learning_rate": 4.986937898168134e-05,
      "loss": 0.4606,
      "num_input_tokens_seen": 3210576,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 0.32839295268058777,
      "learning_rate": 4.986820905357148e-05,
      "loss": 0.4154,
      "num_input_tokens_seen": 3213648,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 0.2830808460712433,
      "learning_rate": 4.9867033923315995e-05,
      "loss": 0.6952,
      "num_input_tokens_seen": 3216592,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 0.665322482585907,
      "learning_rate": 4.9865853591160724e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 3219760,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 0.616901159286499,
      "learning_rate": 4.9864668057352556e-05,
      "loss": 0.7058,
      "num_input_tokens_seen": 3223600,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 0.6219388246536255,
      "learning_rate": 4.9863477322139517e-05,
      "loss": 0.642,
      "num_input_tokens_seen": 3226416,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 0.32239410281181335,
      "learning_rate": 4.986228138577067e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 3230288,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 1.4084066152572632,
      "learning_rate": 4.986108024849619e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 3233104,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 0.33493727445602417,
      "learning_rate": 4.985987391056737e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 3237872,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 0.30834850668907166,
      "learning_rate": 4.9858662372236544e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 3242064,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 0.5538910031318665,
      "learning_rate": 4.985744563375714e-05,
      "loss": 0.4836,
      "num_input_tokens_seen": 3245712,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 0.5979676842689514,
      "learning_rate": 4.985622369538371e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 3248560,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 0.39819857478141785,
      "learning_rate": 4.985499655737185e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 3251440,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 0.5283166766166687,
      "learning_rate": 4.985376421997827e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 3254640,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 0.3493654131889343,
      "learning_rate": 4.985252668346076e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 3258320,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 0.416557639837265,
      "learning_rate": 4.9851283948078206e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 3261168,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 0.4626879394054413,
      "learning_rate": 4.985003601409057e-05,
      "loss": 0.574,
      "num_input_tokens_seen": 3264816,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 0.41623491048812866,
      "learning_rate": 4.9848782881758906e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 3268080,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 0.32221800088882446,
      "learning_rate": 4.9847524551345345e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 3272528,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 0.3979912996292114,
      "learning_rate": 4.984626102311313e-05,
      "loss": 0.5243,
      "num_input_tokens_seen": 3275888,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 0.9074351191520691,
      "learning_rate": 4.984499229732656e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 3281072,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.3066386580467224,
      "learning_rate": 4.984371837425106e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 3285808,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 0.3236061632633209,
      "learning_rate": 4.9842439254153106e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 3289168,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 0.5286319255828857,
      "learning_rate": 4.984115493730027e-05,
      "loss": 0.674,
      "num_input_tokens_seen": 3292336,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 0.44055142998695374,
      "learning_rate": 4.983986542396123e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 3295024,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 1.2955776453018188,
      "learning_rate": 4.983857071440573e-05,
      "loss": 0.695,
      "num_input_tokens_seen": 3297776,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 0.34860265254974365,
      "learning_rate": 4.9837270808904616e-05,
      "loss": 0.4113,
      "num_input_tokens_seen": 3300560,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 0.4148505628108978,
      "learning_rate": 4.9835965707729807e-05,
      "loss": 0.5167,
      "num_input_tokens_seen": 3304432,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 0.3943074345588684,
      "learning_rate": 4.983465541115431e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 3307344,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 0.4199698269367218,
      "learning_rate": 4.983333991945223e-05,
      "loss": 0.5148,
      "num_input_tokens_seen": 3311056,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 0.5405905842781067,
      "learning_rate": 4.983201923289875e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 3313712,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 0.40062591433525085,
      "learning_rate": 4.983069335177014e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 3317232,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 0.4365343153476715,
      "learning_rate": 4.982936227634378e-05,
      "loss": 0.6237,
      "num_input_tokens_seen": 3320048,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.35281866788864136,
      "learning_rate": 4.982802600689809e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 3324720,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 0.5330135226249695,
      "learning_rate": 4.982668454371261e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 3327696,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 0.352119505405426,
      "learning_rate": 4.9825337887067966e-05,
      "loss": 0.3861,
      "num_input_tokens_seen": 3330608,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 0.6225443482398987,
      "learning_rate": 4.982398603724586e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 3333296,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 0.6955031752586365,
      "learning_rate": 4.982262899452909e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 3337552,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 0.38987571001052856,
      "learning_rate": 4.982126675920151e-05,
      "loss": 0.4924,
      "num_input_tokens_seen": 3341296,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 0.6185567378997803,
      "learning_rate": 4.98198993315481e-05,
      "loss": 0.5551,
      "num_input_tokens_seen": 3343920,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 0.4023878574371338,
      "learning_rate": 4.981852671185492e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 3346736,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 0.42421218752861023,
      "learning_rate": 4.9817148900409093e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 3350320,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 0.3800064027309418,
      "learning_rate": 4.981576589749885e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 3353552,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 0.5042321085929871,
      "learning_rate": 4.981437770341348e-05,
      "loss": 0.5376,
      "num_input_tokens_seen": 3357136,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 0.7996291518211365,
      "learning_rate": 4.9812984318443414e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 3360048,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.44023340940475464,
      "learning_rate": 4.98115857428801e-05,
      "loss": 0.4359,
      "num_input_tokens_seen": 3363280,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 0.6463964581489563,
      "learning_rate": 4.9810181977016104e-05,
      "loss": 0.6629,
      "num_input_tokens_seen": 3366000,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 0.28588244318962097,
      "learning_rate": 4.9808773021145096e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 3369648,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 0.660980224609375,
      "learning_rate": 4.98073588755618e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 3372528,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 0.7064247131347656,
      "learning_rate": 4.980593954056205e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 3375728,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 0.6619511842727661,
      "learning_rate": 4.9804515016442754e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 3378864,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 0.5653471946716309,
      "learning_rate": 4.9803085303501896e-05,
      "loss": 0.4294,
      "num_input_tokens_seen": 3381936,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 0.464399129152298,
      "learning_rate": 4.9801650402038555e-05,
      "loss": 0.5338,
      "num_input_tokens_seen": 3384464,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 0.4525461792945862,
      "learning_rate": 4.980021031235291e-05,
      "loss": 0.5762,
      "num_input_tokens_seen": 3387312,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 0.5548064112663269,
      "learning_rate": 4.9798765034746206e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 3391376,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 0.5177909135818481,
      "learning_rate": 4.979731456952077e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 3394352,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 0.39349690079689026,
      "learning_rate": 4.9795858916980034e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 3397456,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.5166923403739929,
      "learning_rate": 4.9794398077428495e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 3403792,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 0.7733566164970398,
      "learning_rate": 4.979293205117175e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 3407184,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 0.4750390350818634,
      "learning_rate": 4.979146083851647e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 3410256,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 0.6115018725395203,
      "learning_rate": 4.978998443977042e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 3413168,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 0.5857140421867371,
      "learning_rate": 4.9788502855242436e-05,
      "loss": 0.6923,
      "num_input_tokens_seen": 3415728,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 1.4551762342453003,
      "learning_rate": 4.978701608524247e-05,
      "loss": 0.4328,
      "num_input_tokens_seen": 3419568,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 0.38358378410339355,
      "learning_rate": 4.978552413008152e-05,
      "loss": 0.3917,
      "num_input_tokens_seen": 3422736,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 0.4199564456939697,
      "learning_rate": 4.9784026990071687e-05,
      "loss": 0.4513,
      "num_input_tokens_seen": 3427472,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 0.43216875195503235,
      "learning_rate": 4.978252466552615e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 3430256,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 0.41332340240478516,
      "learning_rate": 4.9781017156759205e-05,
      "loss": 0.613,
      "num_input_tokens_seen": 3433328,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 0.6460624933242798,
      "learning_rate": 4.977950446408618e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 3437456,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 0.4485129117965698,
      "learning_rate": 4.9777986587823523e-05,
      "loss": 0.3887,
      "num_input_tokens_seen": 3440240,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.7136372923851013,
      "learning_rate": 4.977646352828875e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 3443824,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 0.39156174659729004,
      "learning_rate": 4.9774935285800476e-05,
      "loss": 0.5632,
      "num_input_tokens_seen": 3447184,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 0.6427199244499207,
      "learning_rate": 4.977340186067838e-05,
      "loss": 0.4515,
      "num_input_tokens_seen": 3450320,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 0.266813188791275,
      "learning_rate": 4.977186325324325e-05,
      "loss": 0.6172,
      "num_input_tokens_seen": 3454640,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.6843109130859375,
      "learning_rate": 4.977031946381694e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 3458000,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 0.506111204624176,
      "learning_rate": 4.9768770492722394e-05,
      "loss": 0.628,
      "num_input_tokens_seen": 3461296,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 0.5194320678710938,
      "learning_rate": 4.976721634028363e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 3464336,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 0.5841869711875916,
      "learning_rate": 4.976565700682578e-05,
      "loss": 0.6075,
      "num_input_tokens_seen": 3467184,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 0.48447296023368835,
      "learning_rate": 4.976409249267502e-05,
      "loss": 0.5567,
      "num_input_tokens_seen": 3470320,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 0.58154296875,
      "learning_rate": 4.976252279815863e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 3474128,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 0.5019590854644775,
      "learning_rate": 4.976094792360498e-05,
      "loss": 0.6543,
      "num_input_tokens_seen": 3476336,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 0.4297713339328766,
      "learning_rate": 4.975936786934351e-05,
      "loss": 0.5442,
      "num_input_tokens_seen": 3480016,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 1.4102427959442139,
      "learning_rate": 4.975778263570475e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 3482608,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 0.502859354019165,
      "learning_rate": 4.975619222302032e-05,
      "loss": 0.718,
      "num_input_tokens_seen": 3485200,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 0.501939594745636,
      "learning_rate": 4.975459663162291e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 3488336,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 0.6361467242240906,
      "learning_rate": 4.9752995861846296e-05,
      "loss": 0.5783,
      "num_input_tokens_seen": 3492080,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 0.33858221769332886,
      "learning_rate": 4.975138991402535e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 3495952,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 0.4528180658817291,
      "learning_rate": 4.974977878849601e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 3498224,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 0.854593813419342,
      "learning_rate": 4.9748162485595303e-05,
      "loss": 0.5882,
      "num_input_tokens_seen": 3501872,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 0.4193255305290222,
      "learning_rate": 4.974654100566135e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 3505008,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 0.6107773184776306,
      "learning_rate": 4.974491434903335e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 3508016,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 0.38363903760910034,
      "learning_rate": 4.9743282516051556e-05,
      "loss": 0.5406,
      "num_input_tokens_seen": 3512912,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 0.5100380182266235,
      "learning_rate": 4.974164550705736e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 3516112,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 0.46824634075164795,
      "learning_rate": 4.974000332239319e-05,
      "loss": 0.6063,
      "num_input_tokens_seen": 3518800,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.830548882484436,
      "learning_rate": 4.973835596240257e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 3522608,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 0.702268123626709,
      "learning_rate": 4.9736703427430114e-05,
      "loss": 0.5205,
      "num_input_tokens_seen": 3526512,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 0.48529523611068726,
      "learning_rate": 4.9735045717821514e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 3529456,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 0.7182712554931641,
      "learning_rate": 4.973338283392354e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 3532464,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 0.7334575057029724,
      "learning_rate": 4.9731714776084046e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 3535824,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 0.3039034307003021,
      "learning_rate": 4.973004154465198e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 3539696,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 0.5565024614334106,
      "learning_rate": 4.972836313997736e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 3543408,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 2.3725225925445557,
      "learning_rate": 4.9726679562411285e-05,
      "loss": 0.3942,
      "num_input_tokens_seen": 3546864,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 0.5935547947883606,
      "learning_rate": 4.972499081230595e-05,
      "loss": 0.4831,
      "num_input_tokens_seen": 3549904,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 0.34240293502807617,
      "learning_rate": 4.9723296890014605e-05,
      "loss": 0.4813,
      "num_input_tokens_seen": 3553072,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 0.33507785201072693,
      "learning_rate": 4.9721597795891605e-05,
      "loss": 0.3636,
      "num_input_tokens_seen": 3556784,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 0.19405494630336761,
      "learning_rate": 4.9719893530292396e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 3561232,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 0.32393765449523926,
      "learning_rate": 4.9718184093573475e-05,
      "loss": 0.5311,
      "num_input_tokens_seen": 3564048,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 0.5718541145324707,
      "learning_rate": 4.9716469486092435e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 3567664,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 0.37111154198646545,
      "learning_rate": 4.971474970820797e-05,
      "loss": 0.4952,
      "num_input_tokens_seen": 3571344,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 0.44069841504096985,
      "learning_rate": 4.971302476027982e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 3574128,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 0.34091606736183167,
      "learning_rate": 4.971129464266884e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 3577552,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 0.4480034112930298,
      "learning_rate": 4.970955935573693e-05,
      "loss": 0.6994,
      "num_input_tokens_seen": 3581040,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 0.5428034663200378,
      "learning_rate": 4.9707818899847106e-05,
      "loss": 0.4449,
      "num_input_tokens_seen": 3583760,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 0.6479305624961853,
      "learning_rate": 4.970607327536346e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 3587632,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 0.6333338022232056,
      "learning_rate": 4.970432248265114e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 3590960,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 0.36969828605651855,
      "learning_rate": 4.97025665220764e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 3594672,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 0.3020704984664917,
      "learning_rate": 4.970080539400656e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 3597872,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 0.34036439657211304,
      "learning_rate": 4.969903909881004e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 3600528,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 0.4688164293766022,
      "learning_rate": 4.9697267636856317e-05,
      "loss": 0.5399,
      "num_input_tokens_seen": 3603344,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 0.5410938262939453,
      "learning_rate": 4.969549100851597e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 3606448,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 0.4029536247253418,
      "learning_rate": 4.9693709214160645e-05,
      "loss": 0.6853,
      "num_input_tokens_seen": 3609776,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 0.5980963706970215,
      "learning_rate": 4.969192225416307e-05,
      "loss": 0.5292,
      "num_input_tokens_seen": 3612080,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 0.380415678024292,
      "learning_rate": 4.969013012889706e-05,
      "loss": 0.5477,
      "num_input_tokens_seen": 3615504,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 0.9850108623504639,
      "learning_rate": 4.9688332838737504e-05,
      "loss": 0.732,
      "num_input_tokens_seen": 3618096,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 0.6521546840667725,
      "learning_rate": 4.9686530384060386e-05,
      "loss": 0.4447,
      "num_input_tokens_seen": 3621296,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 0.5107407569885254,
      "learning_rate": 4.968472276524275e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 3624784,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 0.36985868215560913,
      "learning_rate": 4.9682909982662725e-05,
      "loss": 0.5772,
      "num_input_tokens_seen": 3628336,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.6318230032920837,
      "learning_rate": 4.968109203669953e-05,
      "loss": 0.6044,
      "num_input_tokens_seen": 3631280,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 0.27138131856918335,
      "learning_rate": 4.967926892773346e-05,
      "loss": 0.6991,
      "num_input_tokens_seen": 3634448,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 0.5329914093017578,
      "learning_rate": 4.967744065614588e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 3637072,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 0.3193679749965668,
      "learning_rate": 4.9675607222319255e-05,
      "loss": 0.655,
      "num_input_tokens_seen": 3640880,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 0.6746119260787964,
      "learning_rate": 4.967376862663711e-05,
      "loss": 0.5357,
      "num_input_tokens_seen": 3643312,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 0.4097360670566559,
      "learning_rate": 4.9671924869484064e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 3647152,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 0.6440555453300476,
      "learning_rate": 4.967007595124581e-05,
      "loss": 0.5236,
      "num_input_tokens_seen": 3649904,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 0.32978591322898865,
      "learning_rate": 4.966822187230911e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 3653680,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 0.4683102071285248,
      "learning_rate": 4.966636263306183e-05,
      "loss": 0.5949,
      "num_input_tokens_seen": 3657328,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 0.38960322737693787,
      "learning_rate": 4.9664498233892895e-05,
      "loss": 0.4969,
      "num_input_tokens_seen": 3661008,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 0.6707809567451477,
      "learning_rate": 4.966262867519232e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 3664144,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 0.7673314213752747,
      "learning_rate": 4.966075395735119e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 3666800,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 0.492931067943573,
      "learning_rate": 4.965887408076167e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 3669936,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 0.3174654543399811,
      "learning_rate": 4.965698904581703e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 3673584,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 0.47112613916397095,
      "learning_rate": 4.965509885291157e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 3676816,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 0.4665606617927551,
      "learning_rate": 4.9653203502440716e-05,
      "loss": 0.5949,
      "num_input_tokens_seen": 3679792,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 0.38153761625289917,
      "learning_rate": 4.965130299480095e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 3682704,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 0.7677950859069824,
      "learning_rate": 4.9649397330389834e-05,
      "loss": 0.4697,
      "num_input_tokens_seen": 3685968,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 0.38145044445991516,
      "learning_rate": 4.964748650960601e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 3688528,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 0.6192970871925354,
      "learning_rate": 4.96455705328492e-05,
      "loss": 0.5092,
      "num_input_tokens_seen": 3692464,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 0.4365138113498688,
      "learning_rate": 4.964364940052021e-05,
      "loss": 0.3932,
      "num_input_tokens_seen": 3695920,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 0.4697045385837555,
      "learning_rate": 4.964172311302091e-05,
      "loss": 0.6451,
      "num_input_tokens_seen": 3699600,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 0.6540926694869995,
      "learning_rate": 4.9639791670754265e-05,
      "loss": 0.5303,
      "num_input_tokens_seen": 3703024,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 0.2659749686717987,
      "learning_rate": 4.9637855074124305e-05,
      "loss": 0.3982,
      "num_input_tokens_seen": 3706544,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 0.4286426603794098,
      "learning_rate": 4.963591332353615e-05,
      "loss": 0.6268,
      "num_input_tokens_seen": 3709584,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 0.35017451643943787,
      "learning_rate": 4.9633966419396e-05,
      "loss": 0.5955,
      "num_input_tokens_seen": 3713616,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 0.2999727725982666,
      "learning_rate": 4.96320143621111e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 3716656,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 0.4484313726425171,
      "learning_rate": 4.963005715208981e-05,
      "loss": 0.3818,
      "num_input_tokens_seen": 3719504,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 0.6380078196525574,
      "learning_rate": 4.962809478974157e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 3722672,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 0.2722185552120209,
      "learning_rate": 4.962612727547686e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 3725104,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 0.3337029218673706,
      "learning_rate": 4.962415460970728e-05,
      "loss": 0.6409,
      "num_input_tokens_seen": 3728272,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 0.44573092460632324,
      "learning_rate": 4.962217679284548e-05,
      "loss": 0.4314,
      "num_input_tokens_seen": 3731344,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.4791009724140167,
      "eval_runtime": 9.8566,
      "eval_samples_per_second": 86.034,
      "eval_steps_per_second": 21.508,
      "num_input_tokens_seen": 3733736,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 0.4741452932357788,
      "learning_rate": 4.962019382530521e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 3734280,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 0.4709184765815735,
      "learning_rate": 4.9618205707501256e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 3736776,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 0.5045482516288757,
      "learning_rate": 4.961621243984954e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 3739496,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 0.5171760320663452,
      "learning_rate": 4.9614214022767014e-05,
      "loss": 0.4221,
      "num_input_tokens_seen": 3741992,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 0.5165025591850281,
      "learning_rate": 4.9612210456671735e-05,
      "loss": 0.5745,
      "num_input_tokens_seen": 3744488,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 0.4827800393104553,
      "learning_rate": 4.961020174198281e-05,
      "loss": 0.5515,
      "num_input_tokens_seen": 3747688,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 0.35671597719192505,
      "learning_rate": 4.9608187879120454e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 3753064,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.2873089015483856,
      "learning_rate": 4.9606168868505936e-05,
      "loss": 0.5481,
      "num_input_tokens_seen": 3756136,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 0.3309374153614044,
      "learning_rate": 4.9604144710561616e-05,
      "loss": 0.5582,
      "num_input_tokens_seen": 3758888,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 0.4627020061016083,
      "learning_rate": 4.960211540571092e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 3761832,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 0.41923287510871887,
      "learning_rate": 4.960008095437836e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 3765032,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 0.820064127445221,
      "learning_rate": 4.959804135698951e-05,
      "loss": 0.5952,
      "num_input_tokens_seen": 3767944,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 0.6665050387382507,
      "learning_rate": 4.959599661397104e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 3771112,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 0.5401120185852051,
      "learning_rate": 4.959394672575068e-05,
      "loss": 0.5925,
      "num_input_tokens_seen": 3773960,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 0.41047587990760803,
      "learning_rate": 4.959189169275725e-05,
      "loss": 0.5641,
      "num_input_tokens_seen": 3777192,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 0.4726620316505432,
      "learning_rate": 4.958983151542065e-05,
      "loss": 0.5676,
      "num_input_tokens_seen": 3780488,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 0.30630767345428467,
      "learning_rate": 4.958776619417182e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 3784328,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 0.5072776079177856,
      "learning_rate": 4.9585695729442814e-05,
      "loss": 0.5389,
      "num_input_tokens_seen": 3786696,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 0.35240861773490906,
      "learning_rate": 4.9583620121666754e-05,
      "loss": 0.5695,
      "num_input_tokens_seen": 3789768,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 0.46600401401519775,
      "learning_rate": 4.958153937127783e-05,
      "loss": 0.5247,
      "num_input_tokens_seen": 3792936,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 0.405105859041214,
      "learning_rate": 4.95794534787113e-05,
      "loss": 0.5552,
      "num_input_tokens_seen": 3796168,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.6426975727081299,
      "learning_rate": 4.957736244440353e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 3798696,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 0.6207857131958008,
      "learning_rate": 4.957526626879193e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 3801992,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 0.3822755217552185,
      "learning_rate": 4.957316495231499e-05,
      "loss": 0.3931,
      "num_input_tokens_seen": 3804648,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 0.42148062586784363,
      "learning_rate": 4.9571058495412294e-05,
      "loss": 0.6575,
      "num_input_tokens_seen": 3808296,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 0.39800605177879333,
      "learning_rate": 4.9568946898524485e-05,
      "loss": 0.5235,
      "num_input_tokens_seen": 3811688,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 0.3709734380245209,
      "learning_rate": 4.956683016209327e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 3817416,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 0.4298456609249115,
      "learning_rate": 4.956470828656147e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 3820168,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 0.44171738624572754,
      "learning_rate": 4.956258127237294e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 3823720,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 0.5664507150650024,
      "learning_rate": 4.956044911997263e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 3827432,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 0.3291866183280945,
      "learning_rate": 4.955831182980657e-05,
      "loss": 0.5783,
      "num_input_tokens_seen": 3829832,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 0.4790506064891815,
      "learning_rate": 4.955616940232185e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 3832040,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 0.455965131521225,
      "learning_rate": 4.9554021837966636e-05,
      "loss": 0.5072,
      "num_input_tokens_seen": 3834984,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 0.4699585735797882,
      "learning_rate": 4.955186913719019e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 3841800,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 0.5377297401428223,
      "learning_rate": 4.954971130044281e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 3844456,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 0.3543914258480072,
      "learning_rate": 4.954754832817592e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 3847912,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 0.3153572678565979,
      "learning_rate": 4.954538022084196e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 3851208,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 0.24187515676021576,
      "learning_rate": 4.9543206978894496e-05,
      "loss": 0.3694,
      "num_input_tokens_seen": 3854920,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 0.3618524372577667,
      "learning_rate": 4.954102860278813e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 3858408,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 1.301621437072754,
      "learning_rate": 4.9538845092978567e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 3861352,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 0.5705944299697876,
      "learning_rate": 4.953665644992256e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 3864680,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 0.5005591511726379,
      "learning_rate": 4.953446267407795e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 3867176,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 0.4101809859275818,
      "learning_rate": 4.953226376590365e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 3869864,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 0.3833608329296112,
      "learning_rate": 4.953005972585965e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 3872744,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 0.44288715720176697,
      "learning_rate": 4.952785055440701e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 3875880,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 0.40246424078941345,
      "learning_rate": 4.9525636252007864e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 3879048,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 0.38069525361061096,
      "learning_rate": 4.9523416819125425e-05,
      "loss": 0.341,
      "num_input_tokens_seen": 3882376,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 0.3485446274280548,
      "learning_rate": 4.952119225622396e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 3885576,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 0.7998993992805481,
      "learning_rate": 4.951896256376882e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 3888808,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 0.6293022036552429,
      "learning_rate": 4.9516727742226444e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 3891976,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 0.39535969495773315,
      "learning_rate": 4.9514487792064323e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 3895592,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 0.3519040048122406,
      "learning_rate": 4.951224271375104e-05,
      "loss": 0.6634,
      "num_input_tokens_seen": 3899336,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 0.5819531083106995,
      "learning_rate": 4.950999250775623e-05,
      "loss": 0.4393,
      "num_input_tokens_seen": 3902280,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 0.2657083570957184,
      "learning_rate": 4.950773717455061e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 3905672,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 0.29197001457214355,
      "learning_rate": 4.950547671460596e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 3908392,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.4970933794975281,
      "learning_rate": 4.950321112839517e-05,
      "loss": 0.5586,
      "num_input_tokens_seen": 3912200,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 0.6033417582511902,
      "learning_rate": 4.9500940416392156e-05,
      "loss": 0.5706,
      "num_input_tokens_seen": 3915304,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 0.6520524024963379,
      "learning_rate": 4.949866457907193e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 3918952,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 0.2870698869228363,
      "learning_rate": 4.9496383616910565e-05,
      "loss": 0.6402,
      "num_input_tokens_seen": 3921864,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 0.3701743483543396,
      "learning_rate": 4.949409753038522e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 3925256,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 0.40136241912841797,
      "learning_rate": 4.949180631997411e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 3927432,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 0.5520853400230408,
      "learning_rate": 4.948950998615654e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 3931368,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 0.46191784739494324,
      "learning_rate": 4.9487208529412876e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 3934504,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 0.47363513708114624,
      "learning_rate": 4.9484901950224546e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 3937384,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 0.6203423738479614,
      "learning_rate": 4.948259024907407e-05,
      "loss": 0.6298,
      "num_input_tokens_seen": 3940168,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 0.32179296016693115,
      "learning_rate": 4.948027342644502e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 3942504,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 0.3626229166984558,
      "learning_rate": 4.947795148282206e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 3945480,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 0.5950639247894287,
      "learning_rate": 4.947562441869091e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 3948520,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 0.3479786217212677,
      "learning_rate": 4.947329223453836e-05,
      "loss": 0.5565,
      "num_input_tokens_seen": 3952904,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 0.4889819025993347,
      "learning_rate": 4.947095493085229e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 3955816,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 1.5180469751358032,
      "learning_rate": 4.9468612508121616e-05,
      "loss": 0.5129,
      "num_input_tokens_seen": 3958120,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 0.5020920634269714,
      "learning_rate": 4.9466264966836365e-05,
      "loss": 0.5472,
      "num_input_tokens_seen": 3960520,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 0.5876959562301636,
      "learning_rate": 4.94639123074876e-05,
      "loss": 0.523,
      "num_input_tokens_seen": 3963208,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 0.6766026020050049,
      "learning_rate": 4.946155453056749e-05,
      "loss": 0.5599,
      "num_input_tokens_seen": 3966568,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 0.6919275522232056,
      "learning_rate": 4.945919163656923e-05,
      "loss": 0.5545,
      "num_input_tokens_seen": 3970184,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 0.4017353951931,
      "learning_rate": 4.945682362598715e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 3973192,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 0.6029285192489624,
      "learning_rate": 4.945445049931656e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 3975656,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 0.3009690046310425,
      "learning_rate": 4.945207225705393e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 3979240,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 0.2086363583803177,
      "learning_rate": 4.944968889969674e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 3983464,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.6278220415115356,
      "learning_rate": 4.944730042774358e-05,
      "loss": 0.5355,
      "num_input_tokens_seen": 3986568,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 0.36763694882392883,
      "learning_rate": 4.944490684169408e-05,
      "loss": 0.4602,
      "num_input_tokens_seen": 3989224,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 0.3566041886806488,
      "learning_rate": 4.9442508142048945e-05,
      "loss": 0.6476,
      "num_input_tokens_seen": 3992648,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 0.2785678505897522,
      "learning_rate": 4.9440104329309956e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 3995528,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 0.43265166878700256,
      "learning_rate": 4.943769540397998e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 3998536,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 0.5151452422142029,
      "learning_rate": 4.943528136656292e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 4001224,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 0.33368027210235596,
      "learning_rate": 4.943286221756377e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 4004328,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 0.5346469283103943,
      "learning_rate": 4.9430437957488586e-05,
      "loss": 0.4722,
      "num_input_tokens_seen": 4007944,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 0.5066050887107849,
      "learning_rate": 4.94280085868445e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 4011144,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 0.49615779519081116,
      "learning_rate": 4.942557410613972e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 4014376,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 0.2922351062297821,
      "learning_rate": 4.942313451588348e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 4018216,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 0.5483660697937012,
      "learning_rate": 4.942068981658614e-05,
      "loss": 0.5216,
      "num_input_tokens_seen": 4021032,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 0.357181191444397,
      "learning_rate": 4.9418240008759097e-05,
      "loss": 0.5953,
      "num_input_tokens_seen": 4024040,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 0.6175004243850708,
      "learning_rate": 4.9415785092914826e-05,
      "loss": 0.3564,
      "num_input_tokens_seen": 4027336,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 0.6356421709060669,
      "learning_rate": 4.941332506956686e-05,
      "loss": 0.5649,
      "num_input_tokens_seen": 4030472,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 0.27268242835998535,
      "learning_rate": 4.941085993922981e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 4034312,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 0.29530641436576843,
      "learning_rate": 4.9408389702419354e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 4038632,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 0.4179142117500305,
      "learning_rate": 4.940591435965224e-05,
      "loss": 0.5553,
      "num_input_tokens_seen": 4042824,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 0.8854504227638245,
      "learning_rate": 4.9403433911446285e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 4045896,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 0.45631298422813416,
      "learning_rate": 4.9400948358320356e-05,
      "loss": 0.5436,
      "num_input_tokens_seen": 4048712,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 0.3529402017593384,
      "learning_rate": 4.939845770079442e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 4051176,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 0.33712753653526306,
      "learning_rate": 4.9395961939389475e-05,
      "loss": 0.3717,
      "num_input_tokens_seen": 4054056,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 0.6748446822166443,
      "learning_rate": 4.939346107462762e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 4058120,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 0.5457088351249695,
      "learning_rate": 4.9390955107032e-05,
      "loss": 0.5595,
      "num_input_tokens_seen": 4060840,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 0.3530311584472656,
      "learning_rate": 4.938844403712685e-05,
      "loss": 0.553,
      "num_input_tokens_seen": 4064136,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 0.36722493171691895,
      "learning_rate": 4.938592786543743e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 4066824,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 0.4773465692996979,
      "learning_rate": 4.938340659249012e-05,
      "loss": 0.6594,
      "num_input_tokens_seen": 4069864,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 0.20149169862270355,
      "learning_rate": 4.938088021881233e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 4072936,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 0.4368317425251007,
      "learning_rate": 4.937834874493255e-05,
      "loss": 0.6455,
      "num_input_tokens_seen": 4075400,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 0.702182948589325,
      "learning_rate": 4.9375812171380334e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 4078696,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 0.6624465584754944,
      "learning_rate": 4.9373270498686306e-05,
      "loss": 0.501,
      "num_input_tokens_seen": 4082824,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 0.5801588892936707,
      "learning_rate": 4.937072372738216e-05,
      "loss": 0.6177,
      "num_input_tokens_seen": 4085448,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 0.5608566403388977,
      "learning_rate": 4.9368171858000635e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 4088424,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 0.4399203360080719,
      "learning_rate": 4.9365614891075576e-05,
      "loss": 0.5668,
      "num_input_tokens_seen": 4091752,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 0.685201108455658,
      "learning_rate": 4.936305282714185e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 4094568,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 0.5300083160400391,
      "learning_rate": 4.936048566673542e-05,
      "loss": 0.5166,
      "num_input_tokens_seen": 4097352,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 0.4456951320171356,
      "learning_rate": 4.935791341039332e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 4100328,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 0.671994686126709,
      "learning_rate": 4.935533605865361e-05,
      "loss": 0.4615,
      "num_input_tokens_seen": 4102728,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 0.3150067627429962,
      "learning_rate": 4.935275361205547e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 4105352,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 0.4131343364715576,
      "learning_rate": 4.93501660711391e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 4108744,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 0.43199601769447327,
      "learning_rate": 4.934757343644578e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 4112168,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 0.5666489601135254,
      "learning_rate": 4.934497570851788e-05,
      "loss": 0.675,
      "num_input_tokens_seen": 4115240,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 0.3995305597782135,
      "learning_rate": 4.9342372887898804e-05,
      "loss": 0.4863,
      "num_input_tokens_seen": 4117864,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 0.7342089414596558,
      "learning_rate": 4.9339764975133026e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 4121960,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 0.37681519985198975,
      "learning_rate": 4.93371519707661e-05,
      "loss": 0.5664,
      "num_input_tokens_seen": 4125160,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 0.5080811381340027,
      "learning_rate": 4.933453387534464e-05,
      "loss": 0.5563,
      "num_input_tokens_seen": 4128168,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 0.5774143934249878,
      "learning_rate": 4.933191068941631e-05,
      "loss": 0.5711,
      "num_input_tokens_seen": 4134632,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 0.29792678356170654,
      "learning_rate": 4.932928241352987e-05,
      "loss": 0.5389,
      "num_input_tokens_seen": 4138088,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.6333169341087341,
      "learning_rate": 4.9326649048235095e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 4141608,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 0.3578973412513733,
      "learning_rate": 4.932401059408287e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 4144712,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 0.6765801906585693,
      "learning_rate": 4.932136705162514e-05,
      "loss": 0.4233,
      "num_input_tokens_seen": 4147624,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 0.4479524493217468,
      "learning_rate": 4.93187184214149e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 4150728,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 0.3465251624584198,
      "learning_rate": 4.9316064704006205e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 4153928,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 0.3720436692237854,
      "learning_rate": 4.931340589995419e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 4157608,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 0.4416874349117279,
      "learning_rate": 4.9310742009815035e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 4160648,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 0.4201961159706116,
      "learning_rate": 4.930807303414601e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 4163720,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 0.5515235066413879,
      "learning_rate": 4.930539897350542e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 4166920,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 0.5462873578071594,
      "learning_rate": 4.930271982845266e-05,
      "loss": 0.3691,
      "num_input_tokens_seen": 4170024,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 0.4291559159755707,
      "learning_rate": 4.930003559954817e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 4173288,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 0.7506584525108337,
      "learning_rate": 4.929734628735346e-05,
      "loss": 0.3679,
      "num_input_tokens_seen": 4175688,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 0.7274985313415527,
      "learning_rate": 4.9294651892431115e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 4178504,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 0.6680442094802856,
      "learning_rate": 4.929195241534476e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 4180840,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 0.3915211856365204,
      "learning_rate": 4.9289247856659095e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 4184328,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 0.3028887212276459,
      "learning_rate": 4.9286538216939883e-05,
      "loss": 0.5846,
      "num_input_tokens_seen": 4188072,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 0.4104146659374237,
      "learning_rate": 4.9283823496753944e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 4191848,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 0.4927137792110443,
      "learning_rate": 4.928110369666918e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 4194440,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 0.4917311370372772,
      "learning_rate": 4.927837881725455e-05,
      "loss": 0.5523,
      "num_input_tokens_seen": 4197256,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 0.488787978887558,
      "learning_rate": 4.927564885908003e-05,
      "loss": 0.4784,
      "num_input_tokens_seen": 4200872,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 0.4232679605484009,
      "learning_rate": 4.9272913822716735e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 4204232,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 0.5566753149032593,
      "learning_rate": 4.927017370873678e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 4207304,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 0.5115963220596313,
      "learning_rate": 4.926742851771338e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 4209832,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 0.3642794191837311,
      "learning_rate": 4.926467825022079e-05,
      "loss": 0.4657,
      "num_input_tokens_seen": 4213128,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.6780531406402588,
      "learning_rate": 4.926192290683434e-05,
      "loss": 0.4012,
      "num_input_tokens_seen": 4215976,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 0.6992653012275696,
      "learning_rate": 4.9259162488130404e-05,
      "loss": 0.6574,
      "num_input_tokens_seen": 4218632,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 1.1872069835662842,
      "learning_rate": 4.925639699468645e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 4221416,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 0.5626680850982666,
      "learning_rate": 4.925362642708097e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 4224392,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 0.6455109119415283,
      "learning_rate": 4.9250850785893536e-05,
      "loss": 0.4539,
      "num_input_tokens_seen": 4227176,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 0.5297772884368896,
      "learning_rate": 4.924807007170479e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 4229960,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 0.6116682291030884,
      "learning_rate": 4.924528428509643e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 4232136,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 0.5471219420433044,
      "learning_rate": 4.9242493426651196e-05,
      "loss": 0.5606,
      "num_input_tokens_seen": 4235624,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 0.3099344074726105,
      "learning_rate": 4.92396974969529e-05,
      "loss": 0.5493,
      "num_input_tokens_seen": 4240616,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 0.47991520166397095,
      "learning_rate": 4.923689649658644e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 4243528,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 0.36250483989715576,
      "learning_rate": 4.923409042613774e-05,
      "loss": 0.5199,
      "num_input_tokens_seen": 4247208,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 0.6477083563804626,
      "learning_rate": 4.92312792861938e-05,
      "loss": 0.3673,
      "num_input_tokens_seen": 4252424,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 0.7618390321731567,
      "learning_rate": 4.922846307734268e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 4255752,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 0.3694562613964081,
      "learning_rate": 4.9225641800173496e-05,
      "loss": 0.5475,
      "num_input_tokens_seen": 4258472,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 0.4929804503917694,
      "learning_rate": 4.922281545527643e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 4260904,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 0.35359805822372437,
      "learning_rate": 4.9219984043242715e-05,
      "loss": 0.5961,
      "num_input_tokens_seen": 4264104,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 0.3257041275501251,
      "learning_rate": 4.921714756466466e-05,
      "loss": 0.5606,
      "num_input_tokens_seen": 4268648,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 0.6955986022949219,
      "learning_rate": 4.921430602013561e-05,
      "loss": 0.5447,
      "num_input_tokens_seen": 4272776,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 0.5110621452331543,
      "learning_rate": 4.9211459410249995e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 4275784,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 0.3989204168319702,
      "learning_rate": 4.92086077356033e-05,
      "loss": 0.6113,
      "num_input_tokens_seen": 4279400,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 0.42001453042030334,
      "learning_rate": 4.920575099679204e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 4282920,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 0.41333502531051636,
      "learning_rate": 4.920288919441383e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 4286856,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 0.2555703818798065,
      "learning_rate": 4.9200022329067316e-05,
      "loss": 0.5264,
      "num_input_tokens_seen": 4290216,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 0.8133170008659363,
      "learning_rate": 4.9197150401352224e-05,
      "loss": 0.6564,
      "num_input_tokens_seen": 4293768,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 0.28183791041374207,
      "learning_rate": 4.919427341186933e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 4297288,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 0.860659658908844,
      "learning_rate": 4.919139136122045e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 4299592,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 0.8487371206283569,
      "learning_rate": 4.918850425000849e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 4302600,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 1.1474285125732422,
      "learning_rate": 4.918561207883741e-05,
      "loss": 0.6093,
      "num_input_tokens_seen": 4305640,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 0.7963553071022034,
      "learning_rate": 4.9182714848312195e-05,
      "loss": 0.4799,
      "num_input_tokens_seen": 4308168,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 0.4614315629005432,
      "learning_rate": 4.917981255903893e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 4311208,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 0.7343990206718445,
      "learning_rate": 4.917690521162474e-05,
      "loss": 0.5832,
      "num_input_tokens_seen": 4314312,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 0.3196585774421692,
      "learning_rate": 4.9173992806677796e-05,
      "loss": 0.5433,
      "num_input_tokens_seen": 4316936,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 0.5958177447319031,
      "learning_rate": 4.917107534480735e-05,
      "loss": 0.4304,
      "num_input_tokens_seen": 4320232,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 0.5062048435211182,
      "learning_rate": 4.916815282662369e-05,
      "loss": 0.5913,
      "num_input_tokens_seen": 4324104,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 0.36006543040275574,
      "learning_rate": 4.91652252527382e-05,
      "loss": 0.6661,
      "num_input_tokens_seen": 4327304,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 0.5207965970039368,
      "learning_rate": 4.9162292623763275e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 4332008,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 0.35560745000839233,
      "learning_rate": 4.9159354940312386e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 4334728,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 0.42235201597213745,
      "learning_rate": 4.915641220300007e-05,
      "loss": 0.4985,
      "num_input_tokens_seen": 4337416,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 0.2985914647579193,
      "learning_rate": 4.9153464412441915e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 4340328,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 0.5211637616157532,
      "learning_rate": 4.915051156925456e-05,
      "loss": 0.4603,
      "num_input_tokens_seen": 4344456,
      "step": 6675
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 0.3971395790576935,
      "learning_rate": 4.91475536740557e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 4347944,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 0.23024111986160278,
      "learning_rate": 4.914459072746411e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 4350760,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 0.30331870913505554,
      "learning_rate": 4.91416227300996e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 4354760,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 0.447955846786499,
      "learning_rate": 4.913864968258303e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 4358120,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 0.28509721159935,
      "learning_rate": 4.913567158553634e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 4361352,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 0.682216227054596,
      "learning_rate": 4.9132688439582505e-05,
      "loss": 0.5683,
      "num_input_tokens_seen": 4363784,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 0.2457946538925171,
      "learning_rate": 4.9129700245345565e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 4367400,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 0.29047438502311707,
      "learning_rate": 4.912670700345063e-05,
      "loss": 0.5611,
      "num_input_tokens_seen": 4372232,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 0.4235396981239319,
      "learning_rate": 4.912370871452383e-05,
      "loss": 0.4602,
      "num_input_tokens_seen": 4375560,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 0.8294984698295593,
      "learning_rate": 4.91207053791924e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 4378216,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 0.7464446425437927,
      "learning_rate": 4.911769699808457e-05,
      "loss": 0.6006,
      "num_input_tokens_seen": 4380744,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 0.3519672155380249,
      "learning_rate": 4.911468357182969e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 4384264,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 0.41782069206237793,
      "learning_rate": 4.9111665101058116e-05,
      "loss": 0.5958,
      "num_input_tokens_seen": 4388200,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 0.5962792038917542,
      "learning_rate": 4.910864158640128e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 4390760,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 1.0770442485809326,
      "learning_rate": 4.910561302849168e-05,
      "loss": 0.5522,
      "num_input_tokens_seen": 4393416,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 0.3392800986766815,
      "learning_rate": 4.910257942796284e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 4396392,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 0.29058653116226196,
      "learning_rate": 4.909954078544936e-05,
      "loss": 0.4562,
      "num_input_tokens_seen": 4399496,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 0.2800462245941162,
      "learning_rate": 4.909649710158689e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 4402440,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 0.4642702639102936,
      "learning_rate": 4.909344837701214e-05,
      "loss": 0.5765,
      "num_input_tokens_seen": 4405672,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 0.7527528405189514,
      "learning_rate": 4.909039461236285e-05,
      "loss": 0.5816,
      "num_input_tokens_seen": 4408872,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 0.6317232251167297,
      "learning_rate": 4.908733580827786e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 4411112,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 0.3954005837440491,
      "learning_rate": 4.9084271965397014e-05,
      "loss": 0.4294,
      "num_input_tokens_seen": 4414248,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 0.4130578637123108,
      "learning_rate": 4.908120308436124e-05,
      "loss": 0.5825,
      "num_input_tokens_seen": 4417160,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 0.47032850980758667,
      "learning_rate": 4.907812916581251e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 4421032,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 0.44987449049949646,
      "learning_rate": 4.907505021039386e-05,
      "loss": 0.6506,
      "num_input_tokens_seen": 4423848,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 0.45727428793907166,
      "learning_rate": 4.907196621874937e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 4427144,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 0.3542638123035431,
      "learning_rate": 4.9068877191524176e-05,
      "loss": 0.5516,
      "num_input_tokens_seen": 4429896,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 0.4160272181034088,
      "learning_rate": 4.906578312936446e-05,
      "loss": 0.4862,
      "num_input_tokens_seen": 4435720,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 0.3260445296764374,
      "learning_rate": 4.906268403291747e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 4439048,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 0.34279540181159973,
      "learning_rate": 4.905957990283151e-05,
      "loss": 0.5191,
      "num_input_tokens_seen": 4441928,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 0.3121514618396759,
      "learning_rate": 4.905647073975592e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 4445160,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 1.1361746788024902,
      "learning_rate": 4.9053356544341095e-05,
      "loss": 0.5168,
      "num_input_tokens_seen": 4448328,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.4227352738380432,
      "learning_rate": 4.90502373172385e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 4451816,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 0.5739877223968506,
      "learning_rate": 4.9047113059100635e-05,
      "loss": 0.5167,
      "num_input_tokens_seen": 4454856,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 0.847475528717041,
      "learning_rate": 4.904398377058106e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 4457544,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 0.35222914814949036,
      "learning_rate": 4.90408494523344e-05,
      "loss": 0.5536,
      "num_input_tokens_seen": 4461160,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 0.531017541885376,
      "learning_rate": 4.90377101050163e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 4465000,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 0.5666961669921875,
      "learning_rate": 4.903456572928348e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 4467944,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 0.5117846131324768,
      "learning_rate": 4.903141632579371e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 4471880,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 0.8972674608230591,
      "learning_rate": 4.902826189520581e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 4474184,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 0.7709152102470398,
      "learning_rate": 4.9025102438179656e-05,
      "loss": 0.4715,
      "num_input_tokens_seen": 4476584,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 0.6736409068107605,
      "learning_rate": 4.902193795537616e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 4479592,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.7231691479682922,
      "learning_rate": 4.90187684474573e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 4482280,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 0.8582698702812195,
      "learning_rate": 4.901559391508611e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 4485224,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 0.6733477115631104,
      "learning_rate": 4.9012414358926644e-05,
      "loss": 0.5664,
      "num_input_tokens_seen": 4488680,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 0.4744929075241089,
      "learning_rate": 4.900922977964405e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 4492104,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 0.3435119688510895,
      "learning_rate": 4.90060401779045e-05,
      "loss": 0.4688,
      "num_input_tokens_seen": 4495240,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 0.49677878618240356,
      "learning_rate": 4.900284555437521e-05,
      "loss": 0.3863,
      "num_input_tokens_seen": 4498088,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 0.3852479159832001,
      "learning_rate": 4.899964590972448e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 4501224,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 0.584337592124939,
      "learning_rate": 4.899644124462163e-05,
      "loss": 0.5836,
      "num_input_tokens_seen": 4503784,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 0.5694564580917358,
      "learning_rate": 4.8993231559737026e-05,
      "loss": 0.4966,
      "num_input_tokens_seen": 4506568,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 0.43857526779174805,
      "learning_rate": 4.8990016855742114e-05,
      "loss": 0.3896,
      "num_input_tokens_seen": 4509096,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 0.46746209263801575,
      "learning_rate": 4.898679713330937e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 4512200,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 0.447221577167511,
      "learning_rate": 4.898357239311232e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 4514984,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 0.452128529548645,
      "learning_rate": 4.898034263582554e-05,
      "loss": 0.4576,
      "num_input_tokens_seen": 4518568,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 0.5253587961196899,
      "learning_rate": 4.897710786212467e-05,
      "loss": 0.8221,
      "num_input_tokens_seen": 4521832,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 0.8802849054336548,
      "learning_rate": 4.897386807268638e-05,
      "loss": 0.5719,
      "num_input_tokens_seen": 4525640,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 0.4891259968280792,
      "learning_rate": 4.897062326818839e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 4529608,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 0.8594638705253601,
      "learning_rate": 4.8967373449309495e-05,
      "loss": 0.5674,
      "num_input_tokens_seen": 4532328,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 0.34063902497291565,
      "learning_rate": 4.89641186167295e-05,
      "loss": 0.5142,
      "num_input_tokens_seen": 4535336,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 0.343233585357666,
      "learning_rate": 4.8960858771129294e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 4538760,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 0.3686409294605255,
      "learning_rate": 4.895759391319079e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 4546376,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 0.6792685985565186,
      "learning_rate": 4.8954324043596954e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 4550376,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 0.3387846052646637,
      "learning_rate": 4.8951049163031826e-05,
      "loss": 0.6169,
      "num_input_tokens_seen": 4553736,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 0.3817882537841797,
      "learning_rate": 4.894776927218045e-05,
      "loss": 0.4449,
      "num_input_tokens_seen": 4556936,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 0.4710918068885803,
      "learning_rate": 4.894448437172896e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 4559816,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 0.7604605555534363,
      "learning_rate": 4.894119446236452e-05,
      "loss": 0.4576,
      "num_input_tokens_seen": 4564520,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 0.40938663482666016,
      "learning_rate": 4.893789954477532e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 4567592,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 0.4273299276828766,
      "learning_rate": 4.893459961965063e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 4570408,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 0.7515050768852234,
      "learning_rate": 4.8931294687680775e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 4573320,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 0.2870927155017853,
      "learning_rate": 4.8927984749557086e-05,
      "loss": 0.5644,
      "num_input_tokens_seen": 4576680,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 0.8407173156738281,
      "learning_rate": 4.892466980597198e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 4579368,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 0.4081735908985138,
      "learning_rate": 4.892134985761889e-05,
      "loss": 0.6413,
      "num_input_tokens_seen": 4582440,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 0.3312624990940094,
      "learning_rate": 4.891802490519232e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 4585128,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 0.7330320477485657,
      "learning_rate": 4.891469494938781e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 4588168,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 0.6765037178993225,
      "learning_rate": 4.891135999090195e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 4591400,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 0.6599324345588684,
      "learning_rate": 4.890802003043238e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 4594280,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 0.418972373008728,
      "learning_rate": 4.890467506867777e-05,
      "loss": 0.3786,
      "num_input_tokens_seen": 4598536,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 0.4265475273132324,
      "learning_rate": 4.890132510633786e-05,
      "loss": 0.5189,
      "num_input_tokens_seen": 4602440,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 0.3864549398422241,
      "learning_rate": 4.8897970144113416e-05,
      "loss": 0.5166,
      "num_input_tokens_seen": 4605352,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 0.3503909409046173,
      "learning_rate": 4.889461018270626e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 4610152,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 0.25889673829078674,
      "learning_rate": 4.889124522281926e-05,
      "loss": 0.4012,
      "num_input_tokens_seen": 4612872,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 0.30205482244491577,
      "learning_rate": 4.888787526515632e-05,
      "loss": 0.5588,
      "num_input_tokens_seen": 4617064,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 0.46980807185173035,
      "learning_rate": 4.8884500310422406e-05,
      "loss": 0.5627,
      "num_input_tokens_seen": 4621608,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 1.041792631149292,
      "learning_rate": 4.8881120359323516e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 4623880,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 0.3281283974647522,
      "learning_rate": 4.887773541256669e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 4627080,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 0.4106604754924774,
      "learning_rate": 4.887434547086004e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 4630536,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 0.7292269468307495,
      "learning_rate": 4.887095053491269e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 4633256,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 0.4238058626651764,
      "learning_rate": 4.8867550605434806e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 4636104,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 0.4164349138736725,
      "learning_rate": 4.8864145683137644e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 4639272,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 0.6404253244400024,
      "learning_rate": 4.886073576873346e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 4641800,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 0.38867831230163574,
      "learning_rate": 4.885732086293556e-05,
      "loss": 0.615,
      "num_input_tokens_seen": 4645160,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 0.5596790313720703,
      "learning_rate": 4.8853900966458335e-05,
      "loss": 0.3739,
      "num_input_tokens_seen": 4648488,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 0.5217393636703491,
      "learning_rate": 4.8850476080017146e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 4651944,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 0.5876966714859009,
      "learning_rate": 4.884704620432848e-05,
      "loss": 0.5205,
      "num_input_tokens_seen": 4655816,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.5832224488258362,
      "learning_rate": 4.8843611340109794e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 4658824,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 0.39895087480545044,
      "learning_rate": 4.884017148807965e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 4662440,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 0.5325875282287598,
      "learning_rate": 4.883672664895761e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 4666280,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 0.3425893783569336,
      "learning_rate": 4.8833276823464296e-05,
      "loss": 0.413,
      "num_input_tokens_seen": 4669544,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 0.38036513328552246,
      "learning_rate": 4.882982201232138e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 4672392,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 0.4486013948917389,
      "learning_rate": 4.882636221625156e-05,
      "loss": 0.5551,
      "num_input_tokens_seen": 4676872,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 0.26903411746025085,
      "learning_rate": 4.8822897435978586e-05,
      "loss": 0.6248,
      "num_input_tokens_seen": 4680520,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 0.28613415360450745,
      "learning_rate": 4.8819427672227264e-05,
      "loss": 0.4718,
      "num_input_tokens_seen": 4683464,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 0.36757877469062805,
      "learning_rate": 4.881595292572341e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 4686728,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.6709343194961548,
      "learning_rate": 4.881247319719392e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 4690376,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 0.7602813243865967,
      "learning_rate": 4.88089884873667e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 4695304,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 0.8202492594718933,
      "learning_rate": 4.880549879697071e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 4698216,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 0.44268661737442017,
      "learning_rate": 4.8802004126735965e-05,
      "loss": 0.5985,
      "num_input_tokens_seen": 4701512,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 0.5387210249900818,
      "learning_rate": 4.87985044773935e-05,
      "loss": 0.5511,
      "num_input_tokens_seen": 4704744,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 0.45165377855300903,
      "learning_rate": 4.879499984967541e-05,
      "loss": 0.6307,
      "num_input_tokens_seen": 4707880,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 0.2508155107498169,
      "learning_rate": 4.8791490244314824e-05,
      "loss": 0.4281,
      "num_input_tokens_seen": 4711080,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 0.2705579996109009,
      "learning_rate": 4.8787975662045904e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 4714760,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 0.406120628118515,
      "learning_rate": 4.878445610360386e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 4718248,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 0.7005692720413208,
      "learning_rate": 4.878093156972494e-05,
      "loss": 0.8156,
      "num_input_tokens_seen": 4721480,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 0.3180386424064636,
      "learning_rate": 4.877740206114645e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 4724328,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 0.420900821685791,
      "learning_rate": 4.8773867578606716e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 4727432,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 0.3047718405723572,
      "learning_rate": 4.877032812284511e-05,
      "loss": 0.5607,
      "num_input_tokens_seen": 4730952,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 0.9578961730003357,
      "learning_rate": 4.876678369460205e-05,
      "loss": 0.5216,
      "num_input_tokens_seen": 4733704,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 0.39443734288215637,
      "learning_rate": 4.8763234294618985e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 4736424,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 0.33636489510536194,
      "learning_rate": 4.875967992363841e-05,
      "loss": 0.4602,
      "num_input_tokens_seen": 4739944,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 0.34710049629211426,
      "learning_rate": 4.875612058240386e-05,
      "loss": 0.4027,
      "num_input_tokens_seen": 4743656,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 0.5051432847976685,
      "learning_rate": 4.875255627165991e-05,
      "loss": 0.448,
      "num_input_tokens_seen": 4746440,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 0.5667838454246521,
      "learning_rate": 4.874898699215217e-05,
      "loss": 0.4854,
      "num_input_tokens_seen": 4749672,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 0.4470306634902954,
      "learning_rate": 4.8745412744627296e-05,
      "loss": 0.5834,
      "num_input_tokens_seen": 4753224,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 0.3728645145893097,
      "learning_rate": 4.874183352983297e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 4756072,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 0.33542245626449585,
      "learning_rate": 4.873824934851794e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 4759368,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 0.3205030858516693,
      "learning_rate": 4.873466020143196e-05,
      "loss": 0.6197,
      "num_input_tokens_seen": 4762280,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 0.38633307814598083,
      "learning_rate": 4.873106608932585e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 4765544,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 0.7236691117286682,
      "learning_rate": 4.872746701295145e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 4768424,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 0.4988119602203369,
      "learning_rate": 4.872386297306165e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 4771464,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 0.3208133578300476,
      "learning_rate": 4.872025397041037e-05,
      "loss": 0.431,
      "num_input_tokens_seen": 4775208,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 0.6583860516548157,
      "learning_rate": 4.871664000575258e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 4777768,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 0.31329166889190674,
      "learning_rate": 4.871302107984427e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 4781864,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 0.42635998129844666,
      "learning_rate": 4.8709397193442474e-05,
      "loss": 0.4211,
      "num_input_tokens_seen": 4785352,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 0.9349870085716248,
      "learning_rate": 4.870576834730529e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 4787848,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 0.6328272223472595,
      "learning_rate": 4.8702134542191816e-05,
      "loss": 0.3329,
      "num_input_tokens_seen": 4790952,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 0.1653795689344406,
      "learning_rate": 4.86984957788622e-05,
      "loss": 0.2949,
      "num_input_tokens_seen": 4794824,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 0.42406895756721497,
      "learning_rate": 4.869485205807763e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 4798120,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 0.43170592188835144,
      "learning_rate": 4.869120338060034e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 4801704,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 0.5882138013839722,
      "learning_rate": 4.8687549747193585e-05,
      "loss": 0.6504,
      "num_input_tokens_seen": 4804776,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.4409152865409851,
      "learning_rate": 4.868389115862166e-05,
      "loss": 0.6416,
      "num_input_tokens_seen": 4807496,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 0.4395294189453125,
      "learning_rate": 4.868022761564991e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 4810152,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 0.46140211820602417,
      "learning_rate": 4.8676559119044696e-05,
      "loss": 0.365,
      "num_input_tokens_seen": 4813736,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 0.7664576768875122,
      "learning_rate": 4.867288566957344e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 4816904,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 0.43679484724998474,
      "learning_rate": 4.866920726800457e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 4820168,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 0.3675648272037506,
      "learning_rate": 4.8665523915107576e-05,
      "loss": 0.6537,
      "num_input_tokens_seen": 4823176,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 0.4246963560581207,
      "learning_rate": 4.866183561165296e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 4826344,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 0.4399937689304352,
      "learning_rate": 4.86581423584123e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 4830120,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.32934457063674927,
      "learning_rate": 4.8654444156158154e-05,
      "loss": 0.4479,
      "num_input_tokens_seen": 4833576,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 0.4689925014972687,
      "learning_rate": 4.865074100566415e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 4836776,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 0.2943822145462036,
      "learning_rate": 4.864703290770495e-05,
      "loss": 0.6037,
      "num_input_tokens_seen": 4840808,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 0.39162999391555786,
      "learning_rate": 4.864331986305626e-05,
      "loss": 0.5646,
      "num_input_tokens_seen": 4847624,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 0.3435775935649872,
      "learning_rate": 4.8639601872494775e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 4850504,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 0.28158146142959595,
      "learning_rate": 4.8635878936798287e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 4853448,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 0.6093993782997131,
      "learning_rate": 4.863215105674557e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 4856040,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 0.2820405960083008,
      "learning_rate": 4.8628418233116457e-05,
      "loss": 0.5178,
      "num_input_tokens_seen": 4859976,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 0.4206104576587677,
      "learning_rate": 4.8624680466691824e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 4863304,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 0.6208459138870239,
      "learning_rate": 4.862093775825356e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 4866568,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 0.47593793272972107,
      "learning_rate": 4.86171901085846e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 4870312,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 0.3787764608860016,
      "learning_rate": 4.861343751846891e-05,
      "loss": 0.6634,
      "num_input_tokens_seen": 4873864,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 0.3125390410423279,
      "learning_rate": 4.860967998869148e-05,
      "loss": 0.5455,
      "num_input_tokens_seen": 4877032,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 0.35936474800109863,
      "learning_rate": 4.860591752003836e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 4880136,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 0.5334280729293823,
      "learning_rate": 4.8602150113296596e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 4883720,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 0.8102671504020691,
      "learning_rate": 4.8598377769254305e-05,
      "loss": 0.6023,
      "num_input_tokens_seen": 4886440,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 0.6662652492523193,
      "learning_rate": 4.85946004887006e-05,
      "loss": 0.5126,
      "num_input_tokens_seen": 4888936,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 1.6196588277816772,
      "learning_rate": 4.859081827242566e-05,
      "loss": 0.526,
      "num_input_tokens_seen": 4890952,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 0.6913980841636658,
      "learning_rate": 4.858703112122068e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 4893736,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 0.3093538284301758,
      "learning_rate": 4.8583239035877876e-05,
      "loss": 0.5724,
      "num_input_tokens_seen": 4897384,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 0.3792651295661926,
      "learning_rate": 4.8579442017190526e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 4900808,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 0.9705555438995361,
      "learning_rate": 4.8575640065952905e-05,
      "loss": 0.7966,
      "num_input_tokens_seen": 4903592,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 0.32043948769569397,
      "learning_rate": 4.857183318296035e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 4907048,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 0.6184468865394592,
      "learning_rate": 4.856802136900922e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 4909960,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 0.39918994903564453,
      "learning_rate": 4.856420462489689e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 4913768,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 0.6544907689094543,
      "learning_rate": 4.856038295142179e-05,
      "loss": 0.5294,
      "num_input_tokens_seen": 4917416,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 0.35089004039764404,
      "learning_rate": 4.855655634938337e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 4922056,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 0.5622029900550842,
      "learning_rate": 4.8552724819582106e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 4925320,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.740645706653595,
      "learning_rate": 4.8548888362819515e-05,
      "loss": 0.6677,
      "num_input_tokens_seen": 4928168,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 0.510738730430603,
      "learning_rate": 4.854504697989814e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 4931176,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 0.3711123466491699,
      "learning_rate": 4.8541200671621554e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 4935304,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 0.41917040944099426,
      "learning_rate": 4.8537349438794354e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 4938600,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 0.48820093274116516,
      "learning_rate": 4.853349328222219e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 4941512,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 0.2790347635746002,
      "learning_rate": 4.8529632202711706e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 4944744,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 0.9699452519416809,
      "learning_rate": 4.852576620107061e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 4948072,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 0.43492984771728516,
      "learning_rate": 4.8521895278107615e-05,
      "loss": 0.5435,
      "num_input_tokens_seen": 4951656,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 0.49646738171577454,
      "learning_rate": 4.8518019434632484e-05,
      "loss": 0.3117,
      "num_input_tokens_seen": 4955400,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 0.417649507522583,
      "learning_rate": 4.8514138671455995e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 4958568,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 0.4278995394706726,
      "learning_rate": 4.851025298938996e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 4961608,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 0.5813256502151489,
      "learning_rate": 4.850636238924722e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 4965288,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 0.43947964906692505,
      "learning_rate": 4.850246687184163e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 4968424,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 0.36767032742500305,
      "learning_rate": 4.849856643798812e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 4972200,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 0.38747212290763855,
      "learning_rate": 4.849466108850259e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 4975560,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.47087612748146057,
      "eval_runtime": 9.8196,
      "eval_samples_per_second": 86.358,
      "eval_steps_per_second": 21.589,
      "num_input_tokens_seen": 4976128,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 0.42847010493278503,
      "learning_rate": 4.8490750824202e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 4977984,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 0.654839038848877,
      "learning_rate": 4.8486835645904324e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 4980896,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 0.4588435888290405,
      "learning_rate": 4.84829155544286e-05,
      "loss": 0.408,
      "num_input_tokens_seen": 4984512,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 0.4068787395954132,
      "learning_rate": 4.847899055059484e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 4988320,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 0.3905598819255829,
      "learning_rate": 4.847506063522412e-05,
      "loss": 0.6539,
      "num_input_tokens_seen": 4991392,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 0.37014126777648926,
      "learning_rate": 4.847112580913854e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 4995680,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 0.4259880483150482,
      "learning_rate": 4.846718607316121e-05,
      "loss": 0.5861,
      "num_input_tokens_seen": 4999424,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 1.009528636932373,
      "learning_rate": 4.846324142811628e-05,
      "loss": 0.5222,
      "num_input_tokens_seen": 5002272,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 0.2964407801628113,
      "learning_rate": 4.8459291874828924e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 5005888,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 0.34014758467674255,
      "learning_rate": 4.845533741412534e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 5009600,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 0.3635585308074951,
      "learning_rate": 4.845137804683277e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 5013568,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 0.606882631778717,
      "learning_rate": 4.844741377377946e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 5016800,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 0.743543267250061,
      "learning_rate": 4.844344459579469e-05,
      "loss": 0.5327,
      "num_input_tokens_seen": 5020032,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 0.43832847476005554,
      "learning_rate": 4.8439470513708766e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 5023168,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 0.478484570980072,
      "learning_rate": 4.8435491528353026e-05,
      "loss": 0.3781,
      "num_input_tokens_seen": 5027040,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 0.7066949009895325,
      "learning_rate": 4.843150764055982e-05,
      "loss": 0.6637,
      "num_input_tokens_seen": 5030112,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 0.7881984114646912,
      "learning_rate": 4.842751885116254e-05,
      "loss": 0.3896,
      "num_input_tokens_seen": 5033600,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 0.6595377922058105,
      "learning_rate": 4.842352516099559e-05,
      "loss": 0.3115,
      "num_input_tokens_seen": 5036416,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 0.45171919465065,
      "learning_rate": 4.84195265708944e-05,
      "loss": 0.5565,
      "num_input_tokens_seen": 5039552,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 0.4283050298690796,
      "learning_rate": 4.841552308169545e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 5044032,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 0.3421851098537445,
      "learning_rate": 4.841151469423619e-05,
      "loss": 0.3774,
      "num_input_tokens_seen": 5047232,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 0.3834134340286255,
      "learning_rate": 4.840750140935516e-05,
      "loss": 0.5227,
      "num_input_tokens_seen": 5050624,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 1.141677737236023,
      "learning_rate": 4.840348322789189e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 5055008,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 0.5627992153167725,
      "learning_rate": 4.839946015068692e-05,
      "loss": 0.45,
      "num_input_tokens_seen": 5058240,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 0.5716073513031006,
      "learning_rate": 4.839543217858184e-05,
      "loss": 0.3509,
      "num_input_tokens_seen": 5060864,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 0.3683169484138489,
      "learning_rate": 4.839139931241926e-05,
      "loss": 0.5606,
      "num_input_tokens_seen": 5064384,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 0.26173147559165955,
      "learning_rate": 4.838736155304281e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 5068096,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 0.6228063702583313,
      "learning_rate": 4.8383318901297135e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 5070944,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 0.49075281620025635,
      "learning_rate": 4.837927135802792e-05,
      "loss": 0.4924,
      "num_input_tokens_seen": 5074624,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 0.6616680026054382,
      "learning_rate": 4.8375218924081856e-05,
      "loss": 0.597,
      "num_input_tokens_seen": 5077216,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 0.5778094530105591,
      "learning_rate": 4.837116160030668e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 5079872,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 0.9278212785720825,
      "learning_rate": 4.8367099387551126e-05,
      "loss": 0.5756,
      "num_input_tokens_seen": 5082592,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 0.31106069684028625,
      "learning_rate": 4.8363032286664964e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 5086144,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 0.6468587517738342,
      "learning_rate": 4.835896029849899e-05,
      "loss": 0.3708,
      "num_input_tokens_seen": 5089696,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 0.3552496135234833,
      "learning_rate": 4.835488342390502e-05,
      "loss": 0.4447,
      "num_input_tokens_seen": 5092672,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 0.3763597905635834,
      "learning_rate": 4.835080166373587e-05,
      "loss": 0.4895,
      "num_input_tokens_seen": 5094848,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 0.5902431011199951,
      "learning_rate": 4.834671501884542e-05,
      "loss": 0.3752,
      "num_input_tokens_seen": 5097408,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 0.5934367775917053,
      "learning_rate": 4.8342623490088543e-05,
      "loss": 0.5112,
      "num_input_tokens_seen": 5099872,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 0.48981085419654846,
      "learning_rate": 4.833852707832114e-05,
      "loss": 0.5547,
      "num_input_tokens_seen": 5103392,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 0.6984291672706604,
      "learning_rate": 4.833442578440013e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 5105696,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 1.2885675430297852,
      "learning_rate": 4.833031960918346e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 5109312,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 1.5653249025344849,
      "learning_rate": 4.8326208553530095e-05,
      "loss": 0.6225,
      "num_input_tokens_seen": 5111840,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 0.5252601504325867,
      "learning_rate": 4.832209261830002e-05,
      "loss": 0.5869,
      "num_input_tokens_seen": 5114752,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 0.6813803911209106,
      "learning_rate": 4.831797180435423e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 5117760,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 0.2718270719051361,
      "learning_rate": 4.831384611255478e-05,
      "loss": 0.5279,
      "num_input_tokens_seen": 5122176,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 0.6442785859107971,
      "learning_rate": 4.8309715543764697e-05,
      "loss": 0.8264,
      "num_input_tokens_seen": 5124256,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 0.8510882258415222,
      "learning_rate": 4.8305580098848046e-05,
      "loss": 0.5717,
      "num_input_tokens_seen": 5126912,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 0.41437315940856934,
      "learning_rate": 4.830143977866993e-05,
      "loss": 0.5491,
      "num_input_tokens_seen": 5130144,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 0.38974031805992126,
      "learning_rate": 4.829729458409645e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 5133440,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 1.2565882205963135,
      "learning_rate": 4.8293144515994725e-05,
      "loss": 0.5648,
      "num_input_tokens_seen": 5140064,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 0.4097154438495636,
      "learning_rate": 4.828898957523292e-05,
      "loss": 0.529,
      "num_input_tokens_seen": 5142720,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 0.46331122517585754,
      "learning_rate": 4.8284829762680183e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 5145632,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 0.35624808073043823,
      "learning_rate": 4.828066507920671e-05,
      "loss": 0.4328,
      "num_input_tokens_seen": 5148224,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 0.45438921451568604,
      "learning_rate": 4.82764955256837e-05,
      "loss": 0.5626,
      "num_input_tokens_seen": 5152096,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 0.24951769411563873,
      "learning_rate": 4.827232110298338e-05,
      "loss": 0.5947,
      "num_input_tokens_seen": 5155968,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 0.4313060939311981,
      "learning_rate": 4.8268141811978995e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 5159168,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 0.36656734347343445,
      "learning_rate": 4.826395765354479e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 5162816,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.4675220549106598,
      "learning_rate": 4.825976862855606e-05,
      "loss": 0.5557,
      "num_input_tokens_seen": 5167232,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 0.7302473187446594,
      "learning_rate": 4.8255574737889094e-05,
      "loss": 0.5169,
      "num_input_tokens_seen": 5170560,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 0.615595281124115,
      "learning_rate": 4.82513759824212e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 5173184,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 0.4029836058616638,
      "learning_rate": 4.824717236303072e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 5176288,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 0.4748290181159973,
      "learning_rate": 4.8242963880597005e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 5181216,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 0.5485652685165405,
      "learning_rate": 4.823875053600041e-05,
      "loss": 0.3792,
      "num_input_tokens_seen": 5184256,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.2692917287349701,
      "learning_rate": 4.823453233012233e-05,
      "loss": 0.3625,
      "num_input_tokens_seen": 5187616,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 0.6809850335121155,
      "learning_rate": 4.8230309263845155e-05,
      "loss": 0.6338,
      "num_input_tokens_seen": 5189952,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 1.0613303184509277,
      "learning_rate": 4.822608133805231e-05,
      "loss": 0.5406,
      "num_input_tokens_seen": 5192608,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 0.5928499698638916,
      "learning_rate": 4.8221848553628226e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 5196256,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 0.3504805266857147,
      "learning_rate": 4.8217610911458346e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 5199840,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 0.47476205229759216,
      "learning_rate": 4.821336841242915e-05,
      "loss": 0.3414,
      "num_input_tokens_seen": 5202688,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 0.45918792486190796,
      "learning_rate": 4.820912105742811e-05,
      "loss": 0.6596,
      "num_input_tokens_seen": 5205248,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 0.6385456919670105,
      "learning_rate": 4.8204868847343736e-05,
      "loss": 0.3952,
      "num_input_tokens_seen": 5207648,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 0.37323978543281555,
      "learning_rate": 4.820061178306553e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 5210272,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 0.30858761072158813,
      "learning_rate": 4.819634986548403e-05,
      "loss": 0.5988,
      "num_input_tokens_seen": 5214848,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 0.37212300300598145,
      "learning_rate": 4.819208309549077e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 5217888,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 0.44064873456954956,
      "learning_rate": 4.8187811473978316e-05,
      "loss": 0.5543,
      "num_input_tokens_seen": 5221952,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 0.8404810428619385,
      "learning_rate": 4.818353500184025e-05,
      "loss": 0.5701,
      "num_input_tokens_seen": 5225120,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 0.8380261063575745,
      "learning_rate": 4.817925367997115e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 5228448,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 0.7331371307373047,
      "learning_rate": 4.8174967509266615e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 5230976,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 0.37707170844078064,
      "learning_rate": 4.817067649062328e-05,
      "loss": 0.5454,
      "num_input_tokens_seen": 5234112,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 0.426516592502594,
      "learning_rate": 4.816638062493876e-05,
      "loss": 0.6243,
      "num_input_tokens_seen": 5238592,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 0.3360387980937958,
      "learning_rate": 4.8162079913111716e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 5241568,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 0.52022385597229,
      "learning_rate": 4.81577743560418e-05,
      "loss": 0.3784,
      "num_input_tokens_seen": 5244704,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 0.2844143807888031,
      "learning_rate": 4.8153463954629684e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 5248000,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 0.36206209659576416,
      "learning_rate": 4.814914870977706e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 5252096,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 0.3850705325603485,
      "learning_rate": 4.8144828622386627e-05,
      "loss": 0.5581,
      "num_input_tokens_seen": 5255360,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 0.2657557427883148,
      "learning_rate": 4.8140503693362095e-05,
      "loss": 0.6186,
      "num_input_tokens_seen": 5259712,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 0.8173118829727173,
      "learning_rate": 4.813617392360818e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 5262912,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 0.5815361738204956,
      "learning_rate": 4.813183931403065e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 5265664,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 0.429995596408844,
      "learning_rate": 4.8127499865536226e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 5268928,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 0.35438668727874756,
      "learning_rate": 4.812315557903269e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 5272032,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 0.46971166133880615,
      "learning_rate": 4.8118806455428806e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 5275520,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 0.5163415670394897,
      "learning_rate": 4.811445249563437e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 5278528,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 0.832115650177002,
      "learning_rate": 4.811009370056017e-05,
      "loss": 0.6689,
      "num_input_tokens_seen": 5281536,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 0.608717143535614,
      "learning_rate": 4.8105730071118016e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 5287168,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 0.3673902153968811,
      "learning_rate": 4.810136160822074e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 5291008,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 0.5111618041992188,
      "learning_rate": 4.8096988312782174e-05,
      "loss": 0.3742,
      "num_input_tokens_seen": 5294144,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 0.4537774622440338,
      "learning_rate": 4.809261018571715e-05,
      "loss": 0.6115,
      "num_input_tokens_seen": 5297184,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 0.41457900404930115,
      "learning_rate": 4.808822722794153e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 5300480,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 0.3229921758174896,
      "learning_rate": 4.8083839440372194e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 5303936,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 0.34862518310546875,
      "learning_rate": 4.8079446823926984e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 5306816,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 0.5645549893379211,
      "learning_rate": 4.8075049379524805e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 5309792,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 0.31172600388526917,
      "learning_rate": 4.807064710808555e-05,
      "loss": 0.6004,
      "num_input_tokens_seen": 5313056,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 0.43635761737823486,
      "learning_rate": 4.806624001053013e-05,
      "loss": 0.3732,
      "num_input_tokens_seen": 5315488,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 0.5274855494499207,
      "learning_rate": 4.8061828087780455e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 5318272,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 0.47700515389442444,
      "learning_rate": 4.805741134075944e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 5321408,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 0.33387649059295654,
      "learning_rate": 4.8052989770391026e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 5325696,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 0.7815818786621094,
      "learning_rate": 4.804856337760016e-05,
      "loss": 0.5662,
      "num_input_tokens_seen": 5328832,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 0.4609498381614685,
      "learning_rate": 4.804413216331278e-05,
      "loss": 0.5176,
      "num_input_tokens_seen": 5332448,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 0.3649585247039795,
      "learning_rate": 4.8039696128455856e-05,
      "loss": 0.5316,
      "num_input_tokens_seen": 5335520,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 0.5527419447898865,
      "learning_rate": 4.803525527395736e-05,
      "loss": 0.3598,
      "num_input_tokens_seen": 5339328,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 0.5123530030250549,
      "learning_rate": 4.803080960074625e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 5342368,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 0.3554716110229492,
      "learning_rate": 4.8026359109752525e-05,
      "loss": 0.6094,
      "num_input_tokens_seen": 5345632,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 0.5673569440841675,
      "learning_rate": 4.8021903801907175e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 5348608,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 0.6068610548973083,
      "learning_rate": 4.801744367814219e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 5351776,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 0.4238012731075287,
      "learning_rate": 4.801297873939059e-05,
      "loss": 0.6222,
      "num_input_tokens_seen": 5354880,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 0.532181441783905,
      "learning_rate": 4.8008508986586384e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 5357824,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 0.5227685570716858,
      "learning_rate": 4.800403442066459e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 5360512,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 0.39742642641067505,
      "learning_rate": 4.799955504256125e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 5364064,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 0.4141683876514435,
      "learning_rate": 4.7995070853213376e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 5367200,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 0.3957103490829468,
      "learning_rate": 4.799058185355903e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 5370272,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 0.5976115465164185,
      "learning_rate": 4.7986088044537235e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 5372896,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 0.5779735445976257,
      "learning_rate": 4.798158942708807e-05,
      "loss": 0.5637,
      "num_input_tokens_seen": 5376800,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 0.45312148332595825,
      "learning_rate": 4.797708600215258e-05,
      "loss": 0.638,
      "num_input_tokens_seen": 5380288,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 0.34937363862991333,
      "learning_rate": 4.7972577770672845e-05,
      "loss": 0.4518,
      "num_input_tokens_seen": 5383072,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 0.5519121289253235,
      "learning_rate": 4.796806473359192e-05,
      "loss": 0.6482,
      "num_input_tokens_seen": 5386400,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 0.4860893189907074,
      "learning_rate": 4.796354689185389e-05,
      "loss": 0.608,
      "num_input_tokens_seen": 5388832,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 0.6236892938613892,
      "learning_rate": 4.795902424640383e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 5391968,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 0.5509361624717712,
      "learning_rate": 4.795449679818782e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 5398624,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 0.2740337550640106,
      "learning_rate": 4.794996454815297e-05,
      "loss": 0.4986,
      "num_input_tokens_seen": 5401824,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.47677507996559143,
      "learning_rate": 4.7945427497247355e-05,
      "loss": 0.4079,
      "num_input_tokens_seen": 5404928,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 0.28629276156425476,
      "learning_rate": 4.7940885646420084e-05,
      "loss": 0.4529,
      "num_input_tokens_seen": 5408320,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 0.40041860938072205,
      "learning_rate": 4.7936338996621264e-05,
      "loss": 0.534,
      "num_input_tokens_seen": 5410752,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 0.45395052433013916,
      "learning_rate": 4.7931787548801996e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 5413312,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 0.32447534799575806,
      "learning_rate": 4.7927231303914394e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 5416992,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 0.32679545879364014,
      "learning_rate": 4.7922670262911565e-05,
      "loss": 0.3637,
      "num_input_tokens_seen": 5420672,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 1.6026368141174316,
      "learning_rate": 4.791810442674764e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 5423360,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 0.40580904483795166,
      "learning_rate": 4.791353379637773e-05,
      "loss": 0.4565,
      "num_input_tokens_seen": 5426336,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 0.7394614219665527,
      "learning_rate": 4.790895837275796e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 5429632,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 0.4049435555934906,
      "learning_rate": 4.790437815684545e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 5433312,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 0.29393038153648376,
      "learning_rate": 4.789979314959835e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 5437248,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 0.5349745154380798,
      "learning_rate": 4.789520335197577e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 5440384,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 0.4771862030029297,
      "learning_rate": 4.789060876493785e-05,
      "loss": 0.5614,
      "num_input_tokens_seen": 5444000,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 0.7090210914611816,
      "learning_rate": 4.788600938944573e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 5447936,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 0.5521873831748962,
      "learning_rate": 4.788140522646155e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 5451488,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 0.6279760003089905,
      "learning_rate": 4.787679627694843e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 5455520,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 0.5728132724761963,
      "learning_rate": 4.787218254187053e-05,
      "loss": 0.6237,
      "num_input_tokens_seen": 5458560,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 0.5266605019569397,
      "learning_rate": 4.7867564022192965e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 5461600,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 0.4201362133026123,
      "learning_rate": 4.786294071888191e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 5465536,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 0.5732332468032837,
      "learning_rate": 4.785831263290449e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 5468416,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 0.861011266708374,
      "learning_rate": 4.7853679765228845e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 5472000,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 0.495303750038147,
      "learning_rate": 4.7849042116824116e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 5475168,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 0.6582961082458496,
      "learning_rate": 4.7844399688660456e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 5477760,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 0.34505030512809753,
      "learning_rate": 4.783975248170901e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 5480544,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 0.417127400636673,
      "learning_rate": 4.783510049694191e-05,
      "loss": 0.3763,
      "num_input_tokens_seen": 5483776,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 0.5500746369361877,
      "learning_rate": 4.78304437353323e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 5485888,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 0.5394411087036133,
      "learning_rate": 4.782578219785433e-05,
      "loss": 0.6174,
      "num_input_tokens_seen": 5489056,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 0.35469478368759155,
      "learning_rate": 4.782111588548312e-05,
      "loss": 0.5595,
      "num_input_tokens_seen": 5492224,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 0.4165830612182617,
      "learning_rate": 4.781644479919484e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 5495872,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 0.22462545335292816,
      "learning_rate": 4.7811768939966605e-05,
      "loss": 0.3651,
      "num_input_tokens_seen": 5499488,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 0.31791189312934875,
      "learning_rate": 4.780708830877656e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 5503104,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 0.4028906524181366,
      "learning_rate": 4.780240290660384e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 5506912,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 0.5375145077705383,
      "learning_rate": 4.7797712734428575e-05,
      "loss": 0.5599,
      "num_input_tokens_seen": 5509312,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 0.7323121428489685,
      "learning_rate": 4.779301779323189e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 5512096,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 0.7015223503112793,
      "learning_rate": 4.778831808399593e-05,
      "loss": 0.5243,
      "num_input_tokens_seen": 5515168,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 0.3845093250274658,
      "learning_rate": 4.77836136077038e-05,
      "loss": 0.5126,
      "num_input_tokens_seen": 5518464,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 0.34326568245887756,
      "learning_rate": 4.777890436533963e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 5521344,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 0.40681174397468567,
      "learning_rate": 4.777419035788855e-05,
      "loss": 0.501,
      "num_input_tokens_seen": 5524416,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 0.5309159755706787,
      "learning_rate": 4.776947158633667e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 5527488,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 0.6173527836799622,
      "learning_rate": 4.7764748051671095e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 5530144,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.31061652302742004,
      "learning_rate": 4.7760019754879946e-05,
      "loss": 0.6109,
      "num_input_tokens_seen": 5533536,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 0.35725685954093933,
      "learning_rate": 4.7755286696952324e-05,
      "loss": 0.5234,
      "num_input_tokens_seen": 5536032,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 0.2891106903553009,
      "learning_rate": 4.775054887887833e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 5539552,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 0.46973443031311035,
      "learning_rate": 4.774580630164906e-05,
      "loss": 0.5041,
      "num_input_tokens_seen": 5542528,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 0.4212796986103058,
      "learning_rate": 4.774105896625661e-05,
      "loss": 0.595,
      "num_input_tokens_seen": 5545408,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 0.6462487578392029,
      "learning_rate": 4.773630687369407e-05,
      "loss": 0.3784,
      "num_input_tokens_seen": 5549120,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 0.5943660140037537,
      "learning_rate": 4.773155002495552e-05,
      "loss": 0.5536,
      "num_input_tokens_seen": 5552352,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 0.31222954392433167,
      "learning_rate": 4.772678842103605e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 5555296,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 0.6828966736793518,
      "learning_rate": 4.772202206293172e-05,
      "loss": 0.3544,
      "num_input_tokens_seen": 5558720,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 0.4830230176448822,
      "learning_rate": 4.7717250951639586e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 5562176,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 0.4208396375179291,
      "learning_rate": 4.771247508815774e-05,
      "loss": 0.6302,
      "num_input_tokens_seen": 5565408,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 0.4939860999584198,
      "learning_rate": 4.770769447348522e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 5568288,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 0.3066740036010742,
      "learning_rate": 4.7702909108622074e-05,
      "loss": 0.7163,
      "num_input_tokens_seen": 5571744,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 0.3133501708507538,
      "learning_rate": 4.7698118994569355e-05,
      "loss": 0.541,
      "num_input_tokens_seen": 5575008,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 0.418265700340271,
      "learning_rate": 4.769332413232909e-05,
      "loss": 0.6337,
      "num_input_tokens_seen": 5577824,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 0.468367338180542,
      "learning_rate": 4.768852452290432e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 5581216,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 0.4744330644607544,
      "learning_rate": 4.768372016729907e-05,
      "loss": 0.4815,
      "num_input_tokens_seen": 5586272,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 0.4506768584251404,
      "learning_rate": 4.767891106651834e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 5588608,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 0.7528835535049438,
      "learning_rate": 4.767409722156816e-05,
      "loss": 0.3376,
      "num_input_tokens_seen": 5591296,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 0.4587674140930176,
      "learning_rate": 4.7669278633455515e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 5593568,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 0.7561524510383606,
      "learning_rate": 4.7664455303188404e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 5596480,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 0.4235033690929413,
      "learning_rate": 4.7659627231775815e-05,
      "loss": 0.5838,
      "num_input_tokens_seen": 5599968,
      "step": 8585
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 0.4385267496109009,
      "learning_rate": 4.765479442022773e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 5603168,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 0.28776970505714417,
      "learning_rate": 4.76499568695551e-05,
      "loss": 0.6279,
      "num_input_tokens_seen": 5606464,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 0.3836793601512909,
      "learning_rate": 4.764511458076991e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 5609216,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 0.3562920093536377,
      "learning_rate": 4.7640267554885085e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 5612800,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 0.6591190099716187,
      "learning_rate": 4.763541579291459e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 5615488,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 0.22294701635837555,
      "learning_rate": 4.763055929587336e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 5619264,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 0.5805323719978333,
      "learning_rate": 4.7625698064777294e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 5623104,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 0.47559496760368347,
      "learning_rate": 4.762083210064332e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 5626144,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 0.38227379322052,
      "learning_rate": 4.761596140448934e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 5629216,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 0.6650921702384949,
      "learning_rate": 4.7611085977334256e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 5632448,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.4131755530834198,
      "learning_rate": 4.7606205820197946e-05,
      "loss": 0.535,
      "num_input_tokens_seen": 5636704,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 0.26299798488616943,
      "learning_rate": 4.760132093410129e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 5641408,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 0.49199432134628296,
      "learning_rate": 4.759643132006614e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 5644096,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 0.3297850489616394,
      "learning_rate": 4.759153697911535e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 5646720,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 0.4365072548389435,
      "learning_rate": 4.758663791227276e-05,
      "loss": 0.566,
      "num_input_tokens_seen": 5649696,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 0.4534454643726349,
      "learning_rate": 4.7581734120563206e-05,
      "loss": 0.4019,
      "num_input_tokens_seen": 5653504,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 0.40212932229042053,
      "learning_rate": 4.757682560501249e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 5656896,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 0.4875714182853699,
      "learning_rate": 4.757191236664744e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 5660928,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 0.4135870933532715,
      "learning_rate": 4.7566994406495825e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 5664064,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 1.2068604230880737,
      "learning_rate": 4.756207172558645e-05,
      "loss": 0.5777,
      "num_input_tokens_seen": 5667232,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 0.5850833058357239,
      "learning_rate": 4.7557144324949065e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 5670176,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 0.7409782409667969,
      "learning_rate": 4.755221220561443e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 5673120,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 0.4140196442604065,
      "learning_rate": 4.7547275368614295e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 5675936,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 0.22725388407707214,
      "learning_rate": 4.754233381498139e-05,
      "loss": 0.3743,
      "num_input_tokens_seen": 5679072,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 0.5086672902107239,
      "learning_rate": 4.753738754574942e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 5682112,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 0.5653823614120483,
      "learning_rate": 4.75324365619531e-05,
      "loss": 0.5947,
      "num_input_tokens_seen": 5685696,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 0.5160789489746094,
      "learning_rate": 4.752748086462811e-05,
      "loss": 0.4844,
      "num_input_tokens_seen": 5688096,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 0.35963764786720276,
      "learning_rate": 4.752252045481113e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 5691072,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 0.4353172183036804,
      "learning_rate": 4.7517555333539834e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 5694592,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 0.41774633526802063,
      "learning_rate": 4.751258550185284e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 5697312,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 0.32755064964294434,
      "learning_rate": 4.7507610960789804e-05,
      "loss": 0.345,
      "num_input_tokens_seen": 5701088,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 0.3942304253578186,
      "learning_rate": 4.7502631711391334e-05,
      "loss": 0.5414,
      "num_input_tokens_seen": 5703904,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 0.469475656747818,
      "learning_rate": 4.749764775469904e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 5706816,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 0.6412190794944763,
      "learning_rate": 4.749265909175549e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 5709632,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 0.8331477642059326,
      "learning_rate": 4.7487665723604275e-05,
      "loss": 0.4984,
      "num_input_tokens_seen": 5713472,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 0.48364752531051636,
      "learning_rate": 4.7482667651289944e-05,
      "loss": 0.6371,
      "num_input_tokens_seen": 5716736,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 0.6060716509819031,
      "learning_rate": 4.747766487585803e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 5720672,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 0.4254765212535858,
      "learning_rate": 4.747265739835507e-05,
      "loss": 0.6919,
      "num_input_tokens_seen": 5724000,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 0.8273310661315918,
      "learning_rate": 4.746764521982856e-05,
      "loss": 0.6747,
      "num_input_tokens_seen": 5726784,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 0.3330961763858795,
      "learning_rate": 4.7462628341326995e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 5730272,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 0.8126578330993652,
      "learning_rate": 4.7457606763899856e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 5733120,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 0.6279008388519287,
      "learning_rate": 4.745258048859758e-05,
      "loss": 0.3063,
      "num_input_tokens_seen": 5737824,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 0.4519859552383423,
      "learning_rate": 4.744754951647162e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 5740384,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 0.5411456227302551,
      "learning_rate": 4.74425138485744e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 5743232,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 0.36076024174690247,
      "learning_rate": 4.743747348595932e-05,
      "loss": 0.592,
      "num_input_tokens_seen": 5746432,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 0.6029363870620728,
      "learning_rate": 4.743242842968077e-05,
      "loss": 0.5544,
      "num_input_tokens_seen": 5749696,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 0.30144020915031433,
      "learning_rate": 4.742737868079411e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 5753152,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 0.3879924714565277,
      "learning_rate": 4.7422324240355696e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 5755712,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 0.3296617567539215,
      "learning_rate": 4.741726510942286e-05,
      "loss": 0.5059,
      "num_input_tokens_seen": 5759040,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 0.6941156387329102,
      "learning_rate": 4.741220128905391e-05,
      "loss": 0.5778,
      "num_input_tokens_seen": 5761408,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 0.657135546207428,
      "learning_rate": 4.740713278030815e-05,
      "loss": 0.5388,
      "num_input_tokens_seen": 5764672,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 0.35858169198036194,
      "learning_rate": 4.740205958424584e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 5768160,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 1.2622079849243164,
      "learning_rate": 4.739698170192824e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 5770400,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 0.3165275752544403,
      "learning_rate": 4.7391899134417596e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 5774080,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 0.8655548691749573,
      "learning_rate": 4.73868118827771e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 5777184,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 0.5711841583251953,
      "learning_rate": 4.738171994807097e-05,
      "loss": 0.5056,
      "num_input_tokens_seen": 5780736,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 0.400444895029068,
      "learning_rate": 4.7376623331364364e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 5784320,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 0.4339938759803772,
      "learning_rate": 4.737152203372345e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 5787840,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 0.4175103008747101,
      "learning_rate": 4.7366416056215345e-05,
      "loss": 0.614,
      "num_input_tokens_seen": 5791040,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 2.1336050033569336,
      "learning_rate": 4.7361305399908185e-05,
      "loss": 0.7821,
      "num_input_tokens_seen": 5794560,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 0.6533498167991638,
      "learning_rate": 4.735619006587103e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 5797504,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 0.3703977167606354,
      "learning_rate": 4.735107005517398e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 5799808,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 0.34753111004829407,
      "learning_rate": 4.734594536888806e-05,
      "loss": 0.3416,
      "num_input_tokens_seen": 5802752,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 0.38350579142570496,
      "learning_rate": 4.734081600808531e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 5806144,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 0.3958990275859833,
      "learning_rate": 4.733568197383872e-05,
      "loss": 0.5851,
      "num_input_tokens_seen": 5809536,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 0.47772014141082764,
      "learning_rate": 4.7330543267222294e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 5813088,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 0.6149101257324219,
      "learning_rate": 4.732539988931097e-05,
      "loss": 0.3574,
      "num_input_tokens_seen": 5815488,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 0.2891046404838562,
      "learning_rate": 4.732025184118068e-05,
      "loss": 0.6056,
      "num_input_tokens_seen": 5818528,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 0.384330689907074,
      "learning_rate": 4.731509912390837e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 5822016,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 0.21934930980205536,
      "learning_rate": 4.730994173857189e-05,
      "loss": 0.6366,
      "num_input_tokens_seen": 5825952,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 0.29179731011390686,
      "learning_rate": 4.7304779686250126e-05,
      "loss": 0.3702,
      "num_input_tokens_seen": 5828416,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 0.38264381885528564,
      "learning_rate": 4.729961296802292e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 5832352,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 0.3838970363140106,
      "learning_rate": 4.7294441584971095e-05,
      "loss": 0.5524,
      "num_input_tokens_seen": 5835488,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 0.37858399748802185,
      "learning_rate": 4.728926553817643e-05,
      "loss": 0.5085,
      "num_input_tokens_seen": 5838624,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 0.3403443694114685,
      "learning_rate": 4.728408482872171e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 5842592,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 0.5032059550285339,
      "learning_rate": 4.727889945769067e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 5846080,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 0.7421110272407532,
      "learning_rate": 4.727370942616803e-05,
      "loss": 0.5196,
      "num_input_tokens_seen": 5849536,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 0.47140780091285706,
      "learning_rate": 4.72685147352395e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 5852032,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 0.7043873071670532,
      "learning_rate": 4.7263315385991735e-05,
      "loss": 0.581,
      "num_input_tokens_seen": 5855072,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 0.30968406796455383,
      "learning_rate": 4.7258111379512384e-05,
      "loss": 0.506,
      "num_input_tokens_seen": 5858272,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 0.34881624579429626,
      "learning_rate": 4.725290271689007e-05,
      "loss": 0.4768,
      "num_input_tokens_seen": 5861856,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 0.3436099886894226,
      "learning_rate": 4.724768939921438e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 5864448,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.28202876448631287,
      "learning_rate": 4.7242471427575884e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 5867744,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 0.5792743563652039,
      "learning_rate": 4.723724880306612e-05,
      "loss": 0.3969,
      "num_input_tokens_seen": 5870336,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 0.2932124137878418,
      "learning_rate": 4.72320215267776e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 5873728,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 0.314768522977829,
      "learning_rate": 4.722678959980381e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 5876320,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 0.8812607526779175,
      "learning_rate": 4.7221553023239216e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 5879552,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 1.1081230640411377,
      "learning_rate": 4.721631179817925e-05,
      "loss": 0.4609,
      "num_input_tokens_seen": 5882432,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 0.782393217086792,
      "learning_rate": 4.721106592572031e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 5885248,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 0.37410983443260193,
      "learning_rate": 4.7205815406959774e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 5887680,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 0.350294291973114,
      "learning_rate": 4.7200560242996e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 5890848,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 0.30908820033073425,
      "learning_rate": 4.719530043492829e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 5893664,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 0.3904319107532501,
      "learning_rate": 4.7190035983856954e-05,
      "loss": 0.523,
      "num_input_tokens_seen": 5897216,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 0.5223408341407776,
      "learning_rate": 4.718476689088325e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 5900992,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 0.354605495929718,
      "learning_rate": 4.71794931571094e-05,
      "loss": 0.5792,
      "num_input_tokens_seen": 5903904,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 0.4035160541534424,
      "learning_rate": 4.717421478363862e-05,
      "loss": 0.5783,
      "num_input_tokens_seen": 5907008,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 0.4346368610858917,
      "learning_rate": 4.7168931771575085e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 5910112,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 0.45077404379844666,
      "learning_rate": 4.716364412202394e-05,
      "loss": 0.4412,
      "num_input_tokens_seen": 5913216,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 0.34216171503067017,
      "learning_rate": 4.7158351836091306e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 5916256,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 0.4011106789112091,
      "learning_rate": 4.7153054914884266e-05,
      "loss": 0.4993,
      "num_input_tokens_seen": 5919040,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 0.43547266721725464,
      "learning_rate": 4.714775335951086e-05,
      "loss": 0.4384,
      "num_input_tokens_seen": 5922944,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 0.7755692601203918,
      "learning_rate": 4.7142447171080137e-05,
      "loss": 0.3805,
      "num_input_tokens_seen": 5926368,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 0.24189572036266327,
      "learning_rate": 4.713713635070208e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 5929632,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 0.3149758279323578,
      "learning_rate": 4.713182089948765e-05,
      "loss": 0.4891,
      "num_input_tokens_seen": 5932832,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 0.3730473518371582,
      "learning_rate": 4.712650081854878e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 5936608,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 0.7133238315582275,
      "learning_rate": 4.712117610899837e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 5940640,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 0.482783704996109,
      "learning_rate": 4.7115846771950286e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 5944416,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 1.896998643875122,
      "learning_rate": 4.711051280851938e-05,
      "loss": 0.498,
      "num_input_tokens_seen": 5948352,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 0.40028098225593567,
      "learning_rate": 4.7105174219821435e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 5952000,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 0.987740159034729,
      "learning_rate": 4.709983100697323e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 5954528,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 0.5146019458770752,
      "learning_rate": 4.709448317109252e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 5957664,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 0.3735421895980835,
      "learning_rate": 4.708913071329798e-05,
      "loss": 0.3761,
      "num_input_tokens_seen": 5960448,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 0.38328319787979126,
      "learning_rate": 4.708377363470932e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 5963072,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 0.415849894285202,
      "learning_rate": 4.7078411936447144e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 5966240,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 0.31101495027542114,
      "learning_rate": 4.707304561963308e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 5969664,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 0.7386294007301331,
      "learning_rate": 4.70676746853897e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 5972320,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 0.3909318745136261,
      "learning_rate": 4.706229913484054e-05,
      "loss": 0.5379,
      "num_input_tokens_seen": 5975264,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 0.45547521114349365,
      "learning_rate": 4.705691896911009e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 5978336,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 0.5495057702064514,
      "learning_rate": 4.705153418932383e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 5981536,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 0.7440153956413269,
      "learning_rate": 4.7046144796608206e-05,
      "loss": 0.5958,
      "num_input_tokens_seen": 5984640,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 0.6337473392486572,
      "learning_rate": 4.70407507920906e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 5987168,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 0.29676368832588196,
      "learning_rate": 4.7035352176899384e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 5990880,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 0.37293368577957153,
      "learning_rate": 4.7029948952163885e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 5993664,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 0.39607855677604675,
      "learning_rate": 4.7024541119014386e-05,
      "loss": 0.5576,
      "num_input_tokens_seen": 5997216,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 0.39434459805488586,
      "learning_rate": 4.701912867858217e-05,
      "loss": 0.3358,
      "num_input_tokens_seen": 5999840,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 0.6408762335777283,
      "learning_rate": 4.7013711631999434e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 6003008,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 0.42843735218048096,
      "learning_rate": 4.700828998039937e-05,
      "loss": 0.5056,
      "num_input_tokens_seen": 6007552,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 0.5438364148139954,
      "learning_rate": 4.700286372491612e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 6011040,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 0.40973204374313354,
      "learning_rate": 4.6997432866684816e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 6013536,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 0.8549785614013672,
      "learning_rate": 4.6991997406841504e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 6015808,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 0.39028191566467285,
      "learning_rate": 4.698655734652324e-05,
      "loss": 0.3646,
      "num_input_tokens_seen": 6019328,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 0.24637345969676971,
      "learning_rate": 4.698111268686801e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 6022496,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 0.5319041013717651,
      "learning_rate": 4.6975663429014785e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 6025408,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 0.3089504837989807,
      "learning_rate": 4.697020957410347e-05,
      "loss": 0.5489,
      "num_input_tokens_seen": 6029056,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 0.4015665650367737,
      "learning_rate": 4.6964751123274974e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 6031584,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 0.6135467886924744,
      "learning_rate": 4.695928807767113e-05,
      "loss": 0.5161,
      "num_input_tokens_seen": 6034432,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 0.44315698742866516,
      "learning_rate": 4.695382043843474e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 6037792,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 0.4882391095161438,
      "learning_rate": 4.694834820670958e-05,
      "loss": 0.497,
      "num_input_tokens_seen": 6040832,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 0.3891458213329315,
      "learning_rate": 4.694287138364037e-05,
      "loss": 0.6078,
      "num_input_tokens_seen": 6044480,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 0.379818320274353,
      "learning_rate": 4.6937389970372815e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 6047520,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 0.3860335350036621,
      "learning_rate": 4.693190396805354e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 6051008,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 0.33686864376068115,
      "learning_rate": 4.692641337783017e-05,
      "loss": 0.5045,
      "num_input_tokens_seen": 6053216,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 0.35862845182418823,
      "learning_rate": 4.6920918200851274e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 6056864,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 0.3847183883190155,
      "learning_rate": 4.691541843826638e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 6059744,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 0.5108839869499207,
      "learning_rate": 4.690991409122597e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 6062624,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 0.445989191532135,
      "learning_rate": 4.690440516088149e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 6066080,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 0.49789029359817505,
      "learning_rate": 4.689889164838536e-05,
      "loss": 0.5517,
      "num_input_tokens_seen": 6069696,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 0.4013923108577728,
      "learning_rate": 4.6893373554890925e-05,
      "loss": 0.5804,
      "num_input_tokens_seen": 6073408,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 0.4303561747074127,
      "learning_rate": 4.688785088155251e-05,
      "loss": 0.2993,
      "num_input_tokens_seen": 6078592,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 0.35412687063217163,
      "learning_rate": 4.688232362952541e-05,
      "loss": 0.3501,
      "num_input_tokens_seen": 6088128,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 0.6319670677185059,
      "learning_rate": 4.687679179996585e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 6090816,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 0.4053601324558258,
      "learning_rate": 4.687125539403103e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 6093856,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 0.3142436444759369,
      "learning_rate": 4.68657144128791e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 6097792,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 0.33261874318122864,
      "learning_rate": 4.6860168857669175e-05,
      "loss": 0.3341,
      "num_input_tokens_seen": 6100160,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.6377231478691101,
      "learning_rate": 4.6854618729561315e-05,
      "loss": 0.5771,
      "num_input_tokens_seen": 6103264,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 0.45079317688941956,
      "learning_rate": 4.684906402971655e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 6105856,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 0.4917146563529968,
      "learning_rate": 4.684350475929685e-05,
      "loss": 0.5647,
      "num_input_tokens_seen": 6108800,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 0.33004337549209595,
      "learning_rate": 4.683794091946517e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 6111552,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 0.2819584310054779,
      "learning_rate": 4.683237251138538e-05,
      "loss": 0.5584,
      "num_input_tokens_seen": 6115488,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 0.43203631043434143,
      "learning_rate": 4.6826799536222346e-05,
      "loss": 0.4075,
      "num_input_tokens_seen": 6119136,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 0.4581955373287201,
      "learning_rate": 4.6821221995141865e-05,
      "loss": 0.6134,
      "num_input_tokens_seen": 6121696,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 0.5495412349700928,
      "learning_rate": 4.681563988931068e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 6124480,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 0.6702452301979065,
      "learning_rate": 4.681005321989652e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 6127008,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 0.5576153993606567,
      "learning_rate": 4.6804461988068055e-05,
      "loss": 0.3066,
      "num_input_tokens_seen": 6130080,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 0.30296796560287476,
      "learning_rate": 4.6798866194994896e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 6133632,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 0.3284260034561157,
      "learning_rate": 4.6793265841847624e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 6136992,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 0.5232042670249939,
      "learning_rate": 4.6787660929797766e-05,
      "loss": 0.3754,
      "num_input_tokens_seen": 6139776,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 0.3544946312904358,
      "learning_rate": 4.678205146001781e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 6142880,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 0.6016643047332764,
      "learning_rate": 4.67764374336812e-05,
      "loss": 0.438,
      "num_input_tokens_seen": 6147200,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 0.5446454882621765,
      "learning_rate": 4.6770818851962305e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 6150016,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 0.3500501811504364,
      "learning_rate": 4.676519571603648e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 6152960,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 0.43462857604026794,
      "learning_rate": 4.6759568027080034e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 6156544,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 0.6046294569969177,
      "learning_rate": 4.67539357862702e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 6159584,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 0.47780925035476685,
      "learning_rate": 4.6748298994785175e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 6162752,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 0.4971506595611572,
      "learning_rate": 4.6742657653804116e-05,
      "loss": 0.3289,
      "num_input_tokens_seen": 6166016,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 1.0451258420944214,
      "learning_rate": 4.673701176450713e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 6169184,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 0.560488224029541,
      "learning_rate": 4.6731361328075286e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 6172960,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 0.3280695676803589,
      "learning_rate": 4.672570634569056e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 6176832,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 0.3784867227077484,
      "learning_rate": 4.672004681853593e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 6179392,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 0.31284716725349426,
      "learning_rate": 4.6714382747795304e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 6182208,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 0.3596586287021637,
      "learning_rate": 4.670871413465354e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 6185376,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 0.310510516166687,
      "learning_rate": 4.670304098029645e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 6188896,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 0.49492520093917847,
      "learning_rate": 4.669736328591078e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 6191840,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 0.4918389618396759,
      "learning_rate": 4.669168105268424e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 6194784,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.9018572568893433,
      "learning_rate": 4.6685994281805514e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 6198048,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 1.0411227941513062,
      "learning_rate": 4.6680302974464175e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 6202688,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 0.30632826685905457,
      "learning_rate": 4.66746071318508e-05,
      "loss": 0.6235,
      "num_input_tokens_seen": 6205952,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 0.33284416794776917,
      "learning_rate": 4.6668906755156896e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 6209504,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 0.4123550057411194,
      "learning_rate": 4.6663201845574906e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 6212928,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 0.41596439480781555,
      "learning_rate": 4.6657492404298234e-05,
      "loss": 0.4154,
      "num_input_tokens_seen": 6216672,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4628125429153442,
      "learning_rate": 4.665177843252124e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 6219592,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.46520644426345825,
      "eval_runtime": 9.8157,
      "eval_samples_per_second": 86.392,
      "eval_steps_per_second": 21.598,
      "num_input_tokens_seen": 6219592,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 0.6025804877281189,
      "learning_rate": 4.664605993143921e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 6222216,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 0.3962200880050659,
      "learning_rate": 4.6640336902248396e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 6224872,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 0.33559271693229675,
      "learning_rate": 4.6634609346145984e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 6228072,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 0.45215749740600586,
      "learning_rate": 4.6628877264330116e-05,
      "loss": 0.556,
      "num_input_tokens_seen": 6230984,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 0.4507220983505249,
      "learning_rate": 4.6623140657999886e-05,
      "loss": 0.5403,
      "num_input_tokens_seen": 6234056,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 0.2883535325527191,
      "learning_rate": 4.661739952835532e-05,
      "loss": 0.3362,
      "num_input_tokens_seen": 6237800,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 0.5694966912269592,
      "learning_rate": 4.66116538765974e-05,
      "loss": 0.662,
      "num_input_tokens_seen": 6240808,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 0.5730969905853271,
      "learning_rate": 4.6605903703928034e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 6244040,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.31205540895462036,
      "learning_rate": 4.6600149011550116e-05,
      "loss": 0.5518,
      "num_input_tokens_seen": 6246952,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 0.3268347680568695,
      "learning_rate": 4.659438980066745e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 6250472,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 0.35517337918281555,
      "learning_rate": 4.65886260724848e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 6253640,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 0.5820360779762268,
      "learning_rate": 4.658285782820787e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 6257160,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 0.39424821734428406,
      "learning_rate": 4.657708506904331e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 6259688,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 0.5174373388290405,
      "learning_rate": 4.6571307796198724e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 6263080,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 0.37274500727653503,
      "learning_rate": 4.656552601088265e-05,
      "loss": 0.2943,
      "num_input_tokens_seen": 6269320,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 0.46007847785949707,
      "learning_rate": 4.655973971430456e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 6272392,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 0.8066403269767761,
      "learning_rate": 4.6553948907674895e-05,
      "loss": 0.5958,
      "num_input_tokens_seen": 6275496,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 0.43622466921806335,
      "learning_rate": 4.654815359220501e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 6277960,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 0.29756370186805725,
      "learning_rate": 4.6542353769107226e-05,
      "loss": 0.8842,
      "num_input_tokens_seen": 6280520,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 0.4470531940460205,
      "learning_rate": 4.653654943959481e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 6283304,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 1.0156078338623047,
      "learning_rate": 4.653074060488195e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 6286536,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 0.46754342317581177,
      "learning_rate": 4.6524927266183796e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 6289928,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 0.33560502529144287,
      "learning_rate": 4.6519109424716426e-05,
      "loss": 0.5433,
      "num_input_tokens_seen": 6293416,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 0.45655587315559387,
      "learning_rate": 4.6513287081696855e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 6297320,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 0.3875739574432373,
      "learning_rate": 4.6507460238343076e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 6300008,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 0.21258924901485443,
      "learning_rate": 4.6501628895873984e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 6304136,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 0.4745815396308899,
      "learning_rate": 4.649579305550942e-05,
      "loss": 0.5806,
      "num_input_tokens_seen": 6306792,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 0.5189903974533081,
      "learning_rate": 4.64899527184702e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 6309800,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 0.7640408873558044,
      "learning_rate": 4.6484107885978035e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 6312360,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 0.28670719265937805,
      "learning_rate": 4.64782585592556e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 6316008,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 0.35681524872779846,
      "learning_rate": 4.647240473952652e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 6318952,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 0.6506062746047974,
      "learning_rate": 4.6466546428015336e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 6322056,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 0.5070216655731201,
      "learning_rate": 4.646068362594753e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 6325000,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 0.5032241940498352,
      "learning_rate": 4.645481633454956e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 6329960,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 0.4739186763763428,
      "learning_rate": 4.644894455504878e-05,
      "loss": 0.3619,
      "num_input_tokens_seen": 6333640,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 1.034719705581665,
      "learning_rate": 4.64430682886735e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 6336488,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 0.6164433360099792,
      "learning_rate": 4.643718753665297e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 6340328,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 0.3835827112197876,
      "learning_rate": 4.643130230021737e-05,
      "loss": 0.5475,
      "num_input_tokens_seen": 6343464,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 0.43882444500923157,
      "learning_rate": 4.6425412580597836e-05,
      "loss": 0.6539,
      "num_input_tokens_seen": 6349352,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 0.27159082889556885,
      "learning_rate": 4.641951837902643e-05,
      "loss": 0.3972,
      "num_input_tokens_seen": 6351784,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 0.4722251892089844,
      "learning_rate": 4.641361969673614e-05,
      "loss": 0.6075,
      "num_input_tokens_seen": 6355368,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 0.31749334931373596,
      "learning_rate": 4.640771653496091e-05,
      "loss": 0.5298,
      "num_input_tokens_seen": 6358312,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 0.49466726183891296,
      "learning_rate": 4.64018088949356e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 6360776,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 0.3550513982772827,
      "learning_rate": 4.6395896777896055e-05,
      "loss": 0.5808,
      "num_input_tokens_seen": 6364424,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 0.3496077358722687,
      "learning_rate": 4.638998018507899e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 6367496,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 0.7420319318771362,
      "learning_rate": 4.63840591177221e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 6370472,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 0.3632505536079407,
      "learning_rate": 4.637813357706401e-05,
      "loss": 0.383,
      "num_input_tokens_seen": 6373832,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 0.3001570403575897,
      "learning_rate": 4.6372203564344264e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 6377672,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 0.5057389140129089,
      "learning_rate": 4.6366269080803364e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 6380712,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 0.5852741003036499,
      "learning_rate": 4.636033012768273e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 6384776,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 0.5225715041160583,
      "learning_rate": 4.635438670622473e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 6387560,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 0.5381527543067932,
      "learning_rate": 4.6348438817672646e-05,
      "loss": 0.6042,
      "num_input_tokens_seen": 6390952,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 0.24318280816078186,
      "learning_rate": 4.634248646327072e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 6394088,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 0.23282569646835327,
      "learning_rate": 4.6336529644264114e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 6397064,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 0.6158338785171509,
      "learning_rate": 4.633056836189893e-05,
      "loss": 0.6588,
      "num_input_tokens_seen": 6400584,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 0.344962477684021,
      "learning_rate": 4.632460261742219e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 6403976,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 0.5315952897071838,
      "learning_rate": 4.631863241208187e-05,
      "loss": 0.487,
      "num_input_tokens_seen": 6406536,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 0.4053274691104889,
      "learning_rate": 4.6312657747126865e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 6410312,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 0.5054208040237427,
      "learning_rate": 4.630667862380701e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 6414184,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 0.7205480933189392,
      "learning_rate": 4.630069504337306e-05,
      "loss": 0.7025,
      "num_input_tokens_seen": 6417224,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 0.32420614361763,
      "learning_rate": 4.629470700707672e-05,
      "loss": 0.5537,
      "num_input_tokens_seen": 6421128,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 0.3983457088470459,
      "learning_rate": 4.628871451617062e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 6424232,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 0.6757380962371826,
      "learning_rate": 4.628271757190832e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 6427464,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 0.4833410680294037,
      "learning_rate": 4.62767161755443e-05,
      "loss": 0.4063,
      "num_input_tokens_seen": 6430504,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 0.5274591445922852,
      "learning_rate": 4.6270710328334004e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 6434056,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 0.5350170135498047,
      "learning_rate": 4.626470003153377e-05,
      "loss": 0.6534,
      "num_input_tokens_seen": 6437256,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 0.5376296639442444,
      "learning_rate": 4.625868528640089e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 6441064,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 0.3492805063724518,
      "learning_rate": 4.6252666094193586e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 6444648,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 0.44110405445098877,
      "learning_rate": 4.6246642456171e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 6447880,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 0.304965615272522,
      "learning_rate": 4.6240614373593196e-05,
      "loss": 0.4621,
      "num_input_tokens_seen": 6450728,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 0.43245378136634827,
      "learning_rate": 4.623458184772119e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 6453640,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 0.29371678829193115,
      "learning_rate": 4.622854487981693e-05,
      "loss": 0.3803,
      "num_input_tokens_seen": 6456840,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 0.470708966255188,
      "learning_rate": 4.622250347114326e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 6459848,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 0.49448320269584656,
      "learning_rate": 4.621645762296399e-05,
      "loss": 0.5446,
      "num_input_tokens_seen": 6463624,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 0.4061048924922943,
      "learning_rate": 4.6210407336543835e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 6467176,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 0.6169980764389038,
      "learning_rate": 4.620435261314844e-05,
      "loss": 0.497,
      "num_input_tokens_seen": 6469896,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 0.39083024859428406,
      "learning_rate": 4.61982934540444e-05,
      "loss": 0.5815,
      "num_input_tokens_seen": 6472904,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 0.3054385483264923,
      "learning_rate": 4.6192229860499206e-05,
      "loss": 0.5157,
      "num_input_tokens_seen": 6477288,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 0.4885846674442291,
      "learning_rate": 4.6186161833781295e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 6480488,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 0.7013831734657288,
      "learning_rate": 4.618008937516003e-05,
      "loss": 0.3715,
      "num_input_tokens_seen": 6483368,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 1.0230399370193481,
      "learning_rate": 4.617401248590572e-05,
      "loss": 0.5313,
      "num_input_tokens_seen": 6485928,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 0.39665523171424866,
      "learning_rate": 4.616793116728955e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 6489768,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 0.5773641467094421,
      "learning_rate": 4.6161845420583685e-05,
      "loss": 0.5013,
      "num_input_tokens_seen": 6493352,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 0.4259793758392334,
      "learning_rate": 4.6155755247061174e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 6495944,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 0.38702309131622314,
      "learning_rate": 4.614966064799603e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 6498472,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 0.7595377564430237,
      "learning_rate": 4.6143561624663166e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 6502088,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 0.34075114130973816,
      "learning_rate": 4.6137458178338425e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 6505032,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 0.5287438631057739,
      "learning_rate": 4.613135031029859e-05,
      "loss": 0.3587,
      "num_input_tokens_seen": 6508488,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 0.44566068053245544,
      "learning_rate": 4.612523802182134e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 6510984,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 0.37729015946388245,
      "learning_rate": 4.611912131418531e-05,
      "loss": 0.5493,
      "num_input_tokens_seen": 6514376,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 0.5411972403526306,
      "learning_rate": 4.611300018867004e-05,
      "loss": 0.5934,
      "num_input_tokens_seen": 6517192,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 0.521286129951477,
      "learning_rate": 4.6106874646556e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 6520424,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 0.2221977412700653,
      "learning_rate": 4.6100744689124584e-05,
      "loss": 0.2578,
      "num_input_tokens_seen": 6523976,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 0.44744083285331726,
      "learning_rate": 4.6094610317658116e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 6526408,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 0.5503380298614502,
      "learning_rate": 4.608847153343982e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 6529288,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 0.5998731255531311,
      "learning_rate": 4.6082328337753874e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 6532072,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 0.371152400970459,
      "learning_rate": 4.6076180731885353e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 6536520,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 0.8237359523773193,
      "learning_rate": 4.6070028717120286e-05,
      "loss": 0.5137,
      "num_input_tokens_seen": 6539880,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 0.7785926461219788,
      "learning_rate": 4.606387229474558e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 6542632,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 0.6639962196350098,
      "learning_rate": 4.60577114660491e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 6545416,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 0.22071118652820587,
      "learning_rate": 4.605154623231962e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 6548680,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 0.30879345536231995,
      "learning_rate": 4.604537659484684e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 6552744,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 0.30443763732910156,
      "learning_rate": 4.603920255492138e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 6557128,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 0.3685283660888672,
      "learning_rate": 4.603302411383477e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 6565352,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 0.5619073510169983,
      "learning_rate": 4.602684127287946e-05,
      "loss": 0.5015,
      "num_input_tokens_seen": 6568680,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 0.4989531338214874,
      "learning_rate": 4.602065403334885e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 6572520,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 0.37104982137680054,
      "learning_rate": 4.601446239653724e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 6576232,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.4291011393070221,
      "learning_rate": 4.600826636373984e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 6579016,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 0.8316132426261902,
      "learning_rate": 4.600206593625279e-05,
      "loss": 0.457,
      "num_input_tokens_seen": 6581864,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 0.2702614367008209,
      "learning_rate": 4.599586111537315e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 6585256,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 0.4250612258911133,
      "learning_rate": 4.59896519023989e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 6587976,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 0.4328373968601227,
      "learning_rate": 4.5983438298628935e-05,
      "loss": 0.6407,
      "num_input_tokens_seen": 6591080,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 0.5404690504074097,
      "learning_rate": 4.597722030536306e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 6594088,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 0.6455286741256714,
      "learning_rate": 4.597099792390202e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 6597544,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 0.5605184435844421,
      "learning_rate": 4.596477115554748e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 6600552,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 0.7766286730766296,
      "learning_rate": 4.5958540001601976e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 6602888,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 0.48304295539855957,
      "learning_rate": 4.595230446336902e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 6606344,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 0.3762851357460022,
      "learning_rate": 4.5946064542153005e-05,
      "loss": 0.5545,
      "num_input_tokens_seen": 6609864,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 0.5423247814178467,
      "learning_rate": 4.593982023925926e-05,
      "loss": 0.3986,
      "num_input_tokens_seen": 6612616,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 0.4583301246166229,
      "learning_rate": 4.5933571555994e-05,
      "loss": 0.4017,
      "num_input_tokens_seen": 6616072,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 0.307921439409256,
      "learning_rate": 4.59273184936644e-05,
      "loss": 0.6495,
      "num_input_tokens_seen": 6619528,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 0.3474122881889343,
      "learning_rate": 4.5921061053578523e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 6622824,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 1.2331756353378296,
      "learning_rate": 4.5914799237045355e-05,
      "loss": 0.5132,
      "num_input_tokens_seen": 6626152,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 0.4178811311721802,
      "learning_rate": 4.5908533045374796e-05,
      "loss": 0.5025,
      "num_input_tokens_seen": 6628808,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 0.5172649025917053,
      "learning_rate": 4.590226247987766e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 6632296,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 0.3244197964668274,
      "learning_rate": 4.589598754186568e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 6635400,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 0.408214271068573,
      "learning_rate": 4.5889708232651494e-05,
      "loss": 0.5347,
      "num_input_tokens_seen": 6639432,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 0.5291860699653625,
      "learning_rate": 4.588342455354867e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 6643336,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 0.4918294847011566,
      "learning_rate": 4.5877136505871685e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 6646088,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 0.5341249108314514,
      "learning_rate": 4.5870844090935914e-05,
      "loss": 0.3219,
      "num_input_tokens_seen": 6649384,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 0.5906021595001221,
      "learning_rate": 4.586454731005768e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 6652424,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 0.4666018486022949,
      "learning_rate": 4.5858246164554175e-05,
      "loss": 0.4675,
      "num_input_tokens_seen": 6655464,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 0.6671901345252991,
      "learning_rate": 4.5851940655743536e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 6659112,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 0.6898037195205688,
      "learning_rate": 4.5845630784944804e-05,
      "loss": 0.5055,
      "num_input_tokens_seen": 6663304,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 0.48646098375320435,
      "learning_rate": 4.583931655347793e-05,
      "loss": 0.3477,
      "num_input_tokens_seen": 6666888,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 0.4871733784675598,
      "learning_rate": 4.5832997962663785e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 6670504,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 0.37175053358078003,
      "learning_rate": 4.582667501382414e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 6673832,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 0.6243154406547546,
      "learning_rate": 4.582034770828169e-05,
      "loss": 0.466,
      "num_input_tokens_seen": 6677160,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 0.24071557819843292,
      "learning_rate": 4.581401604736002e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 6681160,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 0.4687153995037079,
      "learning_rate": 4.580768003238365e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 6684424,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 0.4365558922290802,
      "learning_rate": 4.5801339664678014e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 6687528,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 1.2515480518341064,
      "learning_rate": 4.579499494556943e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 6690120,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 0.8882181644439697,
      "learning_rate": 4.5788645876385147e-05,
      "loss": 0.5323,
      "num_input_tokens_seen": 6693992,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 0.5805412530899048,
      "learning_rate": 4.578229245845331e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 6698568,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 0.5001826286315918,
      "learning_rate": 4.5775934693102994e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 6701320,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 0.659019947052002,
      "learning_rate": 4.5769572581664165e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 6704168,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 0.5066644549369812,
      "learning_rate": 4.57632061254677e-05,
      "loss": 0.5053,
      "num_input_tokens_seen": 6707688,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 0.7794238328933716,
      "learning_rate": 4.5756835325845394e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 6711048,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 1.5153158903121948,
      "learning_rate": 4.575046018412995e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 6715016,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 0.6621873378753662,
      "learning_rate": 4.5744080701654955e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 6717448,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 0.3791181147098541,
      "learning_rate": 4.573769687975495e-05,
      "loss": 0.3225,
      "num_input_tokens_seen": 6721032,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 0.4983465075492859,
      "learning_rate": 4.573130871976534e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 6723304,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 0.3662085235118866,
      "learning_rate": 4.572491622302247e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 6726344,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 0.3184206485748291,
      "learning_rate": 4.571851939086356e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 6730248,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 0.41854047775268555,
      "learning_rate": 4.571211822462676e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 6733352,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 1.4243172407150269,
      "learning_rate": 4.570571272565114e-05,
      "loss": 0.3738,
      "num_input_tokens_seen": 6736104,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.2695348262786865,
      "learning_rate": 4.569930289527663e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 6739336,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 0.4038671553134918,
      "learning_rate": 4.56928887348441e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 6742216,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 0.3285105228424072,
      "learning_rate": 4.568647024569533e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 6745320,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 0.43806442618370056,
      "learning_rate": 4.568004742917299e-05,
      "loss": 0.3684,
      "num_input_tokens_seen": 6748328,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 0.5250583291053772,
      "learning_rate": 4.567362028662065e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 6751240,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 0.640674889087677,
      "learning_rate": 4.566718881938281e-05,
      "loss": 0.3857,
      "num_input_tokens_seen": 6754344,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 0.5687777996063232,
      "learning_rate": 4.5660753028804846e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 6757320,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 0.5070491433143616,
      "learning_rate": 4.5654312916233066e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 6760744,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 0.4847525656223297,
      "learning_rate": 4.564786848301466e-05,
      "loss": 0.5362,
      "num_input_tokens_seen": 6763848,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 0.5952680110931396,
      "learning_rate": 4.564141973049773e-05,
      "loss": 0.557,
      "num_input_tokens_seen": 6766792,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 0.4817945957183838,
      "learning_rate": 4.563496666003128e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 6770024,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 0.5153579115867615,
      "learning_rate": 4.562850927296523e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 6772584,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 0.38956189155578613,
      "learning_rate": 4.5622047570650365e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 6776040,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 0.4756523072719574,
      "learning_rate": 4.561558155443843e-05,
      "loss": 0.4448,
      "num_input_tokens_seen": 6780616,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 0.3045099973678589,
      "learning_rate": 4.560911122568202e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 6783720,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 0.4182150363922119,
      "learning_rate": 4.560263658573467e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 6786824,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 0.4970841407775879,
      "learning_rate": 4.559615763595079e-05,
      "loss": 0.532,
      "num_input_tokens_seen": 6790696,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 0.33891281485557556,
      "learning_rate": 4.558967437768571e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 6794408,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 0.5119839310646057,
      "learning_rate": 4.558318681229565e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 6798184,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 0.4603254795074463,
      "learning_rate": 4.557669494113773e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 6801928,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 0.30374181270599365,
      "learning_rate": 4.557019876556998e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 6805640,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 0.3931446373462677,
      "learning_rate": 4.556369828695133e-05,
      "loss": 0.3601,
      "num_input_tokens_seen": 6808680,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 0.45887938141822815,
      "learning_rate": 4.5557193506641596e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 6811752,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.4031752049922943,
      "learning_rate": 4.555068442600151e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 6814888,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 0.29053881764411926,
      "learning_rate": 4.5544171046392703e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 6819048,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 0.8176886439323425,
      "learning_rate": 4.5537653369177685e-05,
      "loss": 0.35,
      "num_input_tokens_seen": 6822792,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 0.23671866953372955,
      "learning_rate": 4.5531131395719895e-05,
      "loss": 0.5634,
      "num_input_tokens_seen": 6826248,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 0.5586079955101013,
      "learning_rate": 4.552460512738364e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 6830248,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 0.3805790841579437,
      "learning_rate": 4.551807456553416e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 6833160,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 0.49654629826545715,
      "learning_rate": 4.551153971153755e-05,
      "loss": 0.6334,
      "num_input_tokens_seen": 6836392,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 0.32578030228614807,
      "learning_rate": 4.550500056676085e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 6840872,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 0.3825337588787079,
      "learning_rate": 4.5498457132571966e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 6843432,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 0.48191139101982117,
      "learning_rate": 4.5491909410339706e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 6846760,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 0.5398310422897339,
      "learning_rate": 4.548535740143378e-05,
      "loss": 0.6325,
      "num_input_tokens_seen": 6849256,
      "step": 10490
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 0.4169401228427887,
      "learning_rate": 4.54788011072248e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 6853256,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 0.46258026361465454,
      "learning_rate": 4.5472240529084254e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 6859688,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 0.49381116032600403,
      "learning_rate": 4.5465675668384557e-05,
      "loss": 0.3989,
      "num_input_tokens_seen": 6863080,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 0.40017732977867126,
      "learning_rate": 4.545910652649899e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 6865704,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 0.4315170645713806,
      "learning_rate": 4.5452533104801756e-05,
      "loss": 0.5238,
      "num_input_tokens_seen": 6868392,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 0.5157352089881897,
      "learning_rate": 4.544595540466792e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 6871624,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 0.3144085109233856,
      "learning_rate": 4.5439373427473476e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 6875752,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 0.6588296294212341,
      "learning_rate": 4.54327871745953e-05,
      "loss": 0.6273,
      "num_input_tokens_seen": 6878728,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 0.3585309386253357,
      "learning_rate": 4.542619664741116e-05,
      "loss": 0.5268,
      "num_input_tokens_seen": 6882408,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 0.39422160387039185,
      "learning_rate": 4.5419601847299706e-05,
      "loss": 0.6066,
      "num_input_tokens_seen": 6885448,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 0.45022591948509216,
      "learning_rate": 4.5413002775640503e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 6889192,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 0.5173443555831909,
      "learning_rate": 4.540639943381401e-05,
      "loss": 0.521,
      "num_input_tokens_seen": 6892424,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 0.45596426725387573,
      "learning_rate": 4.539979182320157e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 6895368,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 0.2765852212905884,
      "learning_rate": 4.5393179945185404e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 6899784,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 0.5879523754119873,
      "learning_rate": 4.538656380114865e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 6902504,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 0.34312424063682556,
      "learning_rate": 4.537994339247533e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 6906184,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 0.3767506778240204,
      "learning_rate": 4.5373318720550366e-05,
      "loss": 0.4754,
      "num_input_tokens_seen": 6909320,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 0.5069789290428162,
      "learning_rate": 4.536668978675956e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 6912168,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 0.4528123438358307,
      "learning_rate": 4.53600565924896e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 6915784,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 0.6142091751098633,
      "learning_rate": 4.535341913912808e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 6918920,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 0.4167778789997101,
      "learning_rate": 4.534677742806348e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 6921992,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.5981106758117676,
      "learning_rate": 4.5340131460685175e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 6924840,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 0.4484335780143738,
      "learning_rate": 4.533348123838341e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 6927848,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 0.6521962285041809,
      "learning_rate": 4.532682676254936e-05,
      "loss": 0.5921,
      "num_input_tokens_seen": 6930568,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 0.6124480962753296,
      "learning_rate": 4.5320168034575044e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 6933992,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 0.31749993562698364,
      "learning_rate": 4.53135050558534e-05,
      "loss": 0.4393,
      "num_input_tokens_seen": 6936680,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 0.6899250149726868,
      "learning_rate": 4.530683782777825e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 6938792,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 0.3472849428653717,
      "learning_rate": 4.53001663517443e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 6942664,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 0.7039322257041931,
      "learning_rate": 4.529349062914714e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 6944968,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 0.40630248188972473,
      "learning_rate": 4.528681066138326e-05,
      "loss": 0.4858,
      "num_input_tokens_seen": 6948264,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 0.4349670112133026,
      "learning_rate": 4.528012644985005e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 6951304,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 0.24132567644119263,
      "learning_rate": 4.527343799594574e-05,
      "loss": 0.4097,
      "num_input_tokens_seen": 6954184,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 0.40989989042282104,
      "learning_rate": 4.5266745301069494e-05,
      "loss": 0.601,
      "num_input_tokens_seen": 6956552,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 0.37181356549263,
      "learning_rate": 4.526004836662135e-05,
      "loss": 0.5316,
      "num_input_tokens_seen": 6959496,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 0.3071357011795044,
      "learning_rate": 4.525334719400223e-05,
      "loss": 0.6466,
      "num_input_tokens_seen": 6962600,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 0.4051250219345093,
      "learning_rate": 4.524664178461393e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 6965160,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 0.31991034746170044,
      "learning_rate": 4.5239932139859154e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 6967880,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 0.29588717222213745,
      "learning_rate": 4.523321826114149e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 6971624,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 0.3551926910877228,
      "learning_rate": 4.5226500149865395e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 6974696,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 0.7698989510536194,
      "learning_rate": 4.5219777807436224e-05,
      "loss": 0.6449,
      "num_input_tokens_seen": 6977544,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 0.3037516474723816,
      "learning_rate": 4.5213051235260225e-05,
      "loss": 0.2956,
      "num_input_tokens_seen": 6983688,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 0.4057157039642334,
      "learning_rate": 4.52063204347445e-05,
      "loss": 0.4558,
      "num_input_tokens_seen": 6987112,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 0.4163210690021515,
      "learning_rate": 4.519958540729707e-05,
      "loss": 0.6164,
      "num_input_tokens_seen": 6990312,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 0.4215681850910187,
      "learning_rate": 4.519284615432682e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 6992456,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 0.3724684715270996,
      "learning_rate": 4.518610267724352e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 6994792,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 0.435982346534729,
      "learning_rate": 4.517935497745786e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 6997960,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 0.6414126753807068,
      "learning_rate": 4.517260305638133e-05,
      "loss": 0.5832,
      "num_input_tokens_seen": 7001640,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 0.4512043595314026,
      "learning_rate": 4.516584691542639e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 7004456,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 0.741211473941803,
      "learning_rate": 4.515908655600634e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 7006888,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 0.33455631136894226,
      "learning_rate": 4.515232197953537e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 7010600,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 0.4858441650867462,
      "learning_rate": 4.514555318742856e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 7014312,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 0.24017004668712616,
      "learning_rate": 4.513878018110184e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 7017896,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 0.581510066986084,
      "learning_rate": 4.513200296197207e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 7021256,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 0.22210481762886047,
      "learning_rate": 4.512522153145696e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 7024968,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 0.5762447714805603,
      "learning_rate": 4.5118435890975106e-05,
      "loss": 0.5347,
      "num_input_tokens_seen": 7027528,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 0.8711896538734436,
      "learning_rate": 4.511164604194598e-05,
      "loss": 0.4562,
      "num_input_tokens_seen": 7030632,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 0.23820671439170837,
      "learning_rate": 4.5104851985789956e-05,
      "loss": 0.4197,
      "num_input_tokens_seen": 7034568,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 0.37164753675460815,
      "learning_rate": 4.5098053723928254e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 7037544,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 0.480503648519516,
      "learning_rate": 4.509125125778302e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 7041256,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 0.6753732562065125,
      "learning_rate": 4.508444458877722e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 7043528,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 0.3871496021747589,
      "learning_rate": 4.5077633718334755e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 7046696,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.2708323895931244,
      "learning_rate": 4.5070818647880375e-05,
      "loss": 0.6021,
      "num_input_tokens_seen": 7050216,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 0.32733604311943054,
      "learning_rate": 4.506399937883972e-05,
      "loss": 0.6549,
      "num_input_tokens_seen": 7053832,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 0.48602038621902466,
      "learning_rate": 4.5057175912639286e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 7056584,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 0.3002772033214569,
      "learning_rate": 4.5050348250706475e-05,
      "loss": 0.3529,
      "num_input_tokens_seen": 7059240,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 0.5354130864143372,
      "learning_rate": 4.5043516394469566e-05,
      "loss": 0.5212,
      "num_input_tokens_seen": 7062280,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 0.31464722752571106,
      "learning_rate": 4.503668034535768e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 7065288,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 0.32335060834884644,
      "learning_rate": 4.5029840104800876e-05,
      "loss": 0.5137,
      "num_input_tokens_seen": 7069160,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 0.43606480956077576,
      "learning_rate": 4.5022995674230025e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 7072552,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 0.40360113978385925,
      "learning_rate": 4.5016147055076914e-05,
      "loss": 0.367,
      "num_input_tokens_seen": 7076040,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 0.3750578463077545,
      "learning_rate": 4.500929424877419e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 7079368,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 0.47660577297210693,
      "learning_rate": 4.5002437256755394e-05,
      "loss": 0.3989,
      "num_input_tokens_seen": 7082632,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 0.332670658826828,
      "learning_rate": 4.4995576080454924e-05,
      "loss": 0.4982,
      "num_input_tokens_seen": 7085192,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 0.5809047222137451,
      "learning_rate": 4.4988710721308056e-05,
      "loss": 0.3973,
      "num_input_tokens_seen": 7088040,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 0.6459276676177979,
      "learning_rate": 4.498184118075095e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 7090664,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 0.48724937438964844,
      "learning_rate": 4.497496746022063e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 7093224,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 0.40751293301582336,
      "learning_rate": 4.4968089561155015e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 7096008,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 0.47874897718429565,
      "learning_rate": 4.496120748499286e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 7098760,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 0.7497479319572449,
      "learning_rate": 4.4954321233173833e-05,
      "loss": 0.532,
      "num_input_tokens_seen": 7102120,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 0.29109203815460205,
      "learning_rate": 4.494743080713846e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 7105672,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 0.7666071057319641,
      "learning_rate": 4.494053620832812e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 7109288,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 0.5422132015228271,
      "learning_rate": 4.493363743818511e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 7112264,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 1.1134610176086426,
      "learning_rate": 4.492673449815257e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 7114984,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 0.39042672514915466,
      "learning_rate": 4.4919827389674496e-05,
      "loss": 0.4742,
      "num_input_tokens_seen": 7117896,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 0.46121037006378174,
      "learning_rate": 4.491291611419579e-05,
      "loss": 0.3877,
      "num_input_tokens_seen": 7121192,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 0.36498162150382996,
      "learning_rate": 4.4906000673162216e-05,
      "loss": 0.4515,
      "num_input_tokens_seen": 7124680,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 0.48946288228034973,
      "learning_rate": 4.48990810680204e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 7127624,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 0.4273965656757355,
      "learning_rate": 4.4892157300217844e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 7134792,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 0.3988407850265503,
      "learning_rate": 4.488522937120293e-05,
      "loss": 0.3645,
      "num_input_tokens_seen": 7137448,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 0.8319293260574341,
      "learning_rate": 4.4878297282424886e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 7141384,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 0.9086925983428955,
      "learning_rate": 4.4871361035333836e-05,
      "loss": 0.5268,
      "num_input_tokens_seen": 7144296,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 0.4356897473335266,
      "learning_rate": 4.486442063138076e-05,
      "loss": 0.5877,
      "num_input_tokens_seen": 7148328,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 0.2929290533065796,
      "learning_rate": 4.4857476072017516e-05,
      "loss": 0.545,
      "num_input_tokens_seen": 7151304,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 0.27641749382019043,
      "learning_rate": 4.485052735869682e-05,
      "loss": 0.4119,
      "num_input_tokens_seen": 7155240,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 0.37875062227249146,
      "learning_rate": 4.484357449287228e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 7159496,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 0.46320217847824097,
      "learning_rate": 4.483661747599833e-05,
      "loss": 0.3804,
      "num_input_tokens_seen": 7163688,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 0.31893327832221985,
      "learning_rate": 4.482965630953031e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 7166728,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 0.4216320514678955,
      "learning_rate": 4.4822690994924425e-05,
      "loss": 0.5357,
      "num_input_tokens_seen": 7169928,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 0.39312803745269775,
      "learning_rate": 4.481572153363773e-05,
      "loss": 0.5653,
      "num_input_tokens_seen": 7172936,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 1.1583706140518188,
      "learning_rate": 4.480874792712817e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 7175240,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 0.3701629936695099,
      "learning_rate": 4.480177017685452e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 7178216,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 0.49224725365638733,
      "learning_rate": 4.479478828427647e-05,
      "loss": 0.5123,
      "num_input_tokens_seen": 7182056,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 0.2651211619377136,
      "learning_rate": 4.478780225085454e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 7186056,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 0.3562229573726654,
      "learning_rate": 4.4780812078050125e-05,
      "loss": 0.639,
      "num_input_tokens_seen": 7188808,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 0.2081162929534912,
      "learning_rate": 4.47738177673255e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 7192328,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 0.8680129647254944,
      "learning_rate": 4.4766819320143785e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 7195368,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 0.26201146841049194,
      "learning_rate": 4.475981673796899e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 7199752,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 0.6084786653518677,
      "learning_rate": 4.475281002226596e-05,
      "loss": 0.632,
      "num_input_tokens_seen": 7203432,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 0.5270165205001831,
      "learning_rate": 4.4745799174500414e-05,
      "loss": 0.4063,
      "num_input_tokens_seen": 7205864,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 0.3197998106479645,
      "learning_rate": 4.473878419613896e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 7208424,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 0.377016544342041,
      "learning_rate": 4.473176508864905e-05,
      "loss": 0.784,
      "num_input_tokens_seen": 7211464,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 0.44163885712623596,
      "learning_rate": 4.4724741853498995e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 7214472,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 0.4182155728340149,
      "learning_rate": 4.471771449215797e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 7218184,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 0.3218960464000702,
      "learning_rate": 4.471068300609603e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 7221480,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 0.36404234170913696,
      "learning_rate": 4.4703647396784065e-05,
      "loss": 0.3584,
      "num_input_tokens_seen": 7224424,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 0.43543344736099243,
      "learning_rate": 4.469660766569387e-05,
      "loss": 0.5039,
      "num_input_tokens_seen": 7227144,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 0.4799894392490387,
      "learning_rate": 4.4689563814298057e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 7229768,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 0.9861921072006226,
      "learning_rate": 4.468251584407012e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 7233352,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 0.558486819267273,
      "learning_rate": 4.467546375648443e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 7236328,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 0.635134756565094,
      "learning_rate": 4.4668407553016176e-05,
      "loss": 0.5913,
      "num_input_tokens_seen": 7238952,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 0.3108898997306824,
      "learning_rate": 4.466134723514146e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 7242568,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 1.0516692399978638,
      "learning_rate": 4.465428280433721e-05,
      "loss": 0.5824,
      "num_input_tokens_seen": 7245320,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 0.5919654369354248,
      "learning_rate": 4.464721426208123e-05,
      "loss": 0.398,
      "num_input_tokens_seen": 7248264,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 0.6066204905509949,
      "learning_rate": 4.464014160985217e-05,
      "loss": 0.5969,
      "num_input_tokens_seen": 7251176,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 0.636777937412262,
      "learning_rate": 4.463306484912955e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 7253864,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 0.5669050812721252,
      "learning_rate": 4.462598398139375e-05,
      "loss": 0.3689,
      "num_input_tokens_seen": 7257704,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 0.41383153200149536,
      "learning_rate": 4.4618899008126005e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 7260936,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.3433280289173126,
      "learning_rate": 4.461180993080841e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 7264456,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 0.5879329442977905,
      "learning_rate": 4.460471675092393e-05,
      "loss": 0.6751,
      "num_input_tokens_seen": 7267624,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 0.24187609553337097,
      "learning_rate": 4.4597619469956356e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 7270760,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 0.40804487466812134,
      "learning_rate": 4.459051808939038e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 7273608,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 0.3435603380203247,
      "learning_rate": 4.458341261071151e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 7276264,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 0.27170029282569885,
      "learning_rate": 4.457630303540615e-05,
      "loss": 0.4639,
      "num_input_tokens_seen": 7279336,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.2945544123649597,
      "learning_rate": 4.456918936496154e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 7282376,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 0.3231692314147949,
      "learning_rate": 4.456207160086575e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 7285448,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 0.6032158136367798,
      "learning_rate": 4.4554949744607775e-05,
      "loss": 0.5298,
      "num_input_tokens_seen": 7288648,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 0.6798502206802368,
      "learning_rate": 4.4547823797677405e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 7291432,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 0.42502352595329285,
      "learning_rate": 4.45406937615653e-05,
      "loss": 0.3084,
      "num_input_tokens_seen": 7296744,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 0.5432879328727722,
      "learning_rate": 4.4533559637763006e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 7299496,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 0.36720162630081177,
      "learning_rate": 4.452642142776288e-05,
      "loss": 0.4523,
      "num_input_tokens_seen": 7302440,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 0.32751932740211487,
      "learning_rate": 4.451927913305816e-05,
      "loss": 0.346,
      "num_input_tokens_seen": 7305864,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 0.44971442222595215,
      "learning_rate": 4.4512132755142935e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 7309096,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 0.64851313829422,
      "learning_rate": 4.450498229551214e-05,
      "loss": 0.6404,
      "num_input_tokens_seen": 7312648,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 0.4524354040622711,
      "learning_rate": 4.449782775566158e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 7316264,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 0.24564185738563538,
      "learning_rate": 4.449066913708789e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 7320584,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 0.27808070182800293,
      "learning_rate": 4.448350644128858e-05,
      "loss": 0.6671,
      "num_input_tokens_seen": 7323880,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 0.7841021418571472,
      "learning_rate": 4.4476339669762e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 7326984,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 0.5025445818901062,
      "learning_rate": 4.446916882400737e-05,
      "loss": 0.4416,
      "num_input_tokens_seen": 7330504,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 0.34049350023269653,
      "learning_rate": 4.446199390552472e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 7333480,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 0.39351657032966614,
      "learning_rate": 4.445481491581499e-05,
      "loss": 0.3735,
      "num_input_tokens_seen": 7336232,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 0.48444241285324097,
      "learning_rate": 4.444763185637993e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 7338920,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 0.42310845851898193,
      "learning_rate": 4.4440444728722166e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 7341512,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 0.6065847277641296,
      "learning_rate": 4.4433253534345136e-05,
      "loss": 0.5826,
      "num_input_tokens_seen": 7345128,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 0.4120466411113739,
      "learning_rate": 4.442605827475318e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 7348328,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 0.4678307771682739,
      "learning_rate": 4.4418858951451457e-05,
      "loss": 0.6428,
      "num_input_tokens_seen": 7351336,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 0.649122416973114,
      "learning_rate": 4.441165556594599e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 7355208,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 0.3931877613067627,
      "learning_rate": 4.440444811974363e-05,
      "loss": 0.3688,
      "num_input_tokens_seen": 7357768,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 0.7854993939399719,
      "learning_rate": 4.43972366143521e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 7360808,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 0.5055750608444214,
      "learning_rate": 4.4390021051279975e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 7363208,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 0.3381301462650299,
      "learning_rate": 4.438280143203665e-05,
      "loss": 0.3855,
      "num_input_tokens_seen": 7366504,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 0.611066460609436,
      "learning_rate": 4.43755777581324e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 7369800,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 0.49591076374053955,
      "learning_rate": 4.4368350031078334e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 7372680,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 0.3418109714984894,
      "learning_rate": 4.43611182523864e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 7377480,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 0.8753986358642578,
      "learning_rate": 4.4353882423569416e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 7380456,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 0.5564394593238831,
      "learning_rate": 4.434664254614103e-05,
      "loss": 0.546,
      "num_input_tokens_seen": 7383208,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 0.49683842062950134,
      "learning_rate": 4.4339398621615734e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 7385576,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 0.4620334208011627,
      "learning_rate": 4.433215065150889e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 7388296,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 0.4360470771789551,
      "learning_rate": 4.4324898637336674e-05,
      "loss": 0.5831,
      "num_input_tokens_seen": 7391784,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 0.3770003020763397,
      "learning_rate": 4.431764258061615e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 7395016,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 0.26161158084869385,
      "learning_rate": 4.431038248286517e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 7398856,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 0.4619881510734558,
      "learning_rate": 4.430311834560248e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 7401864,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 0.5417146682739258,
      "learning_rate": 4.4295850170347664e-05,
      "loss": 0.6317,
      "num_input_tokens_seen": 7404360,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 0.44018885493278503,
      "learning_rate": 4.428857795862113e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 7407464,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 0.5232911705970764,
      "learning_rate": 4.428130171194414e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 7411144,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 0.37441200017929077,
      "learning_rate": 4.427402143183879e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 7414376,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 0.2989850640296936,
      "learning_rate": 4.426673711982806e-05,
      "loss": 0.562,
      "num_input_tokens_seen": 7417672,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 0.5610741376876831,
      "learning_rate": 4.425944877743573e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 7420392,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 0.2753271758556366,
      "learning_rate": 4.425215640618644e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 7423880,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 0.33285388350486755,
      "learning_rate": 4.424486000760567e-05,
      "loss": 0.5278,
      "num_input_tokens_seen": 7427624,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 0.5545316338539124,
      "learning_rate": 4.423755958321974e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 7430536,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 0.41084024310112,
      "learning_rate": 4.423025513455582e-05,
      "loss": 0.5879,
      "num_input_tokens_seen": 7433192,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 0.45561328530311584,
      "learning_rate": 4.422294666314192e-05,
      "loss": 0.4453,
      "num_input_tokens_seen": 7436104,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 0.6601157784461975,
      "learning_rate": 4.421563417050688e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 7439432,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 0.3591569662094116,
      "learning_rate": 4.42083176581804e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 7441960,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 0.2650398910045624,
      "learning_rate": 4.4200997127693e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 7445576,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 0.3922121524810791,
      "learning_rate": 4.419367258057606e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 7448104,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 0.4156076908111572,
      "learning_rate": 4.418634401836179e-05,
      "loss": 0.5135,
      "num_input_tokens_seen": 7451976,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 0.3042217493057251,
      "learning_rate": 4.417901144258324e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 7456744,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 0.4133911728858948,
      "learning_rate": 4.4171674854774305e-05,
      "loss": 0.5588,
      "num_input_tokens_seen": 7460392,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 0.3633866012096405,
      "learning_rate": 4.416433425646971e-05,
      "loss": 0.4972,
      "num_input_tokens_seen": 7463496,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 0.6754757165908813,
      "learning_rate": 4.415698964920503e-05,
      "loss": 0.616,
      "num_input_tokens_seen": 7466376,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.4613696038722992,
      "eval_runtime": 9.7537,
      "eval_samples_per_second": 86.941,
      "eval_steps_per_second": 21.735,
      "num_input_tokens_seen": 7467968,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 0.46731674671173096,
      "learning_rate": 4.414964103451667e-05,
      "loss": 0.5362,
      "num_input_tokens_seen": 7468896,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 0.3425073027610779,
      "learning_rate": 4.414228841394187e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 7472640,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 0.4460095763206482,
      "learning_rate": 4.413493178901872e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 7476192,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 0.42639511823654175,
      "learning_rate": 4.412757116128614e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 7479712,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 0.35483381152153015,
      "learning_rate": 4.41202065322839e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 7482816,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 0.5004361271858215,
      "learning_rate": 4.411283790355259e-05,
      "loss": 0.5557,
      "num_input_tokens_seen": 7485760,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 0.4614851176738739,
      "learning_rate": 4.410546527663364e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 7489440,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 0.5156275033950806,
      "learning_rate": 4.409808865306931e-05,
      "loss": 0.2989,
      "num_input_tokens_seen": 7492704,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 0.4877435863018036,
      "learning_rate": 4.409070803440273e-05,
      "loss": 0.4963,
      "num_input_tokens_seen": 7495776,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 0.4069351553916931,
      "learning_rate": 4.4083323422177824e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 7498560,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 0.3737350404262543,
      "learning_rate": 4.407593481793937e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 7500992,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 0.6092632412910461,
      "learning_rate": 4.406854222323298e-05,
      "loss": 0.5726,
      "num_input_tokens_seen": 7504192,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 0.42058199644088745,
      "learning_rate": 4.406114563960511e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 7507552,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 0.7371225953102112,
      "learning_rate": 4.4053745068603026e-05,
      "loss": 0.5319,
      "num_input_tokens_seen": 7510592,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.6026697754859924,
      "learning_rate": 4.404634051177485e-05,
      "loss": 0.5537,
      "num_input_tokens_seen": 7513664,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 1.0665524005889893,
      "learning_rate": 4.403893197066953e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 7516320,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 0.6467530727386475,
      "learning_rate": 4.4031519446836855e-05,
      "loss": 0.3938,
      "num_input_tokens_seen": 7518944,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 0.6504625678062439,
      "learning_rate": 4.402410294182743e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 7521376,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 0.24361100792884827,
      "learning_rate": 4.4016682457192714e-05,
      "loss": 0.4659,
      "num_input_tokens_seen": 7526432,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 0.19612212479114532,
      "learning_rate": 4.400925799448497e-05,
      "loss": 0.3668,
      "num_input_tokens_seen": 7532064,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 0.5511358976364136,
      "learning_rate": 4.400182955525733e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 7535360,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 0.2790357172489166,
      "learning_rate": 4.3994397141063734e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 7539232,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 0.5935273766517639,
      "learning_rate": 4.398696075345896e-05,
      "loss": 0.5777,
      "num_input_tokens_seen": 7542464,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 0.5930413007736206,
      "learning_rate": 4.39795203939986e-05,
      "loss": 0.3851,
      "num_input_tokens_seen": 7546176,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 0.45021793246269226,
      "learning_rate": 4.397207606423911e-05,
      "loss": 0.5821,
      "num_input_tokens_seen": 7549696,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 0.25226861238479614,
      "learning_rate": 4.396462776573775e-05,
      "loss": 0.3718,
      "num_input_tokens_seen": 7552544,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 0.4769577383995056,
      "learning_rate": 4.395717550005262e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 7555168,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 0.47916972637176514,
      "learning_rate": 4.394971926874265e-05,
      "loss": 0.651,
      "num_input_tokens_seen": 7558528,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 0.41103777289390564,
      "learning_rate": 4.3942259073367604e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 7561760,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 0.331319123506546,
      "learning_rate": 4.3934794915488065e-05,
      "loss": 0.566,
      "num_input_tokens_seen": 7565312,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 0.4393236041069031,
      "learning_rate": 4.392732679666544e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 7568640,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 0.4589610695838928,
      "learning_rate": 4.3919854718461986e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 7572096,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 0.631891131401062,
      "learning_rate": 4.391237868244078e-05,
      "loss": 0.4989,
      "num_input_tokens_seen": 7574944,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 0.8178549408912659,
      "learning_rate": 4.390489869016571e-05,
      "loss": 0.6623,
      "num_input_tokens_seen": 7578080,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 0.9634783267974854,
      "learning_rate": 4.389741474320152e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 7580128,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 0.5428584814071655,
      "learning_rate": 4.388992684311376e-05,
      "loss": 0.5069,
      "num_input_tokens_seen": 7583968,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 1.8065698146820068,
      "learning_rate": 4.388243499146881e-05,
      "loss": 0.4164,
      "num_input_tokens_seen": 7586368,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 0.390633761882782,
      "learning_rate": 4.3874939189833876e-05,
      "loss": 0.4609,
      "num_input_tokens_seen": 7589728,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 0.5294121503829956,
      "learning_rate": 4.3867439439777005e-05,
      "loss": 0.5844,
      "num_input_tokens_seen": 7592224,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 0.33753636479377747,
      "learning_rate": 4.385993574286705e-05,
      "loss": 0.6494,
      "num_input_tokens_seen": 7595648,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 0.46170586347579956,
      "learning_rate": 4.3852428100673705e-05,
      "loss": 0.3708,
      "num_input_tokens_seen": 7598784,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 0.48424404859542847,
      "learning_rate": 4.3844916514767484e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 7602112,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.645930826663971,
      "learning_rate": 4.383740098671972e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 7608768,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 0.46666419506073,
      "learning_rate": 4.3829881518102576e-05,
      "loss": 0.5905,
      "num_input_tokens_seen": 7612224,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 0.3954099416732788,
      "learning_rate": 4.3822358110489045e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 7616768,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 0.42562273144721985,
      "learning_rate": 4.381483076545294e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 7619840,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 0.33315712213516235,
      "learning_rate": 4.380729948456888e-05,
      "loss": 0.4961,
      "num_input_tokens_seen": 7623104,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 0.3928425908088684,
      "learning_rate": 4.3799764269412336e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 7625856,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 0.4207104444503784,
      "learning_rate": 4.379222512155958e-05,
      "loss": 0.3474,
      "num_input_tokens_seen": 7629600,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 0.5629008412361145,
      "learning_rate": 4.378468204258773e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 7632960,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 0.5307437777519226,
      "learning_rate": 4.37771350340747e-05,
      "loss": 0.4989,
      "num_input_tokens_seen": 7635840,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 0.3072602450847626,
      "learning_rate": 4.376958409759925e-05,
      "loss": 0.6126,
      "num_input_tokens_seen": 7639232,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 0.3631346523761749,
      "learning_rate": 4.376202923474093e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 7645888,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 0.4557037651538849,
      "learning_rate": 4.3754470447080145e-05,
      "loss": 0.4074,
      "num_input_tokens_seen": 7648608,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 0.6826564073562622,
      "learning_rate": 4.374690773619811e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 7651072,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 0.41592565178871155,
      "learning_rate": 4.373934110367685e-05,
      "loss": 0.6605,
      "num_input_tokens_seen": 7654304,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 0.26721909642219543,
      "learning_rate": 4.373177055109922e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 7657568,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 0.34522923827171326,
      "learning_rate": 4.3724196080048896e-05,
      "loss": 0.4032,
      "num_input_tokens_seen": 7661056,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 0.4484800398349762,
      "learning_rate": 4.3716617692110364e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 7667360,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 0.4151390790939331,
      "learning_rate": 4.3709035388868943e-05,
      "loss": 0.413,
      "num_input_tokens_seen": 7671200,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 0.5231016874313354,
      "learning_rate": 4.3701449171910767e-05,
      "loss": 0.6568,
      "num_input_tokens_seen": 7674144,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 1.0030925273895264,
      "learning_rate": 4.3693859042822774e-05,
      "loss": 0.4327,
      "num_input_tokens_seen": 7677408,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 0.43490180373191833,
      "learning_rate": 4.368626500319275e-05,
      "loss": 0.6065,
      "num_input_tokens_seen": 7680160,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 0.4334118366241455,
      "learning_rate": 4.367866705460927e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 7682848,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 0.49811893701553345,
      "learning_rate": 4.367106519866173e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 7685728,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 0.3410610258579254,
      "learning_rate": 4.366345943694037e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 7688960,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 0.5306428670883179,
      "learning_rate": 4.365584977103622e-05,
      "loss": 0.5254,
      "num_input_tokens_seen": 7691872,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 0.6740232110023499,
      "learning_rate": 4.364823620254113e-05,
      "loss": 0.5255,
      "num_input_tokens_seen": 7694336,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 0.5126869082450867,
      "learning_rate": 4.364061873304778e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 7697280,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 0.4917711615562439,
      "learning_rate": 4.363299736414965e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 7700640,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 0.7574264407157898,
      "learning_rate": 4.3625372097441045e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 7704064,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 0.4429014027118683,
      "learning_rate": 4.3617742934517094e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 7707712,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 0.37228310108184814,
      "learning_rate": 4.361010987697373e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 7710656,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 0.3979887068271637,
      "learning_rate": 4.360247292640768e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 7713664,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 0.8370969295501709,
      "learning_rate": 4.359483208441653e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 7716000,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 0.410439133644104,
      "learning_rate": 4.3587187352598646e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 7718432,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 0.7350664138793945,
      "learning_rate": 4.357953873255323e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 7721216,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 0.3498035669326782,
      "learning_rate": 4.357188622588027e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 7724128,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 0.5745529532432556,
      "learning_rate": 4.356422983418059e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 7727744,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 0.562511146068573,
      "learning_rate": 4.3556569559055824e-05,
      "loss": 0.5525,
      "num_input_tokens_seen": 7730880,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 0.8052176237106323,
      "learning_rate": 4.3548905402108406e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 7734048,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 0.7872975468635559,
      "learning_rate": 4.35412373649416e-05,
      "loss": 0.4081,
      "num_input_tokens_seen": 7736832,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 0.3544672131538391,
      "learning_rate": 4.353356544915947e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 7739808,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 0.4887843430042267,
      "learning_rate": 4.352588965636689e-05,
      "loss": 0.427,
      "num_input_tokens_seen": 7743072,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 0.5043343305587769,
      "learning_rate": 4.351820998816954e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 7746336,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 0.492923378944397,
      "learning_rate": 4.3510526446173935e-05,
      "loss": 0.5319,
      "num_input_tokens_seen": 7748864,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.5083150863647461,
      "learning_rate": 4.350283903198737e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 7751424,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 0.48260068893432617,
      "learning_rate": 4.349514774721797e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 7754272,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 0.5200390219688416,
      "learning_rate": 4.348745259347467e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 7757920,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 0.4518603980541229,
      "learning_rate": 4.347975357236721e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 7761792,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 0.6651845574378967,
      "learning_rate": 4.347205068550612e-05,
      "loss": 0.3689,
      "num_input_tokens_seen": 7764832,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 0.32972270250320435,
      "learning_rate": 4.3464343934502774e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 7768640,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 0.33571547269821167,
      "learning_rate": 4.345663332096932e-05,
      "loss": 0.3491,
      "num_input_tokens_seen": 7771360,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 0.3866909146308899,
      "learning_rate": 4.344891884651875e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 7775104,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 0.6324637532234192,
      "learning_rate": 4.344120051276483e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 7782720,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.5565130710601807,
      "learning_rate": 4.3433478321322155e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 7785568,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 0.4194462299346924,
      "learning_rate": 4.3425752273806106e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 7789248,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 0.718464732170105,
      "learning_rate": 4.34180223718329e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 7791744,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 0.36915305256843567,
      "learning_rate": 4.341028861701953e-05,
      "loss": 0.6283,
      "num_input_tokens_seen": 7795008,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 0.3777995705604553,
      "learning_rate": 4.3402551010983825e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 7798144,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 0.5096555948257446,
      "learning_rate": 4.33948095553444e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 7800608,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 0.619767427444458,
      "learning_rate": 4.338706425172067e-05,
      "loss": 0.6003,
      "num_input_tokens_seen": 7804000,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 0.3416856825351715,
      "learning_rate": 4.3379315101732866e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 7807296,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 0.5600833892822266,
      "learning_rate": 4.3371562107002026e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 7810816,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 0.2440904974937439,
      "learning_rate": 4.3363805269149984e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 7814976,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 0.44818657636642456,
      "learning_rate": 4.335604458979939e-05,
      "loss": 0.7276,
      "num_input_tokens_seen": 7818656,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 0.5551496744155884,
      "learning_rate": 4.3348280070573686e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 7821664,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 0.3737507462501526,
      "learning_rate": 4.334051171309712e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 7824800,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 0.34409382939338684,
      "learning_rate": 4.333273951899474e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 7827456,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 0.42006993293762207,
      "learning_rate": 4.33249634898924e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 7832064,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 0.5008464455604553,
      "learning_rate": 4.331718362741677e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 7834976,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 0.5232035517692566,
      "learning_rate": 4.330939993319529e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 7837920,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 0.6010428071022034,
      "learning_rate": 4.330161240885623e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 7840448,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 0.4051438868045807,
      "learning_rate": 4.3293821056028654e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 7843680,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 0.5306525230407715,
      "learning_rate": 4.3286025876342415e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 7847264,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 0.45825380086898804,
      "learning_rate": 4.327822687142819e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 7850144,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 0.47073981165885925,
      "learning_rate": 4.327042404291742e-05,
      "loss": 0.4159,
      "num_input_tokens_seen": 7853696,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 0.5506303906440735,
      "learning_rate": 4.326261739244239e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 7856736,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 0.5068074464797974,
      "learning_rate": 4.325480692163615e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 7860640,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 0.4225921630859375,
      "learning_rate": 4.324699263213256e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 7863680,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 0.5656444430351257,
      "learning_rate": 4.3239174525566285e-05,
      "loss": 0.5896,
      "num_input_tokens_seen": 7866688,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 0.5852987766265869,
      "learning_rate": 4.3231352603572784e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 7869824,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 0.40117147564888,
      "learning_rate": 4.322352686778832e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 7872640,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 0.5569037199020386,
      "learning_rate": 4.3215697319849926e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 7875424,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 0.22291408479213715,
      "learning_rate": 4.3207863961395476e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 7879360,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 0.4979304075241089,
      "learning_rate": 4.320002679406361e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 7882144,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 0.40622618794441223,
      "learning_rate": 4.319218581949378e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 7886176,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 0.5741574764251709,
      "learning_rate": 4.318434103932622e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 7889760,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 0.4215192496776581,
      "learning_rate": 4.317649245520197e-05,
      "loss": 0.5698,
      "num_input_tokens_seen": 7892448,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 0.4420430362224579,
      "learning_rate": 4.3168640068762875e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 7896608,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 0.741730809211731,
      "learning_rate": 4.316078388165155e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 7899584,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 0.4393496811389923,
      "learning_rate": 4.315292389551143e-05,
      "loss": 0.5239,
      "num_input_tokens_seen": 7903008,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 0.5698391199111938,
      "learning_rate": 4.3145060111986726e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 7905696,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 0.35311299562454224,
      "learning_rate": 4.313719253272246e-05,
      "loss": 0.5351,
      "num_input_tokens_seen": 7908832,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 2.016782760620117,
      "learning_rate": 4.3129321159364436e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 7911520,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 0.4694729745388031,
      "learning_rate": 4.312144599355925e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 7914848,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 0.31642699241638184,
      "learning_rate": 4.31135670369543e-05,
      "loss": 0.4732,
      "num_input_tokens_seen": 7918400,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 0.5112200975418091,
      "learning_rate": 4.310568429119779e-05,
      "loss": 0.7433,
      "num_input_tokens_seen": 7921248,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 0.3695200979709625,
      "learning_rate": 4.309779775793868e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 7926240,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 0.3674316108226776,
      "learning_rate": 4.308990743882675e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 7930176,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 0.5144943594932556,
      "learning_rate": 4.308201333551256e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 7933120,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 0.44795313477516174,
      "learning_rate": 4.3074115449647476e-05,
      "loss": 0.566,
      "num_input_tokens_seen": 7935552,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 0.4316038191318512,
      "learning_rate": 4.306621378288364e-05,
      "loss": 0.3263,
      "num_input_tokens_seen": 7937824,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 0.3095933198928833,
      "learning_rate": 4.305830833687399e-05,
      "loss": 0.5804,
      "num_input_tokens_seen": 7940128,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 0.4949963688850403,
      "learning_rate": 4.305039911327225e-05,
      "loss": 0.3066,
      "num_input_tokens_seen": 7943008,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 1.0595272779464722,
      "learning_rate": 4.304248611373296e-05,
      "loss": 0.4139,
      "num_input_tokens_seen": 7945472,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 0.43826937675476074,
      "learning_rate": 4.30345693399114e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 7948608,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 0.48166146874427795,
      "learning_rate": 4.30266487934637e-05,
      "loss": 0.3682,
      "num_input_tokens_seen": 7951296,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 0.3054279088973999,
      "learning_rate": 4.3018724476046714e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 7955488,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 0.669287383556366,
      "learning_rate": 4.301079638931814e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 7958880,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 0.7113078236579895,
      "learning_rate": 4.3002864534936444e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 7961888,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 0.25115761160850525,
      "learning_rate": 4.2994928914560874e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 7964896,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 0.38602906465530396,
      "learning_rate": 4.298698952985147e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 7968032,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 0.43580982089042664,
      "learning_rate": 4.297904638246906e-05,
      "loss": 0.3718,
      "num_input_tokens_seen": 7971040,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 0.5367909073829651,
      "learning_rate": 4.297109947407527e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 7974560,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 1.2336186170578003,
      "learning_rate": 4.2963148806332486e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 7978144,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 0.5086808800697327,
      "learning_rate": 4.2955194380903915e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 7981024,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 0.27985280752182007,
      "learning_rate": 4.2947236199453524e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 7984448,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.5793603658676147,
      "learning_rate": 4.2939274263646065e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 7987520,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 0.49248644709587097,
      "learning_rate": 4.29313085751471e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 7990784,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 0.46234822273254395,
      "learning_rate": 4.292333913562294e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 7993792,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 0.24285727739334106,
      "learning_rate": 4.291536594674073e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 7996768,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 0.6620190143585205,
      "learning_rate": 4.2907389010168344e-05,
      "loss": 0.3685,
      "num_input_tokens_seen": 7999424,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 0.3275860548019409,
      "learning_rate": 4.2899408327574475e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 8003456,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 0.530758798122406,
      "learning_rate": 4.289142390062859e-05,
      "loss": 0.3924,
      "num_input_tokens_seen": 8007360,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 0.3286628723144531,
      "learning_rate": 4.2883435731000946e-05,
      "loss": 0.3336,
      "num_input_tokens_seen": 8010112,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 0.38322708010673523,
      "learning_rate": 4.287544382036256e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 8013024,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 0.5836008787155151,
      "learning_rate": 4.286744817038527e-05,
      "loss": 0.6016,
      "num_input_tokens_seen": 8015584,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 0.384931355714798,
      "learning_rate": 4.2859448782741666e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 8018720,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 0.3057899475097656,
      "learning_rate": 4.2851445659105126e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 8021184,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 0.6877753734588623,
      "learning_rate": 4.284343880114982e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 8024416,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 0.6099950075149536,
      "learning_rate": 4.2835428210550676e-05,
      "loss": 0.3762,
      "num_input_tokens_seen": 8028160,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 0.4215005040168762,
      "learning_rate": 4.282741388898344e-05,
      "loss": 0.3321,
      "num_input_tokens_seen": 8031296,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 0.6234862208366394,
      "learning_rate": 4.281939583812459e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 8034624,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 0.34712594747543335,
      "learning_rate": 4.281137405965143e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 8038176,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 0.3732239902019501,
      "learning_rate": 4.280334855524203e-05,
      "loss": 0.4697,
      "num_input_tokens_seen": 8041920,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 0.3595317304134369,
      "learning_rate": 4.279531932657521e-05,
      "loss": 0.4562,
      "num_input_tokens_seen": 8044992,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 0.6325646042823792,
      "learning_rate": 4.278728637533062e-05,
      "loss": 0.4711,
      "num_input_tokens_seen": 8048352,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 0.6046114563941956,
      "learning_rate": 4.2779249703188626e-05,
      "loss": 0.56,
      "num_input_tokens_seen": 8051168,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 0.48663532733917236,
      "learning_rate": 4.277120931183045e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 8054848,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 0.40872064232826233,
      "learning_rate": 4.276316520293803e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 8057952,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 0.26398152112960815,
      "learning_rate": 4.27551173781941e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 8061344,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 0.6225286722183228,
      "learning_rate": 4.274706583928216e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 8064704,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 0.46763846278190613,
      "learning_rate": 4.273901058788654e-05,
      "loss": 0.529,
      "num_input_tokens_seen": 8067360,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 0.5349946618080139,
      "learning_rate": 4.2730951625692275e-05,
      "loss": 0.5298,
      "num_input_tokens_seen": 8071520,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 0.4417414963245392,
      "learning_rate": 4.27228889543852e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 8075680,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 0.6660247445106506,
      "learning_rate": 4.2714822575651966e-05,
      "loss": 0.8658,
      "num_input_tokens_seen": 8077600,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 0.3988708555698395,
      "learning_rate": 4.270675249117994e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 8079872,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 0.7086073756217957,
      "learning_rate": 4.269867870265731e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 8082720,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 0.48917388916015625,
      "learning_rate": 4.269060121177301e-05,
      "loss": 0.5455,
      "num_input_tokens_seen": 8085728,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 0.4972490072250366,
      "learning_rate": 4.2682520020216764e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 8089280,
      "step": 12400
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 0.37634411454200745,
      "learning_rate": 4.267443512967906e-05,
      "loss": 0.4507,
      "num_input_tokens_seen": 8092576,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 0.44794991612434387,
      "learning_rate": 4.2666346541851165e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 8096032,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 0.6732062697410583,
      "learning_rate": 4.265825425842512e-05,
      "loss": 0.518,
      "num_input_tokens_seen": 8099424,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 0.5958544611930847,
      "learning_rate": 4.265015828109375e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 8101952,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 0.7991732358932495,
      "learning_rate": 4.264205861155063e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 8105472,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 0.5131320357322693,
      "learning_rate": 4.263395525149011e-05,
      "loss": 0.5778,
      "num_input_tokens_seen": 8108256,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 0.42655304074287415,
      "learning_rate": 4.262584820260734e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 8111008,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 0.5285961627960205,
      "learning_rate": 4.261773746659821e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 8114208,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 0.30648288130760193,
      "learning_rate": 4.260962304515939e-05,
      "loss": 0.3666,
      "num_input_tokens_seen": 8117376,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 0.6531763076782227,
      "learning_rate": 4.2601504939988335e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 8120256,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 1.3081309795379639,
      "learning_rate": 4.259338315278325e-05,
      "loss": 0.3095,
      "num_input_tokens_seen": 8122848,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 0.7243139147758484,
      "learning_rate": 4.258525768524313e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 8126336,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 0.3696107566356659,
      "learning_rate": 4.2577128539067725e-05,
      "loss": 0.293,
      "num_input_tokens_seen": 8129760,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 0.4023221433162689,
      "learning_rate": 4.256899571595756e-05,
      "loss": 0.4233,
      "num_input_tokens_seen": 8133248,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 0.6069809794425964,
      "learning_rate": 4.256085921761393e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 8136032,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 0.35023027658462524,
      "learning_rate": 4.25527190457389e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 8140224,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 0.250565767288208,
      "learning_rate": 4.2544575202035285e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 8143424,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 0.4950747787952423,
      "learning_rate": 4.253642768820671e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 8146336,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 1.1214911937713623,
      "learning_rate": 4.252827650595752e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 8148768,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 0.6601420044898987,
      "learning_rate": 4.252012165699286e-05,
      "loss": 0.3289,
      "num_input_tokens_seen": 8151840,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 0.3107975721359253,
      "learning_rate": 4.2511963143018634e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 8154560,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 0.3156600296497345,
      "learning_rate": 4.25038009657415e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 8158144,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 0.6246376037597656,
      "learning_rate": 4.24956351268689e-05,
      "loss": 0.5953,
      "num_input_tokens_seen": 8160672,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 0.3726406693458557,
      "learning_rate": 4.2487465628109036e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 8163744,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 0.3753552734851837,
      "learning_rate": 4.247929247117087e-05,
      "loss": 0.6609,
      "num_input_tokens_seen": 8167104,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 0.4844294488430023,
      "learning_rate": 4.247111565776413e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 8170304,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 0.6673533916473389,
      "learning_rate": 4.246293518959932e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 8173184,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 0.47817185521125793,
      "learning_rate": 4.245475106838769e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 8177472,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 0.548608660697937,
      "learning_rate": 4.2446563295841266e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 8180480,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 0.3044339716434479,
      "learning_rate": 4.243837187367285e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 8185088,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 0.2675422728061676,
      "learning_rate": 4.2430176803595987e-05,
      "loss": 0.5463,
      "num_input_tokens_seen": 8188384,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 0.7982851266860962,
      "learning_rate": 4.2421978087324985e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 8191360,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 0.45247381925582886,
      "learning_rate": 4.241377572657493e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 8194816,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 0.36305859684944153,
      "learning_rate": 4.2405569723061654e-05,
      "loss": 0.3433,
      "num_input_tokens_seen": 8198016,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 0.4404498040676117,
      "learning_rate": 4.239736007850178e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 8200448,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 0.4268653392791748,
      "learning_rate": 4.238914679461264e-05,
      "loss": 0.615,
      "num_input_tokens_seen": 8203488,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 0.7914480566978455,
      "learning_rate": 4.2380929873112386e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 8206464,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 0.6005890369415283,
      "learning_rate": 4.2372709315719896e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 8212928,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 0.35966193675994873,
      "learning_rate": 4.2364485124154806e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 8216128,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 1.5359331369400024,
      "learning_rate": 4.235625730013754e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 8218496,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 0.42908889055252075,
      "learning_rate": 4.234802584538926e-05,
      "loss": 0.3015,
      "num_input_tokens_seen": 8221824,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 0.7773085832595825,
      "learning_rate": 4.2339790761631884e-05,
      "loss": 0.3816,
      "num_input_tokens_seen": 8225376,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 0.4404299259185791,
      "learning_rate": 4.233155205058811e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 8228224,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 0.47415223717689514,
      "learning_rate": 4.232330971398137e-05,
      "loss": 0.5151,
      "num_input_tokens_seen": 8230656,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 0.42735016345977783,
      "learning_rate": 4.2315063753535875e-05,
      "loss": 0.391,
      "num_input_tokens_seen": 8233408,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 0.4288610517978668,
      "learning_rate": 4.230681417097658e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 8237216,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 0.5948435068130493,
      "learning_rate": 4.229856096802921e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 8240000,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 0.46136677265167236,
      "learning_rate": 4.229030414642024e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 8242848,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 0.537696897983551,
      "learning_rate": 4.2282043707876896e-05,
      "loss": 0.4132,
      "num_input_tokens_seen": 8246304,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 0.27490416169166565,
      "learning_rate": 4.2273779654127175e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 8249984,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 0.4466589689254761,
      "learning_rate": 4.226551198689982e-05,
      "loss": 0.5391,
      "num_input_tokens_seen": 8254240,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 0.422111839056015,
      "learning_rate": 4.225724070792434e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 8257312,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 0.26742470264434814,
      "learning_rate": 4.2248965818930975e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 8261216,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 0.4356735050678253,
      "learning_rate": 4.224068732165075e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 8264608,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 0.8508478403091431,
      "learning_rate": 4.223240521781543e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 8268384,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 0.32531243562698364,
      "learning_rate": 4.2224119509157535e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 8272000,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 0.42290419340133667,
      "learning_rate": 4.221583019741033e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 8275776,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 0.7489855885505676,
      "learning_rate": 4.2207537284307876e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 8279008,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 0.2724243104457855,
      "learning_rate": 4.219924077158493e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 8281952,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 0.3454616367816925,
      "learning_rate": 4.219094066097703e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 8285472,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 0.5293160676956177,
      "learning_rate": 4.218263695422047e-05,
      "loss": 0.5425,
      "num_input_tokens_seen": 8288416,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 0.2832769751548767,
      "learning_rate": 4.217432965305229e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 8292064,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 0.49651333689689636,
      "learning_rate": 4.2166018759210286e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 8295008,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3113126754760742,
      "learning_rate": 4.215770427443299e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 8298432,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 0.3376055359840393,
      "learning_rate": 4.214938620045971e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 8302240,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 0.41986826062202454,
      "learning_rate": 4.214106453903048e-05,
      "loss": 0.5395,
      "num_input_tokens_seen": 8304960,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 0.7486227750778198,
      "learning_rate": 4.2132739291886114e-05,
      "loss": 0.3693,
      "num_input_tokens_seen": 8308320,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 0.3825320303440094,
      "learning_rate": 4.212441046076814e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 8311360,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 0.6739636063575745,
      "learning_rate": 4.211607804741887e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 8315680,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 0.47481805086135864,
      "learning_rate": 4.2107742053581344e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 8319200,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 0.5454773902893066,
      "learning_rate": 4.209940248099935e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 8322816,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 0.4028136432170868,
      "learning_rate": 4.2091059331417445e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 8325408,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 0.28732720017433167,
      "learning_rate": 4.208271260658091e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 8329408,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 0.4487246870994568,
      "learning_rate": 4.2074362308235785e-05,
      "loss": 0.6012,
      "num_input_tokens_seen": 8332736,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 0.38174429535865784,
      "learning_rate": 4.206600843812886e-05,
      "loss": 0.305,
      "num_input_tokens_seen": 8338272,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 0.8325777649879456,
      "learning_rate": 4.205765099800767e-05,
      "loss": 0.6301,
      "num_input_tokens_seen": 8341472,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 0.553068220615387,
      "learning_rate": 4.204928998962049e-05,
      "loss": 0.4339,
      "num_input_tokens_seen": 8344288,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 0.31510183215141296,
      "learning_rate": 4.204092541471636e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 8347392,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 0.44395574927330017,
      "learning_rate": 4.203255727504504e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 8350656,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 0.3566124439239502,
      "learning_rate": 4.202418557235706e-05,
      "loss": 0.4421,
      "num_input_tokens_seen": 8353888,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 0.6729859709739685,
      "learning_rate": 4.201581030840368e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 8356736,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 0.39994189143180847,
      "learning_rate": 4.200743148493691e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 8360992,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 0.4246196746826172,
      "learning_rate": 4.199904910370951e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 8364000,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 0.6732097268104553,
      "learning_rate": 4.1990663166474964e-05,
      "loss": 0.4115,
      "num_input_tokens_seen": 8366304,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 0.34400475025177,
      "learning_rate": 4.1982273674987524e-05,
      "loss": 0.3778,
      "num_input_tokens_seen": 8369792,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 0.49062401056289673,
      "learning_rate": 4.197388063100218e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 8372384,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 0.44423186779022217,
      "learning_rate": 4.196548403627465e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 8374720,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 0.5262893438339233,
      "learning_rate": 4.195708389256142e-05,
      "loss": 0.5481,
      "num_input_tokens_seen": 8378144,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 0.47105512022972107,
      "learning_rate": 4.194868020161969e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 8381536,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 0.5256429314613342,
      "learning_rate": 4.1940272965207425e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 8384416,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 0.46273836493492126,
      "learning_rate": 4.1931862185083326e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 8387616,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 0.45271405577659607,
      "learning_rate": 4.192344786300682e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 8390208,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 0.4624634087085724,
      "learning_rate": 4.19150300007381e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 8393792,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 0.44851359724998474,
      "learning_rate": 4.190660860003806e-05,
      "loss": 0.5489,
      "num_input_tokens_seen": 8397120,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 0.6761364340782166,
      "learning_rate": 4.189818366266841e-05,
      "loss": 0.5761,
      "num_input_tokens_seen": 8400160,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 0.4310600757598877,
      "learning_rate": 4.188975519039151e-05,
      "loss": 0.6627,
      "num_input_tokens_seen": 8404032,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 0.4771769940853119,
      "learning_rate": 4.188132318497051e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 8407392,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 0.724538266658783,
      "learning_rate": 4.1872887648169304e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 8410656,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 1.4850014448165894,
      "learning_rate": 4.186444858175248e-05,
      "loss": 0.5369,
      "num_input_tokens_seen": 8413344,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 0.3146548867225647,
      "learning_rate": 4.1856005987485424e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 8416960,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 0.34599825739860535,
      "learning_rate": 4.1847559867134215e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 8420352,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 0.5531566143035889,
      "learning_rate": 4.1839110222465693e-05,
      "loss": 0.4943,
      "num_input_tokens_seen": 8423360,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 0.47680923342704773,
      "learning_rate": 4.183065705524742e-05,
      "loss": 0.299,
      "num_input_tokens_seen": 8426816,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 0.496258944272995,
      "learning_rate": 4.1822200367247696e-05,
      "loss": 0.5944,
      "num_input_tokens_seen": 8429856,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 0.5183811783790588,
      "learning_rate": 4.181374016023559e-05,
      "loss": 0.5988,
      "num_input_tokens_seen": 8433056,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 0.32410237193107605,
      "learning_rate": 4.1805276435980854e-05,
      "loss": 0.5218,
      "num_input_tokens_seen": 8436448,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 0.33269035816192627,
      "learning_rate": 4.1796809196254014e-05,
      "loss": 0.3302,
      "num_input_tokens_seen": 8439200,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 0.8316981196403503,
      "learning_rate": 4.178833844282632e-05,
      "loss": 0.4964,
      "num_input_tokens_seen": 8443072,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 0.29357996582984924,
      "learning_rate": 4.177986417746976e-05,
      "loss": 0.4008,
      "num_input_tokens_seen": 8445952,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 0.43920114636421204,
      "learning_rate": 4.177138640195704e-05,
      "loss": 0.581,
      "num_input_tokens_seen": 8448768,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 0.5210930705070496,
      "learning_rate": 4.176290511806162e-05,
      "loss": 0.3576,
      "num_input_tokens_seen": 8451232,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.5723350644111633,
      "learning_rate": 4.1754420327557706e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 8454688,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 0.5536741018295288,
      "learning_rate": 4.1745932032220195e-05,
      "loss": 0.3418,
      "num_input_tokens_seen": 8458048,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 0.5225355625152588,
      "learning_rate": 4.173744023382474e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 8460960,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 0.5334290862083435,
      "learning_rate": 4.172894493414776e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 8464448,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 0.3691686987876892,
      "learning_rate": 4.172044613496634e-05,
      "loss": 0.3389,
      "num_input_tokens_seen": 8468000,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 0.41245725750923157,
      "learning_rate": 4.171194383805835e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 8472288,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 0.5174728631973267,
      "learning_rate": 4.170343804520237e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 8475616,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 1.253652811050415,
      "learning_rate": 4.169492875817771e-05,
      "loss": 0.6254,
      "num_input_tokens_seen": 8477984,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 0.43688157200813293,
      "learning_rate": 4.1686415978764415e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 8480864,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 0.2645898759365082,
      "learning_rate": 4.167789970874327e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 8483680,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 0.7519309520721436,
      "learning_rate": 4.166937994989578e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 8487296,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 0.39130720496177673,
      "learning_rate": 4.166085670400418e-05,
      "loss": 0.4047,
      "num_input_tokens_seen": 8490528,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 0.5742249488830566,
      "learning_rate": 4.165232997285143e-05,
      "loss": 0.4678,
      "num_input_tokens_seen": 8493536,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 1.2293524742126465,
      "learning_rate": 4.164379975822124e-05,
      "loss": 0.5701,
      "num_input_tokens_seen": 8497472,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 0.8403027653694153,
      "learning_rate": 4.1635266061898015e-05,
      "loss": 0.5434,
      "num_input_tokens_seen": 8500288,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 0.5377358198165894,
      "learning_rate": 4.162672888566691e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 8503136,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 0.3187030255794525,
      "learning_rate": 4.1618188231313816e-05,
      "loss": 0.3146,
      "num_input_tokens_seen": 8507168,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 0.6617472767829895,
      "learning_rate": 4.1609644100625335e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 8509760,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 0.26866671442985535,
      "learning_rate": 4.16010964953888e-05,
      "loss": 0.3544,
      "num_input_tokens_seen": 8512672,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 0.4691186547279358,
      "learning_rate": 4.159254541739227e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 8516320,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 0.47980406880378723,
      "learning_rate": 4.158399086842454e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 8519456,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 1.2725461721420288,
      "learning_rate": 4.157543285027512e-05,
      "loss": 0.5523,
      "num_input_tokens_seen": 8522560,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 0.4321889579296112,
      "learning_rate": 4.156687136473425e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 8525792,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 0.6120133996009827,
      "learning_rate": 4.1558306413592894e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 8528704,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 0.8495357036590576,
      "learning_rate": 4.1549737998642745e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 8532000,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 0.49237802624702454,
      "learning_rate": 4.154116612167622e-05,
      "loss": 0.5355,
      "num_input_tokens_seen": 8535136,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 0.6927123069763184,
      "learning_rate": 4.1532590784486444e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 8538592,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 0.3776777684688568,
      "learning_rate": 4.1524011988867294e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 8541344,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 0.5019674897193909,
      "learning_rate": 4.151542973661336e-05,
      "loss": 0.4902,
      "num_input_tokens_seen": 8544576,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 0.420747846364975,
      "learning_rate": 4.150684402951994e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 8547680,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 0.39264801144599915,
      "learning_rate": 4.149825486938306e-05,
      "loss": 0.5591,
      "num_input_tokens_seen": 8550720,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 0.6725826859474182,
      "learning_rate": 4.148966225799949e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 8554112,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 0.6637663245201111,
      "learning_rate": 4.14810661971667e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 8557856,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 0.4325997233390808,
      "learning_rate": 4.147246668868289e-05,
      "loss": 0.4947,
      "num_input_tokens_seen": 8560960,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 0.7820632457733154,
      "learning_rate": 4.1463863734346964e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 8564288,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 0.26207852363586426,
      "learning_rate": 4.145525733595859e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 8568608,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 0.39044952392578125,
      "learning_rate": 4.14466474953181e-05,
      "loss": 0.4447,
      "num_input_tokens_seen": 8571872,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 0.5244962573051453,
      "learning_rate": 4.14380342142266e-05,
      "loss": 0.5709,
      "num_input_tokens_seen": 8574336,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 0.3425380289554596,
      "learning_rate": 4.142941749448587e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 8577248,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 0.39025330543518066,
      "learning_rate": 4.142079733789843e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 8581120,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 0.7314925789833069,
      "learning_rate": 4.141217374626753e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 8585280,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 0.37422341108322144,
      "learning_rate": 4.140354672139713e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 8590464,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 0.30290621519088745,
      "learning_rate": 4.1394916265091886e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 8593408,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 0.40472960472106934,
      "learning_rate": 4.1386282379157214e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 8596832,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 0.4077950716018677,
      "learning_rate": 4.137764506539921e-05,
      "loss": 0.3437,
      "num_input_tokens_seen": 8600000,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 0.932991623878479,
      "learning_rate": 4.136900432562471e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 8602560,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 0.5453707575798035,
      "learning_rate": 4.136036016164123e-05,
      "loss": 0.5902,
      "num_input_tokens_seen": 8605088,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 0.5105790495872498,
      "learning_rate": 4.135171257525707e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 8608704,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 0.32821154594421387,
      "learning_rate": 4.1343061568281195e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 8612320,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 0.3314850926399231,
      "learning_rate": 4.133440714252328e-05,
      "loss": 0.5126,
      "num_input_tokens_seen": 8615360,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 0.3661743402481079,
      "learning_rate": 4.132574929979375e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 8618464,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 0.3536818325519562,
      "learning_rate": 4.1317088041903714e-05,
      "loss": 0.2953,
      "num_input_tokens_seen": 8621664,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 0.3248482048511505,
      "learning_rate": 4.1308423370665015e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 8624608,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 0.6645115613937378,
      "learning_rate": 4.1299755287890206e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 8626880,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 0.5405120849609375,
      "learning_rate": 4.129108379539255e-05,
      "loss": 0.5391,
      "num_input_tokens_seen": 8630880,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 0.5757979154586792,
      "learning_rate": 4.128240889498602e-05,
      "loss": 0.5192,
      "num_input_tokens_seen": 8633856,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 0.34307700395584106,
      "learning_rate": 4.127373058848531e-05,
      "loss": 0.5326,
      "num_input_tokens_seen": 8636288,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 0.797544002532959,
      "learning_rate": 4.126504887770582e-05,
      "loss": 0.651,
      "num_input_tokens_seen": 8638880,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 0.5105151534080505,
      "learning_rate": 4.1256363764463666e-05,
      "loss": 0.5098,
      "num_input_tokens_seen": 8642080,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 0.45343920588493347,
      "learning_rate": 4.1247675250575676e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 8644672,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 0.4749299883842468,
      "learning_rate": 4.1238983337859396e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 8648160,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 0.28340786695480347,
      "learning_rate": 4.123028802813305e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 8651168,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 0.2842739522457123,
      "learning_rate": 4.1221589323215625e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 8653984,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 0.4640563726425171,
      "learning_rate": 4.121288722492678e-05,
      "loss": 0.3821,
      "num_input_tokens_seen": 8656544,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 0.5005942583084106,
      "learning_rate": 4.120418173508688e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 8659456,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 0.45013707876205444,
      "learning_rate": 4.1195472855517045e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 8662112,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 0.6803982257843018,
      "learning_rate": 4.1186760588039045e-05,
      "loss": 0.5523,
      "num_input_tokens_seen": 8665760,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 0.3886103332042694,
      "learning_rate": 4.11780449344754e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 8669152,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 0.7929462194442749,
      "learning_rate": 4.1169325896649326e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 8671648,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 0.4052620232105255,
      "learning_rate": 4.116060347638474e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 8676192,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 0.5486326813697815,
      "learning_rate": 4.115187767550628e-05,
      "loss": 0.5771,
      "num_input_tokens_seen": 8679008,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 0.4292070269584656,
      "learning_rate": 4.114314849583928e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 8682848,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.6908094882965088,
      "learning_rate": 4.113441593920979e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 8685376,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 0.4651138484477997,
      "learning_rate": 4.112568000744454e-05,
      "loss": 0.6518,
      "num_input_tokens_seen": 8688672,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 0.3644556403160095,
      "learning_rate": 4.1116940702371025e-05,
      "loss": 0.5105,
      "num_input_tokens_seen": 8692256,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 0.5344982743263245,
      "learning_rate": 4.1108198025817376e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 8695968,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 0.4860595762729645,
      "learning_rate": 4.1099451979612476e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 8699808,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 0.5209257006645203,
      "learning_rate": 4.109070256558589e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 8702432,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 0.5576045513153076,
      "learning_rate": 4.108194978556791e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 8705792,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 0.5440188646316528,
      "learning_rate": 4.1073193641389495e-05,
      "loss": 0.487,
      "num_input_tokens_seen": 8709056,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.45847615599632263,
      "eval_runtime": 9.6512,
      "eval_samples_per_second": 87.865,
      "eval_steps_per_second": 21.966,
      "num_input_tokens_seen": 8709168,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 0.21878275275230408,
      "learning_rate": 4.1064434134882345e-05,
      "loss": 0.5027,
      "num_input_tokens_seen": 8711536,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 0.5198709964752197,
      "learning_rate": 4.1055671267878847e-05,
      "loss": 0.5555,
      "num_input_tokens_seen": 8715728,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 0.3765111267566681,
      "learning_rate": 4.1046905042212106e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 8719216,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 0.3829338252544403,
      "learning_rate": 4.103813545971589e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 8722160,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 0.5010459423065186,
      "learning_rate": 4.10293625222247e-05,
      "loss": 0.348,
      "num_input_tokens_seen": 8725584,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 0.550374448299408,
      "learning_rate": 4.102058623157376e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 8728272,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 0.3623282015323639,
      "learning_rate": 4.101180658959894e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 8731568,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 0.7785825729370117,
      "learning_rate": 4.1003023598136866e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 8735120,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 0.4469172954559326,
      "learning_rate": 4.099423725902482e-05,
      "loss": 0.3516,
      "num_input_tokens_seen": 8737552,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 0.4710206389427185,
      "learning_rate": 4.098544757410081e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 8741712,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 0.4781487286090851,
      "learning_rate": 4.097665454520354e-05,
      "loss": 0.5275,
      "num_input_tokens_seen": 8744464,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 0.29384881258010864,
      "learning_rate": 4.096785817417241e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 8748048,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 0.43638062477111816,
      "learning_rate": 4.0959058462847524e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 8751024,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 0.5962733030319214,
      "learning_rate": 4.095025541306968e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 8753264,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 0.49656257033348083,
      "learning_rate": 4.0941449026680374e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 8756240,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 0.7355381846427917,
      "learning_rate": 4.09326393055218e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 8759184,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 0.41670605540275574,
      "learning_rate": 4.0923826251436845e-05,
      "loss": 0.5748,
      "num_input_tokens_seen": 8762864,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 0.8834871053695679,
      "learning_rate": 4.0915009866269115e-05,
      "loss": 0.3456,
      "num_input_tokens_seen": 8767856,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 0.37281128764152527,
      "learning_rate": 4.090619015186288e-05,
      "loss": 0.4943,
      "num_input_tokens_seen": 8771344,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 0.24659943580627441,
      "learning_rate": 4.0897367110063145e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 8775216,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 0.3956364691257477,
      "learning_rate": 4.088854074271557e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 8779536,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 0.43620315194129944,
      "learning_rate": 4.0879711051666534e-05,
      "loss": 0.5827,
      "num_input_tokens_seen": 8782640,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 0.5197932720184326,
      "learning_rate": 4.087087803876313e-05,
      "loss": 0.668,
      "num_input_tokens_seen": 8786384,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 0.40127795934677124,
      "learning_rate": 4.086204170585308e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 8789616,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 0.6755456924438477,
      "learning_rate": 4.085320205478488e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 8792816,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 0.5130012035369873,
      "learning_rate": 4.084435908740768e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 8796784,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 0.4334022104740143,
      "learning_rate": 4.0835512805571306e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 8800016,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 0.428330659866333,
      "learning_rate": 4.082666321112633e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 8803088,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 0.5702440738677979,
      "learning_rate": 4.081781030592396e-05,
      "loss": 0.4119,
      "num_input_tokens_seen": 8806096,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 0.49229806661605835,
      "learning_rate": 4.080895409181613e-05,
      "loss": 0.5023,
      "num_input_tokens_seen": 8808624,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 0.4629845917224884,
      "learning_rate": 4.080009457065547e-05,
      "loss": 0.5924,
      "num_input_tokens_seen": 8811920,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 0.11265578120946884,
      "learning_rate": 4.0791231744295286e-05,
      "loss": 0.3008,
      "num_input_tokens_seen": 8817904,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 0.43044036626815796,
      "learning_rate": 4.078236561458957e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 8820496,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 0.4760361611843109,
      "learning_rate": 4.077349618339302e-05,
      "loss": 0.5016,
      "num_input_tokens_seen": 8823120,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 0.40594011545181274,
      "learning_rate": 4.076462345256103e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 8826192,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 0.3477127254009247,
      "learning_rate": 4.0755747423949655e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 8829456,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 0.42300501465797424,
      "learning_rate": 4.074686809941568e-05,
      "loss": 0.5842,
      "num_input_tokens_seen": 8832112,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 0.9129904508590698,
      "learning_rate": 4.0737985480816546e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 8834832,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 0.29823657870292664,
      "learning_rate": 4.0729099570010395e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 8838480,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 1.6324617862701416,
      "learning_rate": 4.072021036885607e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 8842128,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 0.5314137935638428,
      "learning_rate": 4.0711317879213075e-05,
      "loss": 0.3832,
      "num_input_tokens_seen": 8845872,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 0.3949635326862335,
      "learning_rate": 4.070242210294162e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 8848720,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 0.6575194001197815,
      "learning_rate": 4.069352304190261e-05,
      "loss": 0.3777,
      "num_input_tokens_seen": 8853072,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 0.32624000310897827,
      "learning_rate": 4.068462069795761e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 8856208,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 0.40938684344291687,
      "learning_rate": 4.067571507296891e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 8859600,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 0.6189245581626892,
      "learning_rate": 4.0666806168799455e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 8862576,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 0.4116719961166382,
      "learning_rate": 4.0657893987312876e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 8866704,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 0.6789384484291077,
      "learning_rate": 4.064897853037352e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 8869584,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 0.2979664206504822,
      "learning_rate": 4.064005979984639e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 8873040,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 1.1950459480285645,
      "learning_rate": 4.0631137797597176e-05,
      "loss": 0.5448,
      "num_input_tokens_seen": 8876208,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 0.332932710647583,
      "learning_rate": 4.0622212525492266e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 8880240,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 1.0529390573501587,
      "learning_rate": 4.061328398539872e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 8883280,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 0.24325694143772125,
      "learning_rate": 4.060435217918429e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 8886192,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 0.3996233344078064,
      "learning_rate": 4.0595417108717405e-05,
      "loss": 0.4088,
      "num_input_tokens_seen": 8889456,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 0.8337404131889343,
      "learning_rate": 4.0586478775867185e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 8892528,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 0.4995132088661194,
      "learning_rate": 4.057753718250343e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 8895824,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 0.5877164006233215,
      "learning_rate": 4.056859233049661e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 8898544,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 0.4038326144218445,
      "learning_rate": 4.05596442217179e-05,
      "loss": 0.5857,
      "num_input_tokens_seen": 8902480,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 0.5468617081642151,
      "learning_rate": 4.055069285803912e-05,
      "loss": 0.5688,
      "num_input_tokens_seen": 8905968,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 0.5541536211967468,
      "learning_rate": 4.0541738241332815e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 8909296,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 0.4718356132507324,
      "learning_rate": 4.053278037347217e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 8911792,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 0.5208099484443665,
      "learning_rate": 4.052381925633109e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 8915280,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 0.6407458186149597,
      "learning_rate": 4.051485489178412e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 8918736,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 0.6466467380523682,
      "learning_rate": 4.05058872817065e-05,
      "loss": 0.5294,
      "num_input_tokens_seen": 8921776,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.28220927715301514,
      "learning_rate": 4.0496916427974177e-05,
      "loss": 0.5484,
      "num_input_tokens_seen": 8925264,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 0.3971620500087738,
      "learning_rate": 4.048794233246373e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 8928400,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 0.33941298723220825,
      "learning_rate": 4.0478964997052435e-05,
      "loss": 0.5841,
      "num_input_tokens_seen": 8931792,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 0.3896082937717438,
      "learning_rate": 4.046998442361827e-05,
      "loss": 0.3801,
      "num_input_tokens_seen": 8934768,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 0.523908257484436,
      "learning_rate": 4.046100061403984e-05,
      "loss": 0.2599,
      "num_input_tokens_seen": 8937552,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 0.3669172525405884,
      "learning_rate": 4.045201357019648e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 8941520,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 0.3596925437450409,
      "learning_rate": 4.044302329396816e-05,
      "loss": 0.3299,
      "num_input_tokens_seen": 8945616,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 0.3637830913066864,
      "learning_rate": 4.0434029787235564e-05,
      "loss": 0.5364,
      "num_input_tokens_seen": 8948592,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 0.6042154431343079,
      "learning_rate": 4.042503305188e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 8951344,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 0.5794976353645325,
      "learning_rate": 4.041603308978351e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 8955056,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 0.3382394313812256,
      "learning_rate": 4.040702990282878e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 8958128,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 0.4872964322566986,
      "learning_rate": 4.039802349289914e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 8961840,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 0.3965621888637543,
      "learning_rate": 4.038901386187868e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 8965040,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 0.7539095282554626,
      "learning_rate": 4.038000101165208e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 8969200,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 0.7701829075813293,
      "learning_rate": 4.0370984944104725e-05,
      "loss": 0.6445,
      "num_input_tokens_seen": 8972624,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 0.4527011513710022,
      "learning_rate": 4.0361965661122695e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 8975728,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 0.3500555157661438,
      "learning_rate": 4.035294316459271e-05,
      "loss": 0.5946,
      "num_input_tokens_seen": 8979312,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 0.36695000529289246,
      "learning_rate": 4.034391745640215e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 8982032,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 1.586883306503296,
      "learning_rate": 4.033488853843913e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 8984432,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 0.46480223536491394,
      "learning_rate": 4.0325856412592363e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 8988624,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.7248356342315674,
      "learning_rate": 4.031682108075129e-05,
      "loss": 0.6825,
      "num_input_tokens_seen": 8991664,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 0.462067574262619,
      "learning_rate": 4.030778254480599e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 8994928,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 0.6331159472465515,
      "learning_rate": 4.029874080664722e-05,
      "loss": 0.6151,
      "num_input_tokens_seen": 8997872,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 0.3994855582714081,
      "learning_rate": 4.0289695868166414e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 9002384,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 0.9811415672302246,
      "learning_rate": 4.028064773125566e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 9005296,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 0.3279908299446106,
      "learning_rate": 4.0271596397807734e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 9009968,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 0.7277731895446777,
      "learning_rate": 4.026254186971607e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 9013776,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 0.45084673166275024,
      "learning_rate": 4.025348414887477e-05,
      "loss": 0.3789,
      "num_input_tokens_seen": 9017072,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 0.6336908936500549,
      "learning_rate": 4.024442323717861e-05,
      "loss": 0.7119,
      "num_input_tokens_seen": 9020272,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 0.3824273347854614,
      "learning_rate": 4.023535913652302e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 9024528,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 0.47585684061050415,
      "learning_rate": 4.022629184880412e-05,
      "loss": 0.4021,
      "num_input_tokens_seen": 9027376,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 0.6268630623817444,
      "learning_rate": 4.021722137591867e-05,
      "loss": 0.5179,
      "num_input_tokens_seen": 9030832,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 0.583100438117981,
      "learning_rate": 4.020814771976412e-05,
      "loss": 0.6022,
      "num_input_tokens_seen": 9034064,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 0.55560702085495,
      "learning_rate": 4.019907088223857e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 9037072,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 0.30405324697494507,
      "learning_rate": 4.018999086524079e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 9040624,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 0.5628478527069092,
      "learning_rate": 4.018090767067022e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 9043792,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 0.5966551303863525,
      "learning_rate": 4.017182130042696e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 9046640,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 1.024075984954834,
      "learning_rate": 4.0162731756411764e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 9049328,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 0.45790916681289673,
      "learning_rate": 4.015363904052608e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 9052944,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 0.511111855506897,
      "learning_rate": 4.014454315467199e-05,
      "loss": 0.3073,
      "num_input_tokens_seen": 9055344,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 0.5156336426734924,
      "learning_rate": 4.013544410075224e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 9058928,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 0.6606308817863464,
      "learning_rate": 4.0126341880670273e-05,
      "loss": 0.5551,
      "num_input_tokens_seen": 9062384,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 0.4921649396419525,
      "learning_rate": 4.011723649633014e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 9066288,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 0.4874810576438904,
      "learning_rate": 4.0108127949636606e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 9070000,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 0.6003185510635376,
      "learning_rate": 4.009901624249507e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 9072784,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 0.28625667095184326,
      "learning_rate": 4.008990137681159e-05,
      "loss": 0.3575,
      "num_input_tokens_seen": 9075632,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 0.6166829466819763,
      "learning_rate": 4.00807833544929e-05,
      "loss": 0.5865,
      "num_input_tokens_seen": 9078448,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 0.5978792309761047,
      "learning_rate": 4.0071662177446376e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 9082352,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 0.6701647043228149,
      "learning_rate": 4.006253784758008e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 9085104,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 0.6846621036529541,
      "learning_rate": 4.0053410366802704e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 9087856,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 0.3081233501434326,
      "learning_rate": 4.004427973702362e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 9091216,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 0.7183588147163391,
      "learning_rate": 4.003514596015285e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 9094256,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 0.22420364618301392,
      "learning_rate": 4.002600903810108e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 9098192,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 0.5283665657043457,
      "learning_rate": 4.001686897277963e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 9100432,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 0.3146721422672272,
      "learning_rate": 4.000772576610052e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 9104016,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 0.44752269983291626,
      "learning_rate": 3.99985794199764e-05,
      "loss": 0.859,
      "num_input_tokens_seen": 9107088,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 0.518165111541748,
      "learning_rate": 3.998942993632058e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 9110064,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 0.4017578065395355,
      "learning_rate": 3.998027731704702e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 9113040,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 0.40811437368392944,
      "learning_rate": 3.9971121564070354e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 9116208,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 0.4992024302482605,
      "learning_rate": 3.996196267930585e-05,
      "loss": 0.4495,
      "num_input_tokens_seen": 9119280,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 0.4546494781970978,
      "learning_rate": 3.995280066466945e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 9124048,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 0.3099794089794159,
      "learning_rate": 3.9943635522077746e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 9128080,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 0.37389227747917175,
      "learning_rate": 3.993446725344797e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 9131120,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 0.3196879029273987,
      "learning_rate": 3.992529586069803e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 9133776,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 0.3415301442146301,
      "learning_rate": 3.991612134574648e-05,
      "loss": 0.5284,
      "num_input_tokens_seen": 9137680,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 0.30571356415748596,
      "learning_rate": 3.99069437105125e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 9140912,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 0.696430504322052,
      "learning_rate": 3.9897762956915976e-05,
      "loss": 0.4405,
      "num_input_tokens_seen": 9144432,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 0.6054428815841675,
      "learning_rate": 3.9888579086877403e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 9147504,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 0.36066100001335144,
      "learning_rate": 3.987939210231794e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 9150256,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 0.4007929265499115,
      "learning_rate": 3.987020200515941e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 9153552,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 0.5087806582450867,
      "learning_rate": 3.9861008797324255e-05,
      "loss": 0.3497,
      "num_input_tokens_seen": 9156624,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 0.35188865661621094,
      "learning_rate": 3.9851812480735605e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 9159792,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.32587602734565735,
      "learning_rate": 3.9842613057317235e-05,
      "loss": 0.4182,
      "num_input_tokens_seen": 9162960,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 0.6465988755226135,
      "learning_rate": 3.9833410528993534e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 9166416,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 0.9037682414054871,
      "learning_rate": 3.9824204897689585e-05,
      "loss": 0.5513,
      "num_input_tokens_seen": 9169712,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 0.3950389325618744,
      "learning_rate": 3.98149961653311e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 9172208,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 0.6383299827575684,
      "learning_rate": 3.980578433384442e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 9174640,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 0.41329413652420044,
      "learning_rate": 3.979656940515659e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 9178416,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 0.427664577960968,
      "learning_rate": 3.9787351381195235e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 9181040,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 0.4140705168247223,
      "learning_rate": 3.977813026388868e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 9184048,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 0.4504900276660919,
      "learning_rate": 3.976890605516587e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 9187344,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 0.5423492193222046,
      "learning_rate": 3.97596787569564e-05,
      "loss": 0.3733,
      "num_input_tokens_seen": 9190608,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 0.7619875073432922,
      "learning_rate": 3.975044837119053e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 9194160,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 0.5254924297332764,
      "learning_rate": 3.974121489979914e-05,
      "loss": 0.4318,
      "num_input_tokens_seen": 9196688,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 0.30464968085289,
      "learning_rate": 3.973197834471377e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 9200880,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 0.44606688618659973,
      "learning_rate": 3.9722738707866595e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 9204720,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 0.5666280388832092,
      "learning_rate": 3.971349599119046e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 9207184,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 0.5698174238204956,
      "learning_rate": 3.970425019661881e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 9210224,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 0.8933576941490173,
      "learning_rate": 3.9695001326085785e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 9213776,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 0.3680756092071533,
      "learning_rate": 3.968574938152613e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 9216368,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 0.505315363407135,
      "learning_rate": 3.9676494364875247e-05,
      "loss": 0.6767,
      "num_input_tokens_seen": 9220112,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 0.6138200163841248,
      "learning_rate": 3.966723627806918e-05,
      "loss": 0.5065,
      "num_input_tokens_seen": 9223120,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 0.5503950715065002,
      "learning_rate": 3.9657975123044624e-05,
      "loss": 0.5515,
      "num_input_tokens_seen": 9225744,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 0.4624130129814148,
      "learning_rate": 3.964871090173888e-05,
      "loss": 0.4109,
      "num_input_tokens_seen": 9229200,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 0.2871043086051941,
      "learning_rate": 3.963944361608995e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 9232752,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 0.4479358196258545,
      "learning_rate": 3.963017326803644e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 9238704,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 0.3820052146911621,
      "learning_rate": 3.962089985951758e-05,
      "loss": 0.6964,
      "num_input_tokens_seen": 9241648,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 0.5723292231559753,
      "learning_rate": 3.961162339247327e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 9244944,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 0.7040912508964539,
      "learning_rate": 3.9602343868844055e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 9248144,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 0.41292113065719604,
      "learning_rate": 3.959306129057109e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 9251792,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 0.3564741313457489,
      "learning_rate": 3.958377565959619e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 9255248,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 0.5228577256202698,
      "learning_rate": 3.95744869778618e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 9258096,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 0.6952520608901978,
      "learning_rate": 3.956519524731101e-05,
      "loss": 0.4438,
      "num_input_tokens_seen": 9260816,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 0.9140357375144958,
      "learning_rate": 3.9555900469887555e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 9263472,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 0.2781620919704437,
      "learning_rate": 3.954660264753577e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 9266832,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 0.6040882468223572,
      "learning_rate": 3.953730178220067e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 9270512,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 0.9463573694229126,
      "learning_rate": 3.952799787582789e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 9274032,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 0.5521678328514099,
      "learning_rate": 3.9518690930363694e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 9276720,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 0.5467528104782104,
      "learning_rate": 3.9509380947754994e-05,
      "loss": 0.3834,
      "num_input_tokens_seen": 9279504,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 0.39390456676483154,
      "learning_rate": 3.9500067929949336e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 9282864,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 0.29292652010917664,
      "learning_rate": 3.9490751878894895e-05,
      "loss": 0.5388,
      "num_input_tokens_seen": 9285520,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 0.38556525111198425,
      "learning_rate": 3.948143279654047e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 9289840,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 0.5965864658355713,
      "learning_rate": 3.947211068483553e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 9292016,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 0.4471603333950043,
      "learning_rate": 3.946278554573013e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 9295312,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 0.9597305655479431,
      "learning_rate": 3.9453457381175e-05,
      "loss": 0.5808,
      "num_input_tokens_seen": 9298032,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 0.4952390193939209,
      "learning_rate": 3.9444126193121474e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 9301552,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 0.8442344069480896,
      "learning_rate": 3.943479198352153e-05,
      "loss": 0.3744,
      "num_input_tokens_seen": 9304688,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 0.2848140299320221,
      "learning_rate": 3.94254547543278e-05,
      "loss": 0.5476,
      "num_input_tokens_seen": 9307664,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 0.44323334097862244,
      "learning_rate": 3.9416114507493504e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 9310736,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 0.5347921252250671,
      "learning_rate": 3.940677124497252e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 9315056,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 0.5617276430130005,
      "learning_rate": 3.939742496871935e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 9318672,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 0.6411775350570679,
      "learning_rate": 3.9388075680689136e-05,
      "loss": 0.5534,
      "num_input_tokens_seen": 9325040,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 0.42453649640083313,
      "learning_rate": 3.9378723382837634e-05,
      "loss": 0.3681,
      "num_input_tokens_seen": 9328496,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 0.31935858726501465,
      "learning_rate": 3.936936807712124e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 9331824,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 0.4193993806838989,
      "learning_rate": 3.936000976549698e-05,
      "loss": 0.3486,
      "num_input_tokens_seen": 9334768,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 0.3361550271511078,
      "learning_rate": 3.935064844992251e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 9337616,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.2858111262321472,
      "learning_rate": 3.93412841323561e-05,
      "loss": 0.543,
      "num_input_tokens_seen": 9340304,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 0.40019112825393677,
      "learning_rate": 3.933191681475666e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 9343440,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 0.5387533903121948,
      "learning_rate": 3.932254649908373e-05,
      "loss": 0.5219,
      "num_input_tokens_seen": 9346864,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 0.6410014033317566,
      "learning_rate": 3.931317318729748e-05,
      "loss": 0.5426,
      "num_input_tokens_seen": 9350384,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 0.5794733762741089,
      "learning_rate": 3.930379688135869e-05,
      "loss": 0.4038,
      "num_input_tokens_seen": 9354256,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 0.5050621032714844,
      "learning_rate": 3.929441758322878e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 9357456,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 0.4379618763923645,
      "learning_rate": 3.928503529486979e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 9360560,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 1.0238665342330933,
      "learning_rate": 3.927565001824438e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 9363248,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 0.3481389880180359,
      "learning_rate": 3.926626175531586e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 9367088,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 0.2702740430831909,
      "learning_rate": 3.925687050804814e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 9370960,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 0.8035674095153809,
      "learning_rate": 3.924747627840575e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 9373936,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 0.3614383935928345,
      "learning_rate": 3.923807906835388e-05,
      "loss": 0.3648,
      "num_input_tokens_seen": 9377264,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 0.5745519399642944,
      "learning_rate": 3.92286788798583e-05,
      "loss": 0.5292,
      "num_input_tokens_seen": 9380240,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 0.3894645869731903,
      "learning_rate": 3.921927571488543e-05,
      "loss": 0.3237,
      "num_input_tokens_seen": 9385136,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 0.4405093491077423,
      "learning_rate": 3.9209869575402296e-05,
      "loss": 0.5126,
      "num_input_tokens_seen": 9387760,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 0.5452393889427185,
      "learning_rate": 3.9200460463376566e-05,
      "loss": 0.5439,
      "num_input_tokens_seen": 9391184,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 0.2820315957069397,
      "learning_rate": 3.9191048380776515e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 9393776,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 0.35102659463882446,
      "learning_rate": 3.918163332957104e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 9397008,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.6922059059143066,
      "learning_rate": 3.9172215311729665e-05,
      "loss": 0.5948,
      "num_input_tokens_seen": 9400144,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 0.5788007378578186,
      "learning_rate": 3.916279432922254e-05,
      "loss": 0.3388,
      "num_input_tokens_seen": 9403728,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 0.739231526851654,
      "learning_rate": 3.915337038402041e-05,
      "loss": 0.5855,
      "num_input_tokens_seen": 9408112,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 0.5071427226066589,
      "learning_rate": 3.914394347809467e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 9411152,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 0.5513225197792053,
      "learning_rate": 3.9134513613417315e-05,
      "loss": 0.3633,
      "num_input_tokens_seen": 9414416,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 0.4638565480709076,
      "learning_rate": 3.912508079196097e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 9417168,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 0.396399587392807,
      "learning_rate": 3.9115645015698866e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 9420880,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 0.3700977861881256,
      "learning_rate": 3.9106206286604865e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 9424144,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 0.5796762704849243,
      "learning_rate": 3.909676460665345e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 9426448,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 0.5149005055427551,
      "learning_rate": 3.908731997781969e-05,
      "loss": 0.5181,
      "num_input_tokens_seen": 9429456,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 0.47606250643730164,
      "learning_rate": 3.907787240207932e-05,
      "loss": 0.5172,
      "num_input_tokens_seen": 9432720,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 0.44130444526672363,
      "learning_rate": 3.9068421881408645e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 9435344,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 0.37637805938720703,
      "learning_rate": 3.9058968417784614e-05,
      "loss": 0.5738,
      "num_input_tokens_seen": 9438416,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 0.37381991744041443,
      "learning_rate": 3.9049512013184785e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 9440912,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 0.16940155625343323,
      "learning_rate": 3.9040052669587325e-05,
      "loss": 0.2718,
      "num_input_tokens_seen": 9444272,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 0.3538668751716614,
      "learning_rate": 3.9030590388971026e-05,
      "loss": 0.598,
      "num_input_tokens_seen": 9446704,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 0.8086124062538147,
      "learning_rate": 3.902112517331529e-05,
      "loss": 0.4615,
      "num_input_tokens_seen": 9450128,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 0.5694828033447266,
      "learning_rate": 3.901165702460012e-05,
      "loss": 0.5516,
      "num_input_tokens_seen": 9453072,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 0.7164643406867981,
      "learning_rate": 3.900218594480616e-05,
      "loss": 0.376,
      "num_input_tokens_seen": 9455856,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 0.3722194731235504,
      "learning_rate": 3.899271193591465e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 9458736,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 0.2877996563911438,
      "learning_rate": 3.8983234999907435e-05,
      "loss": 0.333,
      "num_input_tokens_seen": 9461840,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 0.3485550284385681,
      "learning_rate": 3.897375513876699e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 9468368,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 0.6780244708061218,
      "learning_rate": 3.8964272354476396e-05,
      "loss": 0.6057,
      "num_input_tokens_seen": 9470928,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 0.40672948956489563,
      "learning_rate": 3.895478664901933e-05,
      "loss": 0.6992,
      "num_input_tokens_seen": 9475024,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 0.23264344036579132,
      "learning_rate": 3.89452980243801e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 9479408,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 0.3496885299682617,
      "learning_rate": 3.893580648254362e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 9482480,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 0.47825807332992554,
      "learning_rate": 3.8926312025495414e-05,
      "loss": 0.3691,
      "num_input_tokens_seen": 9484816,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 0.6872493028640747,
      "learning_rate": 3.891681465522161e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 9487728,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 0.5458141565322876,
      "learning_rate": 3.890731437370896e-05,
      "loss": 0.6545,
      "num_input_tokens_seen": 9490448,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 0.4426446259021759,
      "learning_rate": 3.88978111829448e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 9493360,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 0.448268324136734,
      "learning_rate": 3.88883050849171e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 9496560,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 0.5813836455345154,
      "learning_rate": 3.8878796081614417e-05,
      "loss": 0.3433,
      "num_input_tokens_seen": 9499632,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 0.35173743963241577,
      "learning_rate": 3.8869284175025935e-05,
      "loss": 0.4925,
      "num_input_tokens_seen": 9503696,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 0.5554933547973633,
      "learning_rate": 3.885976936714143e-05,
      "loss": 0.5642,
      "num_input_tokens_seen": 9506768,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 0.4009455144405365,
      "learning_rate": 3.885025165995129e-05,
      "loss": 0.4113,
      "num_input_tokens_seen": 9509488,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 0.3229037821292877,
      "learning_rate": 3.884073105544652e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 9512336,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 0.497298926115036,
      "learning_rate": 3.8831207555618723e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 9515376,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 0.4001314342021942,
      "learning_rate": 3.882168116246009e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 9518160,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 0.33561205863952637,
      "learning_rate": 3.881215187796344e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 9520912,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 0.5047945976257324,
      "learning_rate": 3.8802619704122204e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 9523728,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 0.6701164841651917,
      "learning_rate": 3.8793084642930387e-05,
      "loss": 0.6087,
      "num_input_tokens_seen": 9526416,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 0.7468752264976501,
      "learning_rate": 3.878354669638263e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 9529936,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 0.3748522400856018,
      "learning_rate": 3.8774005866474155e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 9533232,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 0.27546608448028564,
      "learning_rate": 3.876446215520079e-05,
      "loss": 0.4086,
      "num_input_tokens_seen": 9536656,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 0.38927215337753296,
      "learning_rate": 3.875491556455898e-05,
      "loss": 0.5189,
      "num_input_tokens_seen": 9539344,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 0.5048738121986389,
      "learning_rate": 3.8745366096545754e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 9542832,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 0.727057158946991,
      "learning_rate": 3.873581375315876e-05,
      "loss": 0.5048,
      "num_input_tokens_seen": 9546640,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 0.5544822812080383,
      "learning_rate": 3.872625853639623e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 9550000,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 0.4839385151863098,
      "learning_rate": 3.871670044825701e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 9553424,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 0.3629872798919678,
      "learning_rate": 3.870713949074056e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 9556144,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 0.6097553968429565,
      "learning_rate": 3.86975756658469e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 9559024,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 0.3652924597263336,
      "learning_rate": 3.868800897557667e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 9562640,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 0.6209380626678467,
      "learning_rate": 3.867843942193113e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 9565360,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 0.356768935918808,
      "learning_rate": 3.866886700691212e-05,
      "loss": 0.4813,
      "num_input_tokens_seen": 9568688,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 0.5324148535728455,
      "learning_rate": 3.8659291732522075e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 9571568,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 0.4652564227581024,
      "learning_rate": 3.8649713600764026e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 9574128,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 0.3683914840221405,
      "learning_rate": 3.864013261364162e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 9577680,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 0.4107219874858856,
      "learning_rate": 3.8630548773159084e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 9580656,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 0.3253234028816223,
      "learning_rate": 3.862096208132125e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 9584496,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 0.28650808334350586,
      "learning_rate": 3.861137254013355e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 9591056,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 0.5703085660934448,
      "learning_rate": 3.8601780151601996e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 9594320,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 0.4478149712085724,
      "learning_rate": 3.859218491773321e-05,
      "loss": 0.4619,
      "num_input_tokens_seen": 9597616,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 0.6351073384284973,
      "learning_rate": 3.858258684053441e-05,
      "loss": 0.5023,
      "num_input_tokens_seen": 9601072,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 0.35604873299598694,
      "learning_rate": 3.8572985922013405e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 9604752,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 0.4520896375179291,
      "learning_rate": 3.856338216417859e-05,
      "loss": 0.4406,
      "num_input_tokens_seen": 9607600,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 0.4050475060939789,
      "learning_rate": 3.855377556903897e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 9611184,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 0.4502826929092407,
      "learning_rate": 3.8544166138604124e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 9614288,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 0.43248531222343445,
      "learning_rate": 3.8534553874884244e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 9617680,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 0.48805299401283264,
      "learning_rate": 3.852493877989011e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 9620944,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 0.595735490322113,
      "learning_rate": 3.851532085563308e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 9627536,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 0.3852924704551697,
      "learning_rate": 3.850570010412511e-05,
      "loss": 0.551,
      "num_input_tokens_seen": 9631216,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 0.3453332185745239,
      "learning_rate": 3.8496076527378766e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 9634160,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.560143232345581,
      "learning_rate": 3.848645012740718e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 9637456,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 0.3012564480304718,
      "learning_rate": 3.8476820906224096e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 9641520,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 0.3691401779651642,
      "learning_rate": 3.846718886584382e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 9644912,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 0.35181111097335815,
      "learning_rate": 3.845755400828128e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 9647568,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 0.29550665616989136,
      "learning_rate": 3.844791633555198e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 9651600,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 0.4818986654281616,
      "learning_rate": 3.8438275849671996e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 9654576,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 1.0349210500717163,
      "learning_rate": 3.842863255265802e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 9658000,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 0.36424437165260315,
      "learning_rate": 3.841898644652732e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 9660880,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 0.5635945200920105,
      "learning_rate": 3.840933753329774e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 9663920,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 0.4856499135494232,
      "learning_rate": 3.839968581498773e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 9667792,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 0.489053875207901,
      "learning_rate": 3.839003129361632e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 9671120,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 0.2624848783016205,
      "learning_rate": 3.8380373971203134e-05,
      "loss": 0.3914,
      "num_input_tokens_seen": 9674608,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 0.45930707454681396,
      "learning_rate": 3.8370713849768346e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 9677904,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 0.4920855462551117,
      "learning_rate": 3.8361050931332774e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 9681648,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 0.4719294607639313,
      "learning_rate": 3.835138521791778e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 9684208,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 0.2917705476284027,
      "learning_rate": 3.8341716711545314e-05,
      "loss": 0.5191,
      "num_input_tokens_seen": 9686800,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.7966768145561218,
      "learning_rate": 3.833204541423793e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 9689424,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 1.1776983737945557,
      "learning_rate": 3.832237132801874e-05,
      "loss": 0.4024,
      "num_input_tokens_seen": 9692368,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 0.34621286392211914,
      "learning_rate": 3.831269445491146e-05,
      "loss": 0.3747,
      "num_input_tokens_seen": 9695856,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 0.36381828784942627,
      "learning_rate": 3.8303014796940375e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 9699280,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 0.43745478987693787,
      "learning_rate": 3.829333235613036e-05,
      "loss": 0.5923,
      "num_input_tokens_seen": 9701840,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 0.3760600686073303,
      "learning_rate": 3.828364713450688e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 9704400,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 0.4342934489250183,
      "learning_rate": 3.827395913409597e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 9707664,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 0.4649755656719208,
      "learning_rate": 3.826426835692424e-05,
      "loss": 0.4931,
      "num_input_tokens_seen": 9710672,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 0.5137372016906738,
      "learning_rate": 3.825457480501891e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 9713872,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 0.35079342126846313,
      "learning_rate": 3.824487848040773e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 9717008,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 0.612442672252655,
      "learning_rate": 3.823517938511908e-05,
      "loss": 0.5858,
      "num_input_tokens_seen": 9720560,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 0.3244573771953583,
      "learning_rate": 3.8225477521181906e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 9723792,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 0.33242520689964294,
      "learning_rate": 3.821577289062571e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 9727184,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 0.5579648613929749,
      "learning_rate": 3.82060654954806e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 9730832,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 0.7176161408424377,
      "learning_rate": 3.819635533777726e-05,
      "loss": 0.4594,
      "num_input_tokens_seen": 9734640,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 0.6350197792053223,
      "learning_rate": 3.8186642419546924e-05,
      "loss": 0.3271,
      "num_input_tokens_seen": 9737520,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 0.8297871351242065,
      "learning_rate": 3.817692674282145e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 9740880,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 0.6431869864463806,
      "learning_rate": 3.816720830963322e-05,
      "loss": 0.3439,
      "num_input_tokens_seen": 9744080,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 0.4934634864330292,
      "learning_rate": 3.815748712201523e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 9747184,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 0.6911168098449707,
      "learning_rate": 3.814776318200105e-05,
      "loss": 0.5785,
      "num_input_tokens_seen": 9749648,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 0.7041099667549133,
      "learning_rate": 3.8138036491624805e-05,
      "loss": 0.5222,
      "num_input_tokens_seen": 9751984,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 0.37166252732276917,
      "learning_rate": 3.812830705292122e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 9755152,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 0.4654366374015808,
      "learning_rate": 3.8118574867925564e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 9758448,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 0.3725086450576782,
      "learning_rate": 3.810883993867371e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 9761264,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 0.6228917837142944,
      "learning_rate": 3.80991022672021e-05,
      "loss": 0.5551,
      "num_input_tokens_seen": 9764976,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 0.4924604594707489,
      "learning_rate": 3.8089361855547744e-05,
      "loss": 0.5873,
      "num_input_tokens_seen": 9768208,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 0.5297139883041382,
      "learning_rate": 3.8079618705748214e-05,
      "loss": 0.4219,
      "num_input_tokens_seen": 9771248,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 0.38467732071876526,
      "learning_rate": 3.806987281984167e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 9773776,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 0.4775305390357971,
      "learning_rate": 3.806012419986684e-05,
      "loss": 0.5212,
      "num_input_tokens_seen": 9777360,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 0.5295791625976562,
      "learning_rate": 3.805037284786303e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 9779984,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 1.6865969896316528,
      "learning_rate": 3.804061876587009e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 9782480,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 0.5831887125968933,
      "learning_rate": 3.80308619559285e-05,
      "loss": 0.5608,
      "num_input_tokens_seen": 9786608,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 0.5102885365486145,
      "learning_rate": 3.802110242007925e-05,
      "loss": 0.5278,
      "num_input_tokens_seen": 9790096,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 0.5638359189033508,
      "learning_rate": 3.801134016036392e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 9793072,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 1.615452766418457,
      "learning_rate": 3.8001575178824665e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 9796528,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 0.6327693462371826,
      "learning_rate": 3.799180747750422e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 9800048,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 0.4227837920188904,
      "learning_rate": 3.798203705844586e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 9802672,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 0.7339602708816528,
      "learning_rate": 3.7972263923693476e-05,
      "loss": 0.5624,
      "num_input_tokens_seen": 9806352,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 0.7215806245803833,
      "learning_rate": 3.796248807529145e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 9809200,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 0.5018795132637024,
      "learning_rate": 3.795270951528482e-05,
      "loss": 0.5928,
      "num_input_tokens_seen": 9811568,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 0.42257997393608093,
      "learning_rate": 3.7942928245719124e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 9814480,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 0.7429598569869995,
      "learning_rate": 3.7933144268640495e-05,
      "loss": 0.4546,
      "num_input_tokens_seen": 9817200,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 0.4533444344997406,
      "learning_rate": 3.7923357586095645e-05,
      "loss": 0.3372,
      "num_input_tokens_seen": 9820336,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 0.34087029099464417,
      "learning_rate": 3.7913568200131823e-05,
      "loss": 0.5135,
      "num_input_tokens_seen": 9823664,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 0.40712785720825195,
      "learning_rate": 3.790377611279685e-05,
      "loss": 0.4756,
      "num_input_tokens_seen": 9827280,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 0.42504775524139404,
      "learning_rate": 3.789398132613914e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 9830192,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 0.8287562727928162,
      "learning_rate": 3.788418384220763e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 9832816,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 0.4415237009525299,
      "learning_rate": 3.7874383663051846e-05,
      "loss": 0.3857,
      "num_input_tokens_seen": 9836080,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 0.5218564867973328,
      "learning_rate": 3.7864580790721894e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 9839152,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 0.3467639684677124,
      "learning_rate": 3.785477522726839e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 9843120,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 0.6462096571922302,
      "learning_rate": 3.784496697474258e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 9846608,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 0.47688302397727966,
      "learning_rate": 3.783515603519622e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 9849552,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 0.5066419839859009,
      "learning_rate": 3.782534241068164e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 9852336,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 0.39994972944259644,
      "learning_rate": 3.781552610325175e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 9855376,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 0.5252594947814941,
      "learning_rate": 3.780570711496001e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 9858608,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 0.4989624619483948,
      "learning_rate": 3.779588544786044e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 9861776,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 1.500832438468933,
      "learning_rate": 3.778606110400761e-05,
      "loss": 0.5138,
      "num_input_tokens_seen": 9864144,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 0.48438364267349243,
      "learning_rate": 3.7776234085456665e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 9866512,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 0.4055767059326172,
      "learning_rate": 3.776640439426332e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 9869936,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 0.4073610007762909,
      "learning_rate": 3.775657203248381e-05,
      "loss": 0.3801,
      "num_input_tokens_seen": 9873168,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 0.5730269551277161,
      "learning_rate": 3.774673700217497e-05,
      "loss": 0.5714,
      "num_input_tokens_seen": 9875888,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 0.4073459208011627,
      "learning_rate": 3.773689930539418e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 9880432,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 0.8396711945533752,
      "learning_rate": 3.7727058944199354e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 9885136,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 0.7649608850479126,
      "learning_rate": 3.7717215920648994e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 9887920,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 0.6203522682189941,
      "learning_rate": 3.770737023680215e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 9892464,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 0.7793831825256348,
      "learning_rate": 3.7697521894718436e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 9895344,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 0.3956615924835205,
      "learning_rate": 3.7687670896458e-05,
      "loss": 0.4947,
      "num_input_tokens_seen": 9898032,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 0.9945388436317444,
      "learning_rate": 3.7677817244081556e-05,
      "loss": 0.3249,
      "num_input_tokens_seen": 9902160,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 0.42806461453437805,
      "learning_rate": 3.766796093965038e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 9906160,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 0.3518437445163727,
      "learning_rate": 3.76581019852263e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 9909296,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 0.492641419172287,
      "learning_rate": 3.7648240382871705e-05,
      "loss": 0.4798,
      "num_input_tokens_seen": 9911952,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 0.45254987478256226,
      "learning_rate": 3.763837613464951e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 9914992,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 0.3848249018192291,
      "learning_rate": 3.7628509242623215e-05,
      "loss": 0.3356,
      "num_input_tokens_seen": 9919184,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 0.8331700563430786,
      "learning_rate": 3.761863970885686e-05,
      "loss": 0.5981,
      "num_input_tokens_seen": 9921744,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 0.5808234810829163,
      "learning_rate": 3.760876753541504e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 9924912,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 0.4613344073295593,
      "learning_rate": 3.7598892724362907e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 9927728,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 0.5272337794303894,
      "learning_rate": 3.758901527776615e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 9930448,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 0.706702470779419,
      "learning_rate": 3.7579135197691015e-05,
      "loss": 0.5039,
      "num_input_tokens_seen": 9934608,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 0.4498330056667328,
      "learning_rate": 3.7569252486204306e-05,
      "loss": 0.4109,
      "num_input_tokens_seen": 9937456,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 0.5384981036186218,
      "learning_rate": 3.755936714537338e-05,
      "loss": 0.5056,
      "num_input_tokens_seen": 9940816,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 0.7794181704521179,
      "learning_rate": 3.754947917726612e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 9944080,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 0.8340044021606445,
      "learning_rate": 3.7539588583951e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 9946480,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 0.5455509424209595,
      "learning_rate": 3.752969536749701e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 9949872,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 0.6135916113853455,
      "learning_rate": 3.7519799529973685e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 9953520,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 0.5711908340454102,
      "learning_rate": 3.7509901073451134e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 9956624,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.45694682002067566,
      "eval_runtime": 9.6952,
      "eval_samples_per_second": 87.466,
      "eval_steps_per_second": 21.867,
      "num_input_tokens_seen": 9958360,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 0.31357279419898987,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 9958968,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 0.37587881088256836,
      "learning_rate": 3.749009631169147e-05,
      "loss": 0.3926,
      "num_input_tokens_seen": 9961624,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 0.5503124594688416,
      "learning_rate": 3.74801900105973e-05,
      "loss": 0.5162,
      "num_input_tokens_seen": 9964600,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 0.6196970343589783,
      "learning_rate": 3.747028109878975e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 9967096,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 0.44470885396003723,
      "learning_rate": 3.746036957834167e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 9969848,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 0.32497653365135193,
      "learning_rate": 3.7450455451326415e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 9973272,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 0.3162376880645752,
      "learning_rate": 3.7440538719817935e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 9975992,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 0.4326592981815338,
      "learning_rate": 3.743061938589068e-05,
      "loss": 0.5387,
      "num_input_tokens_seen": 9979704,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 0.5557297468185425,
      "learning_rate": 3.742069745161966e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 9982712,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 0.6123082041740417,
      "learning_rate": 3.741077291908045e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 9985560,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 0.502640962600708,
      "learning_rate": 3.740084579034913e-05,
      "loss": 0.391,
      "num_input_tokens_seen": 9989272,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 0.4856780767440796,
      "learning_rate": 3.739091606750234e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 9992088,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 0.4536858797073364,
      "learning_rate": 3.7380983752617275e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 9995032,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 0.6973592042922974,
      "learning_rate": 3.737104884777166e-05,
      "loss": 0.334,
      "num_input_tokens_seen": 9998008,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 0.24494269490242004,
      "learning_rate": 3.736111135504376e-05,
      "loss": 0.6021,
      "num_input_tokens_seen": 10001464,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 0.348949134349823,
      "learning_rate": 3.735117127651239e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 10004472,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 1.1160080432891846,
      "learning_rate": 3.734122861425689e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 10007160,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 0.2623809278011322,
      "learning_rate": 3.7331283370357173e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 10010136,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 0.6958244442939758,
      "learning_rate": 3.732133554689365e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 10013560,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 0.41919198632240295,
      "learning_rate": 3.73113851459473e-05,
      "loss": 0.4384,
      "num_input_tokens_seen": 10016280,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 0.7411617636680603,
      "learning_rate": 3.730143216959964e-05,
      "loss": 0.6017,
      "num_input_tokens_seen": 10019096,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 0.4674915671348572,
      "learning_rate": 3.729147661993271e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 10022200,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 0.5969753265380859,
      "learning_rate": 3.728151849902909e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 10025528,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 0.6022756099700928,
      "learning_rate": 3.727155780897193e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 10030104,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 0.5467378497123718,
      "learning_rate": 3.7261594551844865e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 10033112,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 0.5110604763031006,
      "learning_rate": 3.7251628729732116e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 10036280,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 0.329546719789505,
      "learning_rate": 3.724166034471841e-05,
      "loss": 0.3499,
      "num_input_tokens_seen": 10039160,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 0.2535562515258789,
      "learning_rate": 3.723168939888901e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 10043224,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 0.4618567228317261,
      "learning_rate": 3.7221715894329747e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 10045976,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 0.30025291442871094,
      "learning_rate": 3.7211739833126945e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 10048824,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 0.30410563945770264,
      "learning_rate": 3.720176121736748e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 10051512,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 0.4034018814563751,
      "learning_rate": 3.719178004913879e-05,
      "loss": 0.5588,
      "num_input_tokens_seen": 10054712,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 0.6003016829490662,
      "learning_rate": 3.7181796330528796e-05,
      "loss": 0.5545,
      "num_input_tokens_seen": 10058232,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 0.25252747535705566,
      "learning_rate": 3.717181006362599e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 10061880,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 0.3781401515007019,
      "learning_rate": 3.7161821250519384e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 10064984,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 0.39674094319343567,
      "learning_rate": 3.7151829893298515e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 10067384,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 0.26244962215423584,
      "learning_rate": 3.714183599405347e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 10070776,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 0.7313793897628784,
      "learning_rate": 3.713183955487487e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 10074104,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 0.4259229004383087,
      "learning_rate": 3.7121840577853815e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 10077784,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 0.5027136206626892,
      "learning_rate": 3.7111839065082027e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 10082392,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 0.34979569911956787,
      "learning_rate": 3.710183501865168e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 10085656,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 0.39728033542633057,
      "learning_rate": 3.7091828440655514e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 10089688,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 0.4503425657749176,
      "learning_rate": 3.708181933318678e-05,
      "loss": 0.5448,
      "num_input_tokens_seen": 10092632,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.6236687302589417,
      "learning_rate": 3.707180769833929e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 10095384,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 0.6304449439048767,
      "learning_rate": 3.706179353820734e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 10101272,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 0.5799059867858887,
      "learning_rate": 3.7051776854885794e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 10104920,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 0.3684547245502472,
      "learning_rate": 3.7041757650470025e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 10108440,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 0.5708057284355164,
      "learning_rate": 3.703173592705594e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 10111672,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 0.29943200945854187,
      "learning_rate": 3.702171168673996e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 10115448,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 0.26439136266708374,
      "learning_rate": 3.701168493161905e-05,
      "loss": 0.3759,
      "num_input_tokens_seen": 10118328,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 0.6121410727500916,
      "learning_rate": 3.7001655663790676e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 10122328,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 0.4219864308834076,
      "learning_rate": 3.699162388535288e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 10125112,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 0.3812146484851837,
      "learning_rate": 3.698158959840416e-05,
      "loss": 0.3667,
      "num_input_tokens_seen": 10128280,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 0.4270749092102051,
      "learning_rate": 3.6971552805043585e-05,
      "loss": 0.3638,
      "num_input_tokens_seen": 10131320,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 0.4588477909564972,
      "learning_rate": 3.6961513507370754e-05,
      "loss": 0.4496,
      "num_input_tokens_seen": 10135384,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 0.5006206035614014,
      "learning_rate": 3.695147170748576e-05,
      "loss": 0.5861,
      "num_input_tokens_seen": 10138520,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 0.4316962957382202,
      "learning_rate": 3.694142740748923e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 10141528,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 0.31699755787849426,
      "learning_rate": 3.693138060948232e-05,
      "loss": 0.4798,
      "num_input_tokens_seen": 10144568,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 0.42619141936302185,
      "learning_rate": 3.692133131556671e-05,
      "loss": 0.4576,
      "num_input_tokens_seen": 10148504,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 0.3532843291759491,
      "learning_rate": 3.6911279527844596e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 10151288,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 0.3218314051628113,
      "learning_rate": 3.69012252484187e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 10154744,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 0.5577117800712585,
      "learning_rate": 3.6891168479392245e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 10158136,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 0.34597113728523254,
      "learning_rate": 3.688110922286901e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 10161560,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 0.4064127802848816,
      "learning_rate": 3.687104748095328e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 10163960,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 0.5271536707878113,
      "learning_rate": 3.686098325574984e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 10166776,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 0.6379722356796265,
      "learning_rate": 3.685091654936403e-05,
      "loss": 0.346,
      "num_input_tokens_seen": 10169592,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 0.42298781871795654,
      "learning_rate": 3.6840847363901675e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 10172760,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 0.39488697052001953,
      "learning_rate": 3.683077570146914e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 10177496,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 0.8379135131835938,
      "learning_rate": 3.68207015641733e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 10181112,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 0.365355908870697,
      "learning_rate": 3.681062495412155e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 10184632,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 0.3460378348827362,
      "learning_rate": 3.68005458734218e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 10187576,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 0.47765249013900757,
      "learning_rate": 3.679046432418248e-05,
      "loss": 0.5515,
      "num_input_tokens_seen": 10190648,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 0.311210572719574,
      "learning_rate": 3.6780380308512543e-05,
      "loss": 0.3187,
      "num_input_tokens_seen": 10193944,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 0.27414366602897644,
      "learning_rate": 3.677029382852144e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 10196824,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 0.532217264175415,
      "learning_rate": 3.676020488631915e-05,
      "loss": 0.5301,
      "num_input_tokens_seen": 10200312,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 0.35792386531829834,
      "learning_rate": 3.675011348401617e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 10203992,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 0.6240583062171936,
      "learning_rate": 3.6740019623723494e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 10206904,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 0.26419559121131897,
      "learning_rate": 3.672992330755265e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 10209880,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 0.9061166644096375,
      "learning_rate": 3.671982453761568e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 10215864,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 0.7091830968856812,
      "learning_rate": 3.6709723316025114e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 10218904,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 0.5113356113433838,
      "learning_rate": 3.669961964489403e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 10225048,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 0.3538977801799774,
      "learning_rate": 3.668951352633599e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 10227576,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 1.2822932004928589,
      "learning_rate": 3.6679404962465084e-05,
      "loss": 0.6222,
      "num_input_tokens_seen": 10230712,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 0.6659830808639526,
      "learning_rate": 3.66692939553959e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 10233848,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 0.3069051206111908,
      "learning_rate": 3.6659180507243554e-05,
      "loss": 0.4985,
      "num_input_tokens_seen": 10237816,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 0.35985833406448364,
      "learning_rate": 3.664906462012366e-05,
      "loss": 0.4172,
      "num_input_tokens_seen": 10241272,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 0.533422589302063,
      "learning_rate": 3.663894629615234e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 10243864,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 0.3040928840637207,
      "learning_rate": 3.6628825537446246e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 10246712,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 0.2471037209033966,
      "learning_rate": 3.6618702346122525e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 10250648,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 0.5919948220252991,
      "learning_rate": 3.6608576724298817e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 10253624,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 0.6045370697975159,
      "learning_rate": 3.65984486740933e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 10256536,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 1.1868607997894287,
      "learning_rate": 3.6588318197624644e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 10259448,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 0.6361932754516602,
      "learning_rate": 3.6578185297012015e-05,
      "loss": 0.4479,
      "num_input_tokens_seen": 10263768,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 0.8466590642929077,
      "learning_rate": 3.656804997437512e-05,
      "loss": 0.5045,
      "num_input_tokens_seen": 10266712,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 0.5112302303314209,
      "learning_rate": 3.6557912231834155e-05,
      "loss": 0.6569,
      "num_input_tokens_seen": 10270712,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 0.7254679799079895,
      "learning_rate": 3.65477720715098e-05,
      "loss": 0.563,
      "num_input_tokens_seen": 10274456,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 0.28922247886657715,
      "learning_rate": 3.653762949552328e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 10278840,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 0.5099421739578247,
      "learning_rate": 3.6527484505996295e-05,
      "loss": 0.349,
      "num_input_tokens_seen": 10282104,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 0.3585144877433777,
      "learning_rate": 3.651733710505106e-05,
      "loss": 0.5383,
      "num_input_tokens_seen": 10285272,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 1.1280345916748047,
      "learning_rate": 3.6507187294810305e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 10288472,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 0.4130665957927704,
      "learning_rate": 3.649703507739724e-05,
      "loss": 0.5083,
      "num_input_tokens_seen": 10291128,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 0.4310547411441803,
      "learning_rate": 3.6486880454935606e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 10295288,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 0.41963738203048706,
      "learning_rate": 3.647672342954963e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 10298680,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 0.5151297450065613,
      "learning_rate": 3.646656400336405e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 10301496,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 0.5653165578842163,
      "learning_rate": 3.6456402178504094e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 10303992,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 0.3122740089893341,
      "learning_rate": 3.6446237957095505e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 10306808,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 0.2799330949783325,
      "learning_rate": 3.643607134126452e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 10310584,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 0.28774628043174744,
      "learning_rate": 3.642590233313789e-05,
      "loss": 0.3645,
      "num_input_tokens_seen": 10314872,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 0.42708227038383484,
      "learning_rate": 3.6415730934842827e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 10317624,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 0.7077655792236328,
      "learning_rate": 3.64055571485071e-05,
      "loss": 0.3863,
      "num_input_tokens_seen": 10320408,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 0.4931180477142334,
      "learning_rate": 3.639538097625894e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 10323320,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 0.4016321301460266,
      "learning_rate": 3.638520242022708e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 10326776,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 0.6186087727546692,
      "learning_rate": 3.637502148254077e-05,
      "loss": 0.2829,
      "num_input_tokens_seen": 10330584,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 0.5737785696983337,
      "learning_rate": 3.636483816532974e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 10332824,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 0.4898233711719513,
      "learning_rate": 3.6354652470724217e-05,
      "loss": 0.5779,
      "num_input_tokens_seen": 10335480,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.2981787621974945,
      "learning_rate": 3.634446440085494e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 10338424,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 0.40314781665802,
      "learning_rate": 3.633427395785313e-05,
      "loss": 0.4523,
      "num_input_tokens_seen": 10341784,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 0.4630514085292816,
      "learning_rate": 3.6324081143850516e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 10345144,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 0.3189230263233185,
      "learning_rate": 3.6313885960979326e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 10348984,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 0.387997031211853,
      "learning_rate": 3.630368841137226e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 10352152,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 1.0933501720428467,
      "learning_rate": 3.6293488497162554e-05,
      "loss": 0.5102,
      "num_input_tokens_seen": 10354392,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 0.3929661512374878,
      "learning_rate": 3.628328622048388e-05,
      "loss": 0.7325,
      "num_input_tokens_seen": 10357368,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 0.5383184552192688,
      "learning_rate": 3.627308158347046e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 10359864,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 0.43943583965301514,
      "learning_rate": 3.626287458825699e-05,
      "loss": 0.6055,
      "num_input_tokens_seen": 10364536,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 0.40977227687835693,
      "learning_rate": 3.625266523697864e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 10367864,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 0.31370237469673157,
      "learning_rate": 3.62424535317711e-05,
      "loss": 0.5243,
      "num_input_tokens_seen": 10371736,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 0.38208261132240295,
      "learning_rate": 3.623223947477055e-05,
      "loss": 0.3338,
      "num_input_tokens_seen": 10375192,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.3102908432483673,
      "learning_rate": 3.6222023068113635e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 10383288,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 0.4214792251586914,
      "learning_rate": 3.621180431393754e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 10387192,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 0.7457366585731506,
      "learning_rate": 3.6201583214379874e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 10389400,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 0.33065885305404663,
      "learning_rate": 3.61913597715788e-05,
      "loss": 0.5219,
      "num_input_tokens_seen": 10392920,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 0.5012056231498718,
      "learning_rate": 3.6181133987672944e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 10396280,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 1.3421556949615479,
      "learning_rate": 3.617090586480142e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 10399960,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 0.5922398567199707,
      "learning_rate": 3.6160675405103826e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 10403128,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 0.5297924280166626,
      "learning_rate": 3.615044261072027e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 10405976,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 1.134260654449463,
      "learning_rate": 3.614020748379133e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 10408632,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 0.4196633994579315,
      "learning_rate": 3.6129970026458076e-05,
      "loss": 0.5214,
      "num_input_tokens_seen": 10412216,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 0.9192047715187073,
      "learning_rate": 3.611973024086207e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 10414680,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 0.4616706073284149,
      "learning_rate": 3.610948812914536e-05,
      "loss": 0.5277,
      "num_input_tokens_seen": 10417560,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 0.322555810213089,
      "learning_rate": 3.609924369345048e-05,
      "loss": 0.3414,
      "num_input_tokens_seen": 10420056,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 0.332149863243103,
      "learning_rate": 3.6088996935920444e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 10423288,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 0.4386041462421417,
      "learning_rate": 3.607874785869876e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 10425752,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 0.2512848377227783,
      "learning_rate": 3.6068496463929416e-05,
      "loss": 0.3921,
      "num_input_tokens_seen": 10429432,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 0.43404024839401245,
      "learning_rate": 3.605824275375689e-05,
      "loss": 0.4473,
      "num_input_tokens_seen": 10432856,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 0.47769129276275635,
      "learning_rate": 3.604798673032613e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 10436600,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 0.4318760335445404,
      "learning_rate": 3.60377283957826e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 10439352,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 0.5102790594100952,
      "learning_rate": 3.6027467752272203e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 10441976,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 0.4820481240749359,
      "learning_rate": 3.601720480194137e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 10445272,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 0.42403021454811096,
      "learning_rate": 3.600693954693698e-05,
      "loss": 0.5357,
      "num_input_tokens_seen": 10447864,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 0.3728599548339844,
      "learning_rate": 3.5996671989406414e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 10451992,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 0.3883054554462433,
      "learning_rate": 3.598640213149752e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 10454968,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 0.6383684873580933,
      "learning_rate": 3.597612997535864e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 10458392,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 0.5449897050857544,
      "learning_rate": 3.596585552313859e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 10461144,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 0.7143000364303589,
      "learning_rate": 3.595557877698666e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 10463768,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 0.5663264393806458,
      "learning_rate": 3.594529973905266e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 10467128,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 1.0396625995635986,
      "learning_rate": 3.593501841148681e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 10469240,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 0.35015246272087097,
      "learning_rate": 3.5924734796439864e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 10472728,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 1.049986481666565,
      "learning_rate": 3.5914448896063044e-05,
      "loss": 0.4051,
      "num_input_tokens_seen": 10474936,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 0.47517314553260803,
      "learning_rate": 3.590416071250802e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 10478264,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 0.3322654068470001,
      "learning_rate": 3.589387024792699e-05,
      "loss": 0.5194,
      "num_input_tokens_seen": 10481432,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 0.7993855476379395,
      "learning_rate": 3.5883577504472576e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 10483864,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 0.3891352713108063,
      "learning_rate": 3.587328248429792e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 10487480,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 0.48455381393432617,
      "learning_rate": 3.586298518955661e-05,
      "loss": 0.3611,
      "num_input_tokens_seen": 10490616,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 0.7267849445343018,
      "learning_rate": 3.5852685622402746e-05,
      "loss": 0.3685,
      "num_input_tokens_seen": 10493528,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 1.3221166133880615,
      "learning_rate": 3.584238378499085e-05,
      "loss": 0.5157,
      "num_input_tokens_seen": 10496568,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 0.3610331416130066,
      "learning_rate": 3.5832079679475974e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 10499992,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 0.4862571656703949,
      "learning_rate": 3.582177330801359e-05,
      "loss": 0.5096,
      "num_input_tokens_seen": 10502840,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 0.47799625992774963,
      "learning_rate": 3.58114646727597e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 10505752,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 0.4663991630077362,
      "learning_rate": 3.580115377587075e-05,
      "loss": 0.4612,
      "num_input_tokens_seen": 10508664,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 0.4112820327281952,
      "learning_rate": 3.5790840619503654e-05,
      "loss": 0.5156,
      "num_input_tokens_seen": 10511416,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 0.4053691327571869,
      "learning_rate": 3.57805252058158e-05,
      "loss": 0.4176,
      "num_input_tokens_seen": 10513944,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 0.5669575929641724,
      "learning_rate": 3.577020753696507e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 10517208,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 0.5016400218009949,
      "learning_rate": 3.57598876151098e-05,
      "loss": 0.4592,
      "num_input_tokens_seen": 10519704,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 0.31451913714408875,
      "learning_rate": 3.5749565442408796e-05,
      "loss": 0.5719,
      "num_input_tokens_seen": 10523160,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 0.643759548664093,
      "learning_rate": 3.573924102102133e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 10526168,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 0.40326744318008423,
      "learning_rate": 3.572891435310717e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 10528824,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 0.31469812989234924,
      "learning_rate": 3.5718585440826527e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 10531640,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 0.6331575512886047,
      "learning_rate": 3.570825428634009e-05,
      "loss": 0.371,
      "num_input_tokens_seen": 10534808,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 0.662946343421936,
      "learning_rate": 3.569792089180903e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 10537784,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 0.5367668867111206,
      "learning_rate": 3.568758525939497e-05,
      "loss": 0.5133,
      "num_input_tokens_seen": 10540760,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 0.4380442202091217,
      "learning_rate": 3.5677247391259995e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 10543928,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 0.3947208523750305,
      "learning_rate": 3.566690728956668e-05,
      "loss": 0.5895,
      "num_input_tokens_seen": 10546456,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 0.7007541656494141,
      "learning_rate": 3.5656564956478056e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 10549752,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 0.7159861922264099,
      "learning_rate": 3.5646220394157626e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 10553048,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 0.9948469996452332,
      "learning_rate": 3.563587360476933e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 10555768,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 0.647422194480896,
      "learning_rate": 3.562552459047763e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 10559448,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 0.50689297914505,
      "learning_rate": 3.561517335344741e-05,
      "loss": 0.6643,
      "num_input_tokens_seen": 10562008,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.4486434757709503,
      "learning_rate": 3.560481989584402e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 10564760,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 0.5765014290809631,
      "learning_rate": 3.55944642198333e-05,
      "loss": 0.5595,
      "num_input_tokens_seen": 10567512,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 0.52654629945755,
      "learning_rate": 3.5584106327581536e-05,
      "loss": 0.3425,
      "num_input_tokens_seen": 10570104,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 0.48039186000823975,
      "learning_rate": 3.557374622125547e-05,
      "loss": 0.5625,
      "num_input_tokens_seen": 10573304,
      "step": 16215
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 0.5097708106040955,
      "learning_rate": 3.556338390302234e-05,
      "loss": 0.3803,
      "num_input_tokens_seen": 10576408,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 0.389531672000885,
      "learning_rate": 3.55530193750498e-05,
      "loss": 0.3103,
      "num_input_tokens_seen": 10579864,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 0.36453771591186523,
      "learning_rate": 3.5542652639505996e-05,
      "loss": 0.5243,
      "num_input_tokens_seen": 10582840,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 0.33455711603164673,
      "learning_rate": 3.5532283698559556e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 10586392,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 0.28658995032310486,
      "learning_rate": 3.55219125543795e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 10591224,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 0.45805823802948,
      "learning_rate": 3.5511539209135396e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 10594296,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 0.38480380177497864,
      "learning_rate": 3.550116366499721e-05,
      "loss": 0.3864,
      "num_input_tokens_seen": 10597912,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 0.907949686050415,
      "learning_rate": 3.549078592413538e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 10600632,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 0.2927587330341339,
      "learning_rate": 3.548040598872083e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 10603672,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 0.3699330985546112,
      "learning_rate": 3.5470023860924905e-05,
      "loss": 0.3732,
      "num_input_tokens_seen": 10606648,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 0.5209412574768066,
      "learning_rate": 3.5459639542919444e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 10609432,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 0.43659135699272156,
      "learning_rate": 3.5449253036876715e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 10614424,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 0.6223119497299194,
      "learning_rate": 3.543886434496947e-05,
      "loss": 0.5705,
      "num_input_tokens_seen": 10617784,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 0.7443088293075562,
      "learning_rate": 3.542847346937089e-05,
      "loss": 0.5003,
      "num_input_tokens_seen": 10620216,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 0.35825005173683167,
      "learning_rate": 3.541808041225464e-05,
      "loss": 0.4986,
      "num_input_tokens_seen": 10622808,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 0.2698749005794525,
      "learning_rate": 3.540768517579481e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 10625912,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 0.38650038838386536,
      "learning_rate": 3.539728776216599e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 10629016,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 0.367400586605072,
      "learning_rate": 3.5386888173543185e-05,
      "loss": 0.6096,
      "num_input_tokens_seen": 10632920,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 0.6455364227294922,
      "learning_rate": 3.5376486412101875e-05,
      "loss": 0.3709,
      "num_input_tokens_seen": 10635800,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 0.7369980812072754,
      "learning_rate": 3.5366082480017994e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 10638712,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 0.4930366277694702,
      "learning_rate": 3.535567637946792e-05,
      "loss": 0.3713,
      "num_input_tokens_seen": 10641528,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 1.0206434726715088,
      "learning_rate": 3.534526811262848e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 10644152,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 0.3410147726535797,
      "learning_rate": 3.533485768167698e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 10647288,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 0.4945421814918518,
      "learning_rate": 3.532444508879117e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 10650840,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 1.189086675643921,
      "learning_rate": 3.531403033614922e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 10654360,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 0.3344595730304718,
      "learning_rate": 3.530361342592981e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 10659448,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 0.4342429041862488,
      "learning_rate": 3.5293194360312006e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 10662520,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 0.9347803592681885,
      "learning_rate": 3.528277314147538e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 10665176,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 0.6104251146316528,
      "learning_rate": 3.5272349771599925e-05,
      "loss": 0.4226,
      "num_input_tokens_seen": 10668984,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 0.22158104181289673,
      "learning_rate": 3.526192425286609e-05,
      "loss": 0.3584,
      "num_input_tokens_seen": 10673112,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 0.7333953976631165,
      "learning_rate": 3.525149658745476e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 10676504,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 0.3384229838848114,
      "learning_rate": 3.5241066777547315e-05,
      "loss": 0.2968,
      "num_input_tokens_seen": 10679480,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 1.4357314109802246,
      "learning_rate": 3.523063482532554e-05,
      "loss": 0.4515,
      "num_input_tokens_seen": 10682360,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 0.33043572306632996,
      "learning_rate": 3.5220200732971665e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 10686232,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 0.535000205039978,
      "learning_rate": 3.52097645026684e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 10689592,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 0.6412345170974731,
      "learning_rate": 3.519932613659889e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 10693272,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 0.4677431583404541,
      "learning_rate": 3.5188885636946696e-05,
      "loss": 0.346,
      "num_input_tokens_seen": 10697560,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 0.32648569345474243,
      "learning_rate": 3.5178443005895884e-05,
      "loss": 0.6339,
      "num_input_tokens_seen": 10700664,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 0.7144982814788818,
      "learning_rate": 3.51679982456309e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 10704696,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 0.47755110263824463,
      "learning_rate": 3.51575513583367e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 10707512,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 0.5093846917152405,
      "learning_rate": 3.5147102346198635e-05,
      "loss": 0.3556,
      "num_input_tokens_seen": 10711128,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 0.3957797884941101,
      "learning_rate": 3.5136651211402515e-05,
      "loss": 0.5159,
      "num_input_tokens_seen": 10714392,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 0.7090992331504822,
      "learning_rate": 3.5126197956134614e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 10718008,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 0.2708781361579895,
      "learning_rate": 3.5115742582581626e-05,
      "loss": 0.4639,
      "num_input_tokens_seen": 10721240,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 0.5619737505912781,
      "learning_rate": 3.51052850929307e-05,
      "loss": 0.5098,
      "num_input_tokens_seen": 10724312,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 0.9498765468597412,
      "learning_rate": 3.509482548936941e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 10726872,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 0.37809842824935913,
      "learning_rate": 3.5084363774085804e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 10730808,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 0.4639129638671875,
      "learning_rate": 3.507389994926834e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 10733848,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 0.43496429920196533,
      "learning_rate": 3.5063434017105946e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 10736952,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 0.2774607241153717,
      "learning_rate": 3.505296597978795e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 10740024,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 0.4372785985469818,
      "learning_rate": 3.504249583950417e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 10742680,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 0.30905967950820923,
      "learning_rate": 3.503202359844483e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 10746904,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 0.5238490700721741,
      "learning_rate": 3.50215492588006e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 10749656,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 0.5649371147155762,
      "learning_rate": 3.5011072822762605e-05,
      "loss": 0.6224,
      "num_input_tokens_seen": 10752280,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 0.3887130916118622,
      "learning_rate": 3.500059429252238e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 10755672,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 0.3553524315357208,
      "learning_rate": 3.499011367027193e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 10759256,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 0.7149285674095154,
      "learning_rate": 3.4979630958203665e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 10762840,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 0.41372838616371155,
      "learning_rate": 3.496914615851046e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 10765592,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 0.5479944944381714,
      "learning_rate": 3.495865927338562e-05,
      "loss": 0.3864,
      "num_input_tokens_seen": 10767800,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 0.4076075553894043,
      "learning_rate": 3.494817030502286e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 10775384,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 0.6384363174438477,
      "learning_rate": 3.493767925561638e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 10777752,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 0.3132895529270172,
      "learning_rate": 3.492718612736077e-05,
      "loss": 0.3029,
      "num_input_tokens_seen": 10780376,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 0.7035453915596008,
      "learning_rate": 3.491669092245108e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 10784344,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 0.5578194260597229,
      "learning_rate": 3.4906193643082784e-05,
      "loss": 0.5505,
      "num_input_tokens_seen": 10787128,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 0.3499343693256378,
      "learning_rate": 3.4895694291451795e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 10790296,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 0.4434422254562378,
      "learning_rate": 3.4885192869754444e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 10793144,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 0.7466039061546326,
      "learning_rate": 3.4874689380187537e-05,
      "loss": 0.4603,
      "num_input_tokens_seen": 10796280,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 0.3080209195613861,
      "learning_rate": 3.486418382494826e-05,
      "loss": 0.4376,
      "num_input_tokens_seen": 10799000,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.4395920932292938,
      "learning_rate": 3.485367620623426e-05,
      "loss": 0.5205,
      "num_input_tokens_seen": 10802264,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 0.4492672085762024,
      "learning_rate": 3.484316652624362e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 10805656,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 0.35692110657691956,
      "learning_rate": 3.483265478717484e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 10809496,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 0.38955140113830566,
      "learning_rate": 3.482214099122686e-05,
      "loss": 0.4875,
      "num_input_tokens_seen": 10812856,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 0.49799370765686035,
      "learning_rate": 3.481162514059903e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 10815544,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 0.488639235496521,
      "learning_rate": 3.480110723749116e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 10818072,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 0.5501078367233276,
      "learning_rate": 3.479058728410348e-05,
      "loss": 0.4063,
      "num_input_tokens_seen": 10821304,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 0.2528410851955414,
      "learning_rate": 3.478006528263663e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 10824984,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 0.3108175992965698,
      "learning_rate": 3.4769541235291704e-05,
      "loss": 0.5953,
      "num_input_tokens_seen": 10828024,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 0.8600305318832397,
      "learning_rate": 3.475901514427021e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 10831640,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 0.2848237454891205,
      "learning_rate": 3.4748487011774075e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 10835160,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 0.7042626142501831,
      "learning_rate": 3.473795684000569e-05,
      "loss": 0.3642,
      "num_input_tokens_seen": 10838232,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 0.4681222438812256,
      "learning_rate": 3.4727424631167815e-05,
      "loss": 0.3433,
      "num_input_tokens_seen": 10841240,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 0.524385929107666,
      "learning_rate": 3.471689038746369e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 10844312,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 0.3865385353565216,
      "learning_rate": 3.470635411109695e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 10847320,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 0.29328668117523193,
      "learning_rate": 3.469581580427167e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 10850648,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 0.43889978528022766,
      "learning_rate": 3.4685275469192335e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 10853912,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 0.39045217633247375,
      "learning_rate": 3.467473310806387e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 10856824,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 0.3172934353351593,
      "learning_rate": 3.466418872309161e-05,
      "loss": 0.3838,
      "num_input_tokens_seen": 10860280,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 0.3587953746318817,
      "learning_rate": 3.465364231648133e-05,
      "loss": 0.5148,
      "num_input_tokens_seen": 10863832,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 0.4183062016963959,
      "learning_rate": 3.464309389043921e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 10867576,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 0.32736021280288696,
      "learning_rate": 3.4632543447171856e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 10871384,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 0.48284345865249634,
      "learning_rate": 3.4621990988886315e-05,
      "loss": 0.5996,
      "num_input_tokens_seen": 10874584,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 0.40400898456573486,
      "learning_rate": 3.461143651779003e-05,
      "loss": 0.4058,
      "num_input_tokens_seen": 10878584,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 0.5361109375953674,
      "learning_rate": 3.4600880036090876e-05,
      "loss": 0.6668,
      "num_input_tokens_seen": 10881560,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 0.4419104754924774,
      "learning_rate": 3.459032154599715e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 10884600,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 0.32824617624282837,
      "learning_rate": 3.457976104971757e-05,
      "loss": 0.36,
      "num_input_tokens_seen": 10888408,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.369194358587265,
      "learning_rate": 3.4569198549461276e-05,
      "loss": 0.4953,
      "num_input_tokens_seen": 10891736,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 0.5357910394668579,
      "learning_rate": 3.455863404743781e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 10894648,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 0.517416775226593,
      "learning_rate": 3.454806754585716e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 10897624,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 0.39762604236602783,
      "learning_rate": 3.453749904692971e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 10900696,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 0.4750915765762329,
      "learning_rate": 3.452692855286626e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 10904280,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 0.40974128246307373,
      "learning_rate": 3.4516356065878055e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 10907896,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 0.42296215891838074,
      "learning_rate": 3.450578158817671e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 10911192,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 0.5323036909103394,
      "learning_rate": 3.4495205121974325e-05,
      "loss": 0.5882,
      "num_input_tokens_seen": 10914488,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 0.5690569281578064,
      "learning_rate": 3.448462666948335e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 10918456,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 0.3644033968448639,
      "learning_rate": 3.4474046232916674e-05,
      "loss": 0.4855,
      "num_input_tokens_seen": 10922296,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 0.26192203164100647,
      "learning_rate": 3.4463463814487616e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 10925400,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 0.35880815982818604,
      "learning_rate": 3.44528794164099e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 10928120,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 0.6146193146705627,
      "learning_rate": 3.4442293040897646e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 10932088,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 0.4275722801685333,
      "learning_rate": 3.443170469016542e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 10935832,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 0.6036550998687744,
      "learning_rate": 3.442111436642817e-05,
      "loss": 0.5925,
      "num_input_tokens_seen": 10938552,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 0.5274964570999146,
      "learning_rate": 3.441052207190127e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 10942520,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 0.4214929938316345,
      "learning_rate": 3.4399927808800526e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 10946008,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 0.5659907460212708,
      "learning_rate": 3.438933157934211e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 10949144,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 0.5463878512382507,
      "learning_rate": 3.4378733385742653e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 10953400,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 0.399646133184433,
      "learning_rate": 3.436813323021918e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 10956472,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 0.4420223534107208,
      "learning_rate": 3.43575311149891e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 10960312,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 0.4857182800769806,
      "learning_rate": 3.434692704227028e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 10962616,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 0.3778037130832672,
      "learning_rate": 3.4336321014280944e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 10966008,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 0.526974081993103,
      "learning_rate": 3.432571303323977e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 10970040,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 0.5461493134498596,
      "learning_rate": 3.431510310136583e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 10972472,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 0.39425426721572876,
      "learning_rate": 3.4304491220878585e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 10976664,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 0.2931075692176819,
      "learning_rate": 3.429387739399793e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 10979192,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 0.5970881581306458,
      "learning_rate": 3.428326162294416e-05,
      "loss": 0.4233,
      "num_input_tokens_seen": 10981688,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 0.4478064775466919,
      "learning_rate": 3.427264390993795e-05,
      "loss": 0.4654,
      "num_input_tokens_seen": 10985176,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 0.2482537031173706,
      "learning_rate": 3.426202425720044e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 10988536,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 0.6654458045959473,
      "learning_rate": 3.425140266695312e-05,
      "loss": 0.3636,
      "num_input_tokens_seen": 10990936,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 0.37031957507133484,
      "learning_rate": 3.4240779141417905e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 10994200,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 0.3195081353187561,
      "learning_rate": 3.423015368281711e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 10997240,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 0.41948026418685913,
      "learning_rate": 3.421952629337348e-05,
      "loss": 0.4115,
      "num_input_tokens_seen": 11000568,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 0.2972739338874817,
      "learning_rate": 3.420889697531013e-05,
      "loss": 0.3424,
      "num_input_tokens_seen": 11005528,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 0.5092275738716125,
      "learning_rate": 3.419826573085058e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 11009176,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 0.3533327579498291,
      "learning_rate": 3.418763256221879e-05,
      "loss": 0.3722,
      "num_input_tokens_seen": 11012504,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 0.31023818254470825,
      "learning_rate": 3.4176997471639085e-05,
      "loss": 0.4822,
      "num_input_tokens_seen": 11015992,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 0.4373389482498169,
      "learning_rate": 3.4166360461336206e-05,
      "loss": 0.3541,
      "num_input_tokens_seen": 11019928,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 0.39028212428092957,
      "learning_rate": 3.41557215335353e-05,
      "loss": 0.5212,
      "num_input_tokens_seen": 11023096,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 0.5886793732643127,
      "learning_rate": 3.4145080690461887e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 11026232,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 0.6720621585845947,
      "learning_rate": 3.4134437934341936e-05,
      "loss": 0.6426,
      "num_input_tokens_seen": 11029560,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 0.5611003041267395,
      "learning_rate": 3.4123793267401784e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 11032120,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 0.44706273078918457,
      "learning_rate": 3.411314669186816e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 11034968,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 0.6734988689422607,
      "learning_rate": 3.410249820996823e-05,
      "loss": 0.311,
      "num_input_tokens_seen": 11038456,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.6795942783355713,
      "learning_rate": 3.4091847823929513e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 11042712,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 0.5048241019248962,
      "learning_rate": 3.4081195535979946e-05,
      "loss": 0.5402,
      "num_input_tokens_seen": 11046008,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 0.6654280424118042,
      "learning_rate": 3.40705413483479e-05,
      "loss": 0.5617,
      "num_input_tokens_seen": 11049336,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 0.34393173456192017,
      "learning_rate": 3.405988526326207e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 11053272,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 0.4929460883140564,
      "learning_rate": 3.40492272829516e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 11057848,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 0.5156415104866028,
      "learning_rate": 3.403856740964604e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 11060216,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 0.3922269642353058,
      "learning_rate": 3.4027905645575275e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 11063640,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 0.3747295141220093,
      "learning_rate": 3.401724199296966e-05,
      "loss": 0.3731,
      "num_input_tokens_seen": 11066872,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.4717170298099518,
      "learning_rate": 3.400657645405989e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 11069624,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 0.3392971456050873,
      "learning_rate": 3.3995909031077065e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 11073048,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 0.45570603013038635,
      "learning_rate": 3.398523972625272e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 11076312,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 0.5516793131828308,
      "learning_rate": 3.3974568541818723e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 11079416,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 0.35681188106536865,
      "learning_rate": 3.396389548000738e-05,
      "loss": 0.4997,
      "num_input_tokens_seen": 11082168,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 0.6661454439163208,
      "learning_rate": 3.395322054305137e-05,
      "loss": 0.3449,
      "num_input_tokens_seen": 11084568,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 0.2819897532463074,
      "learning_rate": 3.394254373318376e-05,
      "loss": 0.4101,
      "num_input_tokens_seen": 11087736,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 0.35106027126312256,
      "learning_rate": 3.3931865052638026e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 11090648,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 0.4504992961883545,
      "learning_rate": 3.392118450364802e-05,
      "loss": 0.6442,
      "num_input_tokens_seen": 11093240,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 0.469844251871109,
      "learning_rate": 3.391050208844799e-05,
      "loss": 0.5657,
      "num_input_tokens_seen": 11096568,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 0.2986105978488922,
      "learning_rate": 3.389981780927259e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 11099384,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 0.8243834972381592,
      "learning_rate": 3.3889131668356834e-05,
      "loss": 0.2844,
      "num_input_tokens_seen": 11101944,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 0.41205689311027527,
      "learning_rate": 3.387844366793614e-05,
      "loss": 0.6072,
      "num_input_tokens_seen": 11104856,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 0.5776448845863342,
      "learning_rate": 3.3867753810246326e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 11108440,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 0.3983471393585205,
      "learning_rate": 3.3857062097523586e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 11112504,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 0.7811423540115356,
      "learning_rate": 3.38463685320045e-05,
      "loss": 0.5284,
      "num_input_tokens_seen": 11116024,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 0.4509957432746887,
      "learning_rate": 3.3835673115926045e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 11119128,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 0.42605963349342346,
      "learning_rate": 3.382497585152557e-05,
      "loss": 0.5547,
      "num_input_tokens_seen": 11122584,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 0.3331858217716217,
      "learning_rate": 3.3814276741040835e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 11125592,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 0.346168577671051,
      "learning_rate": 3.380357578670996e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 11128312,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 0.3008742332458496,
      "learning_rate": 3.379287299077146e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 11132120,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 0.42539313435554504,
      "learning_rate": 3.3782168355464263e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 11135768,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 0.3924963176250458,
      "learning_rate": 3.377146188302762e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 11139128,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 0.37807461619377136,
      "learning_rate": 3.376075357570124e-05,
      "loss": 0.6437,
      "num_input_tokens_seen": 11142360,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 0.39613619446754456,
      "learning_rate": 3.3750043435725144e-05,
      "loss": 0.3856,
      "num_input_tokens_seen": 11144984,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 0.6207303404808044,
      "learning_rate": 3.373933146533979e-05,
      "loss": 0.5114,
      "num_input_tokens_seen": 11147672,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 0.35695382952690125,
      "learning_rate": 3.3728617666786e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 11150872,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 0.42186328768730164,
      "learning_rate": 3.371790204230497e-05,
      "loss": 0.4768,
      "num_input_tokens_seen": 11154360,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 0.47001275420188904,
      "learning_rate": 3.3707184594138295e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 11156984,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 0.4337555766105652,
      "learning_rate": 3.369646532452795e-05,
      "loss": 0.7103,
      "num_input_tokens_seen": 11160696,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 0.3354576528072357,
      "learning_rate": 3.3685744235716254e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 11163800,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 0.4218921959400177,
      "learning_rate": 3.367502132994598e-05,
      "loss": 0.5301,
      "num_input_tokens_seen": 11167480,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 0.5182685852050781,
      "learning_rate": 3.366429660946019e-05,
      "loss": 0.4947,
      "num_input_tokens_seen": 11170232,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 0.5480759143829346,
      "learning_rate": 3.365357007650241e-05,
      "loss": 0.6944,
      "num_input_tokens_seen": 11173400,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 0.46241551637649536,
      "learning_rate": 3.364284173331649e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 11176472,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 0.426377534866333,
      "learning_rate": 3.3632111582146686e-05,
      "loss": 0.4009,
      "num_input_tokens_seen": 11179288,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 0.5871288776397705,
      "learning_rate": 3.362137962523761e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 11182136,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 0.3979763090610504,
      "learning_rate": 3.361064586483429e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 11186296,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 0.3964279592037201,
      "learning_rate": 3.359991030318208e-05,
      "loss": 0.3326,
      "num_input_tokens_seen": 11189304,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 0.49968862533569336,
      "learning_rate": 3.358917294252675e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 11192024,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 0.4207405745983124,
      "learning_rate": 3.357843378511443e-05,
      "loss": 0.5215,
      "num_input_tokens_seen": 11195160,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 0.7722777128219604,
      "learning_rate": 3.3567692833191635e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 11198264,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 0.36080560088157654,
      "learning_rate": 3.355695008900525e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 11202840,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.4552430212497711,
      "eval_runtime": 9.7561,
      "eval_samples_per_second": 86.92,
      "eval_steps_per_second": 21.73,
      "num_input_tokens_seen": 11204000,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 0.42479485273361206,
      "learning_rate": 3.354620555480251e-05,
      "loss": 0.4972,
      "num_input_tokens_seen": 11206048,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 0.44665810465812683,
      "learning_rate": 3.353545923283108e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 11210464,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 0.588491678237915,
      "learning_rate": 3.352471112533895e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 11213312,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 0.3772318959236145,
      "learning_rate": 3.351396123457451e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 11216704,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 0.4061695337295532,
      "learning_rate": 3.350320956278652e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 11221248,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 0.41621264815330505,
      "learning_rate": 3.349245611222409e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 11223904,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 0.3082529604434967,
      "learning_rate": 3.348170088513673e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 11227648,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 0.5195984244346619,
      "learning_rate": 3.347094388377431e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 11230432,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 0.5620493292808533,
      "learning_rate": 3.3460185110387076e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 11233536,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 0.5436208844184875,
      "learning_rate": 3.344942456722563e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 11236512,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 0.5154839158058167,
      "learning_rate": 3.343866225654096e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 11238656,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 0.2789956331253052,
      "learning_rate": 3.342789818058443e-05,
      "loss": 0.5224,
      "num_input_tokens_seen": 11242400,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 0.5919594168663025,
      "learning_rate": 3.341713234160774e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 11245824,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 0.6163651943206787,
      "learning_rate": 3.3406364741862986e-05,
      "loss": 0.3809,
      "num_input_tokens_seen": 11248736,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 0.48980844020843506,
      "learning_rate": 3.3395595383602644e-05,
      "loss": 0.551,
      "num_input_tokens_seen": 11252096,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 0.4845641851425171,
      "learning_rate": 3.338482426907953e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 11258496,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 0.7105169296264648,
      "learning_rate": 3.337405140054683e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 11260960,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 0.639907956123352,
      "learning_rate": 3.336327678025813e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 11264512,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 0.43056735396385193,
      "learning_rate": 3.3352500410467324e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 11268256,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 0.30225059390068054,
      "learning_rate": 3.3341722293428726e-05,
      "loss": 0.6073,
      "num_input_tokens_seen": 11271328,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 0.44795727729797363,
      "learning_rate": 3.3330942431397e-05,
      "loss": 0.3899,
      "num_input_tokens_seen": 11274048,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.26128965616226196,
      "learning_rate": 3.332016082662715e-05,
      "loss": 0.5414,
      "num_input_tokens_seen": 11277440,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 0.3821469843387604,
      "learning_rate": 3.3309377481374584e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 11281120,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 0.25690150260925293,
      "learning_rate": 3.3298592397895046e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 11284064,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 0.41977646946907043,
      "learning_rate": 3.3287805578444645e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 11288096,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 0.2307824194431305,
      "learning_rate": 3.327701702527989e-05,
      "loss": 0.3582,
      "num_input_tokens_seen": 11292032,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 0.49944350123405457,
      "learning_rate": 3.3266226740657584e-05,
      "loss": 0.3835,
      "num_input_tokens_seen": 11294720,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 0.3770239055156708,
      "learning_rate": 3.3255434726834954e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 11298880,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 0.3917525112628937,
      "learning_rate": 3.324464098606957e-05,
      "loss": 0.629,
      "num_input_tokens_seen": 11302336,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 0.533774197101593,
      "learning_rate": 3.323384552061934e-05,
      "loss": 0.3588,
      "num_input_tokens_seen": 11305216,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 0.44155797362327576,
      "learning_rate": 3.322304833274257e-05,
      "loss": 0.5871,
      "num_input_tokens_seen": 11308512,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 0.5234484076499939,
      "learning_rate": 3.32122494246979e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 11312416,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 0.341035932302475,
      "learning_rate": 3.320144879874434e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 11315904,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 0.09153519570827484,
      "learning_rate": 3.319064645714126e-05,
      "loss": 0.3032,
      "num_input_tokens_seen": 11321664,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 0.36632195115089417,
      "learning_rate": 3.3179842402148386e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 11325728,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 0.42416420578956604,
      "learning_rate": 3.316903663602581e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 11329344,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 0.7399371266365051,
      "learning_rate": 3.315822916103395e-05,
      "loss": 0.4368,
      "num_input_tokens_seen": 11331840,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 0.5971871614456177,
      "learning_rate": 3.3147419979433635e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 11334912,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 0.5796435475349426,
      "learning_rate": 3.313660909348601e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 11337344,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 0.32187673449516296,
      "learning_rate": 3.312579650545259e-05,
      "loss": 0.5909,
      "num_input_tokens_seen": 11340384,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 1.0043821334838867,
      "learning_rate": 3.311498221759524e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 11342976,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 0.7843081951141357,
      "learning_rate": 3.310416623217619e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 11345504,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 0.2943434715270996,
      "learning_rate": 3.309334855145803e-05,
      "loss": 0.4267,
      "num_input_tokens_seen": 11349216,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 0.39955657720565796,
      "learning_rate": 3.3082529177703684e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 11352576,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 0.4242601990699768,
      "learning_rate": 3.3071708113176435e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 11355104,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 0.44109803438186646,
      "learning_rate": 3.306088536013995e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 11361504,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 0.292744517326355,
      "learning_rate": 3.3050060920858206e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 11365120,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 0.3292374908924103,
      "learning_rate": 3.3039234797595544e-05,
      "loss": 0.5234,
      "num_input_tokens_seen": 11368800,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 0.5175161361694336,
      "learning_rate": 3.3028406992616695e-05,
      "loss": 0.6573,
      "num_input_tokens_seen": 11371616,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 0.3203207850456238,
      "learning_rate": 3.301757750818669e-05,
      "loss": 0.5185,
      "num_input_tokens_seen": 11374496,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 0.35195255279541016,
      "learning_rate": 3.300674634657094e-05,
      "loss": 0.4177,
      "num_input_tokens_seen": 11378624,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 0.5596160888671875,
      "learning_rate": 3.2995913510035206e-05,
      "loss": 0.5203,
      "num_input_tokens_seen": 11381312,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 0.6703624129295349,
      "learning_rate": 3.298507900084558e-05,
      "loss": 0.3714,
      "num_input_tokens_seen": 11383808,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 0.37881267070770264,
      "learning_rate": 3.2974242821268524e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 11387840,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 0.3283602297306061,
      "learning_rate": 3.2963404973570855e-05,
      "loss": 0.3672,
      "num_input_tokens_seen": 11392800,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 0.5435294508934021,
      "learning_rate": 3.2952565460019715e-05,
      "loss": 0.3231,
      "num_input_tokens_seen": 11396224,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 0.35081538558006287,
      "learning_rate": 3.29417242828826e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 11399456,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 0.3462441563606262,
      "learning_rate": 3.2930881444427374e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 11402304,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 0.4591725468635559,
      "learning_rate": 3.292003694692223e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 11405440,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 0.41317451000213623,
      "learning_rate": 3.290919079263572e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 11408256,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 0.28163769841194153,
      "learning_rate": 3.2898342983836715e-05,
      "loss": 0.3624,
      "num_input_tokens_seen": 11412288,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 0.8281683325767517,
      "learning_rate": 3.288749352279447e-05,
      "loss": 0.4327,
      "num_input_tokens_seen": 11414816,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 0.36878421902656555,
      "learning_rate": 3.2876642411778573e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 11417632,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 0.3436899483203888,
      "learning_rate": 3.286578965305894e-05,
      "loss": 0.5924,
      "num_input_tokens_seen": 11420384,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 0.38454288244247437,
      "learning_rate": 3.285493524890585e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 11423648,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 0.26414090394973755,
      "learning_rate": 3.284407920158992e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 11426976,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 0.49608927965164185,
      "learning_rate": 3.283322151338211e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 11430304,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 0.8183252811431885,
      "learning_rate": 3.282236218655373e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 11432704,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 0.3258975148200989,
      "learning_rate": 3.281150122337642e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 11436448,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 0.5876390337944031,
      "learning_rate": 3.280063862612216e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 11439872,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 0.9678620100021362,
      "learning_rate": 3.2789774397063314e-05,
      "loss": 0.5509,
      "num_input_tokens_seen": 11442304,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 0.3673587143421173,
      "learning_rate": 3.277890853847253e-05,
      "loss": 0.6147,
      "num_input_tokens_seen": 11445856,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 0.6457005143165588,
      "learning_rate": 3.2768041052622824e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 11452288,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 0.4997417628765106,
      "learning_rate": 3.275717194178755e-05,
      "loss": 0.5563,
      "num_input_tokens_seen": 11456032,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 0.36424002051353455,
      "learning_rate": 3.27463012082404e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 11459392,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 0.6588851809501648,
      "learning_rate": 3.2735428854255416e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 11462336,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 0.9546771049499512,
      "learning_rate": 3.272455488210697e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 11464896,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 0.5354321599006653,
      "learning_rate": 3.271367929406977e-05,
      "loss": 0.4249,
      "num_input_tokens_seen": 11472224,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 0.5646787881851196,
      "learning_rate": 3.270280209241886e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 11474912,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 0.8588024377822876,
      "learning_rate": 3.2691923279429624e-05,
      "loss": 0.6589,
      "num_input_tokens_seen": 11478240,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 0.46713680028915405,
      "learning_rate": 3.2681042857377804e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 11481216,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 0.451274037361145,
      "learning_rate": 3.2670160828539435e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 11484384,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 0.7934378981590271,
      "learning_rate": 3.2659277195190934e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 11487104,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 0.4628658890724182,
      "learning_rate": 3.2648391959609026e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 11489696,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 1.111345648765564,
      "learning_rate": 3.263750512407078e-05,
      "loss": 0.5224,
      "num_input_tokens_seen": 11492512,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 0.3407762348651886,
      "learning_rate": 3.262661669085359e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 11495648,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 0.3694557249546051,
      "learning_rate": 3.2615726662235205e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 11498976,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 0.5129593014717102,
      "learning_rate": 3.2604835040493676e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 11501888,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 0.5176270008087158,
      "learning_rate": 3.259394182790742e-05,
      "loss": 0.5358,
      "num_input_tokens_seen": 11504544,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 0.41450268030166626,
      "learning_rate": 3.2583047026755174e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 11507168,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 0.4496156871318817,
      "learning_rate": 3.2572150639316e-05,
      "loss": 0.3957,
      "num_input_tokens_seen": 11510272,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 0.6893298625946045,
      "learning_rate": 3.256125266786931e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 11513280,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 0.2559066712856293,
      "learning_rate": 3.255035311469481e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 11516384,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.6999375820159912,
      "learning_rate": 3.2539451982072586e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 11520672,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 0.5043221712112427,
      "learning_rate": 3.2528549272283016e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 11523456,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 0.6440033316612244,
      "learning_rate": 3.251764498760683e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 11526240,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 0.38348400592803955,
      "learning_rate": 3.2506739130325084e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 11531040,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 0.36203309893608093,
      "learning_rate": 3.2495831702719146e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 11533984,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 0.5005424618721008,
      "learning_rate": 3.248492270707074e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 11537248,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 0.9867222905158997,
      "learning_rate": 3.247401214566189e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 11540384,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 0.4553893804550171,
      "learning_rate": 3.2463100020774965e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 11543424,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 0.2942312955856323,
      "learning_rate": 3.245218633469267e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 11547136,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 0.7874212861061096,
      "learning_rate": 3.244127108969801e-05,
      "loss": 0.5774,
      "num_input_tokens_seen": 11550528,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 0.3128491938114166,
      "learning_rate": 3.2430354288074314e-05,
      "loss": 0.3436,
      "num_input_tokens_seen": 11553440,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 0.3269175887107849,
      "learning_rate": 3.241943593210529e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 11557184,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 0.6618948578834534,
      "learning_rate": 3.2408516024074916e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 11560416,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 0.37202486395835876,
      "learning_rate": 3.2397594566267505e-05,
      "loss": 0.4649,
      "num_input_tokens_seen": 11563712,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 0.28726908564567566,
      "learning_rate": 3.23866715609677e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 11567776,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 0.3455909192562103,
      "learning_rate": 3.237574701046049e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 11571456,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 0.380852073431015,
      "learning_rate": 3.236482091703115e-05,
      "loss": 0.4862,
      "num_input_tokens_seen": 11575136,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 0.28178462386131287,
      "learning_rate": 3.235389328296528e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 11578880,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 0.3861263692378998,
      "learning_rate": 3.2342964110548845e-05,
      "loss": 0.3244,
      "num_input_tokens_seen": 11581792,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 0.32679325342178345,
      "learning_rate": 3.2332033402068093e-05,
      "loss": 0.5751,
      "num_input_tokens_seen": 11584704,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 0.6852577924728394,
      "learning_rate": 3.2321101159809605e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 11587936,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 0.8411333560943604,
      "learning_rate": 3.231016738606027e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 11591232,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 0.41853779554367065,
      "learning_rate": 3.2299232083107313e-05,
      "loss": 0.6495,
      "num_input_tokens_seen": 11594656,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 0.3008008599281311,
      "learning_rate": 3.228829525323829e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 11598016,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 0.42911648750305176,
      "learning_rate": 3.227735689874104e-05,
      "loss": 0.3717,
      "num_input_tokens_seen": 11601312,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 0.43566447496414185,
      "learning_rate": 3.226641702190375e-05,
      "loss": 0.5532,
      "num_input_tokens_seen": 11604896,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 0.489303320646286,
      "learning_rate": 3.225547562501493e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 11607840,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 0.46069788932800293,
      "learning_rate": 3.2244532710363375e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 11611392,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 0.3335775136947632,
      "learning_rate": 3.2233588280238224e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 11614752,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 0.49853187799453735,
      "learning_rate": 3.2222642336928944e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 11617728,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 0.5365101099014282,
      "learning_rate": 3.2211694882725275e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 11620736,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 0.3806530833244324,
      "learning_rate": 3.220074591991731e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 11623744,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 0.4805380702018738,
      "learning_rate": 3.218979545079546e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 11626496,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 0.44891756772994995,
      "learning_rate": 3.217884347765042e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 11629504,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 0.6081233024597168,
      "learning_rate": 3.216789000277323e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 11632384,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 0.5915145874023438,
      "learning_rate": 3.2156935028455224e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 11635616,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 0.47178539633750916,
      "learning_rate": 3.2145978556988056e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 11638912,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 0.5075383186340332,
      "learning_rate": 3.213502059066371e-05,
      "loss": 0.4964,
      "num_input_tokens_seen": 11642432,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 0.6123031973838806,
      "learning_rate": 3.212406113177445e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 11645120,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 0.4767598509788513,
      "learning_rate": 3.211310018261287e-05,
      "loss": 0.4211,
      "num_input_tokens_seen": 11647456,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 0.32794222235679626,
      "learning_rate": 3.21021377454719e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 11650976,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 0.5100332498550415,
      "learning_rate": 3.209117382264473e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 11654304,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 0.5250400900840759,
      "learning_rate": 3.208020841642489e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 11658592,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 0.5155608654022217,
      "learning_rate": 3.206924152910623e-05,
      "loss": 0.532,
      "num_input_tokens_seen": 11662464,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 0.37577560544013977,
      "learning_rate": 3.205827316298289e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 11665248,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 0.6026126146316528,
      "learning_rate": 3.204730332034934e-05,
      "loss": 0.5066,
      "num_input_tokens_seen": 11668256,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 0.39818906784057617,
      "learning_rate": 3.2036332003500324e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 11671264,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 0.5046424269676208,
      "learning_rate": 3.2025359214730924e-05,
      "loss": 0.6395,
      "num_input_tokens_seen": 11673728,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 0.4388717710971832,
      "learning_rate": 3.201438495633654e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 11676960,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 0.8627055883407593,
      "learning_rate": 3.200340923061284e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 11681440,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 0.5015323758125305,
      "learning_rate": 3.199243203985583e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 11684544,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 0.34582582116127014,
      "learning_rate": 3.1981453386361804e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 11687584,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 0.32921260595321655,
      "learning_rate": 3.1970473272427386e-05,
      "loss": 0.3574,
      "num_input_tokens_seen": 11690944,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 0.4529435336589813,
      "learning_rate": 3.195949170034948e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 11694112,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 0.318999707698822,
      "learning_rate": 3.19485086724253e-05,
      "loss": 0.4659,
      "num_input_tokens_seen": 11697376,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 0.6572704315185547,
      "learning_rate": 3.193752419095239e-05,
      "loss": 0.3695,
      "num_input_tokens_seen": 11700832,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 0.46482911705970764,
      "learning_rate": 3.192653825822857e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 11703520,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 0.4151599407196045,
      "learning_rate": 3.1915550876551964e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 11707008,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 0.44363266229629517,
      "learning_rate": 3.190456204822101e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 11709408,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 0.5922812819480896,
      "learning_rate": 3.1893571775534445e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 11712000,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 0.3361526429653168,
      "learning_rate": 3.1882580060791315e-05,
      "loss": 0.5224,
      "num_input_tokens_seen": 11716064,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 0.4570101201534271,
      "learning_rate": 3.187158690629095e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 11719776,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 0.4963303208351135,
      "learning_rate": 3.1860592314333e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 11722880,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 0.5027836561203003,
      "learning_rate": 3.184959628721742e-05,
      "loss": 0.6203,
      "num_input_tokens_seen": 11725824,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 0.7142764329910278,
      "learning_rate": 3.183859882724443e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 11728480,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 0.2665795683860779,
      "learning_rate": 3.1827599936714584e-05,
      "loss": 0.4303,
      "num_input_tokens_seen": 11731040,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 0.32804223895072937,
      "learning_rate": 3.1816599617928736e-05,
      "loss": 0.3553,
      "num_input_tokens_seen": 11733664,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 0.4540661573410034,
      "learning_rate": 3.1805597873188e-05,
      "loss": 0.3866,
      "num_input_tokens_seen": 11737216,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 0.4587283730506897,
      "learning_rate": 3.179459470479384e-05,
      "loss": 0.4164,
      "num_input_tokens_seen": 11740864,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 0.3113430440425873,
      "learning_rate": 3.1783590115047983e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 11743776,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 0.3682987093925476,
      "learning_rate": 3.177258410625246e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 11747552,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 0.488596111536026,
      "learning_rate": 3.176157668070962e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 11750656,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.5235497951507568,
      "learning_rate": 3.175056784072207e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 11753888,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 0.8016236424446106,
      "learning_rate": 3.173955758859274e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 11757184,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 0.4174121022224426,
      "learning_rate": 3.172854592662486e-05,
      "loss": 0.4061,
      "num_input_tokens_seen": 11759936,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 0.5425742268562317,
      "learning_rate": 3.1717532857121926e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 11762816,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.41373881697654724,
      "learning_rate": 3.170651838238776e-05,
      "loss": 0.3895,
      "num_input_tokens_seen": 11765824,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 0.5385342240333557,
      "learning_rate": 3.169550250472646e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 11769056,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 0.5261901021003723,
      "learning_rate": 3.168448522644243e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 11771808,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 0.37852051854133606,
      "learning_rate": 3.167346654984034e-05,
      "loss": 0.3493,
      "num_input_tokens_seen": 11774720,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 1.0828863382339478,
      "learning_rate": 3.166244647722519e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 11777664,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 0.366608202457428,
      "learning_rate": 3.1651425010902244e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 11781696,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 0.47694098949432373,
      "learning_rate": 3.164040215317706e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 11784672,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 0.40320539474487305,
      "learning_rate": 3.162937790635552e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 11787424,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 0.7897581458091736,
      "learning_rate": 3.1618352272743746e-05,
      "loss": 0.5119,
      "num_input_tokens_seen": 11789824,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 0.4468761682510376,
      "learning_rate": 3.160732525464818e-05,
      "loss": 0.2593,
      "num_input_tokens_seen": 11793120,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 0.4796367585659027,
      "learning_rate": 3.1596296854375554e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 11798720,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 0.4692193567752838,
      "learning_rate": 3.1585267074232884e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 11801600,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 0.7237116098403931,
      "learning_rate": 3.157423591652746e-05,
      "loss": 0.3205,
      "num_input_tokens_seen": 11804640,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 0.4831824004650116,
      "learning_rate": 3.1563203383566894e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 11807296,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 0.4185318350791931,
      "learning_rate": 3.155216947765906e-05,
      "loss": 0.3343,
      "num_input_tokens_seen": 11809920,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 0.4356648325920105,
      "learning_rate": 3.154113420111211e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 11813440,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 0.4024970829486847,
      "learning_rate": 3.1530097556234526e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 11817440,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 0.48012053966522217,
      "learning_rate": 3.151905954533502e-05,
      "loss": 0.5374,
      "num_input_tokens_seen": 11820192,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 0.47257861495018005,
      "learning_rate": 3.150802017072263e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 11824192,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 0.49471044540405273,
      "learning_rate": 3.149697943470668e-05,
      "loss": 0.5912,
      "num_input_tokens_seen": 11826944,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 0.4271036982536316,
      "learning_rate": 3.1485937339596736e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 11830784,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 0.7991412878036499,
      "learning_rate": 3.147489388770271e-05,
      "loss": 0.6163,
      "num_input_tokens_seen": 11833632,
      "step": 18125
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 0.524199903011322,
      "learning_rate": 3.146384908133475e-05,
      "loss": 0.5637,
      "num_input_tokens_seen": 11837056,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 0.5016895532608032,
      "learning_rate": 3.14528029228033e-05,
      "loss": 0.3719,
      "num_input_tokens_seen": 11840736,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 0.5176773071289062,
      "learning_rate": 3.144175541441911e-05,
      "loss": 0.6236,
      "num_input_tokens_seen": 11843328,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 0.33007559180259705,
      "learning_rate": 3.1430706558493166e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 11846784,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 0.3046227693557739,
      "learning_rate": 3.141965635733677e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 11849984,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 0.5192352533340454,
      "learning_rate": 3.140860481326152e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 11852992,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 0.9087638854980469,
      "learning_rate": 3.139755192857925e-05,
      "loss": 0.4263,
      "num_input_tokens_seen": 11855776,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 0.3820173740386963,
      "learning_rate": 3.1386497705602094e-05,
      "loss": 0.3468,
      "num_input_tokens_seen": 11859008,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 0.5916977524757385,
      "learning_rate": 3.137544214664249e-05,
      "loss": 0.3683,
      "num_input_tokens_seen": 11861632,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 0.4997282028198242,
      "learning_rate": 3.136438525401312e-05,
      "loss": 0.3737,
      "num_input_tokens_seen": 11864864,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 0.37722569704055786,
      "learning_rate": 3.135332703002696e-05,
      "loss": 0.508,
      "num_input_tokens_seen": 11868224,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 0.4179837703704834,
      "learning_rate": 3.1342267476997276e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 11871072,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 0.5480092167854309,
      "learning_rate": 3.1331206597237584e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 11873888,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 0.7286959290504456,
      "learning_rate": 3.1320144393061715e-05,
      "loss": 0.3972,
      "num_input_tokens_seen": 11876608,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 0.2774248421192169,
      "learning_rate": 3.130908086678373e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 11880864,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 0.4400698244571686,
      "learning_rate": 3.1298016020718014e-05,
      "loss": 0.4393,
      "num_input_tokens_seen": 11883520,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 0.38563674688339233,
      "learning_rate": 3.128694985717919e-05,
      "loss": 0.3556,
      "num_input_tokens_seen": 11887168,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 0.5326460003852844,
      "learning_rate": 3.1275882378482185e-05,
      "loss": 0.5847,
      "num_input_tokens_seen": 11890976,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 0.5083280205726624,
      "learning_rate": 3.1264813586942185e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 11893888,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 0.5180025696754456,
      "learning_rate": 3.125374348487466e-05,
      "loss": 0.448,
      "num_input_tokens_seen": 11896704,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 0.3829621374607086,
      "learning_rate": 3.1242672074595334e-05,
      "loss": 0.5426,
      "num_input_tokens_seen": 11900736,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 0.34245699644088745,
      "learning_rate": 3.123159935842024e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 11904160,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 0.3002576231956482,
      "learning_rate": 3.122052533866565e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 11907200,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 0.3225947916507721,
      "learning_rate": 3.120945001764813e-05,
      "loss": 0.3809,
      "num_input_tokens_seen": 11910368,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 0.4563891589641571,
      "learning_rate": 3.11983733976845e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 11913472,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 0.42592668533325195,
      "learning_rate": 3.1187295481091876e-05,
      "loss": 0.3713,
      "num_input_tokens_seen": 11917280,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 0.469501793384552,
      "learning_rate": 3.1176216270187625e-05,
      "loss": 0.3914,
      "num_input_tokens_seen": 11919744,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 0.5636067986488342,
      "learning_rate": 3.116513576728939e-05,
      "loss": 0.5522,
      "num_input_tokens_seen": 11924096,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 0.5939868688583374,
      "learning_rate": 3.1154053974715085e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 11927744,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 1.5604735612869263,
      "learning_rate": 3.1142970894782906e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 11930368,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 0.4137036204338074,
      "learning_rate": 3.1131886529811296e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 11933344,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 0.4260804355144501,
      "learning_rate": 3.1120800882118985e-05,
      "loss": 0.3627,
      "num_input_tokens_seen": 11936992,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 0.5417341589927673,
      "learning_rate": 3.1109713954024945e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 11940288,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 0.7070181369781494,
      "learning_rate": 3.109862574784846e-05,
      "loss": 0.3852,
      "num_input_tokens_seen": 11943072,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 0.4938617944717407,
      "learning_rate": 3.108753626590905e-05,
      "loss": 0.3714,
      "num_input_tokens_seen": 11946016,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 0.6762165427207947,
      "learning_rate": 3.1076445510526486e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 11949280,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 0.36432337760925293,
      "learning_rate": 3.106535348402086e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 11953216,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 0.4007055461406708,
      "learning_rate": 3.105426018871247e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 11956320,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 0.5102608799934387,
      "learning_rate": 3.104316562692193e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 11959104,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 0.4459886848926544,
      "learning_rate": 3.103206980097009e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 11962208,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 0.6512212157249451,
      "learning_rate": 3.102097271317805e-05,
      "loss": 0.6259,
      "num_input_tokens_seen": 11964832,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 0.3714168667793274,
      "learning_rate": 3.100987436586721e-05,
      "loss": 0.6601,
      "num_input_tokens_seen": 11968000,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 0.40723904967308044,
      "learning_rate": 3.099877476135923e-05,
      "loss": 0.5737,
      "num_input_tokens_seen": 11970912,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 0.345341295003891,
      "learning_rate": 3.0987673901976e-05,
      "loss": 0.3313,
      "num_input_tokens_seen": 11973472,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 0.6473419070243835,
      "learning_rate": 3.0976571790039697e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 11976864,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 0.5649501085281372,
      "learning_rate": 3.096546842787277e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 11979872,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.3130391538143158,
      "learning_rate": 3.095436381779789e-05,
      "loss": 0.4421,
      "num_input_tokens_seen": 11983168,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 0.5079391002655029,
      "learning_rate": 3.0943257962138045e-05,
      "loss": 0.4576,
      "num_input_tokens_seen": 11986432,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 0.7035865187644958,
      "learning_rate": 3.093215086321644e-05,
      "loss": 0.2873,
      "num_input_tokens_seen": 11988512,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 0.5036592483520508,
      "learning_rate": 3.092104252335654e-05,
      "loss": 0.6828,
      "num_input_tokens_seen": 11991424,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 0.4197594225406647,
      "learning_rate": 3.09099329448821e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 11994208,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 0.5889665484428406,
      "learning_rate": 3.089882213011711e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 11997312,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 0.660692572593689,
      "learning_rate": 3.088771008138583e-05,
      "loss": 0.5798,
      "num_input_tokens_seen": 12000288,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 0.5397717952728271,
      "learning_rate": 3.087659680101278e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 12003328,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 0.9124352335929871,
      "learning_rate": 3.086548229132271e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 12006176,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 0.562472403049469,
      "learning_rate": 3.085436655464067e-05,
      "loss": 0.5722,
      "num_input_tokens_seen": 12009856,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 0.280860960483551,
      "learning_rate": 3.084324959329193e-05,
      "loss": 0.345,
      "num_input_tokens_seen": 12013376,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 0.43809106945991516,
      "learning_rate": 3.083213140960204e-05,
      "loss": 0.614,
      "num_input_tokens_seen": 12016192,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 0.44459250569343567,
      "learning_rate": 3.08210120058968e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 12019680,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 0.2722587287425995,
      "learning_rate": 3.080989138450225e-05,
      "loss": 0.4012,
      "num_input_tokens_seen": 12023200,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 0.4905107319355011,
      "learning_rate": 3.0798769547744713e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 12026880,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 0.5052523016929626,
      "learning_rate": 3.0787646497950736e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 12029472,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 0.48969241976737976,
      "learning_rate": 3.0776522237447146e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 12032128,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 0.28915348649024963,
      "learning_rate": 3.076539676856101e-05,
      "loss": 0.3878,
      "num_input_tokens_seen": 12034912,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 0.3673774003982544,
      "learning_rate": 3.0754270093619634e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 12037888,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 0.7197951078414917,
      "learning_rate": 3.074314221495061e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 12040576,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 0.3797397017478943,
      "learning_rate": 3.0732013134881756e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 12044096,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 0.5979658365249634,
      "learning_rate": 3.072088285574114e-05,
      "loss": 0.3852,
      "num_input_tokens_seen": 12047424,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 0.36565685272216797,
      "learning_rate": 3.070975137985711e-05,
      "loss": 0.3541,
      "num_input_tokens_seen": 12050560,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 0.5095211267471313,
      "learning_rate": 3.0698618709558226e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 12053952,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 0.5164008736610413,
      "learning_rate": 3.068748484717332e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 12056416,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 0.3992116451263428,
      "learning_rate": 3.067634979503147e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 12060448,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 0.2995305061340332,
      "learning_rate": 3.0665213555462e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 12064960,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 0.33559927344322205,
      "learning_rate": 3.0654076130794494e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 12068192,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 0.7202004194259644,
      "learning_rate": 3.064293752335876e-05,
      "loss": 0.3868,
      "num_input_tokens_seen": 12071200,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 0.556877076625824,
      "learning_rate": 3.063179773548487e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 12074112,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 1.8017425537109375,
      "learning_rate": 3.0620656769503164e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 12078048,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 0.5106170773506165,
      "learning_rate": 3.060951462774417e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 12080896,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 0.34584349393844604,
      "learning_rate": 3.0598371312538714e-05,
      "loss": 0.3832,
      "num_input_tokens_seen": 12083744,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 0.28068748116493225,
      "learning_rate": 3.058722682621785e-05,
      "loss": 0.4132,
      "num_input_tokens_seen": 12086496,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 0.40249139070510864,
      "learning_rate": 3.0576081171112875e-05,
      "loss": 0.3942,
      "num_input_tokens_seen": 12089728,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 0.6573852300643921,
      "learning_rate": 3.056493434955534e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 12093120,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 0.4894905388355255,
      "learning_rate": 3.0553786363877027e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 12095680,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 0.5274435877799988,
      "learning_rate": 3.054263721640996e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 12099040,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 0.2905617356300354,
      "learning_rate": 3.053148690948642e-05,
      "loss": 0.3378,
      "num_input_tokens_seen": 12102112,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 0.5007463097572327,
      "learning_rate": 3.0520335445438935e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 12104544,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 0.48516976833343506,
      "learning_rate": 3.0509182826600246e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 12107712,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 0.3739664852619171,
      "learning_rate": 3.049802905530337e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 12110528,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 0.5281677842140198,
      "learning_rate": 3.0486874133881533e-05,
      "loss": 0.4406,
      "num_input_tokens_seen": 12113600,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 0.34179967641830444,
      "learning_rate": 3.047571806466823e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 12116896,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 0.38173335790634155,
      "learning_rate": 3.0464560849997177e-05,
      "loss": 0.3728,
      "num_input_tokens_seen": 12120832,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 0.2869454026222229,
      "learning_rate": 3.0453402492202325e-05,
      "loss": 0.3166,
      "num_input_tokens_seen": 12124704,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 0.5900119543075562,
      "learning_rate": 3.0442242993617904e-05,
      "loss": 0.3677,
      "num_input_tokens_seen": 12129760,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 1.1802841424942017,
      "learning_rate": 3.0431082356578334e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 12132096,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 0.3883656859397888,
      "learning_rate": 3.0419920583418294e-05,
      "loss": 0.5513,
      "num_input_tokens_seen": 12135008,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 0.7886462211608887,
      "learning_rate": 3.0408757676472694e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 12138144,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 0.9444787502288818,
      "learning_rate": 3.0397593638076705e-05,
      "loss": 0.5784,
      "num_input_tokens_seen": 12141568,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 0.37118279933929443,
      "learning_rate": 3.0386428470565702e-05,
      "loss": 0.5428,
      "num_input_tokens_seen": 12145248,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 0.4159270226955414,
      "learning_rate": 3.0375262176275308e-05,
      "loss": 0.3942,
      "num_input_tokens_seen": 12149472,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 0.28463324904441833,
      "learning_rate": 3.03640947575414e-05,
      "loss": 0.5415,
      "num_input_tokens_seen": 12152992,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 0.5945909023284912,
      "learning_rate": 3.0352926216700067e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 12155712,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 0.6215947866439819,
      "learning_rate": 3.034175655608763e-05,
      "loss": 0.5319,
      "num_input_tokens_seen": 12158688,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 0.786897599697113,
      "learning_rate": 3.0330585778040677e-05,
      "loss": 0.7629,
      "num_input_tokens_seen": 12161824,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 0.5733551383018494,
      "learning_rate": 3.031941388489598e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 12164992,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 0.40603160858154297,
      "learning_rate": 3.0308240878990584e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 12167808,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 0.44609004259109497,
      "learning_rate": 3.0297066762661773e-05,
      "loss": 0.2957,
      "num_input_tokens_seen": 12170752,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 0.6247959136962891,
      "learning_rate": 3.0285891538247015e-05,
      "loss": 0.3429,
      "num_input_tokens_seen": 12173664,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 0.46652042865753174,
      "learning_rate": 3.0274715208084053e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 12176960,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 0.7613792419433594,
      "learning_rate": 3.0263537774510847e-05,
      "loss": 0.616,
      "num_input_tokens_seen": 12180096,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 0.5678331851959229,
      "learning_rate": 3.0252359239865585e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 12183424,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 1.039775013923645,
      "learning_rate": 3.0241179606486704e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 12186592,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 0.7254601716995239,
      "learning_rate": 3.0229998876712827e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 12190240,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 1.0929440259933472,
      "learning_rate": 3.0218817052882858e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 12193024,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 0.4643692374229431,
      "learning_rate": 3.020763413733591e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 12196128,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 0.37686654925346375,
      "learning_rate": 3.0196450132411302e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 12198880,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 0.4972955882549286,
      "learning_rate": 3.0185265040448616e-05,
      "loss": 0.4017,
      "num_input_tokens_seen": 12201760,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 0.2728675603866577,
      "learning_rate": 3.017407886378764e-05,
      "loss": 0.6846,
      "num_input_tokens_seen": 12206272,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 0.30911338329315186,
      "learning_rate": 3.016289160476839e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 12209472,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.49392539262771606,
      "learning_rate": 3.0151703265731125e-05,
      "loss": 0.4965,
      "num_input_tokens_seen": 12212256,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 0.2702524662017822,
      "learning_rate": 3.0140513849016305e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 12215456,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 0.30758631229400635,
      "learning_rate": 3.0129323356964627e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 12219936,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 0.5223352313041687,
      "learning_rate": 3.0118131791917032e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 12223200,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 0.5991958975791931,
      "learning_rate": 3.010693915621466e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 12226432,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 0.4128972291946411,
      "learning_rate": 3.009574545219887e-05,
      "loss": 0.4214,
      "num_input_tokens_seen": 12229120,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 0.2596225440502167,
      "learning_rate": 3.0084550682211272e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 12233248,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 0.3049580454826355,
      "learning_rate": 3.0073354848593677e-05,
      "loss": 0.5514,
      "num_input_tokens_seen": 12237696,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 0.373282790184021,
      "learning_rate": 3.0062157953688125e-05,
      "loss": 0.308,
      "num_input_tokens_seen": 12241504,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 1.1055893898010254,
      "learning_rate": 3.0050959999836887e-05,
      "loss": 0.459,
      "num_input_tokens_seen": 12244832,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 0.6465285420417786,
      "learning_rate": 3.003976098938244e-05,
      "loss": 0.4323,
      "num_input_tokens_seen": 12247776,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 0.6597641706466675,
      "learning_rate": 3.0028560924667487e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 12250560,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 0.3788335919380188,
      "learning_rate": 3.001735980803496e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 12253504,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 0.48996299505233765,
      "learning_rate": 3.0006157641828004e-05,
      "loss": 0.4177,
      "num_input_tokens_seen": 12256128,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 0.281606525182724,
      "learning_rate": 2.9994954428389977e-05,
      "loss": 0.5629,
      "num_input_tokens_seen": 12260032,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 0.33817851543426514,
      "learning_rate": 2.9983750170064472e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 12263136,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 1.235662817955017,
      "learning_rate": 2.9972544869195292e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 12266496,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 0.5821531414985657,
      "learning_rate": 2.9961338528126447e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 12268768,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 0.2858089506626129,
      "learning_rate": 2.9950131149202192e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 12272160,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 0.4285712242126465,
      "learning_rate": 2.9938922734766973e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 12275776,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 0.510491669178009,
      "learning_rate": 2.992771328716546e-05,
      "loss": 0.4179,
      "num_input_tokens_seen": 12278880,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 0.44356870651245117,
      "learning_rate": 2.991650280874255e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 12281536,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 0.33177465200424194,
      "learning_rate": 2.990529130184334e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 12284416,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 0.2599055767059326,
      "learning_rate": 2.989407876881316e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 12288224,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 0.4166613221168518,
      "learning_rate": 2.988286521199753e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 12291520,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 0.6988981366157532,
      "learning_rate": 2.9871650633742198e-05,
      "loss": 0.5417,
      "num_input_tokens_seen": 12295264,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 0.29345420002937317,
      "learning_rate": 2.9860435036393142e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 12297984,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 0.8005891442298889,
      "learning_rate": 2.9849218422296527e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 12301184,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 0.443805456161499,
      "learning_rate": 2.983800079379873e-05,
      "loss": 0.488,
      "num_input_tokens_seen": 12303968,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 0.5773389935493469,
      "learning_rate": 2.982678215324638e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 12308256,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 0.5423310995101929,
      "learning_rate": 2.9815562502986256e-05,
      "loss": 0.5565,
      "num_input_tokens_seen": 12310816,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 0.3461560010910034,
      "learning_rate": 2.9804341845365398e-05,
      "loss": 0.4442,
      "num_input_tokens_seen": 12314752,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 0.3071780800819397,
      "learning_rate": 2.9793120182731037e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 12318272,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 0.4023772180080414,
      "learning_rate": 2.9781897517430617e-05,
      "loss": 0.5114,
      "num_input_tokens_seen": 12322336,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 0.3892959952354431,
      "learning_rate": 2.977067385181179e-05,
      "loss": 0.4176,
      "num_input_tokens_seen": 12325760,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 0.43225905299186707,
      "learning_rate": 2.9759449188222426e-05,
      "loss": 0.5568,
      "num_input_tokens_seen": 12330304,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 0.5454850196838379,
      "learning_rate": 2.9748223529010582e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 12332960,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 0.3987858295440674,
      "learning_rate": 2.9736996876524546e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 12337536,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 0.81558758020401,
      "learning_rate": 2.9725769233112816e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 12341120,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 0.5032902359962463,
      "learning_rate": 2.9714540601124074e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 12343904,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 0.7651951909065247,
      "learning_rate": 2.9703310982907218e-05,
      "loss": 0.5626,
      "num_input_tokens_seen": 12346912,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 0.7924271821975708,
      "learning_rate": 2.9692080380811365e-05,
      "loss": 0.3778,
      "num_input_tokens_seen": 12350272,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 1.1161117553710938,
      "learning_rate": 2.9680848797185828e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 12352800,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 0.35012295842170715,
      "learning_rate": 2.966961623438012e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 12355616,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 0.8354390263557434,
      "learning_rate": 2.9658382694743973e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 12358784,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 0.43143439292907715,
      "learning_rate": 2.9647148180627317e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 12361920,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 0.3888009488582611,
      "learning_rate": 2.9635912694380275e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 12365600,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 0.43774303793907166,
      "learning_rate": 2.962467623835319e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 12368000,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 0.4925232529640198,
      "learning_rate": 2.961343881489659e-05,
      "loss": 0.4627,
      "num_input_tokens_seen": 12370592,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 0.6107484698295593,
      "learning_rate": 2.9602200426361226e-05,
      "loss": 0.4993,
      "num_input_tokens_seen": 12373760,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 0.41555023193359375,
      "learning_rate": 2.9590961075098038e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 12377600,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 0.941331684589386,
      "learning_rate": 2.957972076345817e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 12381152,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 0.5081667304039001,
      "learning_rate": 2.9568479493792967e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 12384096,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 0.5800533294677734,
      "learning_rate": 2.955723726845397e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 12386848,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 0.7544443607330322,
      "learning_rate": 2.9545994089792928e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 12389344,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 0.46071338653564453,
      "learning_rate": 2.9534749960161795e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 12392640,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 1.550345540046692,
      "learning_rate": 2.9523504881912694e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 12397248,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 0.44675618410110474,
      "learning_rate": 2.9512258857397978e-05,
      "loss": 0.5085,
      "num_input_tokens_seen": 12400352,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 0.5744008421897888,
      "learning_rate": 2.9501011888970203e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 12402944,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 0.6339061856269836,
      "learning_rate": 2.9489763978982075e-05,
      "loss": 0.4369,
      "num_input_tokens_seen": 12405952,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.5320835709571838,
      "learning_rate": 2.9478515129786556e-05,
      "loss": 0.595,
      "num_input_tokens_seen": 12409376,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 0.6202113032341003,
      "learning_rate": 2.946726534373676e-05,
      "loss": 0.4932,
      "num_input_tokens_seen": 12413216,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 0.45903855562210083,
      "learning_rate": 2.945601462318601e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 12416512,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 0.29559019207954407,
      "learning_rate": 2.944476297048786e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 12420192,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 1.5439789295196533,
      "learning_rate": 2.9433510387995987e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 12422464,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 0.4768322706222534,
      "learning_rate": 2.9422256878064325e-05,
      "loss": 0.5715,
      "num_input_tokens_seen": 12425248,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 0.5880764126777649,
      "learning_rate": 2.9411002443046982e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 12428704,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 0.5150425434112549,
      "learning_rate": 2.9399747085298245e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 12432480,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 0.3921326994895935,
      "learning_rate": 2.9388490807172614e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 12434912,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 0.6223648190498352,
      "learning_rate": 2.9377233611024775e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 12438016,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 0.4863746166229248,
      "learning_rate": 2.9365975499209596e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 12440672,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 0.4236801862716675,
      "learning_rate": 2.9354716474082154e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 12443840,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.7078751921653748,
      "learning_rate": 2.9343456537997714e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 12446408,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.4542488157749176,
      "eval_runtime": 9.8355,
      "eval_samples_per_second": 86.218,
      "eval_steps_per_second": 21.554,
      "num_input_tokens_seen": 12446408,
      "step": 19080
    },
    {
      "epoch": 10.002620545073375,
      "grad_norm": 0.23807600140571594,
      "learning_rate": 2.9332195693311714e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 12450472,
      "step": 19085
    },
    {
      "epoch": 10.00524109014675,
      "grad_norm": 0.5745911002159119,
      "learning_rate": 2.9320933942379803e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 12454056,
      "step": 19090
    },
    {
      "epoch": 10.007861635220126,
      "grad_norm": 0.47766220569610596,
      "learning_rate": 2.9309671287557805e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 12457576,
      "step": 19095
    },
    {
      "epoch": 10.0104821802935,
      "grad_norm": 0.6476589441299438,
      "learning_rate": 2.9298407731201737e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 12461608,
      "step": 19100
    },
    {
      "epoch": 10.013102725366876,
      "grad_norm": 0.39752909541130066,
      "learning_rate": 2.928714327566781e-05,
      "loss": 0.3856,
      "num_input_tokens_seen": 12464968,
      "step": 19105
    },
    {
      "epoch": 10.015723270440251,
      "grad_norm": 0.7215709686279297,
      "learning_rate": 2.9275877923312427e-05,
      "loss": 0.5114,
      "num_input_tokens_seen": 12468648,
      "step": 19110
    },
    {
      "epoch": 10.018343815513626,
      "grad_norm": 0.49893712997436523,
      "learning_rate": 2.9264611676492165e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 12471784,
      "step": 19115
    },
    {
      "epoch": 10.020964360587001,
      "grad_norm": 0.5543431043624878,
      "learning_rate": 2.9253344537563782e-05,
      "loss": 0.5928,
      "num_input_tokens_seen": 12474632,
      "step": 19120
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 0.34879589080810547,
      "learning_rate": 2.924207650888426e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 12477832,
      "step": 19125
    },
    {
      "epoch": 10.026205450733753,
      "grad_norm": 0.36939510703086853,
      "learning_rate": 2.9230807592810723e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 12482312,
      "step": 19130
    },
    {
      "epoch": 10.028825995807129,
      "grad_norm": 0.33240807056427,
      "learning_rate": 2.92195377917005e-05,
      "loss": 0.4249,
      "num_input_tokens_seen": 12485768,
      "step": 19135
    },
    {
      "epoch": 10.031446540880504,
      "grad_norm": 0.45270201563835144,
      "learning_rate": 2.920826710791112e-05,
      "loss": 0.3617,
      "num_input_tokens_seen": 12488776,
      "step": 19140
    },
    {
      "epoch": 10.034067085953879,
      "grad_norm": 0.31854167580604553,
      "learning_rate": 2.919699554380026e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 12491464,
      "step": 19145
    },
    {
      "epoch": 10.036687631027254,
      "grad_norm": 0.3948211669921875,
      "learning_rate": 2.9185723101725825e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 12494600,
      "step": 19150
    },
    {
      "epoch": 10.03930817610063,
      "grad_norm": 0.41206055879592896,
      "learning_rate": 2.9174449784045867e-05,
      "loss": 0.6292,
      "num_input_tokens_seen": 12498696,
      "step": 19155
    },
    {
      "epoch": 10.041928721174004,
      "grad_norm": 0.6635914444923401,
      "learning_rate": 2.916317559311863e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 12501960,
      "step": 19160
    },
    {
      "epoch": 10.04454926624738,
      "grad_norm": 0.6095110177993774,
      "learning_rate": 2.9151900531302566e-05,
      "loss": 0.6136,
      "num_input_tokens_seen": 12505224,
      "step": 19165
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 0.3905150890350342,
      "learning_rate": 2.9140624600956272e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 12508744,
      "step": 19170
    },
    {
      "epoch": 10.04979035639413,
      "grad_norm": 0.5428135991096497,
      "learning_rate": 2.912934780443855e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 12511464,
      "step": 19175
    },
    {
      "epoch": 10.052410901467505,
      "grad_norm": 0.8557369112968445,
      "learning_rate": 2.9118070144108388e-05,
      "loss": 0.5007,
      "num_input_tokens_seen": 12513736,
      "step": 19180
    },
    {
      "epoch": 10.05503144654088,
      "grad_norm": 0.40027543902397156,
      "learning_rate": 2.9106791622324924e-05,
      "loss": 0.5085,
      "num_input_tokens_seen": 12516584,
      "step": 19185
    },
    {
      "epoch": 10.057651991614255,
      "grad_norm": 0.31833165884017944,
      "learning_rate": 2.909551224144751e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 12520680,
      "step": 19190
    },
    {
      "epoch": 10.06027253668763,
      "grad_norm": 0.7658445239067078,
      "learning_rate": 2.9084232003835653e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 12523624,
      "step": 19195
    },
    {
      "epoch": 10.062893081761006,
      "grad_norm": 0.4507674276828766,
      "learning_rate": 2.9072950911849058e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 12526472,
      "step": 19200
    },
    {
      "epoch": 10.065513626834381,
      "grad_norm": 0.5234936475753784,
      "learning_rate": 2.90616689678476e-05,
      "loss": 0.5255,
      "num_input_tokens_seen": 12529768,
      "step": 19205
    },
    {
      "epoch": 10.068134171907756,
      "grad_norm": 0.33625704050064087,
      "learning_rate": 2.9050386174191337e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 12532680,
      "step": 19210
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 0.3905457556247711,
      "learning_rate": 2.903910253324048e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 12535944,
      "step": 19215
    },
    {
      "epoch": 10.073375262054507,
      "grad_norm": 0.6507329940795898,
      "learning_rate": 2.9027818047355477e-05,
      "loss": 0.5895,
      "num_input_tokens_seen": 12540008,
      "step": 19220
    },
    {
      "epoch": 10.075995807127883,
      "grad_norm": 0.604023814201355,
      "learning_rate": 2.9016532718896873e-05,
      "loss": 0.3533,
      "num_input_tokens_seen": 12542952,
      "step": 19225
    },
    {
      "epoch": 10.078616352201259,
      "grad_norm": 0.6019741296768188,
      "learning_rate": 2.900524655022545e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 12545672,
      "step": 19230
    },
    {
      "epoch": 10.081236897274634,
      "grad_norm": 1.3049132823944092,
      "learning_rate": 2.899395954370215e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 12548808,
      "step": 19235
    },
    {
      "epoch": 10.083857442348009,
      "grad_norm": 0.7230603098869324,
      "learning_rate": 2.898267170168807e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 12551944,
      "step": 19240
    },
    {
      "epoch": 10.086477987421384,
      "grad_norm": 0.4949077069759369,
      "learning_rate": 2.8971383026544507e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 12555048,
      "step": 19245
    },
    {
      "epoch": 10.08909853249476,
      "grad_norm": 0.40614721179008484,
      "learning_rate": 2.896009352063293e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 12558312,
      "step": 19250
    },
    {
      "epoch": 10.091719077568134,
      "grad_norm": 0.24597227573394775,
      "learning_rate": 2.894880318631496e-05,
      "loss": 0.4705,
      "num_input_tokens_seen": 12561992,
      "step": 19255
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 0.2664828300476074,
      "learning_rate": 2.8937512025952417e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 12565544,
      "step": 19260
    },
    {
      "epoch": 10.096960167714885,
      "grad_norm": 0.6397587060928345,
      "learning_rate": 2.8926220041907283e-05,
      "loss": 0.5167,
      "num_input_tokens_seen": 12568232,
      "step": 19265
    },
    {
      "epoch": 10.09958071278826,
      "grad_norm": 1.0566091537475586,
      "learning_rate": 2.8914927236541706e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 12570568,
      "step": 19270
    },
    {
      "epoch": 10.102201257861635,
      "grad_norm": 0.39872390031814575,
      "learning_rate": 2.8903633612218012e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 12573448,
      "step": 19275
    },
    {
      "epoch": 10.10482180293501,
      "grad_norm": 0.3834522068500519,
      "learning_rate": 2.8892339171298703e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 12577256,
      "step": 19280
    },
    {
      "epoch": 10.107442348008385,
      "grad_norm": 0.375691294670105,
      "learning_rate": 2.888104391614645e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 12580808,
      "step": 19285
    },
    {
      "epoch": 10.11006289308176,
      "grad_norm": 0.24665135145187378,
      "learning_rate": 2.886974784912408e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 12585896,
      "step": 19290
    },
    {
      "epoch": 10.112683438155136,
      "grad_norm": 0.32247307896614075,
      "learning_rate": 2.8858450972594612e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 12589000,
      "step": 19295
    },
    {
      "epoch": 10.115303983228511,
      "grad_norm": 0.3882722854614258,
      "learning_rate": 2.8847153288921226e-05,
      "loss": 0.4606,
      "num_input_tokens_seen": 12592104,
      "step": 19300
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 0.260164737701416,
      "learning_rate": 2.883585480046725e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 12595240,
      "step": 19305
    },
    {
      "epoch": 10.120545073375261,
      "grad_norm": 0.7646326422691345,
      "learning_rate": 2.8824555509596223e-05,
      "loss": 0.349,
      "num_input_tokens_seen": 12597480,
      "step": 19310
    },
    {
      "epoch": 10.123165618448636,
      "grad_norm": 0.4403691589832306,
      "learning_rate": 2.8813255418671815e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 12600328,
      "step": 19315
    },
    {
      "epoch": 10.125786163522013,
      "grad_norm": 0.9142723679542542,
      "learning_rate": 2.8801954530057874e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 12602152,
      "step": 19320
    },
    {
      "epoch": 10.128406708595389,
      "grad_norm": 0.8902075886726379,
      "learning_rate": 2.879065284611842e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 12605352,
      "step": 19325
    },
    {
      "epoch": 10.131027253668764,
      "grad_norm": 0.4352945387363434,
      "learning_rate": 2.8779350369217634e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 12607784,
      "step": 19330
    },
    {
      "epoch": 10.133647798742139,
      "grad_norm": 0.4344830811023712,
      "learning_rate": 2.876804710171987e-05,
      "loss": 0.5955,
      "num_input_tokens_seen": 12614056,
      "step": 19335
    },
    {
      "epoch": 10.136268343815514,
      "grad_norm": 0.6418483257293701,
      "learning_rate": 2.8756743045989643e-05,
      "loss": 0.5121,
      "num_input_tokens_seen": 12616552,
      "step": 19340
    },
    {
      "epoch": 10.13888888888889,
      "grad_norm": 0.6002222895622253,
      "learning_rate": 2.8745438204391623e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 12619976,
      "step": 19345
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 0.5524416565895081,
      "learning_rate": 2.8734132579290675e-05,
      "loss": 0.4393,
      "num_input_tokens_seen": 12623272,
      "step": 19350
    },
    {
      "epoch": 10.14412997903564,
      "grad_norm": 0.7002711892127991,
      "learning_rate": 2.8722826173051782e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 12625960,
      "step": 19355
    },
    {
      "epoch": 10.146750524109015,
      "grad_norm": 0.45924466848373413,
      "learning_rate": 2.871151898804012e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 12630184,
      "step": 19360
    },
    {
      "epoch": 10.14937106918239,
      "grad_norm": 0.2968181371688843,
      "learning_rate": 2.8700211026621043e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 12633224,
      "step": 19365
    },
    {
      "epoch": 10.151991614255765,
      "grad_norm": 0.43375563621520996,
      "learning_rate": 2.8688902291160025e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 12636584,
      "step": 19370
    },
    {
      "epoch": 10.15461215932914,
      "grad_norm": 0.3383088707923889,
      "learning_rate": 2.867759278402273e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 12639944,
      "step": 19375
    },
    {
      "epoch": 10.157232704402515,
      "grad_norm": 0.469119131565094,
      "learning_rate": 2.8666282507574994e-05,
      "loss": 0.5063,
      "num_input_tokens_seen": 12643624,
      "step": 19380
    },
    {
      "epoch": 10.15985324947589,
      "grad_norm": 0.23921212553977966,
      "learning_rate": 2.8654971464182773e-05,
      "loss": 0.3156,
      "num_input_tokens_seen": 12646792,
      "step": 19385
    },
    {
      "epoch": 10.162473794549266,
      "grad_norm": 0.5370135307312012,
      "learning_rate": 2.8643659656212222e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 12650696,
      "step": 19390
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 0.3504379093647003,
      "learning_rate": 2.8632347086029642e-05,
      "loss": 0.5761,
      "num_input_tokens_seen": 12653512,
      "step": 19395
    },
    {
      "epoch": 10.167714884696016,
      "grad_norm": 0.3608400523662567,
      "learning_rate": 2.8621033756001486e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 12657160,
      "step": 19400
    },
    {
      "epoch": 10.170335429769391,
      "grad_norm": 0.42491599917411804,
      "learning_rate": 2.8609719668494394e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 12661096,
      "step": 19405
    },
    {
      "epoch": 10.172955974842766,
      "grad_norm": 0.2935706675052643,
      "learning_rate": 2.8598404825875113e-05,
      "loss": 0.3782,
      "num_input_tokens_seen": 12664168,
      "step": 19410
    },
    {
      "epoch": 10.175576519916143,
      "grad_norm": 0.7707458734512329,
      "learning_rate": 2.8587089230510604e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 12667304,
      "step": 19415
    },
    {
      "epoch": 10.178197064989519,
      "grad_norm": 0.21104875206947327,
      "learning_rate": 2.857577288476795e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 12670760,
      "step": 19420
    },
    {
      "epoch": 10.180817610062894,
      "grad_norm": 0.4601619243621826,
      "learning_rate": 2.85644557910144e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 12673384,
      "step": 19425
    },
    {
      "epoch": 10.183438155136269,
      "grad_norm": 0.813778817653656,
      "learning_rate": 2.8553137951617366e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 12676168,
      "step": 19430
    },
    {
      "epoch": 10.186058700209644,
      "grad_norm": 0.3084438145160675,
      "learning_rate": 2.854181936894441e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 12678984,
      "step": 19435
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 0.6241058111190796,
      "learning_rate": 2.8530500045363247e-05,
      "loss": 0.598,
      "num_input_tokens_seen": 12682600,
      "step": 19440
    },
    {
      "epoch": 10.191299790356394,
      "grad_norm": 0.4033466875553131,
      "learning_rate": 2.851917998324175e-05,
      "loss": 0.3779,
      "num_input_tokens_seen": 12685704,
      "step": 19445
    },
    {
      "epoch": 10.19392033542977,
      "grad_norm": 0.4428327679634094,
      "learning_rate": 2.8507859184947957e-05,
      "loss": 0.3798,
      "num_input_tokens_seen": 12689256,
      "step": 19450
    },
    {
      "epoch": 10.196540880503145,
      "grad_norm": 0.26671531796455383,
      "learning_rate": 2.8496537652850043e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 12692968,
      "step": 19455
    },
    {
      "epoch": 10.19916142557652,
      "grad_norm": 0.43624624609947205,
      "learning_rate": 2.8485215389316333e-05,
      "loss": 0.3896,
      "num_input_tokens_seen": 12696712,
      "step": 19460
    },
    {
      "epoch": 10.201781970649895,
      "grad_norm": 0.7666231989860535,
      "learning_rate": 2.8473892396715328e-05,
      "loss": 0.4869,
      "num_input_tokens_seen": 12699848,
      "step": 19465
    },
    {
      "epoch": 10.20440251572327,
      "grad_norm": 0.5485559701919556,
      "learning_rate": 2.846256867741567e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 12703176,
      "step": 19470
    },
    {
      "epoch": 10.207023060796645,
      "grad_norm": 0.2955032289028168,
      "learning_rate": 2.845124423378614e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 12706280,
      "step": 19475
    },
    {
      "epoch": 10.20964360587002,
      "grad_norm": 0.5020673274993896,
      "learning_rate": 2.8439919068195697e-05,
      "loss": 0.5642,
      "num_input_tokens_seen": 12709224,
      "step": 19480
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 0.4360915422439575,
      "learning_rate": 2.842859318301343e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 12713064,
      "step": 19485
    },
    {
      "epoch": 10.214884696016771,
      "grad_norm": 0.49836263060569763,
      "learning_rate": 2.841726658060858e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 12716424,
      "step": 19490
    },
    {
      "epoch": 10.217505241090146,
      "grad_norm": 0.45158660411834717,
      "learning_rate": 2.8405939263350555e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 12719816,
      "step": 19495
    },
    {
      "epoch": 10.220125786163521,
      "grad_norm": 0.6199604272842407,
      "learning_rate": 2.8394611233608886e-05,
      "loss": 0.537,
      "num_input_tokens_seen": 12723624,
      "step": 19500
    },
    {
      "epoch": 10.222746331236896,
      "grad_norm": 0.5977474451065063,
      "learning_rate": 2.8383282493753283e-05,
      "loss": 0.4585,
      "num_input_tokens_seen": 12727240,
      "step": 19505
    },
    {
      "epoch": 10.225366876310272,
      "grad_norm": 0.7933608293533325,
      "learning_rate": 2.8371953046153575e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 12730792,
      "step": 19510
    },
    {
      "epoch": 10.227987421383649,
      "grad_norm": 0.700352668762207,
      "learning_rate": 2.8360622893179757e-05,
      "loss": 0.4659,
      "num_input_tokens_seen": 12733928,
      "step": 19515
    },
    {
      "epoch": 10.230607966457024,
      "grad_norm": 0.4129725396633148,
      "learning_rate": 2.834929203720197e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 12737992,
      "step": 19520
    },
    {
      "epoch": 10.233228511530399,
      "grad_norm": 0.5921358466148376,
      "learning_rate": 2.83379604805905e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 12740456,
      "step": 19525
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 0.723558783531189,
      "learning_rate": 2.832662822571578e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 12743496,
      "step": 19530
    },
    {
      "epoch": 10.23846960167715,
      "grad_norm": 0.7285535335540771,
      "learning_rate": 2.8315295274948395e-05,
      "loss": 0.4004,
      "num_input_tokens_seen": 12746952,
      "step": 19535
    },
    {
      "epoch": 10.241090146750524,
      "grad_norm": 0.5201640129089355,
      "learning_rate": 2.8303961630659047e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 12749800,
      "step": 19540
    },
    {
      "epoch": 10.2437106918239,
      "grad_norm": 0.45347490906715393,
      "learning_rate": 2.8292627295218627e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 12752872,
      "step": 19545
    },
    {
      "epoch": 10.246331236897275,
      "grad_norm": 1.899949073791504,
      "learning_rate": 2.828129227099815e-05,
      "loss": 0.5767,
      "num_input_tokens_seen": 12755944,
      "step": 19550
    },
    {
      "epoch": 10.24895178197065,
      "grad_norm": 0.8601281046867371,
      "learning_rate": 2.826995656036875e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 12758344,
      "step": 19555
    },
    {
      "epoch": 10.251572327044025,
      "grad_norm": 0.2718953788280487,
      "learning_rate": 2.8258620165701747e-05,
      "loss": 0.3274,
      "num_input_tokens_seen": 12762312,
      "step": 19560
    },
    {
      "epoch": 10.2541928721174,
      "grad_norm": 0.5444430708885193,
      "learning_rate": 2.8247283089368586e-05,
      "loss": 0.4379,
      "num_input_tokens_seen": 12765416,
      "step": 19565
    },
    {
      "epoch": 10.256813417190775,
      "grad_norm": 0.432084858417511,
      "learning_rate": 2.8235945333740842e-05,
      "loss": 0.3816,
      "num_input_tokens_seen": 12768552,
      "step": 19570
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 0.8218435645103455,
      "learning_rate": 2.822460690119026e-05,
      "loss": 0.4116,
      "num_input_tokens_seen": 12770920,
      "step": 19575
    },
    {
      "epoch": 10.262054507337526,
      "grad_norm": 0.3397331237792969,
      "learning_rate": 2.82132677940887e-05,
      "loss": 0.4304,
      "num_input_tokens_seen": 12774184,
      "step": 19580
    },
    {
      "epoch": 10.264675052410901,
      "grad_norm": 0.5368154644966125,
      "learning_rate": 2.820192801480817e-05,
      "loss": 0.4982,
      "num_input_tokens_seen": 12777288,
      "step": 19585
    },
    {
      "epoch": 10.267295597484276,
      "grad_norm": 0.583961546421051,
      "learning_rate": 2.8190587565720842e-05,
      "loss": 0.4164,
      "num_input_tokens_seen": 12780136,
      "step": 19590
    },
    {
      "epoch": 10.269916142557651,
      "grad_norm": 0.3828243017196655,
      "learning_rate": 2.8179246449198986e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 12783144,
      "step": 19595
    },
    {
      "epoch": 10.272536687631026,
      "grad_norm": 0.3843878507614136,
      "learning_rate": 2.8167904667615048e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 12785992,
      "step": 19600
    },
    {
      "epoch": 10.275157232704402,
      "grad_norm": 0.4472501873970032,
      "learning_rate": 2.8156562223341592e-05,
      "loss": 0.5422,
      "num_input_tokens_seen": 12789000,
      "step": 19605
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 0.3709612488746643,
      "learning_rate": 2.814521911875133e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 12791976,
      "step": 19610
    },
    {
      "epoch": 10.280398322851154,
      "grad_norm": 0.6082771420478821,
      "learning_rate": 2.8133875356217114e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 12795976,
      "step": 19615
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 0.35267093777656555,
      "learning_rate": 2.8122530938111923e-05,
      "loss": 0.5162,
      "num_input_tokens_seen": 12798984,
      "step": 19620
    },
    {
      "epoch": 10.285639412997904,
      "grad_norm": 0.7366939783096313,
      "learning_rate": 2.8111185866808888e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 12801864,
      "step": 19625
    },
    {
      "epoch": 10.28825995807128,
      "grad_norm": 0.7465375065803528,
      "learning_rate": 2.809984014468126e-05,
      "loss": 0.5065,
      "num_input_tokens_seen": 12804968,
      "step": 19630
    },
    {
      "epoch": 10.290880503144654,
      "grad_norm": 0.4823147654533386,
      "learning_rate": 2.8088493774102443e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 12807848,
      "step": 19635
    },
    {
      "epoch": 10.29350104821803,
      "grad_norm": 0.30403047800064087,
      "learning_rate": 2.807714675744596e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 12810632,
      "step": 19640
    },
    {
      "epoch": 10.296121593291405,
      "grad_norm": 0.44421878457069397,
      "learning_rate": 2.8065799097085475e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 12814024,
      "step": 19645
    },
    {
      "epoch": 10.29874213836478,
      "grad_norm": 0.39977923035621643,
      "learning_rate": 2.8054450795394804e-05,
      "loss": 0.4061,
      "num_input_tokens_seen": 12817800,
      "step": 19650
    },
    {
      "epoch": 10.301362683438155,
      "grad_norm": 0.6841985583305359,
      "learning_rate": 2.8043101854747878e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 12820456,
      "step": 19655
    },
    {
      "epoch": 10.30398322851153,
      "grad_norm": 0.5054211020469666,
      "learning_rate": 2.8031752277518748e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 12822792,
      "step": 19660
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 0.3878526985645294,
      "learning_rate": 2.802040206608164e-05,
      "loss": 0.602,
      "num_input_tokens_seen": 12826888,
      "step": 19665
    },
    {
      "epoch": 10.30922431865828,
      "grad_norm": 0.2616793215274811,
      "learning_rate": 2.8009051222810877e-05,
      "loss": 0.4639,
      "num_input_tokens_seen": 12829896,
      "step": 19670
    },
    {
      "epoch": 10.311844863731656,
      "grad_norm": 0.4246371388435364,
      "learning_rate": 2.7997699750080918e-05,
      "loss": 0.5649,
      "num_input_tokens_seen": 12833480,
      "step": 19675
    },
    {
      "epoch": 10.314465408805031,
      "grad_norm": 0.34035417437553406,
      "learning_rate": 2.7986347650266386e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 12836424,
      "step": 19680
    },
    {
      "epoch": 10.317085953878406,
      "grad_norm": 0.8962550759315491,
      "learning_rate": 2.797499492574199e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 12839464,
      "step": 19685
    },
    {
      "epoch": 10.319706498951781,
      "grad_norm": 0.3537777066230774,
      "learning_rate": 2.7963641578882594e-05,
      "loss": 0.3387,
      "num_input_tokens_seen": 12841896,
      "step": 19690
    },
    {
      "epoch": 10.322327044025156,
      "grad_norm": 0.470975786447525,
      "learning_rate": 2.7952287612063193e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 12845768,
      "step": 19695
    },
    {
      "epoch": 10.324947589098532,
      "grad_norm": 0.5153923034667969,
      "learning_rate": 2.79409330276589e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 12848680,
      "step": 19700
    },
    {
      "epoch": 10.327568134171909,
      "grad_norm": 0.8431916236877441,
      "learning_rate": 2.7929577828044974e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 12851464,
      "step": 19705
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 0.5144768953323364,
      "learning_rate": 2.7918222015596784e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 12854504,
      "step": 19710
    },
    {
      "epoch": 10.332809224318659,
      "grad_norm": 0.5970658659934998,
      "learning_rate": 2.7906865592689847e-05,
      "loss": 0.3698,
      "num_input_tokens_seen": 12857736,
      "step": 19715
    },
    {
      "epoch": 10.335429769392034,
      "grad_norm": 0.6984923481941223,
      "learning_rate": 2.789550856169979e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 12860360,
      "step": 19720
    },
    {
      "epoch": 10.33805031446541,
      "grad_norm": 0.5460546612739563,
      "learning_rate": 2.7884150925002374e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 12863016,
      "step": 19725
    },
    {
      "epoch": 10.340670859538784,
      "grad_norm": 1.540684700012207,
      "learning_rate": 2.787279268497349e-05,
      "loss": 0.5822,
      "num_input_tokens_seen": 12865768,
      "step": 19730
    },
    {
      "epoch": 10.34329140461216,
      "grad_norm": 0.45448029041290283,
      "learning_rate": 2.786143384398915e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 12869256,
      "step": 19735
    },
    {
      "epoch": 10.345911949685535,
      "grad_norm": 0.2907451391220093,
      "learning_rate": 2.785007440442549e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 12872456,
      "step": 19740
    },
    {
      "epoch": 10.34853249475891,
      "grad_norm": 0.6161002516746521,
      "learning_rate": 2.7838714368658785e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 12875720,
      "step": 19745
    },
    {
      "epoch": 10.351153039832285,
      "grad_norm": 0.44085603952407837,
      "learning_rate": 2.7827353739065427e-05,
      "loss": 0.3749,
      "num_input_tokens_seen": 12878088,
      "step": 19750
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 1.1437386274337769,
      "learning_rate": 2.781599251802191e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 12880616,
      "step": 19755
    },
    {
      "epoch": 10.356394129979035,
      "grad_norm": 0.5906187295913696,
      "learning_rate": 2.780463070790489e-05,
      "loss": 0.6076,
      "num_input_tokens_seen": 12883144,
      "step": 19760
    },
    {
      "epoch": 10.35901467505241,
      "grad_norm": 0.5488401651382446,
      "learning_rate": 2.7793268311091125e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 12885832,
      "step": 19765
    },
    {
      "epoch": 10.361635220125786,
      "grad_norm": 0.4167284667491913,
      "learning_rate": 2.7781905329957496e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 12889000,
      "step": 19770
    },
    {
      "epoch": 10.364255765199161,
      "grad_norm": 0.4495694935321808,
      "learning_rate": 2.7770541766881008e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 12892616,
      "step": 19775
    },
    {
      "epoch": 10.366876310272536,
      "grad_norm": 0.612069845199585,
      "learning_rate": 2.7759177624238786e-05,
      "loss": 0.5181,
      "num_input_tokens_seen": 12895112,
      "step": 19780
    },
    {
      "epoch": 10.369496855345911,
      "grad_norm": 0.4098947048187256,
      "learning_rate": 2.7747812904408094e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 12899176,
      "step": 19785
    },
    {
      "epoch": 10.372117400419286,
      "grad_norm": 0.29646065831184387,
      "learning_rate": 2.7736447609766286e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 12902184,
      "step": 19790
    },
    {
      "epoch": 10.374737945492662,
      "grad_norm": 0.7180443406105042,
      "learning_rate": 2.772508174269085e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 12904712,
      "step": 19795
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 0.35297155380249023,
      "learning_rate": 2.7713715305559414e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 12907560,
      "step": 19800
    },
    {
      "epoch": 10.379979035639414,
      "grad_norm": 0.40541496872901917,
      "learning_rate": 2.7702348300749692e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 12910568,
      "step": 19805
    },
    {
      "epoch": 10.382599580712789,
      "grad_norm": 0.9091391563415527,
      "learning_rate": 2.7690980730639533e-05,
      "loss": 0.6775,
      "num_input_tokens_seen": 12913608,
      "step": 19810
    },
    {
      "epoch": 10.385220125786164,
      "grad_norm": 0.4203943908214569,
      "learning_rate": 2.7679612597606902e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 12917288,
      "step": 19815
    },
    {
      "epoch": 10.38784067085954,
      "grad_norm": 0.584373950958252,
      "learning_rate": 2.7668243904029894e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 12920232,
      "step": 19820
    },
    {
      "epoch": 10.390461215932914,
      "grad_norm": 0.47909456491470337,
      "learning_rate": 2.76568746522867e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 12923432,
      "step": 19825
    },
    {
      "epoch": 10.39308176100629,
      "grad_norm": 0.32819312810897827,
      "learning_rate": 2.7645504844755637e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 12927912,
      "step": 19830
    },
    {
      "epoch": 10.395702306079665,
      "grad_norm": 0.39073652029037476,
      "learning_rate": 2.7634134483815143e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 12931304,
      "step": 19835
    },
    {
      "epoch": 10.39832285115304,
      "grad_norm": 0.5449885129928589,
      "learning_rate": 2.762276357184377e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 12934248,
      "step": 19840
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 0.3583555221557617,
      "learning_rate": 2.7611392111220176e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 12937608,
      "step": 19845
    },
    {
      "epoch": 10.40356394129979,
      "grad_norm": 0.46365851163864136,
      "learning_rate": 2.7600020104323154e-05,
      "loss": 0.3635,
      "num_input_tokens_seen": 12940360,
      "step": 19850
    },
    {
      "epoch": 10.406184486373165,
      "grad_norm": 0.6615519523620605,
      "learning_rate": 2.7588647553531576e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 12943048,
      "step": 19855
    },
    {
      "epoch": 10.40880503144654,
      "grad_norm": 0.7316948771476746,
      "learning_rate": 2.7577274461224463e-05,
      "loss": 0.3151,
      "num_input_tokens_seen": 12945704,
      "step": 19860
    },
    {
      "epoch": 10.411425576519916,
      "grad_norm": 0.5096989274024963,
      "learning_rate": 2.756590082978095e-05,
      "loss": 0.4154,
      "num_input_tokens_seen": 12948616,
      "step": 19865
    },
    {
      "epoch": 10.414046121593291,
      "grad_norm": 0.7443439960479736,
      "learning_rate": 2.7554526661580243e-05,
      "loss": 0.3308,
      "num_input_tokens_seen": 12951880,
      "step": 19870
    },
    {
      "epoch": 10.416666666666666,
      "grad_norm": 0.31162816286087036,
      "learning_rate": 2.7543151959001708e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 12954984,
      "step": 19875
    },
    {
      "epoch": 10.419287211740041,
      "grad_norm": 0.5237964987754822,
      "learning_rate": 2.7531776724424797e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 12957640,
      "step": 19880
    },
    {
      "epoch": 10.421907756813416,
      "grad_norm": 0.6645166277885437,
      "learning_rate": 2.752040096022907e-05,
      "loss": 0.3675,
      "num_input_tokens_seen": 12960488,
      "step": 19885
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 0.4847505986690521,
      "learning_rate": 2.750902466879423e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 12963560,
      "step": 19890
    },
    {
      "epoch": 10.427148846960169,
      "grad_norm": 0.30000704526901245,
      "learning_rate": 2.7497647852500048e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 12967208,
      "step": 19895
    },
    {
      "epoch": 10.429769392033544,
      "grad_norm": 0.3530392050743103,
      "learning_rate": 2.7486270513726424e-05,
      "loss": 0.3226,
      "num_input_tokens_seen": 12970248,
      "step": 19900
    },
    {
      "epoch": 10.432389937106919,
      "grad_norm": 0.9842698574066162,
      "learning_rate": 2.7474892654853384e-05,
      "loss": 0.4294,
      "num_input_tokens_seen": 12973480,
      "step": 19905
    },
    {
      "epoch": 10.435010482180294,
      "grad_norm": 0.2600981891155243,
      "learning_rate": 2.7463514278261034e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 12976904,
      "step": 19910
    },
    {
      "epoch": 10.43763102725367,
      "grad_norm": 0.9475554823875427,
      "learning_rate": 2.7452135386329596e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 12979912,
      "step": 19915
    },
    {
      "epoch": 10.440251572327044,
      "grad_norm": 0.6171470284461975,
      "learning_rate": 2.7440755981439414e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 12982728,
      "step": 19920
    },
    {
      "epoch": 10.44287211740042,
      "grad_norm": 0.482144832611084,
      "learning_rate": 2.7429376065970924e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 12986472,
      "step": 19925
    },
    {
      "epoch": 10.445492662473795,
      "grad_norm": 0.4225584864616394,
      "learning_rate": 2.7417995642304678e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 12990600,
      "step": 19930
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 0.45219293236732483,
      "learning_rate": 2.7406614712821333e-05,
      "loss": 0.5567,
      "num_input_tokens_seen": 12994216,
      "step": 19935
    },
    {
      "epoch": 10.450733752620545,
      "grad_norm": 0.41697457432746887,
      "learning_rate": 2.739523327990164e-05,
      "loss": 0.4754,
      "num_input_tokens_seen": 12997256,
      "step": 19940
    },
    {
      "epoch": 10.45335429769392,
      "grad_norm": 0.4445503354072571,
      "learning_rate": 2.7383851345926477e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 13001384,
      "step": 19945
    },
    {
      "epoch": 10.455974842767295,
      "grad_norm": 0.6358289122581482,
      "learning_rate": 2.7372468913276804e-05,
      "loss": 0.6258,
      "num_input_tokens_seen": 13004232,
      "step": 19950
    },
    {
      "epoch": 10.45859538784067,
      "grad_norm": 0.49572232365608215,
      "learning_rate": 2.7361085984333705e-05,
      "loss": 0.4304,
      "num_input_tokens_seen": 13007208,
      "step": 19955
    },
    {
      "epoch": 10.461215932914046,
      "grad_norm": 0.27722278237342834,
      "learning_rate": 2.7349702561478356e-05,
      "loss": 0.4615,
      "num_input_tokens_seen": 13010696,
      "step": 19960
    },
    {
      "epoch": 10.463836477987421,
      "grad_norm": 0.5844735503196716,
      "learning_rate": 2.733831864709203e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 13013416,
      "step": 19965
    },
    {
      "epoch": 10.466457023060796,
      "grad_norm": 0.3290785253047943,
      "learning_rate": 2.732693424355613e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 13016904,
      "step": 19970
    },
    {
      "epoch": 10.469077568134171,
      "grad_norm": 0.6760270595550537,
      "learning_rate": 2.731554935325213e-05,
      "loss": 0.4639,
      "num_input_tokens_seen": 13019496,
      "step": 19975
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 0.45746177434921265,
      "learning_rate": 2.7304163978561613e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 13022184,
      "step": 19980
    },
    {
      "epoch": 10.474318658280922,
      "grad_norm": 0.9866836071014404,
      "learning_rate": 2.729277812186629e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 13024968,
      "step": 19985
    },
    {
      "epoch": 10.476939203354299,
      "grad_norm": 0.5770145058631897,
      "learning_rate": 2.7281391785547945e-05,
      "loss": 0.4166,
      "num_input_tokens_seen": 13028328,
      "step": 19990
    },
    {
      "epoch": 10.479559748427674,
      "grad_norm": 0.5539757609367371,
      "learning_rate": 2.7270004971988456e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 13032040,
      "step": 19995
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 0.7837700247764587,
      "learning_rate": 2.7258617683569825e-05,
      "loss": 0.4379,
      "num_input_tokens_seen": 13035944,
      "step": 20000
    },
    {
      "epoch": 10.484800838574424,
      "grad_norm": 0.8014627695083618,
      "learning_rate": 2.7247229922674143e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 13042856,
      "step": 20005
    },
    {
      "epoch": 10.4874213836478,
      "grad_norm": 0.7184135317802429,
      "learning_rate": 2.7235841691683605e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 13046792,
      "step": 20010
    },
    {
      "epoch": 10.490041928721174,
      "grad_norm": 0.2811380922794342,
      "learning_rate": 2.7224452992980487e-05,
      "loss": 0.4061,
      "num_input_tokens_seen": 13050920,
      "step": 20015
    },
    {
      "epoch": 10.49266247379455,
      "grad_norm": 0.37662726640701294,
      "learning_rate": 2.7213063828947182e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 13058664,
      "step": 20020
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 0.32448530197143555,
      "learning_rate": 2.7201674201966176e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 13062952,
      "step": 20025
    },
    {
      "epoch": 10.4979035639413,
      "grad_norm": 0.40793827176094055,
      "learning_rate": 2.7190284114420035e-05,
      "loss": 0.2855,
      "num_input_tokens_seen": 13065768,
      "step": 20030
    },
    {
      "epoch": 10.500524109014675,
      "grad_norm": 0.30104541778564453,
      "learning_rate": 2.717889356869146e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 13069160,
      "step": 20035
    },
    {
      "epoch": 10.50314465408805,
      "grad_norm": 0.45372113585472107,
      "learning_rate": 2.7167502567163194e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 13072168,
      "step": 20040
    },
    {
      "epoch": 10.505765199161425,
      "grad_norm": 0.3759719729423523,
      "learning_rate": 2.715611111221813e-05,
      "loss": 0.3727,
      "num_input_tokens_seen": 13074888,
      "step": 20045
    },
    {
      "epoch": 10.5083857442348,
      "grad_norm": 0.4386923015117645,
      "learning_rate": 2.7144719206239217e-05,
      "loss": 0.6741,
      "num_input_tokens_seen": 13078280,
      "step": 20050
    },
    {
      "epoch": 10.511006289308176,
      "grad_norm": 0.4234353005886078,
      "learning_rate": 2.7133326851609508e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 13081544,
      "step": 20055
    },
    {
      "epoch": 10.51362683438155,
      "grad_norm": 0.41828614473342896,
      "learning_rate": 2.7121934050712166e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 13084360,
      "step": 20060
    },
    {
      "epoch": 10.516247379454926,
      "grad_norm": 0.54926598072052,
      "learning_rate": 2.711054080593043e-05,
      "loss": 0.5593,
      "num_input_tokens_seen": 13087464,
      "step": 20065
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 0.280032217502594,
      "learning_rate": 2.7099147119647627e-05,
      "loss": 0.3729,
      "num_input_tokens_seen": 13091112,
      "step": 20070
    },
    {
      "epoch": 10.521488469601676,
      "grad_norm": 1.5209646224975586,
      "learning_rate": 2.7087752994247206e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 13094824,
      "step": 20075
    },
    {
      "epoch": 10.524109014675052,
      "grad_norm": 0.3639526665210724,
      "learning_rate": 2.707635843211267e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 13097768,
      "step": 20080
    },
    {
      "epoch": 10.526729559748428,
      "grad_norm": 0.39553773403167725,
      "learning_rate": 2.7064963435627637e-05,
      "loss": 0.5179,
      "num_input_tokens_seen": 13100712,
      "step": 20085
    },
    {
      "epoch": 10.529350104821804,
      "grad_norm": 0.43706944584846497,
      "learning_rate": 2.7053568007175823e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 13103464,
      "step": 20090
    },
    {
      "epoch": 10.531970649895179,
      "grad_norm": 0.6363603472709656,
      "learning_rate": 2.7042172149140994e-05,
      "loss": 0.3914,
      "num_input_tokens_seen": 13106312,
      "step": 20095
    },
    {
      "epoch": 10.534591194968554,
      "grad_norm": 0.4867595434188843,
      "learning_rate": 2.703077586390706e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 13108936,
      "step": 20100
    },
    {
      "epoch": 10.53721174004193,
      "grad_norm": 0.2823026776313782,
      "learning_rate": 2.7019379153857982e-05,
      "loss": 0.3768,
      "num_input_tokens_seen": 13113544,
      "step": 20105
    },
    {
      "epoch": 10.539832285115304,
      "grad_norm": 0.4218772351741791,
      "learning_rate": 2.700798202137782e-05,
      "loss": 0.5558,
      "num_input_tokens_seen": 13116648,
      "step": 20110
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 0.3101717233657837,
      "learning_rate": 2.6996584468850732e-05,
      "loss": 0.2805,
      "num_input_tokens_seen": 13119816,
      "step": 20115
    },
    {
      "epoch": 10.545073375262055,
      "grad_norm": 0.5034247040748596,
      "learning_rate": 2.6985186498660946e-05,
      "loss": 0.708,
      "num_input_tokens_seen": 13123656,
      "step": 20120
    },
    {
      "epoch": 10.54769392033543,
      "grad_norm": 0.4606434106826782,
      "learning_rate": 2.69737881131928e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 13126120,
      "step": 20125
    },
    {
      "epoch": 10.550314465408805,
      "grad_norm": 0.32377856969833374,
      "learning_rate": 2.6962389314830692e-05,
      "loss": 0.5484,
      "num_input_tokens_seen": 13128968,
      "step": 20130
    },
    {
      "epoch": 10.55293501048218,
      "grad_norm": 0.8357192277908325,
      "learning_rate": 2.6950990105959135e-05,
      "loss": 0.3696,
      "num_input_tokens_seen": 13132168,
      "step": 20135
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 0.3968350887298584,
      "learning_rate": 2.6939590488962703e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 13135560,
      "step": 20140
    },
    {
      "epoch": 10.55817610062893,
      "grad_norm": 0.3364139795303345,
      "learning_rate": 2.6928190466226067e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 13139400,
      "step": 20145
    },
    {
      "epoch": 10.560796645702306,
      "grad_norm": 0.3277980089187622,
      "learning_rate": 2.691679004013398e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 13144232,
      "step": 20150
    },
    {
      "epoch": 10.56341719077568,
      "grad_norm": 0.49791914224624634,
      "learning_rate": 2.6905389213071285e-05,
      "loss": 0.3423,
      "num_input_tokens_seen": 13147432,
      "step": 20155
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 0.5861920118331909,
      "learning_rate": 2.6893987987422907e-05,
      "loss": 0.344,
      "num_input_tokens_seen": 13150632,
      "step": 20160
    },
    {
      "epoch": 10.568658280922431,
      "grad_norm": 0.41364845633506775,
      "learning_rate": 2.6882586365573854e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 13154152,
      "step": 20165
    },
    {
      "epoch": 10.571278825995806,
      "grad_norm": 0.28755372762680054,
      "learning_rate": 2.6871184349909206e-05,
      "loss": 0.3302,
      "num_input_tokens_seen": 13160552,
      "step": 20170
    },
    {
      "epoch": 10.573899371069182,
      "grad_norm": 0.7079183459281921,
      "learning_rate": 2.685978194281414e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 13163208,
      "step": 20175
    },
    {
      "epoch": 10.576519916142558,
      "grad_norm": 0.5283378958702087,
      "learning_rate": 2.6848379146673914e-05,
      "loss": 0.5157,
      "num_input_tokens_seen": 13166056,
      "step": 20180
    },
    {
      "epoch": 10.579140461215934,
      "grad_norm": 0.6353793740272522,
      "learning_rate": 2.6836975963873843e-05,
      "loss": 0.3957,
      "num_input_tokens_seen": 13168712,
      "step": 20185
    },
    {
      "epoch": 10.581761006289309,
      "grad_norm": 0.6146846413612366,
      "learning_rate": 2.6825572396799377e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 13171400,
      "step": 20190
    },
    {
      "epoch": 10.584381551362684,
      "grad_norm": 0.49863341450691223,
      "learning_rate": 2.6814168447835986e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 13174888,
      "step": 20195
    },
    {
      "epoch": 10.58700209643606,
      "grad_norm": 0.5813689231872559,
      "learning_rate": 2.6802764119369245e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 13178920,
      "step": 20200
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 0.2569715976715088,
      "learning_rate": 2.6791359413784828e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 13183176,
      "step": 20205
    },
    {
      "epoch": 10.59224318658281,
      "grad_norm": 0.8023406863212585,
      "learning_rate": 2.677995433346845e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 13186760,
      "step": 20210
    },
    {
      "epoch": 10.594863731656185,
      "grad_norm": 0.47118422389030457,
      "learning_rate": 2.6768548880805934e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 13189544,
      "step": 20215
    },
    {
      "epoch": 10.59748427672956,
      "grad_norm": 0.4752652645111084,
      "learning_rate": 2.6757143058183177e-05,
      "loss": 0.4473,
      "num_input_tokens_seen": 13193096,
      "step": 20220
    },
    {
      "epoch": 10.600104821802935,
      "grad_norm": 0.5180539488792419,
      "learning_rate": 2.6745736867986132e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 13195944,
      "step": 20225
    },
    {
      "epoch": 10.60272536687631,
      "grad_norm": 0.3842012286186218,
      "learning_rate": 2.673433031260085e-05,
      "loss": 0.3853,
      "num_input_tokens_seen": 13198760,
      "step": 20230
    },
    {
      "epoch": 10.605345911949685,
      "grad_norm": 0.38859066367149353,
      "learning_rate": 2.6722923394413456e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 13201704,
      "step": 20235
    },
    {
      "epoch": 10.60796645702306,
      "grad_norm": 0.40413880348205566,
      "learning_rate": 2.6711516115810136e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 13204872,
      "step": 20240
    },
    {
      "epoch": 10.610587002096436,
      "grad_norm": 0.43680644035339355,
      "learning_rate": 2.6700108479177188e-05,
      "loss": 0.4166,
      "num_input_tokens_seen": 13208200,
      "step": 20245
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 0.48084670305252075,
      "learning_rate": 2.6688700486900926e-05,
      "loss": 0.3738,
      "num_input_tokens_seen": 13211848,
      "step": 20250
    },
    {
      "epoch": 10.615828092243186,
      "grad_norm": 0.5830458402633667,
      "learning_rate": 2.6677292141367787e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 13214760,
      "step": 20255
    },
    {
      "epoch": 10.618448637316561,
      "grad_norm": 0.5563913583755493,
      "learning_rate": 2.6665883444964278e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 13219016,
      "step": 20260
    },
    {
      "epoch": 10.621069182389936,
      "grad_norm": 0.4210817813873291,
      "learning_rate": 2.665447440007695e-05,
      "loss": 0.3551,
      "num_input_tokens_seen": 13221800,
      "step": 20265
    },
    {
      "epoch": 10.623689727463312,
      "grad_norm": 0.43505141139030457,
      "learning_rate": 2.664306500909246e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 13224936,
      "step": 20270
    },
    {
      "epoch": 10.626310272536688,
      "grad_norm": 0.5391950011253357,
      "learning_rate": 2.663165527439751e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 13227560,
      "step": 20275
    },
    {
      "epoch": 10.628930817610064,
      "grad_norm": 0.5982903242111206,
      "learning_rate": 2.662024519837889e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 13230888,
      "step": 20280
    },
    {
      "epoch": 10.631551362683439,
      "grad_norm": 0.3808901011943817,
      "learning_rate": 2.6608834783423465e-05,
      "loss": 0.5048,
      "num_input_tokens_seen": 13234152,
      "step": 20285
    },
    {
      "epoch": 10.634171907756814,
      "grad_norm": 0.3703930974006653,
      "learning_rate": 2.6597424031918156e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 13237352,
      "step": 20290
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 0.681954562664032,
      "learning_rate": 2.6586012946249954e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 13240264,
      "step": 20295
    },
    {
      "epoch": 10.639412997903564,
      "grad_norm": 0.7343990802764893,
      "learning_rate": 2.6574601528805953e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 13243528,
      "step": 20300
    },
    {
      "epoch": 10.64203354297694,
      "grad_norm": 0.401250958442688,
      "learning_rate": 2.6563189781973268e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 13248712,
      "step": 20305
    },
    {
      "epoch": 10.644654088050315,
      "grad_norm": 0.5432491302490234,
      "learning_rate": 2.655177770813912e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 13252712,
      "step": 20310
    },
    {
      "epoch": 10.64727463312369,
      "grad_norm": 0.6913458108901978,
      "learning_rate": 2.654036530969077e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 13256392,
      "step": 20315
    },
    {
      "epoch": 10.649895178197065,
      "grad_norm": 0.4602138102054596,
      "learning_rate": 2.6528952589015576e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 13259240,
      "step": 20320
    },
    {
      "epoch": 10.65251572327044,
      "grad_norm": 0.3367140591144562,
      "learning_rate": 2.6517539548500947e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 13263176,
      "step": 20325
    },
    {
      "epoch": 10.655136268343815,
      "grad_norm": 0.4687321186065674,
      "learning_rate": 2.6506126190534354e-05,
      "loss": 0.5109,
      "num_input_tokens_seen": 13266696,
      "step": 20330
    },
    {
      "epoch": 10.65775681341719,
      "grad_norm": 0.8354809880256653,
      "learning_rate": 2.649471251750334e-05,
      "loss": 0.3981,
      "num_input_tokens_seen": 13269512,
      "step": 20335
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 0.32764530181884766,
      "learning_rate": 2.6483298531795527e-05,
      "loss": 0.329,
      "num_input_tokens_seen": 13272072,
      "step": 20340
    },
    {
      "epoch": 10.66299790356394,
      "grad_norm": 0.35942235589027405,
      "learning_rate": 2.6471884235798583e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 13275848,
      "step": 20345
    },
    {
      "epoch": 10.665618448637316,
      "grad_norm": 1.8904039859771729,
      "learning_rate": 2.6460469631900247e-05,
      "loss": 0.5585,
      "num_input_tokens_seen": 13278440,
      "step": 20350
    },
    {
      "epoch": 10.668238993710691,
      "grad_norm": 0.3674987852573395,
      "learning_rate": 2.644905472248832e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 13281576,
      "step": 20355
    },
    {
      "epoch": 10.670859538784066,
      "grad_norm": 0.5115389823913574,
      "learning_rate": 2.6437639509950686e-05,
      "loss": 0.4728,
      "num_input_tokens_seen": 13284168,
      "step": 20360
    },
    {
      "epoch": 10.673480083857442,
      "grad_norm": 0.39357811212539673,
      "learning_rate": 2.6426223996675264e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 13286760,
      "step": 20365
    },
    {
      "epoch": 10.676100628930818,
      "grad_norm": 0.86423659324646,
      "learning_rate": 2.641480818505005e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 13289672,
      "step": 20370
    },
    {
      "epoch": 10.678721174004194,
      "grad_norm": 0.5686913728713989,
      "learning_rate": 2.6403392077463108e-05,
      "loss": 0.5516,
      "num_input_tokens_seen": 13292616,
      "step": 20375
    },
    {
      "epoch": 10.681341719077569,
      "grad_norm": 0.3689523935317993,
      "learning_rate": 2.639197567630256e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 13296424,
      "step": 20380
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 0.8212447166442871,
      "learning_rate": 2.638055898395657e-05,
      "loss": 0.3876,
      "num_input_tokens_seen": 13299048,
      "step": 20385
    },
    {
      "epoch": 10.68658280922432,
      "grad_norm": 0.2904132008552551,
      "learning_rate": 2.6369142002813403e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 13302152,
      "step": 20390
    },
    {
      "epoch": 10.689203354297694,
      "grad_norm": 0.5488532185554504,
      "learning_rate": 2.6357724735261335e-05,
      "loss": 0.3519,
      "num_input_tokens_seen": 13304840,
      "step": 20395
    },
    {
      "epoch": 10.69182389937107,
      "grad_norm": 0.3968956172466278,
      "learning_rate": 2.6346307183688752e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 13307976,
      "step": 20400
    },
    {
      "epoch": 10.694444444444445,
      "grad_norm": 0.4341953992843628,
      "learning_rate": 2.6334889350484064e-05,
      "loss": 0.4303,
      "num_input_tokens_seen": 13310856,
      "step": 20405
    },
    {
      "epoch": 10.69706498951782,
      "grad_norm": 0.41327786445617676,
      "learning_rate": 2.632347123803575e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 13313448,
      "step": 20410
    },
    {
      "epoch": 10.699685534591195,
      "grad_norm": 0.514306366443634,
      "learning_rate": 2.6312052848732353e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 13316776,
      "step": 20415
    },
    {
      "epoch": 10.70230607966457,
      "grad_norm": 0.2627403736114502,
      "learning_rate": 2.6300634184962475e-05,
      "loss": 0.5434,
      "num_input_tokens_seen": 13319560,
      "step": 20420
    },
    {
      "epoch": 10.704926624737945,
      "grad_norm": 0.5286449193954468,
      "learning_rate": 2.628921524911475e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 13322440,
      "step": 20425
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 0.5113546848297119,
      "learning_rate": 2.6277796043577924e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 13325448,
      "step": 20430
    },
    {
      "epoch": 10.710167714884696,
      "grad_norm": 0.45663467049598694,
      "learning_rate": 2.6266376570740725e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 13328520,
      "step": 20435
    },
    {
      "epoch": 10.71278825995807,
      "grad_norm": 0.49519386887550354,
      "learning_rate": 2.6254956832992005e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 13331560,
      "step": 20440
    },
    {
      "epoch": 10.715408805031446,
      "grad_norm": 0.5152158141136169,
      "learning_rate": 2.6243536832720627e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 13334216,
      "step": 20445
    },
    {
      "epoch": 10.718029350104821,
      "grad_norm": 0.9288917779922485,
      "learning_rate": 2.623211657231553e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 13337768,
      "step": 20450
    },
    {
      "epoch": 10.720649895178196,
      "grad_norm": 0.45441776514053345,
      "learning_rate": 2.6220696054165704e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 13341064,
      "step": 20455
    },
    {
      "epoch": 10.723270440251572,
      "grad_norm": 0.3096106946468353,
      "learning_rate": 2.6209275280660194e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 13343688,
      "step": 20460
    },
    {
      "epoch": 10.725890985324948,
      "grad_norm": 0.3880907893180847,
      "learning_rate": 2.6197854254188085e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 13347048,
      "step": 20465
    },
    {
      "epoch": 10.728511530398324,
      "grad_norm": 0.27886471152305603,
      "learning_rate": 2.618643297713854e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 13350952,
      "step": 20470
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 0.586216390132904,
      "learning_rate": 2.617501145190075e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 13353672,
      "step": 20475
    },
    {
      "epoch": 10.733752620545074,
      "grad_norm": 0.35683152079582214,
      "learning_rate": 2.616358968086397e-05,
      "loss": 0.5519,
      "num_input_tokens_seen": 13357544,
      "step": 20480
    },
    {
      "epoch": 10.73637316561845,
      "grad_norm": 0.40879881381988525,
      "learning_rate": 2.6152167666417493e-05,
      "loss": 0.3863,
      "num_input_tokens_seen": 13361544,
      "step": 20485
    },
    {
      "epoch": 10.738993710691824,
      "grad_norm": 0.30840498208999634,
      "learning_rate": 2.6140745410950694e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 13365480,
      "step": 20490
    },
    {
      "epoch": 10.7416142557652,
      "grad_norm": 0.33165162801742554,
      "learning_rate": 2.6129322916852973e-05,
      "loss": 0.3642,
      "num_input_tokens_seen": 13368072,
      "step": 20495
    },
    {
      "epoch": 10.744234800838575,
      "grad_norm": 0.512101948261261,
      "learning_rate": 2.611790018651377e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 13370920,
      "step": 20500
    },
    {
      "epoch": 10.74685534591195,
      "grad_norm": 0.711812436580658,
      "learning_rate": 2.6106477222322613e-05,
      "loss": 0.7299,
      "num_input_tokens_seen": 13373832,
      "step": 20505
    },
    {
      "epoch": 10.749475890985325,
      "grad_norm": 0.4063056707382202,
      "learning_rate": 2.6095054026669037e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 13376712,
      "step": 20510
    },
    {
      "epoch": 10.7520964360587,
      "grad_norm": 0.8141680359840393,
      "learning_rate": 2.608363060194265e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 13379496,
      "step": 20515
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 0.3043820559978485,
      "learning_rate": 2.6072206950533117e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 13383688,
      "step": 20520
    },
    {
      "epoch": 10.75733752620545,
      "grad_norm": 0.2610624134540558,
      "learning_rate": 2.606078307483011e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 13387112,
      "step": 20525
    },
    {
      "epoch": 10.759958071278826,
      "grad_norm": 0.27080482244491577,
      "learning_rate": 2.6049358977223392e-05,
      "loss": 0.4432,
      "num_input_tokens_seen": 13391048,
      "step": 20530
    },
    {
      "epoch": 10.7625786163522,
      "grad_norm": 0.39136457443237305,
      "learning_rate": 2.6037934660102747e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 13393448,
      "step": 20535
    },
    {
      "epoch": 10.765199161425576,
      "grad_norm": 0.45963987708091736,
      "learning_rate": 2.6026510125858006e-05,
      "loss": 0.5475,
      "num_input_tokens_seen": 13396040,
      "step": 20540
    },
    {
      "epoch": 10.767819706498951,
      "grad_norm": 0.5779476165771484,
      "learning_rate": 2.6015085376879063e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 13398824,
      "step": 20545
    },
    {
      "epoch": 10.770440251572326,
      "grad_norm": 0.4994368851184845,
      "learning_rate": 2.600366041555584e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 13403368,
      "step": 20550
    },
    {
      "epoch": 10.773060796645701,
      "grad_norm": 0.3152039647102356,
      "learning_rate": 2.5992235244278302e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 13406376,
      "step": 20555
    },
    {
      "epoch": 10.775681341719078,
      "grad_norm": 0.42220643162727356,
      "learning_rate": 2.5980809865436485e-05,
      "loss": 0.5499,
      "num_input_tokens_seen": 13411208,
      "step": 20560
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 0.4173280596733093,
      "learning_rate": 2.5969384281420424e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 13413704,
      "step": 20565
    },
    {
      "epoch": 10.780922431865829,
      "grad_norm": 0.5495432615280151,
      "learning_rate": 2.5957958494620227e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 13417128,
      "step": 20570
    },
    {
      "epoch": 10.783542976939204,
      "grad_norm": 0.6223535537719727,
      "learning_rate": 2.594653250742605e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 13420136,
      "step": 20575
    },
    {
      "epoch": 10.786163522012579,
      "grad_norm": 0.7383907437324524,
      "learning_rate": 2.5935106322228065e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 13422472,
      "step": 20580
    },
    {
      "epoch": 10.788784067085954,
      "grad_norm": 0.5826482176780701,
      "learning_rate": 2.5923679941416514e-05,
      "loss": 0.3321,
      "num_input_tokens_seen": 13426280,
      "step": 20585
    },
    {
      "epoch": 10.79140461215933,
      "grad_norm": 0.6281793117523193,
      "learning_rate": 2.591225336738165e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 13429416,
      "step": 20590
    },
    {
      "epoch": 10.794025157232705,
      "grad_norm": 0.5339989066123962,
      "learning_rate": 2.5900826602513782e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 13432776,
      "step": 20595
    },
    {
      "epoch": 10.79664570230608,
      "grad_norm": 0.7619349956512451,
      "learning_rate": 2.5889399649203277e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 13435976,
      "step": 20600
    },
    {
      "epoch": 10.799266247379455,
      "grad_norm": 0.4472929835319519,
      "learning_rate": 2.5877972509840514e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 13439208,
      "step": 20605
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 0.5876714587211609,
      "learning_rate": 2.5866545186815912e-05,
      "loss": 0.3935,
      "num_input_tokens_seen": 13441928,
      "step": 20610
    },
    {
      "epoch": 10.804507337526205,
      "grad_norm": 0.4473322927951813,
      "learning_rate": 2.5855117682519954e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 13445000,
      "step": 20615
    },
    {
      "epoch": 10.80712788259958,
      "grad_norm": 1.350942611694336,
      "learning_rate": 2.584368999934312e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 13449032,
      "step": 20620
    },
    {
      "epoch": 10.809748427672956,
      "grad_norm": 0.34139886498451233,
      "learning_rate": 2.5832262139675966e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 13452392,
      "step": 20625
    },
    {
      "epoch": 10.81236897274633,
      "grad_norm": 0.3170621395111084,
      "learning_rate": 2.582083410590907e-05,
      "loss": 0.6839,
      "num_input_tokens_seen": 13455784,
      "step": 20630
    },
    {
      "epoch": 10.814989517819706,
      "grad_norm": 0.4052482545375824,
      "learning_rate": 2.5809405900433036e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 13459368,
      "step": 20635
    },
    {
      "epoch": 10.817610062893081,
      "grad_norm": 0.4008921682834625,
      "learning_rate": 2.579797752563853e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 13462664,
      "step": 20640
    },
    {
      "epoch": 10.820230607966456,
      "grad_norm": 0.492328017950058,
      "learning_rate": 2.5786548983916232e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 13464872,
      "step": 20645
    },
    {
      "epoch": 10.822851153039831,
      "grad_norm": 0.7000446915626526,
      "learning_rate": 2.577512027765685e-05,
      "loss": 0.3027,
      "num_input_tokens_seen": 13467624,
      "step": 20650
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 0.5242775082588196,
      "learning_rate": 2.5763691409251155e-05,
      "loss": 0.5213,
      "num_input_tokens_seen": 13470856,
      "step": 20655
    },
    {
      "epoch": 10.828092243186584,
      "grad_norm": 0.4866035580635071,
      "learning_rate": 2.5752262381089932e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 13474024,
      "step": 20660
    },
    {
      "epoch": 10.830712788259959,
      "grad_norm": 0.5449416637420654,
      "learning_rate": 2.5740833195563996e-05,
      "loss": 0.4558,
      "num_input_tokens_seen": 13478376,
      "step": 20665
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 0.49889352917671204,
      "learning_rate": 2.572940385506421e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 13481480,
      "step": 20670
    },
    {
      "epoch": 10.835953878406709,
      "grad_norm": 0.7347068190574646,
      "learning_rate": 2.5717974361981457e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 13484456,
      "step": 20675
    },
    {
      "epoch": 10.838574423480084,
      "grad_norm": 0.6652183532714844,
      "learning_rate": 2.5706544718706665e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 13487560,
      "step": 20680
    },
    {
      "epoch": 10.84119496855346,
      "grad_norm": 0.20922622084617615,
      "learning_rate": 2.569511492763077e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 13491272,
      "step": 20685
    },
    {
      "epoch": 10.843815513626835,
      "grad_norm": 0.5450721979141235,
      "learning_rate": 2.5683684991144764e-05,
      "loss": 0.3781,
      "num_input_tokens_seen": 13493992,
      "step": 20690
    },
    {
      "epoch": 10.84643605870021,
      "grad_norm": 0.613920271396637,
      "learning_rate": 2.5672254911639666e-05,
      "loss": 0.5199,
      "num_input_tokens_seen": 13496968,
      "step": 20695
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 0.37988704442977905,
      "learning_rate": 2.5660824691506496e-05,
      "loss": 0.5754,
      "num_input_tokens_seen": 13501032,
      "step": 20700
    },
    {
      "epoch": 10.85167714884696,
      "grad_norm": 0.5891987681388855,
      "learning_rate": 2.5649394333136356e-05,
      "loss": 0.6534,
      "num_input_tokens_seen": 13502984,
      "step": 20705
    },
    {
      "epoch": 10.854297693920335,
      "grad_norm": 0.3073768615722656,
      "learning_rate": 2.5637963838920316e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 13507816,
      "step": 20710
    },
    {
      "epoch": 10.85691823899371,
      "grad_norm": 0.4855647683143616,
      "learning_rate": 2.5626533211249525e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 13511368,
      "step": 20715
    },
    {
      "epoch": 10.859538784067086,
      "grad_norm": 0.6703751087188721,
      "learning_rate": 2.5615102452515138e-05,
      "loss": 0.5083,
      "num_input_tokens_seen": 13515720,
      "step": 20720
    },
    {
      "epoch": 10.86215932914046,
      "grad_norm": 0.43763095140457153,
      "learning_rate": 2.5603671565108328e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 13519784,
      "step": 20725
    },
    {
      "epoch": 10.864779874213836,
      "grad_norm": 0.32997921109199524,
      "learning_rate": 2.5592240551420317e-05,
      "loss": 0.3842,
      "num_input_tokens_seen": 13523304,
      "step": 20730
    },
    {
      "epoch": 10.867400419287211,
      "grad_norm": 1.142174243927002,
      "learning_rate": 2.5580809413842343e-05,
      "loss": 0.7169,
      "num_input_tokens_seen": 13526120,
      "step": 20735
    },
    {
      "epoch": 10.870020964360586,
      "grad_norm": 0.5686646103858948,
      "learning_rate": 2.5569378154765655e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 13529128,
      "step": 20740
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 0.2753567695617676,
      "learning_rate": 2.5557946776581572e-05,
      "loss": 0.5119,
      "num_input_tokens_seen": 13532584,
      "step": 20745
    },
    {
      "epoch": 10.875262054507338,
      "grad_norm": 0.43575069308280945,
      "learning_rate": 2.554651528168137e-05,
      "loss": 0.413,
      "num_input_tokens_seen": 13536104,
      "step": 20750
    },
    {
      "epoch": 10.877882599580714,
      "grad_norm": 0.3675479292869568,
      "learning_rate": 2.5535083672456407e-05,
      "loss": 0.3493,
      "num_input_tokens_seen": 13539528,
      "step": 20755
    },
    {
      "epoch": 10.880503144654089,
      "grad_norm": 0.584627091884613,
      "learning_rate": 2.5523651951298055e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 13542600,
      "step": 20760
    },
    {
      "epoch": 10.883123689727464,
      "grad_norm": 0.5084503293037415,
      "learning_rate": 2.5512220120597686e-05,
      "loss": 0.3324,
      "num_input_tokens_seen": 13545704,
      "step": 20765
    },
    {
      "epoch": 10.885744234800839,
      "grad_norm": 0.5363572239875793,
      "learning_rate": 2.550078818274671e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 13549096,
      "step": 20770
    },
    {
      "epoch": 10.888364779874214,
      "grad_norm": 0.5583590269088745,
      "learning_rate": 2.548935614013656e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 13552232,
      "step": 20775
    },
    {
      "epoch": 10.89098532494759,
      "grad_norm": 0.7388966083526611,
      "learning_rate": 2.5477923995158683e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 13554984,
      "step": 20780
    },
    {
      "epoch": 10.893605870020965,
      "grad_norm": 0.5168837904930115,
      "learning_rate": 2.5466491750204562e-05,
      "loss": 0.2725,
      "num_input_tokens_seen": 13561288,
      "step": 20785
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 0.34021973609924316,
      "learning_rate": 2.5455059407665693e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 13564072,
      "step": 20790
    },
    {
      "epoch": 10.898846960167715,
      "grad_norm": 0.492851585149765,
      "learning_rate": 2.544362696993358e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 13566856,
      "step": 20795
    },
    {
      "epoch": 10.90146750524109,
      "grad_norm": 0.4878513813018799,
      "learning_rate": 2.5432194439399775e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 13570696,
      "step": 20800
    },
    {
      "epoch": 10.904088050314465,
      "grad_norm": 0.3100266754627228,
      "learning_rate": 2.542076181845582e-05,
      "loss": 0.3432,
      "num_input_tokens_seen": 13573416,
      "step": 20805
    },
    {
      "epoch": 10.90670859538784,
      "grad_norm": 0.5069925785064697,
      "learning_rate": 2.540932910949329e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 13576392,
      "step": 20810
    },
    {
      "epoch": 10.909329140461216,
      "grad_norm": 0.6650694012641907,
      "learning_rate": 2.5397896314903785e-05,
      "loss": 0.5209,
      "num_input_tokens_seen": 13579048,
      "step": 20815
    },
    {
      "epoch": 10.91194968553459,
      "grad_norm": 1.028156042098999,
      "learning_rate": 2.5386463437078905e-05,
      "loss": 0.5817,
      "num_input_tokens_seen": 13582920,
      "step": 20820
    },
    {
      "epoch": 10.914570230607966,
      "grad_norm": 0.4164457619190216,
      "learning_rate": 2.5375030478410282e-05,
      "loss": 0.3533,
      "num_input_tokens_seen": 13585576,
      "step": 20825
    },
    {
      "epoch": 10.917190775681341,
      "grad_norm": 0.767359733581543,
      "learning_rate": 2.536359744128957e-05,
      "loss": 0.4267,
      "num_input_tokens_seen": 13588360,
      "step": 20830
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 0.35187283158302307,
      "learning_rate": 2.5352164328108413e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 13591272,
      "step": 20835
    },
    {
      "epoch": 10.922431865828091,
      "grad_norm": 0.4489682912826538,
      "learning_rate": 2.53407311412585e-05,
      "loss": 0.3659,
      "num_input_tokens_seen": 13594184,
      "step": 20840
    },
    {
      "epoch": 10.925052410901468,
      "grad_norm": 0.6729080677032471,
      "learning_rate": 2.5329297883131526e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 13597352,
      "step": 20845
    },
    {
      "epoch": 10.927672955974844,
      "grad_norm": 0.335610032081604,
      "learning_rate": 2.5317864556119182e-05,
      "loss": 0.5112,
      "num_input_tokens_seen": 13600008,
      "step": 20850
    },
    {
      "epoch": 10.930293501048219,
      "grad_norm": 0.3767024278640747,
      "learning_rate": 2.53064311626132e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 13603656,
      "step": 20855
    },
    {
      "epoch": 10.932914046121594,
      "grad_norm": 0.4128919243812561,
      "learning_rate": 2.5294997705005318e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 13607176,
      "step": 20860
    },
    {
      "epoch": 10.935534591194969,
      "grad_norm": 0.5745609998703003,
      "learning_rate": 2.5283564185687282e-05,
      "loss": 0.5962,
      "num_input_tokens_seen": 13610600,
      "step": 20865
    },
    {
      "epoch": 10.938155136268344,
      "grad_norm": 0.4607028067111969,
      "learning_rate": 2.5272130607050848e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 13614472,
      "step": 20870
    },
    {
      "epoch": 10.94077568134172,
      "grad_norm": 0.48879092931747437,
      "learning_rate": 2.5260696971487802e-05,
      "loss": 0.369,
      "num_input_tokens_seen": 13617384,
      "step": 20875
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 0.473403662443161,
      "learning_rate": 2.5249263281389923e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 13620968,
      "step": 20880
    },
    {
      "epoch": 10.94601677148847,
      "grad_norm": 0.5845133662223816,
      "learning_rate": 2.5237829539149004e-05,
      "loss": 0.5346,
      "num_input_tokens_seen": 13623752,
      "step": 20885
    },
    {
      "epoch": 10.948637316561845,
      "grad_norm": 0.4347004294395447,
      "learning_rate": 2.5226395747156866e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 13629000,
      "step": 20890
    },
    {
      "epoch": 10.95125786163522,
      "grad_norm": 0.4983164668083191,
      "learning_rate": 2.5214961907805317e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 13632200,
      "step": 20895
    },
    {
      "epoch": 10.953878406708595,
      "grad_norm": 0.5261877775192261,
      "learning_rate": 2.520352802348619e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 13634536,
      "step": 20900
    },
    {
      "epoch": 10.95649895178197,
      "grad_norm": 0.35419467091560364,
      "learning_rate": 2.5192094096591323e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 13638760,
      "step": 20905
    },
    {
      "epoch": 10.959119496855346,
      "grad_norm": 0.4989893138408661,
      "learning_rate": 2.5180660129512556e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 13641256,
      "step": 20910
    },
    {
      "epoch": 10.96174004192872,
      "grad_norm": 0.4551429748535156,
      "learning_rate": 2.516922612464176e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 13644808,
      "step": 20915
    },
    {
      "epoch": 10.964360587002096,
      "grad_norm": 0.3598991930484772,
      "learning_rate": 2.5157792084370783e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 13648168,
      "step": 20920
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 0.33699899911880493,
      "learning_rate": 2.5146358011091502e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 13652520,
      "step": 20925
    },
    {
      "epoch": 10.969601677148846,
      "grad_norm": 0.5152308344841003,
      "learning_rate": 2.5134923907195805e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 13655656,
      "step": 20930
    },
    {
      "epoch": 10.972222222222221,
      "grad_norm": 0.7914795279502869,
      "learning_rate": 2.5123489775075558e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 13658440,
      "step": 20935
    },
    {
      "epoch": 10.974842767295598,
      "grad_norm": 0.5735278129577637,
      "learning_rate": 2.511205561712266e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 13661160,
      "step": 20940
    },
    {
      "epoch": 10.977463312368974,
      "grad_norm": 0.3905673921108246,
      "learning_rate": 2.5100621435729004e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 13664936,
      "step": 20945
    },
    {
      "epoch": 10.980083857442349,
      "grad_norm": 0.522681474685669,
      "learning_rate": 2.5089187233286492e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 13668072,
      "step": 20950
    },
    {
      "epoch": 10.982704402515724,
      "grad_norm": 0.3151540160179138,
      "learning_rate": 2.5077753012187026e-05,
      "loss": 0.3894,
      "num_input_tokens_seen": 13670856,
      "step": 20955
    },
    {
      "epoch": 10.985324947589099,
      "grad_norm": 0.3698515295982361,
      "learning_rate": 2.506631877482252e-05,
      "loss": 0.4036,
      "num_input_tokens_seen": 13674792,
      "step": 20960
    },
    {
      "epoch": 10.987945492662474,
      "grad_norm": 0.4618019759654999,
      "learning_rate": 2.505488452358488e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 13677544,
      "step": 20965
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 0.23664651811122894,
      "learning_rate": 2.504345026086603e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 13681896,
      "step": 20970
    },
    {
      "epoch": 10.993186582809225,
      "grad_norm": 0.46652159094810486,
      "learning_rate": 2.5032015989057877e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 13685256,
      "step": 20975
    },
    {
      "epoch": 10.9958071278826,
      "grad_norm": 1.0353630781173706,
      "learning_rate": 2.5020581710552348e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 13688072,
      "step": 20980
    },
    {
      "epoch": 10.998427672955975,
      "grad_norm": 0.4278191030025482,
      "learning_rate": 2.5009147427741358e-05,
      "loss": 0.3796,
      "num_input_tokens_seen": 13690824,
      "step": 20985
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.45386403799057007,
      "eval_runtime": 9.8645,
      "eval_samples_per_second": 85.965,
      "eval_steps_per_second": 21.491,
      "num_input_tokens_seen": 13691904,
      "step": 20988
    },
    {
      "epoch": 11.00104821802935,
      "grad_norm": 0.4357921779155731,
      "learning_rate": 2.4997713143016822e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 13692896,
      "step": 20990
    },
    {
      "epoch": 11.003668763102725,
      "grad_norm": 0.48829731345176697,
      "learning_rate": 2.498627885877068e-05,
      "loss": 0.3573,
      "num_input_tokens_seen": 13696160,
      "step": 20995
    },
    {
      "epoch": 11.0062893081761,
      "grad_norm": 0.6004542708396912,
      "learning_rate": 2.4974844577394846e-05,
      "loss": 0.5174,
      "num_input_tokens_seen": 13699136,
      "step": 21000
    },
    {
      "epoch": 11.008909853249476,
      "grad_norm": 0.4690842032432556,
      "learning_rate": 2.496341030128123e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 13702496,
      "step": 21005
    },
    {
      "epoch": 11.01153039832285,
      "grad_norm": 0.29571333527565,
      "learning_rate": 2.495197603282177e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 13706048,
      "step": 21010
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 0.36921200156211853,
      "learning_rate": 2.4940541774408368e-05,
      "loss": 0.5478,
      "num_input_tokens_seen": 13708640,
      "step": 21015
    },
    {
      "epoch": 11.016771488469601,
      "grad_norm": 0.7296498417854309,
      "learning_rate": 2.4929107528432953e-05,
      "loss": 0.3484,
      "num_input_tokens_seen": 13712000,
      "step": 21020
    },
    {
      "epoch": 11.019392033542976,
      "grad_norm": 0.4804331958293915,
      "learning_rate": 2.4917673297287446e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 13715232,
      "step": 21025
    },
    {
      "epoch": 11.022012578616351,
      "grad_norm": 0.5379234552383423,
      "learning_rate": 2.4906239083363728e-05,
      "loss": 0.8048,
      "num_input_tokens_seen": 13718368,
      "step": 21030
    },
    {
      "epoch": 11.024633123689727,
      "grad_norm": 0.7564116716384888,
      "learning_rate": 2.4894804889053732e-05,
      "loss": 0.4649,
      "num_input_tokens_seen": 13721824,
      "step": 21035
    },
    {
      "epoch": 11.027253668763104,
      "grad_norm": 0.38731566071510315,
      "learning_rate": 2.4883370716749347e-05,
      "loss": 0.4813,
      "num_input_tokens_seen": 13724480,
      "step": 21040
    },
    {
      "epoch": 11.029874213836479,
      "grad_norm": 0.5725563168525696,
      "learning_rate": 2.4871936568842486e-05,
      "loss": 0.3741,
      "num_input_tokens_seen": 13726656,
      "step": 21045
    },
    {
      "epoch": 11.032494758909854,
      "grad_norm": 0.5038725733757019,
      "learning_rate": 2.4860502447725033e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 13729184,
      "step": 21050
    },
    {
      "epoch": 11.035115303983229,
      "grad_norm": 0.3791976273059845,
      "learning_rate": 2.4849068355788873e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 13733600,
      "step": 21055
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 0.3668937385082245,
      "learning_rate": 2.4837634295425892e-05,
      "loss": 0.5548,
      "num_input_tokens_seen": 13736832,
      "step": 21060
    },
    {
      "epoch": 11.04035639412998,
      "grad_norm": 0.6850131154060364,
      "learning_rate": 2.4826200269027957e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 13739360,
      "step": 21065
    },
    {
      "epoch": 11.042976939203355,
      "grad_norm": 0.5353417992591858,
      "learning_rate": 2.481476627898695e-05,
      "loss": 0.3645,
      "num_input_tokens_seen": 13743296,
      "step": 21070
    },
    {
      "epoch": 11.04559748427673,
      "grad_norm": 0.2922656238079071,
      "learning_rate": 2.4803332327694724e-05,
      "loss": 0.354,
      "num_input_tokens_seen": 13747072,
      "step": 21075
    },
    {
      "epoch": 11.048218029350105,
      "grad_norm": 0.42603185772895813,
      "learning_rate": 2.4791898417543136e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 13750208,
      "step": 21080
    },
    {
      "epoch": 11.05083857442348,
      "grad_norm": 0.5158711075782776,
      "learning_rate": 2.4780464550924014e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 13752928,
      "step": 21085
    },
    {
      "epoch": 11.053459119496855,
      "grad_norm": 0.5349435210227966,
      "learning_rate": 2.4769030730229213e-05,
      "loss": 0.3734,
      "num_input_tokens_seen": 13756736,
      "step": 21090
    },
    {
      "epoch": 11.05607966457023,
      "grad_norm": 0.7046539187431335,
      "learning_rate": 2.475759695785054e-05,
      "loss": 0.4495,
      "num_input_tokens_seen": 13759520,
      "step": 21095
    },
    {
      "epoch": 11.058700209643606,
      "grad_norm": 0.4722307324409485,
      "learning_rate": 2.4746163236179822e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 13762592,
      "step": 21100
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 0.41430139541625977,
      "learning_rate": 2.4734729567608858e-05,
      "loss": 0.5534,
      "num_input_tokens_seen": 13766464,
      "step": 21105
    },
    {
      "epoch": 11.063941299790356,
      "grad_norm": 0.10387632995843887,
      "learning_rate": 2.472329595452945e-05,
      "loss": 0.4115,
      "num_input_tokens_seen": 13772672,
      "step": 21110
    },
    {
      "epoch": 11.066561844863731,
      "grad_norm": 0.46042659878730774,
      "learning_rate": 2.4711862399333367e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 13776000,
      "step": 21115
    },
    {
      "epoch": 11.069182389937106,
      "grad_norm": 0.7732405662536621,
      "learning_rate": 2.4700428904412377e-05,
      "loss": 0.7294,
      "num_input_tokens_seen": 13778464,
      "step": 21120
    },
    {
      "epoch": 11.071802935010481,
      "grad_norm": 0.4419375956058502,
      "learning_rate": 2.4688995472158245e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 13781536,
      "step": 21125
    },
    {
      "epoch": 11.074423480083857,
      "grad_norm": 0.5478854775428772,
      "learning_rate": 2.4677562104962712e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 13784800,
      "step": 21130
    },
    {
      "epoch": 11.077044025157234,
      "grad_norm": 0.23007477819919586,
      "learning_rate": 2.4666128805217513e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 13788384,
      "step": 21135
    },
    {
      "epoch": 11.079664570230609,
      "grad_norm": 0.4149163067340851,
      "learning_rate": 2.465469557531436e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 13791744,
      "step": 21140
    },
    {
      "epoch": 11.082285115303984,
      "grad_norm": 0.276321679353714,
      "learning_rate": 2.4643262417644944e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 13798304,
      "step": 21145
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 0.36855387687683105,
      "learning_rate": 2.463182933460096e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 13801184,
      "step": 21150
    },
    {
      "epoch": 11.087526205450734,
      "grad_norm": 0.37798479199409485,
      "learning_rate": 2.4620396328574087e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 13805120,
      "step": 21155
    },
    {
      "epoch": 11.09014675052411,
      "grad_norm": 0.23706971108913422,
      "learning_rate": 2.4608963401955965e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 13808448,
      "step": 21160
    },
    {
      "epoch": 11.092767295597485,
      "grad_norm": 0.5577272772789001,
      "learning_rate": 2.4597530557138247e-05,
      "loss": 0.4139,
      "num_input_tokens_seen": 13812320,
      "step": 21165
    },
    {
      "epoch": 11.09538784067086,
      "grad_norm": 0.30490773916244507,
      "learning_rate": 2.4586097796512538e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 13815168,
      "step": 21170
    },
    {
      "epoch": 11.098008385744235,
      "grad_norm": 0.6865130662918091,
      "learning_rate": 2.457466512247045e-05,
      "loss": 0.5118,
      "num_input_tokens_seen": 13817696,
      "step": 21175
    },
    {
      "epoch": 11.10062893081761,
      "grad_norm": 0.3446144759654999,
      "learning_rate": 2.4563232537403563e-05,
      "loss": 0.5181,
      "num_input_tokens_seen": 13820320,
      "step": 21180
    },
    {
      "epoch": 11.103249475890985,
      "grad_norm": 0.49824395775794983,
      "learning_rate": 2.4551800043703442e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 13823680,
      "step": 21185
    },
    {
      "epoch": 11.10587002096436,
      "grad_norm": 0.7807856798171997,
      "learning_rate": 2.454036764376165e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 13826016,
      "step": 21190
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 0.8906300663948059,
      "learning_rate": 2.4528935339969705e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 13828416,
      "step": 21195
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 0.6060546636581421,
      "learning_rate": 2.45175031347191e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 13830976,
      "step": 21200
    },
    {
      "epoch": 11.113731656184486,
      "grad_norm": 0.4707571268081665,
      "learning_rate": 2.4506071030401344e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 13833952,
      "step": 21205
    },
    {
      "epoch": 11.116352201257861,
      "grad_norm": 0.614512026309967,
      "learning_rate": 2.4494639029407884e-05,
      "loss": 0.3648,
      "num_input_tokens_seen": 13836224,
      "step": 21210
    },
    {
      "epoch": 11.118972746331236,
      "grad_norm": 0.5126772522926331,
      "learning_rate": 2.448320713413018e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 13840416,
      "step": 21215
    },
    {
      "epoch": 11.121593291404611,
      "grad_norm": 1.3995593786239624,
      "learning_rate": 2.4471775346959654e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 13843072,
      "step": 21220
    },
    {
      "epoch": 11.124213836477987,
      "grad_norm": 0.5354810357093811,
      "learning_rate": 2.446034367028769e-05,
      "loss": 0.5756,
      "num_input_tokens_seen": 13845728,
      "step": 21225
    },
    {
      "epoch": 11.126834381551364,
      "grad_norm": 0.37015312910079956,
      "learning_rate": 2.444891210650567e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 13849472,
      "step": 21230
    },
    {
      "epoch": 11.129454926624739,
      "grad_norm": 0.4721488952636719,
      "learning_rate": 2.4437480658004958e-05,
      "loss": 0.3931,
      "num_input_tokens_seen": 13852704,
      "step": 21235
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 0.4473842680454254,
      "learning_rate": 2.4426049327176868e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 13856000,
      "step": 21240
    },
    {
      "epoch": 11.134696016771489,
      "grad_norm": 0.47667035460472107,
      "learning_rate": 2.4414618116412716e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 13858944,
      "step": 21245
    },
    {
      "epoch": 11.137316561844864,
      "grad_norm": 0.6128430962562561,
      "learning_rate": 2.440318702810378e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 13861504,
      "step": 21250
    },
    {
      "epoch": 11.13993710691824,
      "grad_norm": 0.37388697266578674,
      "learning_rate": 2.43917560646413e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 13865088,
      "step": 21255
    },
    {
      "epoch": 11.142557651991615,
      "grad_norm": 0.5218276977539062,
      "learning_rate": 2.4380325228416518e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 13867488,
      "step": 21260
    },
    {
      "epoch": 11.14517819706499,
      "grad_norm": 0.36953479051589966,
      "learning_rate": 2.4368894521820624e-05,
      "loss": 0.4853,
      "num_input_tokens_seen": 13870912,
      "step": 21265
    },
    {
      "epoch": 11.147798742138365,
      "grad_norm": 0.545555055141449,
      "learning_rate": 2.43574639472448e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 13874080,
      "step": 21270
    },
    {
      "epoch": 11.15041928721174,
      "grad_norm": 0.6202980875968933,
      "learning_rate": 2.4346033507080178e-05,
      "loss": 0.552,
      "num_input_tokens_seen": 13877920,
      "step": 21275
    },
    {
      "epoch": 11.153039832285115,
      "grad_norm": 0.34491443634033203,
      "learning_rate": 2.4334603203717903e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 13880608,
      "step": 21280
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 1.5706945657730103,
      "learning_rate": 2.432317303954904e-05,
      "loss": 0.6112,
      "num_input_tokens_seen": 13883936,
      "step": 21285
    },
    {
      "epoch": 11.158280922431866,
      "grad_norm": 0.2913244664669037,
      "learning_rate": 2.4311743016964648e-05,
      "loss": 0.5552,
      "num_input_tokens_seen": 13887680,
      "step": 21290
    },
    {
      "epoch": 11.16090146750524,
      "grad_norm": 0.3634314239025116,
      "learning_rate": 2.4300313138355772e-05,
      "loss": 0.487,
      "num_input_tokens_seen": 13890592,
      "step": 21295
    },
    {
      "epoch": 11.163522012578616,
      "grad_norm": 0.5523890852928162,
      "learning_rate": 2.4288883406113396e-05,
      "loss": 0.5083,
      "num_input_tokens_seen": 13892704,
      "step": 21300
    },
    {
      "epoch": 11.166142557651991,
      "grad_norm": 0.5256287455558777,
      "learning_rate": 2.4277453822628512e-05,
      "loss": 0.3534,
      "num_input_tokens_seen": 13895616,
      "step": 21305
    },
    {
      "epoch": 11.168763102725366,
      "grad_norm": 0.45684128999710083,
      "learning_rate": 2.4266024390292037e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 13898688,
      "step": 21310
    },
    {
      "epoch": 11.171383647798741,
      "grad_norm": 0.35545027256011963,
      "learning_rate": 2.425459511149489e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 13901760,
      "step": 21315
    },
    {
      "epoch": 11.174004192872117,
      "grad_norm": 0.6345577836036682,
      "learning_rate": 2.424316598862793e-05,
      "loss": 0.5715,
      "num_input_tokens_seen": 13904416,
      "step": 21320
    },
    {
      "epoch": 11.176624737945493,
      "grad_norm": 0.29294338822364807,
      "learning_rate": 2.4231737024082023e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 13908192,
      "step": 21325
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 0.31824955344200134,
      "learning_rate": 2.4220308220247954e-05,
      "loss": 0.5598,
      "num_input_tokens_seen": 13911392,
      "step": 21330
    },
    {
      "epoch": 11.181865828092244,
      "grad_norm": 0.443101167678833,
      "learning_rate": 2.4208879579516528e-05,
      "loss": 0.3231,
      "num_input_tokens_seen": 13914368,
      "step": 21335
    },
    {
      "epoch": 11.184486373165619,
      "grad_norm": 0.43039441108703613,
      "learning_rate": 2.419745110427846e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 13916544,
      "step": 21340
    },
    {
      "epoch": 11.187106918238994,
      "grad_norm": 0.31673163175582886,
      "learning_rate": 2.418602279692446e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 13919520,
      "step": 21345
    },
    {
      "epoch": 11.18972746331237,
      "grad_norm": 0.7396457195281982,
      "learning_rate": 2.417459465984521e-05,
      "loss": 0.3734,
      "num_input_tokens_seen": 13922496,
      "step": 21350
    },
    {
      "epoch": 11.192348008385745,
      "grad_norm": 0.5412261486053467,
      "learning_rate": 2.4163166695431336e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 13925440,
      "step": 21355
    },
    {
      "epoch": 11.19496855345912,
      "grad_norm": 0.46402472257614136,
      "learning_rate": 2.4151738906073446e-05,
      "loss": 0.4739,
      "num_input_tokens_seen": 13928544,
      "step": 21360
    },
    {
      "epoch": 11.197589098532495,
      "grad_norm": 0.6467875838279724,
      "learning_rate": 2.4140311294162105e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 13931200,
      "step": 21365
    },
    {
      "epoch": 11.20020964360587,
      "grad_norm": 0.4121837019920349,
      "learning_rate": 2.412888386208782e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 13933984,
      "step": 21370
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 0.3510471284389496,
      "learning_rate": 2.4117456612241103e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 13937184,
      "step": 21375
    },
    {
      "epoch": 11.20545073375262,
      "grad_norm": 0.3288041651248932,
      "learning_rate": 2.410602954701239e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 13940864,
      "step": 21380
    },
    {
      "epoch": 11.208071278825996,
      "grad_norm": 0.48853009939193726,
      "learning_rate": 2.409460266879209e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 13943904,
      "step": 21385
    },
    {
      "epoch": 11.21069182389937,
      "grad_norm": 0.45728152990341187,
      "learning_rate": 2.4083175979970592e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 13948192,
      "step": 21390
    },
    {
      "epoch": 11.213312368972746,
      "grad_norm": 0.4528047442436218,
      "learning_rate": 2.4071749482938226e-05,
      "loss": 0.2936,
      "num_input_tokens_seen": 13950912,
      "step": 21395
    },
    {
      "epoch": 11.215932914046121,
      "grad_norm": 0.7141419053077698,
      "learning_rate": 2.4060323180085266e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 13953664,
      "step": 21400
    },
    {
      "epoch": 11.218553459119496,
      "grad_norm": 0.31721359491348267,
      "learning_rate": 2.404889707380198e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 13956576,
      "step": 21405
    },
    {
      "epoch": 11.221174004192871,
      "grad_norm": 0.529963493347168,
      "learning_rate": 2.403747116647857e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 13959520,
      "step": 21410
    },
    {
      "epoch": 11.223794549266247,
      "grad_norm": 0.2872457206249237,
      "learning_rate": 2.4026045460505216e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 13963552,
      "step": 21415
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 0.37789613008499146,
      "learning_rate": 2.4014619958272047e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 13967072,
      "step": 21420
    },
    {
      "epoch": 11.229035639412999,
      "grad_norm": 0.34642356634140015,
      "learning_rate": 2.4003194662169132e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 13970592,
      "step": 21425
    },
    {
      "epoch": 11.231656184486374,
      "grad_norm": 0.47744688391685486,
      "learning_rate": 2.3991769574586522e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 13973696,
      "step": 21430
    },
    {
      "epoch": 11.234276729559749,
      "grad_norm": 0.5242628455162048,
      "learning_rate": 2.3980344697914215e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 13977248,
      "step": 21435
    },
    {
      "epoch": 11.236897274633124,
      "grad_norm": 0.4235936403274536,
      "learning_rate": 2.3968920034542168e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 13980704,
      "step": 21440
    },
    {
      "epoch": 11.2395178197065,
      "grad_norm": 0.4710446894168854,
      "learning_rate": 2.3957495586860278e-05,
      "loss": 0.517,
      "num_input_tokens_seen": 13982976,
      "step": 21445
    },
    {
      "epoch": 11.242138364779874,
      "grad_norm": 2.672663450241089,
      "learning_rate": 2.394607135725843e-05,
      "loss": 0.5981,
      "num_input_tokens_seen": 13985888,
      "step": 21450
    },
    {
      "epoch": 11.24475890985325,
      "grad_norm": 0.43695345520973206,
      "learning_rate": 2.3934647348126426e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 13990048,
      "step": 21455
    },
    {
      "epoch": 11.247379454926625,
      "grad_norm": 0.4106239676475525,
      "learning_rate": 2.3923223561854035e-05,
      "loss": 0.4124,
      "num_input_tokens_seen": 13993152,
      "step": 21460
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.48395058512687683,
      "learning_rate": 2.391180000083099e-05,
      "loss": 0.5481,
      "num_input_tokens_seen": 13995776,
      "step": 21465
    },
    {
      "epoch": 11.252620545073375,
      "grad_norm": 0.36988574266433716,
      "learning_rate": 2.390037666744698e-05,
      "loss": 0.5952,
      "num_input_tokens_seen": 13999168,
      "step": 21470
    },
    {
      "epoch": 11.25524109014675,
      "grad_norm": 0.6131606101989746,
      "learning_rate": 2.388895356409162e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 14002048,
      "step": 21475
    },
    {
      "epoch": 11.257861635220126,
      "grad_norm": 0.5067524909973145,
      "learning_rate": 2.3877530693154506e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 14004864,
      "step": 21480
    },
    {
      "epoch": 11.2604821802935,
      "grad_norm": 0.46775028109550476,
      "learning_rate": 2.3866108057025162e-05,
      "loss": 0.3329,
      "num_input_tokens_seen": 14008352,
      "step": 21485
    },
    {
      "epoch": 11.263102725366876,
      "grad_norm": 0.37109118700027466,
      "learning_rate": 2.385468565809307e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 14011232,
      "step": 21490
    },
    {
      "epoch": 11.265723270440251,
      "grad_norm": 0.34434425830841064,
      "learning_rate": 2.3843263498747675e-05,
      "loss": 0.4412,
      "num_input_tokens_seen": 14014848,
      "step": 21495
    },
    {
      "epoch": 11.268343815513626,
      "grad_norm": 0.506712019443512,
      "learning_rate": 2.383184158137836e-05,
      "loss": 0.413,
      "num_input_tokens_seen": 14017536,
      "step": 21500
    },
    {
      "epoch": 11.270964360587001,
      "grad_norm": 0.5114637613296509,
      "learning_rate": 2.3820419908374465e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 14021376,
      "step": 21505
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 0.3872476816177368,
      "learning_rate": 2.3808998482125265e-05,
      "loss": 0.4875,
      "num_input_tokens_seen": 14024224,
      "step": 21510
    },
    {
      "epoch": 11.276205450733753,
      "grad_norm": 0.47521939873695374,
      "learning_rate": 2.3797577305019987e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 14026848,
      "step": 21515
    },
    {
      "epoch": 11.278825995807129,
      "grad_norm": 0.26864349842071533,
      "learning_rate": 2.378615637944783e-05,
      "loss": 0.3603,
      "num_input_tokens_seen": 14030720,
      "step": 21520
    },
    {
      "epoch": 11.281446540880504,
      "grad_norm": 0.3348752558231354,
      "learning_rate": 2.3774735707797895e-05,
      "loss": 0.3857,
      "num_input_tokens_seen": 14033632,
      "step": 21525
    },
    {
      "epoch": 11.284067085953879,
      "grad_norm": 0.44897058606147766,
      "learning_rate": 2.376331529245928e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 14036960,
      "step": 21530
    },
    {
      "epoch": 11.286687631027254,
      "grad_norm": 0.36227673292160034,
      "learning_rate": 2.3751895135821006e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 14039648,
      "step": 21535
    },
    {
      "epoch": 11.28930817610063,
      "grad_norm": 0.4295143187046051,
      "learning_rate": 2.374047524027202e-05,
      "loss": 0.5511,
      "num_input_tokens_seen": 14043840,
      "step": 21540
    },
    {
      "epoch": 11.291928721174004,
      "grad_norm": 0.42028480768203735,
      "learning_rate": 2.3729055608201243e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 14047200,
      "step": 21545
    },
    {
      "epoch": 11.29454926624738,
      "grad_norm": 0.7300304174423218,
      "learning_rate": 2.3717636241997537e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 14050336,
      "step": 21550
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 0.3762809932231903,
      "learning_rate": 2.3706217144049696e-05,
      "loss": 0.5053,
      "num_input_tokens_seen": 14053312,
      "step": 21555
    },
    {
      "epoch": 11.29979035639413,
      "grad_norm": 0.3340291678905487,
      "learning_rate": 2.369479831674647e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 14056256,
      "step": 21560
    },
    {
      "epoch": 11.302410901467505,
      "grad_norm": 0.5503702759742737,
      "learning_rate": 2.3683379762476555e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 14059168,
      "step": 21565
    },
    {
      "epoch": 11.30503144654088,
      "grad_norm": 0.34528908133506775,
      "learning_rate": 2.3671961483628562e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 14062720,
      "step": 21570
    },
    {
      "epoch": 11.307651991614255,
      "grad_norm": 0.4183398187160492,
      "learning_rate": 2.3660543482591078e-05,
      "loss": 0.4124,
      "num_input_tokens_seen": 14065952,
      "step": 21575
    },
    {
      "epoch": 11.31027253668763,
      "grad_norm": 0.6761372089385986,
      "learning_rate": 2.364912576175261e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 14069664,
      "step": 21580
    },
    {
      "epoch": 11.312893081761006,
      "grad_norm": 0.7161133885383606,
      "learning_rate": 2.363770832350163e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 14073024,
      "step": 21585
    },
    {
      "epoch": 11.315513626834381,
      "grad_norm": 0.45691120624542236,
      "learning_rate": 2.362629117022653e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 14076576,
      "step": 21590
    },
    {
      "epoch": 11.318134171907756,
      "grad_norm": 0.6466216444969177,
      "learning_rate": 2.3614874304315633e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 14080128,
      "step": 21595
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 0.27249717712402344,
      "learning_rate": 2.3603457728157237e-05,
      "loss": 0.4609,
      "num_input_tokens_seen": 14084320,
      "step": 21600
    },
    {
      "epoch": 11.323375262054507,
      "grad_norm": 0.33192548155784607,
      "learning_rate": 2.3592041444139547e-05,
      "loss": 0.3583,
      "num_input_tokens_seen": 14087008,
      "step": 21605
    },
    {
      "epoch": 11.325995807127883,
      "grad_norm": 0.3696141839027405,
      "learning_rate": 2.358062545465073e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 14090208,
      "step": 21610
    },
    {
      "epoch": 11.328616352201259,
      "grad_norm": 0.43026483058929443,
      "learning_rate": 2.3569209762078866e-05,
      "loss": 0.3159,
      "num_input_tokens_seen": 14093184,
      "step": 21615
    },
    {
      "epoch": 11.331236897274634,
      "grad_norm": 0.4451253116130829,
      "learning_rate": 2.355779436881201e-05,
      "loss": 0.4609,
      "num_input_tokens_seen": 14100064,
      "step": 21620
    },
    {
      "epoch": 11.333857442348009,
      "grad_norm": 0.35607168078422546,
      "learning_rate": 2.3546379277238107e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 14103104,
      "step": 21625
    },
    {
      "epoch": 11.336477987421384,
      "grad_norm": 0.3335048258304596,
      "learning_rate": 2.353496448974508e-05,
      "loss": 0.3774,
      "num_input_tokens_seen": 14106496,
      "step": 21630
    },
    {
      "epoch": 11.33909853249476,
      "grad_norm": 0.41662338376045227,
      "learning_rate": 2.3523550008720763e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 14110368,
      "step": 21635
    },
    {
      "epoch": 11.341719077568134,
      "grad_norm": 0.2634771764278412,
      "learning_rate": 2.3512135836552945e-05,
      "loss": 0.6267,
      "num_input_tokens_seen": 14114944,
      "step": 21640
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 0.654684841632843,
      "learning_rate": 2.350072197562933e-05,
      "loss": 0.3954,
      "num_input_tokens_seen": 14118272,
      "step": 21645
    },
    {
      "epoch": 11.346960167714885,
      "grad_norm": 0.2761531472206116,
      "learning_rate": 2.3489308428337586e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 14121472,
      "step": 21650
    },
    {
      "epoch": 11.34958071278826,
      "grad_norm": 0.4244985282421112,
      "learning_rate": 2.3477895197065275e-05,
      "loss": 0.366,
      "num_input_tokens_seen": 14123808,
      "step": 21655
    },
    {
      "epoch": 11.352201257861635,
      "grad_norm": 1.2474441528320312,
      "learning_rate": 2.346648228419992e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 14126144,
      "step": 21660
    },
    {
      "epoch": 11.35482180293501,
      "grad_norm": 0.5159876942634583,
      "learning_rate": 2.3455069692128982e-05,
      "loss": 0.4652,
      "num_input_tokens_seen": 14129888,
      "step": 21665
    },
    {
      "epoch": 11.357442348008385,
      "grad_norm": 0.372322142124176,
      "learning_rate": 2.3443657423239832e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 14132448,
      "step": 21670
    },
    {
      "epoch": 11.36006289308176,
      "grad_norm": 0.46511605381965637,
      "learning_rate": 2.34322454799198e-05,
      "loss": 0.3777,
      "num_input_tokens_seen": 14136736,
      "step": 21675
    },
    {
      "epoch": 11.362683438155136,
      "grad_norm": 0.3445449471473694,
      "learning_rate": 2.3420833864556133e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 14139840,
      "step": 21680
    },
    {
      "epoch": 11.365303983228511,
      "grad_norm": 0.4308314323425293,
      "learning_rate": 2.3409422579535995e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 14142528,
      "step": 21685
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 0.5504768490791321,
      "learning_rate": 2.339801162724651e-05,
      "loss": 0.3805,
      "num_input_tokens_seen": 14145984,
      "step": 21690
    },
    {
      "epoch": 11.370545073375261,
      "grad_norm": 0.2648014724254608,
      "learning_rate": 2.338660101007471e-05,
      "loss": 0.5272,
      "num_input_tokens_seen": 14149120,
      "step": 21695
    },
    {
      "epoch": 11.373165618448636,
      "grad_norm": 0.4443986117839813,
      "learning_rate": 2.3375190730407578e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 14151936,
      "step": 21700
    },
    {
      "epoch": 11.375786163522013,
      "grad_norm": 0.5960084199905396,
      "learning_rate": 2.336378079063202e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 14155040,
      "step": 21705
    },
    {
      "epoch": 11.378406708595389,
      "grad_norm": 0.39548173546791077,
      "learning_rate": 2.335237119313484e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 14158336,
      "step": 21710
    },
    {
      "epoch": 11.381027253668764,
      "grad_norm": 1.1222646236419678,
      "learning_rate": 2.3340961940302807e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 14161248,
      "step": 21715
    },
    {
      "epoch": 11.383647798742139,
      "grad_norm": 0.47422662377357483,
      "learning_rate": 2.3329553034522616e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 14164960,
      "step": 21720
    },
    {
      "epoch": 11.386268343815514,
      "grad_norm": 0.48013773560523987,
      "learning_rate": 2.3318144478180866e-05,
      "loss": 0.3899,
      "num_input_tokens_seen": 14167584,
      "step": 21725
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.47684645652770996,
      "learning_rate": 2.330673627366411e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 14169888,
      "step": 21730
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 0.27951064705848694,
      "learning_rate": 2.329532842335881e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 14173024,
      "step": 21735
    },
    {
      "epoch": 11.39412997903564,
      "grad_norm": 0.33746594190597534,
      "learning_rate": 2.328392092965135e-05,
      "loss": 0.3714,
      "num_input_tokens_seen": 14176576,
      "step": 21740
    },
    {
      "epoch": 11.396750524109015,
      "grad_norm": 0.5144140124320984,
      "learning_rate": 2.3272513794928054e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 14180064,
      "step": 21745
    },
    {
      "epoch": 11.39937106918239,
      "grad_norm": 0.7570464015007019,
      "learning_rate": 2.326110702157516e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 14183456,
      "step": 21750
    },
    {
      "epoch": 11.401991614255765,
      "grad_norm": 0.2886662781238556,
      "learning_rate": 2.3249700611978847e-05,
      "loss": 0.3654,
      "num_input_tokens_seen": 14186752,
      "step": 21755
    },
    {
      "epoch": 11.40461215932914,
      "grad_norm": 0.3477708399295807,
      "learning_rate": 2.323829456852519e-05,
      "loss": 0.311,
      "num_input_tokens_seen": 14189280,
      "step": 21760
    },
    {
      "epoch": 11.407232704402515,
      "grad_norm": 0.4298478364944458,
      "learning_rate": 2.3226888893600223e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 14193120,
      "step": 21765
    },
    {
      "epoch": 11.40985324947589,
      "grad_norm": 0.5320766568183899,
      "learning_rate": 2.3215483589589865e-05,
      "loss": 0.5582,
      "num_input_tokens_seen": 14196352,
      "step": 21770
    },
    {
      "epoch": 11.412473794549266,
      "grad_norm": 0.5914115309715271,
      "learning_rate": 2.3204078658879974e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 14199456,
      "step": 21775
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 0.5879579782485962,
      "learning_rate": 2.3192674103856337e-05,
      "loss": 0.4442,
      "num_input_tokens_seen": 14201600,
      "step": 21780
    },
    {
      "epoch": 11.417714884696016,
      "grad_norm": 0.4476831555366516,
      "learning_rate": 2.318126992690467e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 14204672,
      "step": 21785
    },
    {
      "epoch": 11.420335429769391,
      "grad_norm": 0.43657660484313965,
      "learning_rate": 2.3169866130410585e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 14208928,
      "step": 21790
    },
    {
      "epoch": 11.422955974842766,
      "grad_norm": 0.41067951917648315,
      "learning_rate": 2.315846271675962e-05,
      "loss": 0.2919,
      "num_input_tokens_seen": 14213888,
      "step": 21795
    },
    {
      "epoch": 11.425576519916142,
      "grad_norm": 0.9622601270675659,
      "learning_rate": 2.3147059688337244e-05,
      "loss": 0.3455,
      "num_input_tokens_seen": 14216704,
      "step": 21800
    },
    {
      "epoch": 11.428197064989519,
      "grad_norm": 0.7637841701507568,
      "learning_rate": 2.313565704752884e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 14219616,
      "step": 21805
    },
    {
      "epoch": 11.430817610062894,
      "grad_norm": 0.39147382974624634,
      "learning_rate": 2.3124254796719718e-05,
      "loss": 0.3329,
      "num_input_tokens_seen": 14222368,
      "step": 21810
    },
    {
      "epoch": 11.433438155136269,
      "grad_norm": 0.28745073080062866,
      "learning_rate": 2.3112852938295088e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 14229888,
      "step": 21815
    },
    {
      "epoch": 11.436058700209644,
      "grad_norm": 0.4617130756378174,
      "learning_rate": 2.3101451474640104e-05,
      "loss": 0.4376,
      "num_input_tokens_seen": 14233440,
      "step": 21820
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 0.717075526714325,
      "learning_rate": 2.3090050408139808e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 14235488,
      "step": 21825
    },
    {
      "epoch": 11.441299790356394,
      "grad_norm": 0.5066440105438232,
      "learning_rate": 2.307864974117917e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 14240128,
      "step": 21830
    },
    {
      "epoch": 11.44392033542977,
      "grad_norm": 0.3882460296154022,
      "learning_rate": 2.3067249476143087e-05,
      "loss": 0.3327,
      "num_input_tokens_seen": 14243264,
      "step": 21835
    },
    {
      "epoch": 11.446540880503145,
      "grad_norm": 0.2986859977245331,
      "learning_rate": 2.3055849615416363e-05,
      "loss": 0.5319,
      "num_input_tokens_seen": 14247712,
      "step": 21840
    },
    {
      "epoch": 11.44916142557652,
      "grad_norm": 0.36449965834617615,
      "learning_rate": 2.304445016138372e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 14251360,
      "step": 21845
    },
    {
      "epoch": 11.451781970649895,
      "grad_norm": 0.4113316833972931,
      "learning_rate": 2.303305111642981e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 14254432,
      "step": 21850
    },
    {
      "epoch": 11.45440251572327,
      "grad_norm": 0.6398021578788757,
      "learning_rate": 2.3021652482939144e-05,
      "loss": 0.3735,
      "num_input_tokens_seen": 14257664,
      "step": 21855
    },
    {
      "epoch": 11.457023060796645,
      "grad_norm": 0.37867817282676697,
      "learning_rate": 2.3010254263296214e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 14260960,
      "step": 21860
    },
    {
      "epoch": 11.45964360587002,
      "grad_norm": 0.4863424599170685,
      "learning_rate": 2.2998856459885396e-05,
      "loss": 0.5717,
      "num_input_tokens_seen": 14264960,
      "step": 21865
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 0.7899954319000244,
      "learning_rate": 2.298745907509097e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 14267840,
      "step": 21870
    },
    {
      "epoch": 11.464884696016771,
      "grad_norm": 0.3906029164791107,
      "learning_rate": 2.2976062111297156e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 14270368,
      "step": 21875
    },
    {
      "epoch": 11.467505241090146,
      "grad_norm": 0.4204643666744232,
      "learning_rate": 2.296466557088805e-05,
      "loss": 0.3692,
      "num_input_tokens_seen": 14273248,
      "step": 21880
    },
    {
      "epoch": 11.470125786163521,
      "grad_norm": 0.38251936435699463,
      "learning_rate": 2.2953269456247688e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 14281472,
      "step": 21885
    },
    {
      "epoch": 11.472746331236896,
      "grad_norm": 0.7072123289108276,
      "learning_rate": 2.2941873769760006e-05,
      "loss": 0.5331,
      "num_input_tokens_seen": 14284864,
      "step": 21890
    },
    {
      "epoch": 11.475366876310272,
      "grad_norm": 0.3816370964050293,
      "learning_rate": 2.2930478513808844e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 14287968,
      "step": 21895
    },
    {
      "epoch": 11.477987421383649,
      "grad_norm": 0.5000025033950806,
      "learning_rate": 2.2919083690777976e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 14291392,
      "step": 21900
    },
    {
      "epoch": 11.480607966457024,
      "grad_norm": 0.4502526521682739,
      "learning_rate": 2.2907689303051062e-05,
      "loss": 0.4117,
      "num_input_tokens_seen": 14294432,
      "step": 21905
    },
    {
      "epoch": 11.483228511530399,
      "grad_norm": 0.3337111473083496,
      "learning_rate": 2.289629535301167e-05,
      "loss": 0.3478,
      "num_input_tokens_seen": 14299328,
      "step": 21910
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 0.5035531520843506,
      "learning_rate": 2.288490184304329e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 14302272,
      "step": 21915
    },
    {
      "epoch": 11.48846960167715,
      "grad_norm": 0.3241134285926819,
      "learning_rate": 2.2873508775529308e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 14306432,
      "step": 21920
    },
    {
      "epoch": 11.491090146750524,
      "grad_norm": 0.3350403308868408,
      "learning_rate": 2.2862116152853034e-05,
      "loss": 0.5711,
      "num_input_tokens_seen": 14311680,
      "step": 21925
    },
    {
      "epoch": 11.4937106918239,
      "grad_norm": 0.4045839309692383,
      "learning_rate": 2.285072397739767e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 14314208,
      "step": 21930
    },
    {
      "epoch": 11.496331236897275,
      "grad_norm": 0.4974867105484009,
      "learning_rate": 2.283933225154634e-05,
      "loss": 0.6306,
      "num_input_tokens_seen": 14317696,
      "step": 21935
    },
    {
      "epoch": 11.49895178197065,
      "grad_norm": 0.5160799026489258,
      "learning_rate": 2.2827940977682045e-05,
      "loss": 0.5008,
      "num_input_tokens_seen": 14320864,
      "step": 21940
    },
    {
      "epoch": 11.501572327044025,
      "grad_norm": 1.0254653692245483,
      "learning_rate": 2.2816550158187717e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 14323520,
      "step": 21945
    },
    {
      "epoch": 11.5041928721174,
      "grad_norm": 0.31872159242630005,
      "learning_rate": 2.2805159795446182e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 14326400,
      "step": 21950
    },
    {
      "epoch": 11.506813417190775,
      "grad_norm": 0.8033477067947388,
      "learning_rate": 2.2793769891840184e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 14328992,
      "step": 21955
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 0.47216299176216125,
      "learning_rate": 2.278238044975236e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 14331584,
      "step": 21960
    },
    {
      "epoch": 11.512054507337526,
      "grad_norm": 0.446824312210083,
      "learning_rate": 2.2770991471565236e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 14335360,
      "step": 21965
    },
    {
      "epoch": 11.514675052410901,
      "grad_norm": 0.3454779088497162,
      "learning_rate": 2.2759602959661265e-05,
      "loss": 0.5592,
      "num_input_tokens_seen": 14339360,
      "step": 21970
    },
    {
      "epoch": 11.517295597484276,
      "grad_norm": 0.6184045672416687,
      "learning_rate": 2.2748214916422796e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 14342112,
      "step": 21975
    },
    {
      "epoch": 11.519916142557651,
      "grad_norm": 0.3835223913192749,
      "learning_rate": 2.2736827344232082e-05,
      "loss": 0.3662,
      "num_input_tokens_seen": 14344960,
      "step": 21980
    },
    {
      "epoch": 11.522536687631026,
      "grad_norm": 0.5924032330513,
      "learning_rate": 2.272544024547126e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 14348384,
      "step": 21985
    },
    {
      "epoch": 11.525157232704402,
      "grad_norm": 1.174119472503662,
      "learning_rate": 2.27140536225224e-05,
      "loss": 0.6078,
      "num_input_tokens_seen": 14351616,
      "step": 21990
    },
    {
      "epoch": 11.527777777777779,
      "grad_norm": 0.6053532958030701,
      "learning_rate": 2.270266747776744e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 14354848,
      "step": 21995
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 0.8042640089988708,
      "learning_rate": 2.2691281813588226e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 14358496,
      "step": 22000
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 0.5130794644355774,
      "learning_rate": 2.267989663236652e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 14361728,
      "step": 22005
    },
    {
      "epoch": 11.535639412997904,
      "grad_norm": 0.4877740442752838,
      "learning_rate": 2.2668511936483975e-05,
      "loss": 0.3436,
      "num_input_tokens_seen": 14365248,
      "step": 22010
    },
    {
      "epoch": 11.53825995807128,
      "grad_norm": 0.4386318027973175,
      "learning_rate": 2.2657127728322124e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 14368416,
      "step": 22015
    },
    {
      "epoch": 11.540880503144654,
      "grad_norm": 0.41553768515586853,
      "learning_rate": 2.264574401026244e-05,
      "loss": 0.5544,
      "num_input_tokens_seen": 14371424,
      "step": 22020
    },
    {
      "epoch": 11.54350104821803,
      "grad_norm": 0.2991483509540558,
      "learning_rate": 2.2634360784686242e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 14374592,
      "step": 22025
    },
    {
      "epoch": 11.546121593291405,
      "grad_norm": 0.49566975235939026,
      "learning_rate": 2.2622978053974776e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 14377504,
      "step": 22030
    },
    {
      "epoch": 11.54874213836478,
      "grad_norm": 0.3872275948524475,
      "learning_rate": 2.2611595820509186e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 14380960,
      "step": 22035
    },
    {
      "epoch": 11.551362683438155,
      "grad_norm": 0.38895007967948914,
      "learning_rate": 2.2600214086670494e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 14384608,
      "step": 22040
    },
    {
      "epoch": 11.55398322851153,
      "grad_norm": 0.4018469452857971,
      "learning_rate": 2.2588832854839644e-05,
      "loss": 0.3357,
      "num_input_tokens_seen": 14388160,
      "step": 22045
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 0.6879119873046875,
      "learning_rate": 2.2577452127397463e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 14391296,
      "step": 22050
    },
    {
      "epoch": 11.55922431865828,
      "grad_norm": 0.3310818076133728,
      "learning_rate": 2.256607190672464e-05,
      "loss": 0.3,
      "num_input_tokens_seen": 14394752,
      "step": 22055
    },
    {
      "epoch": 11.561844863731656,
      "grad_norm": 0.3961906433105469,
      "learning_rate": 2.2554692195201817e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 14397376,
      "step": 22060
    },
    {
      "epoch": 11.564465408805031,
      "grad_norm": 0.8632062673568726,
      "learning_rate": 2.254331299520948e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 14399872,
      "step": 22065
    },
    {
      "epoch": 11.567085953878406,
      "grad_norm": 0.3424314558506012,
      "learning_rate": 2.253193430912805e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 14403552,
      "step": 22070
    },
    {
      "epoch": 11.569706498951781,
      "grad_norm": 0.5232119560241699,
      "learning_rate": 2.2520556139337803e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 14407552,
      "step": 22075
    },
    {
      "epoch": 11.572327044025156,
      "grad_norm": 0.3564062714576721,
      "learning_rate": 2.250917848821892e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 14410912,
      "step": 22080
    },
    {
      "epoch": 11.574947589098532,
      "grad_norm": 0.5417959094047546,
      "learning_rate": 2.2497801358151472e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 14413440,
      "step": 22085
    },
    {
      "epoch": 11.577568134171909,
      "grad_norm": 0.27282649278640747,
      "learning_rate": 2.248642475151545e-05,
      "loss": 0.3466,
      "num_input_tokens_seen": 14417568,
      "step": 22090
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 0.38479045033454895,
      "learning_rate": 2.2475048670690678e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 14420736,
      "step": 22095
    },
    {
      "epoch": 11.582809224318659,
      "grad_norm": 0.38129422068595886,
      "learning_rate": 2.2463673118056927e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 14424224,
      "step": 22100
    },
    {
      "epoch": 11.585429769392034,
      "grad_norm": 0.36970365047454834,
      "learning_rate": 2.245229809599383e-05,
      "loss": 0.4005,
      "num_input_tokens_seen": 14427264,
      "step": 22105
    },
    {
      "epoch": 11.58805031446541,
      "grad_norm": 0.7440237402915955,
      "learning_rate": 2.2440923606880896e-05,
      "loss": 0.5509,
      "num_input_tokens_seen": 14429504,
      "step": 22110
    },
    {
      "epoch": 11.590670859538784,
      "grad_norm": 0.6431719064712524,
      "learning_rate": 2.242954965309755e-05,
      "loss": 0.506,
      "num_input_tokens_seen": 14432672,
      "step": 22115
    },
    {
      "epoch": 11.59329140461216,
      "grad_norm": 0.9794046878814697,
      "learning_rate": 2.2418176237023088e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 14435072,
      "step": 22120
    },
    {
      "epoch": 11.595911949685535,
      "grad_norm": 0.428591251373291,
      "learning_rate": 2.2406803361036706e-05,
      "loss": 0.3519,
      "num_input_tokens_seen": 14439424,
      "step": 22125
    },
    {
      "epoch": 11.59853249475891,
      "grad_norm": 0.5604376792907715,
      "learning_rate": 2.2395431027517475e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 14441888,
      "step": 22130
    },
    {
      "epoch": 11.601153039832285,
      "grad_norm": 1.104600429534912,
      "learning_rate": 2.2384059238844364e-05,
      "loss": 0.4897,
      "num_input_tokens_seen": 14444896,
      "step": 22135
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 0.3339206576347351,
      "learning_rate": 2.2372687997396215e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 14447904,
      "step": 22140
    },
    {
      "epoch": 11.606394129979035,
      "grad_norm": 0.6888051629066467,
      "learning_rate": 2.2361317305551756e-05,
      "loss": 0.4715,
      "num_input_tokens_seen": 14450624,
      "step": 22145
    },
    {
      "epoch": 11.60901467505241,
      "grad_norm": 0.36484575271606445,
      "learning_rate": 2.2349947165689615e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 14454240,
      "step": 22150
    },
    {
      "epoch": 11.611635220125786,
      "grad_norm": 0.6322431564331055,
      "learning_rate": 2.2338577580188296e-05,
      "loss": 0.3864,
      "num_input_tokens_seen": 14457088,
      "step": 22155
    },
    {
      "epoch": 11.614255765199161,
      "grad_norm": 0.621921956539154,
      "learning_rate": 2.2327208551426188e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 14460608,
      "step": 22160
    },
    {
      "epoch": 11.616876310272536,
      "grad_norm": 0.34360867738723755,
      "learning_rate": 2.2315840081781546e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 14464064,
      "step": 22165
    },
    {
      "epoch": 11.619496855345911,
      "grad_norm": 0.4747287333011627,
      "learning_rate": 2.230447217363255e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 14467584,
      "step": 22170
    },
    {
      "epoch": 11.622117400419286,
      "grad_norm": 0.5309551954269409,
      "learning_rate": 2.2293104829357212e-05,
      "loss": 0.4048,
      "num_input_tokens_seen": 14471328,
      "step": 22175
    },
    {
      "epoch": 11.624737945492662,
      "grad_norm": 0.26318246126174927,
      "learning_rate": 2.2281738051333467e-05,
      "loss": 0.295,
      "num_input_tokens_seen": 14475072,
      "step": 22180
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 0.6885122060775757,
      "learning_rate": 2.2270371841939105e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 14477984,
      "step": 22185
    },
    {
      "epoch": 11.629979035639414,
      "grad_norm": 0.21921099722385406,
      "learning_rate": 2.2259006203551824e-05,
      "loss": 0.2528,
      "num_input_tokens_seen": 14481984,
      "step": 22190
    },
    {
      "epoch": 11.632599580712789,
      "grad_norm": 0.4872589707374573,
      "learning_rate": 2.2247641138549168e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 14485760,
      "step": 22195
    },
    {
      "epoch": 11.635220125786164,
      "grad_norm": 0.40825319290161133,
      "learning_rate": 2.2236276649308583e-05,
      "loss": 0.4931,
      "num_input_tokens_seen": 14489664,
      "step": 22200
    },
    {
      "epoch": 11.63784067085954,
      "grad_norm": 0.3702206313610077,
      "learning_rate": 2.2224912738207395e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 14493056,
      "step": 22205
    },
    {
      "epoch": 11.640461215932914,
      "grad_norm": 0.5762009620666504,
      "learning_rate": 2.2213549407622796e-05,
      "loss": 0.3283,
      "num_input_tokens_seen": 14495680,
      "step": 22210
    },
    {
      "epoch": 11.64308176100629,
      "grad_norm": 0.5286040306091309,
      "learning_rate": 2.220218665993188e-05,
      "loss": 0.4019,
      "num_input_tokens_seen": 14498592,
      "step": 22215
    },
    {
      "epoch": 11.645702306079665,
      "grad_norm": 0.36643680930137634,
      "learning_rate": 2.21908244975116e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 14501888,
      "step": 22220
    },
    {
      "epoch": 11.64832285115304,
      "grad_norm": 0.3722730278968811,
      "learning_rate": 2.2179462922738776e-05,
      "loss": 0.4948,
      "num_input_tokens_seen": 14505088,
      "step": 22225
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 0.35367706418037415,
      "learning_rate": 2.2168101937990135e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 14507744,
      "step": 22230
    },
    {
      "epoch": 11.65356394129979,
      "grad_norm": 0.578855574131012,
      "learning_rate": 2.2156741545642253e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 14510560,
      "step": 22235
    },
    {
      "epoch": 11.656184486373165,
      "grad_norm": 0.3299800455570221,
      "learning_rate": 2.2145381748071606e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 14514432,
      "step": 22240
    },
    {
      "epoch": 11.65880503144654,
      "grad_norm": 0.5658776760101318,
      "learning_rate": 2.2134022547654538e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 14517664,
      "step": 22245
    },
    {
      "epoch": 11.661425576519916,
      "grad_norm": 0.41426679491996765,
      "learning_rate": 2.212266394676725e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 14520704,
      "step": 22250
    },
    {
      "epoch": 11.664046121593291,
      "grad_norm": 0.42603185772895813,
      "learning_rate": 2.2111305947785836e-05,
      "loss": 0.3646,
      "num_input_tokens_seen": 14525152,
      "step": 22255
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.3125109076499939,
      "learning_rate": 2.2099948553086263e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 14529824,
      "step": 22260
    },
    {
      "epoch": 11.669287211740041,
      "grad_norm": 0.5038139224052429,
      "learning_rate": 2.2088591765044365e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 14532416,
      "step": 22265
    },
    {
      "epoch": 11.671907756813416,
      "grad_norm": 0.4285168945789337,
      "learning_rate": 2.207723558603586e-05,
      "loss": 0.5222,
      "num_input_tokens_seen": 14535776,
      "step": 22270
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 0.7013987898826599,
      "learning_rate": 2.206588001843633e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 14538624,
      "step": 22275
    },
    {
      "epoch": 11.677148846960169,
      "grad_norm": 0.5593305230140686,
      "learning_rate": 2.2054525064621216e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 14542112,
      "step": 22280
    },
    {
      "epoch": 11.679769392033544,
      "grad_norm": 0.7367529273033142,
      "learning_rate": 2.2043170726965858e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 14545024,
      "step": 22285
    },
    {
      "epoch": 11.682389937106919,
      "grad_norm": 0.29029449820518494,
      "learning_rate": 2.2031817007845447e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 14549440,
      "step": 22290
    },
    {
      "epoch": 11.685010482180294,
      "grad_norm": 0.5126214623451233,
      "learning_rate": 2.202046390963506e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 14552224,
      "step": 22295
    },
    {
      "epoch": 11.68763102725367,
      "grad_norm": 0.5529841184616089,
      "learning_rate": 2.2009111434709627e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 14555040,
      "step": 22300
    },
    {
      "epoch": 11.690251572327044,
      "grad_norm": 0.47349095344543457,
      "learning_rate": 2.1997759585443968e-05,
      "loss": 0.3963,
      "num_input_tokens_seen": 14558016,
      "step": 22305
    },
    {
      "epoch": 11.69287211740042,
      "grad_norm": 0.29850995540618896,
      "learning_rate": 2.198640836421275e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 14560640,
      "step": 22310
    },
    {
      "epoch": 11.695492662473795,
      "grad_norm": 0.556698203086853,
      "learning_rate": 2.1975057773390515e-05,
      "loss": 0.3538,
      "num_input_tokens_seen": 14563264,
      "step": 22315
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 0.7581281661987305,
      "learning_rate": 2.1963707815351683e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 14566432,
      "step": 22320
    },
    {
      "epoch": 11.700733752620545,
      "grad_norm": 0.5294463634490967,
      "learning_rate": 2.1952358492470544e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 14570176,
      "step": 22325
    },
    {
      "epoch": 11.70335429769392,
      "grad_norm": 0.2879001796245575,
      "learning_rate": 2.194100980712124e-05,
      "loss": 0.5241,
      "num_input_tokens_seen": 14573536,
      "step": 22330
    },
    {
      "epoch": 11.705974842767295,
      "grad_norm": 0.4485529661178589,
      "learning_rate": 2.1929661761677774e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 14576672,
      "step": 22335
    },
    {
      "epoch": 11.70859538784067,
      "grad_norm": 0.29722440242767334,
      "learning_rate": 2.191831435851405e-05,
      "loss": 0.408,
      "num_input_tokens_seen": 14579584,
      "step": 22340
    },
    {
      "epoch": 11.711215932914046,
      "grad_norm": 0.5306529402732849,
      "learning_rate": 2.1906967600003793e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 14582528,
      "step": 22345
    },
    {
      "epoch": 11.713836477987421,
      "grad_norm": 0.8421977758407593,
      "learning_rate": 2.1895621488520636e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 14586464,
      "step": 22350
    },
    {
      "epoch": 11.716457023060796,
      "grad_norm": 0.34419694542884827,
      "learning_rate": 2.1884276026438044e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 14589440,
      "step": 22355
    },
    {
      "epoch": 11.719077568134171,
      "grad_norm": 0.4780057668685913,
      "learning_rate": 2.187293121612937e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 14593280,
      "step": 22360
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 0.48958268761634827,
      "learning_rate": 2.1861587059967804e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 14596800,
      "step": 22365
    },
    {
      "epoch": 11.724318658280922,
      "grad_norm": 0.490355908870697,
      "learning_rate": 2.1850243560326414e-05,
      "loss": 0.5842,
      "num_input_tokens_seen": 14599552,
      "step": 22370
    },
    {
      "epoch": 11.726939203354299,
      "grad_norm": 1.2294343709945679,
      "learning_rate": 2.1838900719578143e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 14601952,
      "step": 22375
    },
    {
      "epoch": 11.729559748427674,
      "grad_norm": 0.43734878301620483,
      "learning_rate": 2.1827558540095772e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 14606336,
      "step": 22380
    },
    {
      "epoch": 11.732180293501049,
      "grad_norm": 0.9417218565940857,
      "learning_rate": 2.1816217024251966e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 14608864,
      "step": 22385
    },
    {
      "epoch": 11.734800838574424,
      "grad_norm": 0.3217083513736725,
      "learning_rate": 2.1804876174419242e-05,
      "loss": 0.3233,
      "num_input_tokens_seen": 14611488,
      "step": 22390
    },
    {
      "epoch": 11.7374213836478,
      "grad_norm": 0.5592781901359558,
      "learning_rate": 2.179353599296996e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 14615296,
      "step": 22395
    },
    {
      "epoch": 11.740041928721174,
      "grad_norm": 0.5385827422142029,
      "learning_rate": 2.1782196482276368e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 14620096,
      "step": 22400
    },
    {
      "epoch": 11.74266247379455,
      "grad_norm": 0.49443477392196655,
      "learning_rate": 2.1770857644710567e-05,
      "loss": 0.6546,
      "num_input_tokens_seen": 14622912,
      "step": 22405
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 0.39201804995536804,
      "learning_rate": 2.1759519482644495e-05,
      "loss": 0.5387,
      "num_input_tokens_seen": 14625568,
      "step": 22410
    },
    {
      "epoch": 11.7479035639413,
      "grad_norm": 0.5327494740486145,
      "learning_rate": 2.174818199844999e-05,
      "loss": 0.521,
      "num_input_tokens_seen": 14628576,
      "step": 22415
    },
    {
      "epoch": 11.750524109014675,
      "grad_norm": 0.49351903796195984,
      "learning_rate": 2.173684519449872e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 14632096,
      "step": 22420
    },
    {
      "epoch": 11.75314465408805,
      "grad_norm": 0.9033586978912354,
      "learning_rate": 2.1725509073162188e-05,
      "loss": 0.2762,
      "num_input_tokens_seen": 14636192,
      "step": 22425
    },
    {
      "epoch": 11.755765199161425,
      "grad_norm": 1.0010815858840942,
      "learning_rate": 2.171417363681181e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 14638944,
      "step": 22430
    },
    {
      "epoch": 11.7583857442348,
      "grad_norm": 0.974818229675293,
      "learning_rate": 2.1702838887818816e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 14641088,
      "step": 22435
    },
    {
      "epoch": 11.761006289308176,
      "grad_norm": 0.5663519501686096,
      "learning_rate": 2.1691504828554315e-05,
      "loss": 0.3315,
      "num_input_tokens_seen": 14644992,
      "step": 22440
    },
    {
      "epoch": 11.76362683438155,
      "grad_norm": 0.745776891708374,
      "learning_rate": 2.1680171461389265e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 14648000,
      "step": 22445
    },
    {
      "epoch": 11.766247379454926,
      "grad_norm": 0.5108440518379211,
      "learning_rate": 2.166883878869446e-05,
      "loss": 0.5902,
      "num_input_tokens_seen": 14651168,
      "step": 22450
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 0.3722027540206909,
      "learning_rate": 2.1657506812840582e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 14654528,
      "step": 22455
    },
    {
      "epoch": 11.771488469601676,
      "grad_norm": 0.40117162466049194,
      "learning_rate": 2.164617553619814e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 14657088,
      "step": 22460
    },
    {
      "epoch": 11.774109014675052,
      "grad_norm": 0.3606220781803131,
      "learning_rate": 2.1634844961137517e-05,
      "loss": 0.408,
      "num_input_tokens_seen": 14660352,
      "step": 22465
    },
    {
      "epoch": 11.776729559748428,
      "grad_norm": 0.3540193438529968,
      "learning_rate": 2.162351509002893e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 14664288,
      "step": 22470
    },
    {
      "epoch": 11.779350104821804,
      "grad_norm": 0.46384721994400024,
      "learning_rate": 2.161218592524248e-05,
      "loss": 0.6901,
      "num_input_tokens_seen": 14668000,
      "step": 22475
    },
    {
      "epoch": 11.781970649895179,
      "grad_norm": 0.4400887191295624,
      "learning_rate": 2.160085746914807e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 14671328,
      "step": 22480
    },
    {
      "epoch": 11.784591194968554,
      "grad_norm": 0.6530197262763977,
      "learning_rate": 2.1589529724115503e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 14674272,
      "step": 22485
    },
    {
      "epoch": 11.78721174004193,
      "grad_norm": 0.5666998624801636,
      "learning_rate": 2.1578202692514407e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 14676896,
      "step": 22490
    },
    {
      "epoch": 11.789832285115304,
      "grad_norm": 1.8677726984024048,
      "learning_rate": 2.1566876376714274e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 14679840,
      "step": 22495
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 0.3114660382270813,
      "learning_rate": 2.155555077908443e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 14683456,
      "step": 22500
    },
    {
      "epoch": 11.795073375262055,
      "grad_norm": 0.522325336933136,
      "learning_rate": 2.1544225901994077e-05,
      "loss": 0.5514,
      "num_input_tokens_seen": 14686112,
      "step": 22505
    },
    {
      "epoch": 11.79769392033543,
      "grad_norm": 0.48225340247154236,
      "learning_rate": 2.1532901747812234e-05,
      "loss": 0.3166,
      "num_input_tokens_seen": 14689024,
      "step": 22510
    },
    {
      "epoch": 11.800314465408805,
      "grad_norm": 0.4868148863315582,
      "learning_rate": 2.1521578318907787e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 14691712,
      "step": 22515
    },
    {
      "epoch": 11.80293501048218,
      "grad_norm": 0.6215001940727234,
      "learning_rate": 2.1510255617649476e-05,
      "loss": 0.5845,
      "num_input_tokens_seen": 14695104,
      "step": 22520
    },
    {
      "epoch": 11.805555555555555,
      "grad_norm": 0.6537094712257385,
      "learning_rate": 2.149893364640587e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 14697952,
      "step": 22525
    },
    {
      "epoch": 11.80817610062893,
      "grad_norm": 0.46920642256736755,
      "learning_rate": 2.1487612407545417e-05,
      "loss": 0.3878,
      "num_input_tokens_seen": 14700704,
      "step": 22530
    },
    {
      "epoch": 11.810796645702306,
      "grad_norm": 0.4756525456905365,
      "learning_rate": 2.147629190343637e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 14703488,
      "step": 22535
    },
    {
      "epoch": 11.81341719077568,
      "grad_norm": 0.6517846584320068,
      "learning_rate": 2.1464972136446852e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 14706752,
      "step": 22540
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 0.5962201356887817,
      "learning_rate": 2.1453653108944838e-05,
      "loss": 0.3513,
      "num_input_tokens_seen": 14710176,
      "step": 22545
    },
    {
      "epoch": 11.818658280922431,
      "grad_norm": 0.4635109603404999,
      "learning_rate": 2.1442334823298128e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 14713376,
      "step": 22550
    },
    {
      "epoch": 11.821278825995806,
      "grad_norm": 0.42206352949142456,
      "learning_rate": 2.1431017281874386e-05,
      "loss": 0.4678,
      "num_input_tokens_seen": 14716864,
      "step": 22555
    },
    {
      "epoch": 11.823899371069182,
      "grad_norm": 0.5299577713012695,
      "learning_rate": 2.1419700487041123e-05,
      "loss": 0.2997,
      "num_input_tokens_seen": 14719776,
      "step": 22560
    },
    {
      "epoch": 11.826519916142558,
      "grad_norm": 0.26725855469703674,
      "learning_rate": 2.140838444116566e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 14724512,
      "step": 22565
    },
    {
      "epoch": 11.829140461215934,
      "grad_norm": 0.3974665701389313,
      "learning_rate": 2.13970691466152e-05,
      "loss": 0.4496,
      "num_input_tokens_seen": 14727168,
      "step": 22570
    },
    {
      "epoch": 11.831761006289309,
      "grad_norm": 0.4070724546909332,
      "learning_rate": 2.1385754605756773e-05,
      "loss": 0.3123,
      "num_input_tokens_seen": 14730112,
      "step": 22575
    },
    {
      "epoch": 11.834381551362684,
      "grad_norm": 0.7875333428382874,
      "learning_rate": 2.137444082095724e-05,
      "loss": 0.4394,
      "num_input_tokens_seen": 14733312,
      "step": 22580
    },
    {
      "epoch": 11.83700209643606,
      "grad_norm": 0.41175732016563416,
      "learning_rate": 2.1363127794583335e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 14737440,
      "step": 22585
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 0.5229924321174622,
      "learning_rate": 2.1351815529001605e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 14740672,
      "step": 22590
    },
    {
      "epoch": 11.84224318658281,
      "grad_norm": 0.5164172649383545,
      "learning_rate": 2.1340504026578438e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 14743552,
      "step": 22595
    },
    {
      "epoch": 11.844863731656185,
      "grad_norm": 0.3879239857196808,
      "learning_rate": 2.1329193289680084e-05,
      "loss": 0.4597,
      "num_input_tokens_seen": 14746816,
      "step": 22600
    },
    {
      "epoch": 11.84748427672956,
      "grad_norm": 0.5245533585548401,
      "learning_rate": 2.1317883320672608e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 14749696,
      "step": 22605
    },
    {
      "epoch": 11.850104821802935,
      "grad_norm": 0.31681710481643677,
      "learning_rate": 2.1306574121921942e-05,
      "loss": 0.6238,
      "num_input_tokens_seen": 14753376,
      "step": 22610
    },
    {
      "epoch": 11.85272536687631,
      "grad_norm": 0.5344341397285461,
      "learning_rate": 2.1295265695793836e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 14756576,
      "step": 22615
    },
    {
      "epoch": 11.855345911949685,
      "grad_norm": 0.40262371301651,
      "learning_rate": 2.128395804465387e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 14759552,
      "step": 22620
    },
    {
      "epoch": 11.85796645702306,
      "grad_norm": 0.49007412791252136,
      "learning_rate": 2.1272651170867487e-05,
      "loss": 0.4831,
      "num_input_tokens_seen": 14763424,
      "step": 22625
    },
    {
      "epoch": 11.860587002096436,
      "grad_norm": 0.7858318090438843,
      "learning_rate": 2.1261345076799955e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 14767392,
      "step": 22630
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 0.5469961762428284,
      "learning_rate": 2.125003976481638e-05,
      "loss": 0.572,
      "num_input_tokens_seen": 14770272,
      "step": 22635
    },
    {
      "epoch": 11.865828092243186,
      "grad_norm": 0.582789421081543,
      "learning_rate": 2.123873523728171e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 14773664,
      "step": 22640
    },
    {
      "epoch": 11.868448637316561,
      "grad_norm": 0.8151178956031799,
      "learning_rate": 2.122743149656072e-05,
      "loss": 0.3545,
      "num_input_tokens_seen": 14776640,
      "step": 22645
    },
    {
      "epoch": 11.871069182389936,
      "grad_norm": 0.3805333375930786,
      "learning_rate": 2.1216128545018012e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 14779584,
      "step": 22650
    },
    {
      "epoch": 11.873689727463312,
      "grad_norm": 0.5996405482292175,
      "learning_rate": 2.1204826385018052e-05,
      "loss": 0.556,
      "num_input_tokens_seen": 14782784,
      "step": 22655
    },
    {
      "epoch": 11.876310272536688,
      "grad_norm": 0.4477546513080597,
      "learning_rate": 2.1193525018925106e-05,
      "loss": 0.6092,
      "num_input_tokens_seen": 14787040,
      "step": 22660
    },
    {
      "epoch": 11.878930817610064,
      "grad_norm": 0.62208092212677,
      "learning_rate": 2.1182224449103304e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 14790560,
      "step": 22665
    },
    {
      "epoch": 11.881551362683439,
      "grad_norm": 0.5179007649421692,
      "learning_rate": 2.117092467791659e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 14794976,
      "step": 22670
    },
    {
      "epoch": 11.884171907756814,
      "grad_norm": 0.3080943524837494,
      "learning_rate": 2.115962570772876e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 14798368,
      "step": 22675
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 0.3182421922683716,
      "learning_rate": 2.1148327540903408e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 14801536,
      "step": 22680
    },
    {
      "epoch": 11.889412997903564,
      "grad_norm": 0.5338253974914551,
      "learning_rate": 2.113703017980399e-05,
      "loss": 0.3062,
      "num_input_tokens_seen": 14804544,
      "step": 22685
    },
    {
      "epoch": 11.89203354297694,
      "grad_norm": 0.4419715702533722,
      "learning_rate": 2.1125733626793793e-05,
      "loss": 0.3352,
      "num_input_tokens_seen": 14807328,
      "step": 22690
    },
    {
      "epoch": 11.894654088050315,
      "grad_norm": 0.4662877321243286,
      "learning_rate": 2.1114437884235915e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 14810464,
      "step": 22695
    },
    {
      "epoch": 11.89727463312369,
      "grad_norm": 0.629718542098999,
      "learning_rate": 2.1103142954493313e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 14814080,
      "step": 22700
    },
    {
      "epoch": 11.899895178197065,
      "grad_norm": 0.5776451826095581,
      "learning_rate": 2.1091848839928745e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 14818368,
      "step": 22705
    },
    {
      "epoch": 11.90251572327044,
      "grad_norm": 0.2537129521369934,
      "learning_rate": 2.1080555542904805e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 14821344,
      "step": 22710
    },
    {
      "epoch": 11.905136268343815,
      "grad_norm": 0.5049937963485718,
      "learning_rate": 2.106926306578393e-05,
      "loss": 0.3535,
      "num_input_tokens_seen": 14824352,
      "step": 22715
    },
    {
      "epoch": 11.90775681341719,
      "grad_norm": 0.48736581206321716,
      "learning_rate": 2.1057971410928383e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 14828704,
      "step": 22720
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 1.0331522226333618,
      "learning_rate": 2.104668058070024e-05,
      "loss": 0.4868,
      "num_input_tokens_seen": 14831456,
      "step": 22725
    },
    {
      "epoch": 11.91299790356394,
      "grad_norm": 0.48104578256607056,
      "learning_rate": 2.1035390577461427e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 14833856,
      "step": 22730
    },
    {
      "epoch": 11.915618448637316,
      "grad_norm": 0.484369695186615,
      "learning_rate": 2.102410140357367e-05,
      "loss": 0.3097,
      "num_input_tokens_seen": 14837120,
      "step": 22735
    },
    {
      "epoch": 11.918238993710691,
      "grad_norm": 0.5147477984428406,
      "learning_rate": 2.1012813061398534e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 14841056,
      "step": 22740
    },
    {
      "epoch": 11.920859538784066,
      "grad_norm": 0.38110968470573425,
      "learning_rate": 2.1001525553297422e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 14844064,
      "step": 22745
    },
    {
      "epoch": 11.923480083857442,
      "grad_norm": 0.6071344017982483,
      "learning_rate": 2.099023888163154e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 14846656,
      "step": 22750
    },
    {
      "epoch": 11.926100628930818,
      "grad_norm": 0.43175020813941956,
      "learning_rate": 2.0978953048761947e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 14849248,
      "step": 22755
    },
    {
      "epoch": 11.928721174004194,
      "grad_norm": 0.41146764159202576,
      "learning_rate": 2.0967668057049507e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 14852384,
      "step": 22760
    },
    {
      "epoch": 11.931341719077569,
      "grad_norm": 0.5764368176460266,
      "learning_rate": 2.0956383908854895e-05,
      "loss": 0.3836,
      "num_input_tokens_seen": 14854720,
      "step": 22765
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 0.41826292872428894,
      "learning_rate": 2.0945100606538644e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 14857696,
      "step": 22770
    },
    {
      "epoch": 11.93658280922432,
      "grad_norm": 0.6215028762817383,
      "learning_rate": 2.0933818152461077e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 14861216,
      "step": 22775
    },
    {
      "epoch": 11.939203354297694,
      "grad_norm": 0.3320992588996887,
      "learning_rate": 2.092253654898237e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 14864064,
      "step": 22780
    },
    {
      "epoch": 11.94182389937107,
      "grad_norm": 0.7288159728050232,
      "learning_rate": 2.0911255798462494e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 14866784,
      "step": 22785
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.3829754889011383,
      "learning_rate": 2.0899975903261266e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 14869664,
      "step": 22790
    },
    {
      "epoch": 11.94706498951782,
      "grad_norm": 0.43760910630226135,
      "learning_rate": 2.0888696865738295e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 14872640,
      "step": 22795
    },
    {
      "epoch": 11.949685534591195,
      "grad_norm": 0.6729031205177307,
      "learning_rate": 2.087741868825304e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 14876512,
      "step": 22800
    },
    {
      "epoch": 11.95230607966457,
      "grad_norm": 0.6325030326843262,
      "learning_rate": 2.0866141373164758e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 14879776,
      "step": 22805
    },
    {
      "epoch": 11.954926624737945,
      "grad_norm": 0.5141189694404602,
      "learning_rate": 2.0854864922832544e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 14883552,
      "step": 22810
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 1.4828814268112183,
      "learning_rate": 2.0843589339615304e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 14885984,
      "step": 22815
    },
    {
      "epoch": 11.960167714884696,
      "grad_norm": 0.3623350262641907,
      "learning_rate": 2.0832314625871742e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 14888992,
      "step": 22820
    },
    {
      "epoch": 11.96278825995807,
      "grad_norm": 0.4269334077835083,
      "learning_rate": 2.0821040783960423e-05,
      "loss": 0.4047,
      "num_input_tokens_seen": 14891744,
      "step": 22825
    },
    {
      "epoch": 11.965408805031446,
      "grad_norm": 0.38268500566482544,
      "learning_rate": 2.0809767816239686e-05,
      "loss": 0.6202,
      "num_input_tokens_seen": 14894880,
      "step": 22830
    },
    {
      "epoch": 11.968029350104821,
      "grad_norm": 0.347921758890152,
      "learning_rate": 2.0798495725067725e-05,
      "loss": 0.4421,
      "num_input_tokens_seen": 14897760,
      "step": 22835
    },
    {
      "epoch": 11.970649895178196,
      "grad_norm": 0.47671979665756226,
      "learning_rate": 2.078722451280252e-05,
      "loss": 0.4496,
      "num_input_tokens_seen": 14900736,
      "step": 22840
    },
    {
      "epoch": 11.973270440251572,
      "grad_norm": 0.46299663186073303,
      "learning_rate": 2.0775954181801893e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 14904544,
      "step": 22845
    },
    {
      "epoch": 11.975890985324948,
      "grad_norm": 0.3729782998561859,
      "learning_rate": 2.0764684734423455e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 14907648,
      "step": 22850
    },
    {
      "epoch": 11.978511530398324,
      "grad_norm": 0.4058825671672821,
      "learning_rate": 2.075341617302465e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 14910816,
      "step": 22855
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 0.4586960971355438,
      "learning_rate": 2.074214849996274e-05,
      "loss": 0.3654,
      "num_input_tokens_seen": 14913632,
      "step": 22860
    },
    {
      "epoch": 11.983752620545074,
      "grad_norm": 0.358315110206604,
      "learning_rate": 2.0730881717594776e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 14916768,
      "step": 22865
    },
    {
      "epoch": 11.98637316561845,
      "grad_norm": 0.4148043990135193,
      "learning_rate": 2.0719615828277657e-05,
      "loss": 0.3655,
      "num_input_tokens_seen": 14920896,
      "step": 22870
    },
    {
      "epoch": 11.988993710691824,
      "grad_norm": 0.3498198390007019,
      "learning_rate": 2.0708350834368083e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 14924192,
      "step": 22875
    },
    {
      "epoch": 11.9916142557652,
      "grad_norm": 0.31246259808540344,
      "learning_rate": 2.069708673822255e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 14926784,
      "step": 22880
    },
    {
      "epoch": 11.994234800838575,
      "grad_norm": 0.4922376275062561,
      "learning_rate": 2.068582354219737e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 14930432,
      "step": 22885
    },
    {
      "epoch": 11.99685534591195,
      "grad_norm": 0.6042543053627014,
      "learning_rate": 2.0674561248648697e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 14933664,
      "step": 22890
    },
    {
      "epoch": 11.999475890985325,
      "grad_norm": 0.3623757064342499,
      "learning_rate": 2.0663299859932456e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 14937088,
      "step": 22895
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.4534948766231537,
      "eval_runtime": 9.7785,
      "eval_samples_per_second": 86.721,
      "eval_steps_per_second": 21.68,
      "num_input_tokens_seen": 14937216,
      "step": 22896
    },
    {
      "epoch": 12.0020964360587,
      "grad_norm": 0.5790804028511047,
      "learning_rate": 2.065203937840442e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 14939712,
      "step": 22900
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 0.5454241037368774,
      "learning_rate": 2.0640779806420136e-05,
      "loss": 0.4739,
      "num_input_tokens_seen": 14942272,
      "step": 22905
    },
    {
      "epoch": 12.00733752620545,
      "grad_norm": 0.5872121453285217,
      "learning_rate": 2.0629521146334978e-05,
      "loss": 0.5176,
      "num_input_tokens_seen": 14945632,
      "step": 22910
    },
    {
      "epoch": 12.009958071278826,
      "grad_norm": 0.3761124312877655,
      "learning_rate": 2.061826340050414e-05,
      "loss": 0.6571,
      "num_input_tokens_seen": 14949376,
      "step": 22915
    },
    {
      "epoch": 12.0125786163522,
      "grad_norm": 0.3955667018890381,
      "learning_rate": 2.0607006571282605e-05,
      "loss": 0.5045,
      "num_input_tokens_seen": 14951840,
      "step": 22920
    },
    {
      "epoch": 12.015199161425576,
      "grad_norm": 0.3694641590118408,
      "learning_rate": 2.0595750661025183e-05,
      "loss": 0.3303,
      "num_input_tokens_seen": 14956768,
      "step": 22925
    },
    {
      "epoch": 12.017819706498951,
      "grad_norm": 1.1564958095550537,
      "learning_rate": 2.0584495672086483e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 14960384,
      "step": 22930
    },
    {
      "epoch": 12.020440251572326,
      "grad_norm": 0.3881152868270874,
      "learning_rate": 2.05732416068209e-05,
      "loss": 0.359,
      "num_input_tokens_seen": 14963392,
      "step": 22935
    },
    {
      "epoch": 12.023060796645701,
      "grad_norm": 0.7326445579528809,
      "learning_rate": 2.0561988467582676e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 14967392,
      "step": 22940
    },
    {
      "epoch": 12.025681341719078,
      "grad_norm": 0.30050423741340637,
      "learning_rate": 2.0550736256725828e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 14972704,
      "step": 22945
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 0.5648408532142639,
      "learning_rate": 2.0539484976604196e-05,
      "loss": 0.3877,
      "num_input_tokens_seen": 14975744,
      "step": 22950
    },
    {
      "epoch": 12.030922431865829,
      "grad_norm": 0.2891768515110016,
      "learning_rate": 2.052823462957142e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 14979104,
      "step": 22955
    },
    {
      "epoch": 12.033542976939204,
      "grad_norm": 0.4357060492038727,
      "learning_rate": 2.0516985217980953e-05,
      "loss": 0.5399,
      "num_input_tokens_seen": 14981600,
      "step": 22960
    },
    {
      "epoch": 12.036163522012579,
      "grad_norm": 0.6139094829559326,
      "learning_rate": 2.050573674418602e-05,
      "loss": 0.4339,
      "num_input_tokens_seen": 14985152,
      "step": 22965
    },
    {
      "epoch": 12.038784067085954,
      "grad_norm": 0.4020206034183502,
      "learning_rate": 2.0494489210539693e-05,
      "loss": 0.3233,
      "num_input_tokens_seen": 14989568,
      "step": 22970
    },
    {
      "epoch": 12.04140461215933,
      "grad_norm": 0.471404492855072,
      "learning_rate": 2.0483242619394817e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 14992352,
      "step": 22975
    },
    {
      "epoch": 12.044025157232705,
      "grad_norm": 0.30630967020988464,
      "learning_rate": 2.0471996973104063e-05,
      "loss": 0.3549,
      "num_input_tokens_seen": 14995168,
      "step": 22980
    },
    {
      "epoch": 12.04664570230608,
      "grad_norm": 0.5612263083457947,
      "learning_rate": 2.0460752274019886e-05,
      "loss": 0.6099,
      "num_input_tokens_seen": 14998240,
      "step": 22985
    },
    {
      "epoch": 12.049266247379455,
      "grad_norm": 0.7006593346595764,
      "learning_rate": 2.0449508524494537e-05,
      "loss": 0.5013,
      "num_input_tokens_seen": 15001248,
      "step": 22990
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 0.38211575150489807,
      "learning_rate": 2.0438265726880092e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 15004352,
      "step": 22995
    },
    {
      "epoch": 12.054507337526205,
      "grad_norm": 0.9327338337898254,
      "learning_rate": 2.042702388352841e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 15007968,
      "step": 23000
    },
    {
      "epoch": 12.05712788259958,
      "grad_norm": 0.4157036542892456,
      "learning_rate": 2.041578299679117e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 15011648,
      "step": 23005
    },
    {
      "epoch": 12.059748427672956,
      "grad_norm": 0.508234441280365,
      "learning_rate": 2.0404543069019818e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 15014176,
      "step": 23010
    },
    {
      "epoch": 12.06236897274633,
      "grad_norm": 0.453438937664032,
      "learning_rate": 2.0393304102565642e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 15017184,
      "step": 23015
    },
    {
      "epoch": 12.064989517819706,
      "grad_norm": 0.8019669651985168,
      "learning_rate": 2.0382066099779685e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 15020416,
      "step": 23020
    },
    {
      "epoch": 12.067610062893081,
      "grad_norm": 0.4751894176006317,
      "learning_rate": 2.0370829063012808e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 15024800,
      "step": 23025
    },
    {
      "epoch": 12.070230607966456,
      "grad_norm": 0.9110729694366455,
      "learning_rate": 2.035959299461568e-05,
      "loss": 0.3606,
      "num_input_tokens_seen": 15027456,
      "step": 23030
    },
    {
      "epoch": 12.072851153039831,
      "grad_norm": 0.3710189461708069,
      "learning_rate": 2.0348357896938762e-05,
      "loss": 0.5676,
      "num_input_tokens_seen": 15030208,
      "step": 23035
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 0.5462782382965088,
      "learning_rate": 2.03371237723323e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 15033120,
      "step": 23040
    },
    {
      "epoch": 12.078092243186584,
      "grad_norm": 0.3979801833629608,
      "learning_rate": 2.032589062314636e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 15036960,
      "step": 23045
    },
    {
      "epoch": 12.080712788259959,
      "grad_norm": 0.40092387795448303,
      "learning_rate": 2.0314658451730767e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 15040992,
      "step": 23050
    },
    {
      "epoch": 12.083333333333334,
      "grad_norm": 0.5011147856712341,
      "learning_rate": 2.030342726043517e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 15044096,
      "step": 23055
    },
    {
      "epoch": 12.085953878406709,
      "grad_norm": 0.6009672284126282,
      "learning_rate": 2.0292197051609013e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 15047680,
      "step": 23060
    },
    {
      "epoch": 12.088574423480084,
      "grad_norm": 0.4150668978691101,
      "learning_rate": 2.0280967827601518e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 15051328,
      "step": 23065
    },
    {
      "epoch": 12.09119496855346,
      "grad_norm": 0.4993797242641449,
      "learning_rate": 2.0269739590761723e-05,
      "loss": 0.3792,
      "num_input_tokens_seen": 15054560,
      "step": 23070
    },
    {
      "epoch": 12.093815513626835,
      "grad_norm": 0.3970782160758972,
      "learning_rate": 2.0258512343438442e-05,
      "loss": 0.4732,
      "num_input_tokens_seen": 15057888,
      "step": 23075
    },
    {
      "epoch": 12.09643605870021,
      "grad_norm": 0.5082402229309082,
      "learning_rate": 2.0247286087980276e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 15061120,
      "step": 23080
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 0.5347899794578552,
      "learning_rate": 2.0236060826735644e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 15064224,
      "step": 23085
    },
    {
      "epoch": 12.10167714884696,
      "grad_norm": 0.39219313859939575,
      "learning_rate": 2.0224836562052732e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 15067360,
      "step": 23090
    },
    {
      "epoch": 12.104297693920335,
      "grad_norm": 0.48487192392349243,
      "learning_rate": 2.0213613296279533e-05,
      "loss": 0.3838,
      "num_input_tokens_seen": 15070688,
      "step": 23095
    },
    {
      "epoch": 12.10691823899371,
      "grad_norm": 0.5386830568313599,
      "learning_rate": 2.0202391031763844e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 15074304,
      "step": 23100
    },
    {
      "epoch": 12.109538784067086,
      "grad_norm": 0.5975913405418396,
      "learning_rate": 2.0191169770853195e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 15076768,
      "step": 23105
    },
    {
      "epoch": 12.11215932914046,
      "grad_norm": 0.5390138626098633,
      "learning_rate": 2.0179949515894975e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 15079232,
      "step": 23110
    },
    {
      "epoch": 12.114779874213836,
      "grad_norm": 0.40120819211006165,
      "learning_rate": 2.016873026923633e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 15082688,
      "step": 23115
    },
    {
      "epoch": 12.117400419287211,
      "grad_norm": 0.2901837229728699,
      "learning_rate": 2.0157512033224187e-05,
      "loss": 0.3761,
      "num_input_tokens_seen": 15086400,
      "step": 23120
    },
    {
      "epoch": 12.120020964360586,
      "grad_norm": 0.35353901982307434,
      "learning_rate": 2.0146294810205287e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 15089184,
      "step": 23125
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 0.4029199182987213,
      "learning_rate": 2.0135078602526146e-05,
      "loss": 0.3373,
      "num_input_tokens_seen": 15092384,
      "step": 23130
    },
    {
      "epoch": 12.125262054507338,
      "grad_norm": 0.5719616413116455,
      "learning_rate": 2.0123863412533046e-05,
      "loss": 0.4024,
      "num_input_tokens_seen": 15095776,
      "step": 23135
    },
    {
      "epoch": 12.127882599580714,
      "grad_norm": 0.4957464039325714,
      "learning_rate": 2.0112649242572095e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 15098624,
      "step": 23140
    },
    {
      "epoch": 12.130503144654089,
      "grad_norm": 0.4013603627681732,
      "learning_rate": 2.0101436094989164e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 15100992,
      "step": 23145
    },
    {
      "epoch": 12.133123689727464,
      "grad_norm": 0.5434017181396484,
      "learning_rate": 2.0090223972129918e-05,
      "loss": 0.3548,
      "num_input_tokens_seen": 15104352,
      "step": 23150
    },
    {
      "epoch": 12.135744234800839,
      "grad_norm": 0.7363807559013367,
      "learning_rate": 2.0079012876339797e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 15107200,
      "step": 23155
    },
    {
      "epoch": 12.138364779874214,
      "grad_norm": 0.6625010371208191,
      "learning_rate": 2.0067802809964055e-05,
      "loss": 0.541,
      "num_input_tokens_seen": 15111008,
      "step": 23160
    },
    {
      "epoch": 12.14098532494759,
      "grad_norm": 0.5184260010719299,
      "learning_rate": 2.0056593775347686e-05,
      "loss": 0.4524,
      "num_input_tokens_seen": 15113440,
      "step": 23165
    },
    {
      "epoch": 12.143605870020965,
      "grad_norm": 0.6138588786125183,
      "learning_rate": 2.00453857748355e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 15116256,
      "step": 23170
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 0.3519432544708252,
      "learning_rate": 2.0034178810772082e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 15119296,
      "step": 23175
    },
    {
      "epoch": 12.148846960167715,
      "grad_norm": 0.34344029426574707,
      "learning_rate": 2.002297288550181e-05,
      "loss": 0.3288,
      "num_input_tokens_seen": 15123584,
      "step": 23180
    },
    {
      "epoch": 12.15146750524109,
      "grad_norm": 0.4177723824977875,
      "learning_rate": 2.0011768001368835e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 15127008,
      "step": 23185
    },
    {
      "epoch": 12.154088050314465,
      "grad_norm": 0.4898570775985718,
      "learning_rate": 2.0000564160717076e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 15129728,
      "step": 23190
    },
    {
      "epoch": 12.15670859538784,
      "grad_norm": 0.44535019993782043,
      "learning_rate": 1.9989361365890263e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 15133152,
      "step": 23195
    },
    {
      "epoch": 12.159329140461216,
      "grad_norm": 0.3484140634536743,
      "learning_rate": 1.997815961923188e-05,
      "loss": 0.4875,
      "num_input_tokens_seen": 15136960,
      "step": 23200
    },
    {
      "epoch": 12.16194968553459,
      "grad_norm": 0.45377007126808167,
      "learning_rate": 1.996695892308522e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 15140128,
      "step": 23205
    },
    {
      "epoch": 12.164570230607966,
      "grad_norm": 0.26711663603782654,
      "learning_rate": 1.9955759279793327e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 15144128,
      "step": 23210
    },
    {
      "epoch": 12.167190775681341,
      "grad_norm": 0.5267496705055237,
      "learning_rate": 1.9944560691699057e-05,
      "loss": 0.45,
      "num_input_tokens_seen": 15147968,
      "step": 23215
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 0.44116100668907166,
      "learning_rate": 1.9933363161145006e-05,
      "loss": 0.3766,
      "num_input_tokens_seen": 15150976,
      "step": 23220
    },
    {
      "epoch": 12.172431865828091,
      "grad_norm": 0.5826636552810669,
      "learning_rate": 1.9922166690473575e-05,
      "loss": 0.6015,
      "num_input_tokens_seen": 15154368,
      "step": 23225
    },
    {
      "epoch": 12.175052410901468,
      "grad_norm": 0.6897755265235901,
      "learning_rate": 1.991097128202695e-05,
      "loss": 0.4928,
      "num_input_tokens_seen": 15157984,
      "step": 23230
    },
    {
      "epoch": 12.177672955974844,
      "grad_norm": 0.7001205086708069,
      "learning_rate": 1.9899776938147065e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 15162208,
      "step": 23235
    },
    {
      "epoch": 12.180293501048219,
      "grad_norm": 0.534866988658905,
      "learning_rate": 1.9888583661175665e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 15165088,
      "step": 23240
    },
    {
      "epoch": 12.182914046121594,
      "grad_norm": 0.49846312403678894,
      "learning_rate": 1.9877391453454253e-05,
      "loss": 0.3379,
      "num_input_tokens_seen": 15167200,
      "step": 23245
    },
    {
      "epoch": 12.185534591194969,
      "grad_norm": 0.4251430630683899,
      "learning_rate": 1.98662003173241e-05,
      "loss": 0.5026,
      "num_input_tokens_seen": 15170304,
      "step": 23250
    },
    {
      "epoch": 12.188155136268344,
      "grad_norm": 0.7250073552131653,
      "learning_rate": 1.9855010255126266e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 15173024,
      "step": 23255
    },
    {
      "epoch": 12.19077568134172,
      "grad_norm": 1.116719126701355,
      "learning_rate": 1.9843821269201598e-05,
      "loss": 0.5488,
      "num_input_tokens_seen": 15176960,
      "step": 23260
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 0.5464887619018555,
      "learning_rate": 1.983263336189069e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 15184576,
      "step": 23265
    },
    {
      "epoch": 12.19601677148847,
      "grad_norm": 0.5567009449005127,
      "learning_rate": 1.9821446535533937e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 15187840,
      "step": 23270
    },
    {
      "epoch": 12.198637316561845,
      "grad_norm": 0.2641857862472534,
      "learning_rate": 1.9810260792471492e-05,
      "loss": 0.3848,
      "num_input_tokens_seen": 15191072,
      "step": 23275
    },
    {
      "epoch": 12.20125786163522,
      "grad_norm": 0.722868025302887,
      "learning_rate": 1.979907613504327e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 15194592,
      "step": 23280
    },
    {
      "epoch": 12.203878406708595,
      "grad_norm": 0.7031617760658264,
      "learning_rate": 1.978789256558899e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 15198240,
      "step": 23285
    },
    {
      "epoch": 12.20649895178197,
      "grad_norm": 0.7847262620925903,
      "learning_rate": 1.977671008644812e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 15201056,
      "step": 23290
    },
    {
      "epoch": 12.209119496855346,
      "grad_norm": 0.5470415949821472,
      "learning_rate": 1.9765528699959912e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 15204320,
      "step": 23295
    },
    {
      "epoch": 12.21174004192872,
      "grad_norm": 0.45433512330055237,
      "learning_rate": 1.975434840846339e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 15207264,
      "step": 23300
    },
    {
      "epoch": 12.214360587002096,
      "grad_norm": 0.46565452218055725,
      "learning_rate": 1.9743169214297326e-05,
      "loss": 0.3495,
      "num_input_tokens_seen": 15210112,
      "step": 23305
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 0.32898929715156555,
      "learning_rate": 1.973199111980029e-05,
      "loss": 0.3698,
      "num_input_tokens_seen": 15213504,
      "step": 23310
    },
    {
      "epoch": 12.219601677148846,
      "grad_norm": 0.3922995328903198,
      "learning_rate": 1.9720814127310608e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 15215808,
      "step": 23315
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 0.740490198135376,
      "learning_rate": 1.9709638239166386e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 15218304,
      "step": 23320
    },
    {
      "epoch": 12.224842767295598,
      "grad_norm": 0.5704590678215027,
      "learning_rate": 1.9698463457705485e-05,
      "loss": 0.5879,
      "num_input_tokens_seen": 15221024,
      "step": 23325
    },
    {
      "epoch": 12.227463312368974,
      "grad_norm": 0.572600245475769,
      "learning_rate": 1.968728978526556e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 15223872,
      "step": 23330
    },
    {
      "epoch": 12.230083857442349,
      "grad_norm": 0.5056528449058533,
      "learning_rate": 1.9676117224183986e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 15227168,
      "step": 23335
    },
    {
      "epoch": 12.232704402515724,
      "grad_norm": 0.6349392533302307,
      "learning_rate": 1.9664945776797956e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 15230816,
      "step": 23340
    },
    {
      "epoch": 12.235324947589099,
      "grad_norm": 0.7006680965423584,
      "learning_rate": 1.96537754454444e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 15233568,
      "step": 23345
    },
    {
      "epoch": 12.237945492662474,
      "grad_norm": 0.6713096499443054,
      "learning_rate": 1.964260623246003e-05,
      "loss": 0.3331,
      "num_input_tokens_seen": 15236064,
      "step": 23350
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 0.4612477719783783,
      "learning_rate": 1.9631438140181324e-05,
      "loss": 0.3961,
      "num_input_tokens_seen": 15239168,
      "step": 23355
    },
    {
      "epoch": 12.243186582809225,
      "grad_norm": 0.3476896584033966,
      "learning_rate": 1.9620271170944503e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 15241984,
      "step": 23360
    },
    {
      "epoch": 12.2458071278826,
      "grad_norm": 0.6562462449073792,
      "learning_rate": 1.960910532708558e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 15244800,
      "step": 23365
    },
    {
      "epoch": 12.248427672955975,
      "grad_norm": 0.5007543563842773,
      "learning_rate": 1.9597940610940315e-05,
      "loss": 0.3641,
      "num_input_tokens_seen": 15248416,
      "step": 23370
    },
    {
      "epoch": 12.25104821802935,
      "grad_norm": 0.7544858455657959,
      "learning_rate": 1.958677702484425e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 15251008,
      "step": 23375
    },
    {
      "epoch": 12.253668763102725,
      "grad_norm": 0.5367448925971985,
      "learning_rate": 1.957561457113267e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 15256992,
      "step": 23380
    },
    {
      "epoch": 12.2562893081761,
      "grad_norm": 0.8874644637107849,
      "learning_rate": 1.9564453252140648e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 15259712,
      "step": 23385
    },
    {
      "epoch": 12.258909853249476,
      "grad_norm": 0.4415163993835449,
      "learning_rate": 1.9553293070202984e-05,
      "loss": 0.551,
      "num_input_tokens_seen": 15262880,
      "step": 23390
    },
    {
      "epoch": 12.26153039832285,
      "grad_norm": 0.502659797668457,
      "learning_rate": 1.9542134027654267e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 15266048,
      "step": 23395
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 0.6084691286087036,
      "learning_rate": 1.953097612682885e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 15268928,
      "step": 23400
    },
    {
      "epoch": 12.266771488469601,
      "grad_norm": 0.4098556637763977,
      "learning_rate": 1.9519819370060832e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 15271872,
      "step": 23405
    },
    {
      "epoch": 12.269392033542976,
      "grad_norm": 0.24342083930969238,
      "learning_rate": 1.9508663759684074e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 15276704,
      "step": 23410
    },
    {
      "epoch": 12.272012578616351,
      "grad_norm": 0.4181840717792511,
      "learning_rate": 1.9497509298032224e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 15279648,
      "step": 23415
    },
    {
      "epoch": 12.274633123689728,
      "grad_norm": 0.8743410110473633,
      "learning_rate": 1.9486355987438647e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 15282176,
      "step": 23420
    },
    {
      "epoch": 12.277253668763104,
      "grad_norm": 0.5273066759109497,
      "learning_rate": 1.947520383023649e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 15284640,
      "step": 23425
    },
    {
      "epoch": 12.279874213836479,
      "grad_norm": 0.43036964535713196,
      "learning_rate": 1.9464052828758666e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 15287712,
      "step": 23430
    },
    {
      "epoch": 12.282494758909854,
      "grad_norm": 0.37901991605758667,
      "learning_rate": 1.945290298533783e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 15291296,
      "step": 23435
    },
    {
      "epoch": 12.285115303983229,
      "grad_norm": 0.4031318724155426,
      "learning_rate": 1.9441754302306406e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 15294464,
      "step": 23440
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 0.38703107833862305,
      "learning_rate": 1.9430606781996578e-05,
      "loss": 0.2711,
      "num_input_tokens_seen": 15299392,
      "step": 23445
    },
    {
      "epoch": 12.29035639412998,
      "grad_norm": 0.7586690783500671,
      "learning_rate": 1.941946042674026e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 15302496,
      "step": 23450
    },
    {
      "epoch": 12.292976939203355,
      "grad_norm": 0.38528600335121155,
      "learning_rate": 1.940831523886916e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 15306464,
      "step": 23455
    },
    {
      "epoch": 12.29559748427673,
      "grad_norm": 0.759675920009613,
      "learning_rate": 1.9397171220714714e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 15309792,
      "step": 23460
    },
    {
      "epoch": 12.298218029350105,
      "grad_norm": 0.5338413119316101,
      "learning_rate": 1.9386028374608134e-05,
      "loss": 0.4715,
      "num_input_tokens_seen": 15312576,
      "step": 23465
    },
    {
      "epoch": 12.30083857442348,
      "grad_norm": 0.5726226568222046,
      "learning_rate": 1.9374886702880372e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 15315296,
      "step": 23470
    },
    {
      "epoch": 12.303459119496855,
      "grad_norm": 0.3119830787181854,
      "learning_rate": 1.936374620786213e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 15318112,
      "step": 23475
    },
    {
      "epoch": 12.30607966457023,
      "grad_norm": 0.383835107088089,
      "learning_rate": 1.9352606891883885e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 15320896,
      "step": 23480
    },
    {
      "epoch": 12.308700209643606,
      "grad_norm": 0.6752082705497742,
      "learning_rate": 1.9341468757275844e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 15323904,
      "step": 23485
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 0.3443674147129059,
      "learning_rate": 1.933033180636798e-05,
      "loss": 0.528,
      "num_input_tokens_seen": 15326944,
      "step": 23490
    },
    {
      "epoch": 12.313941299790356,
      "grad_norm": 0.44811901450157166,
      "learning_rate": 1.9319196041490027e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 15331776,
      "step": 23495
    },
    {
      "epoch": 12.316561844863731,
      "grad_norm": 0.2937861979007721,
      "learning_rate": 1.930806146497146e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 15334624,
      "step": 23500
    },
    {
      "epoch": 12.319182389937106,
      "grad_norm": 0.42346426844596863,
      "learning_rate": 1.9296928079141485e-05,
      "loss": 0.4438,
      "num_input_tokens_seen": 15339072,
      "step": 23505
    },
    {
      "epoch": 12.321802935010481,
      "grad_norm": 0.513841986656189,
      "learning_rate": 1.9285795886329093e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 15342240,
      "step": 23510
    },
    {
      "epoch": 12.324423480083858,
      "grad_norm": 0.4056822657585144,
      "learning_rate": 1.927466488886301e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 15346528,
      "step": 23515
    },
    {
      "epoch": 12.327044025157234,
      "grad_norm": 0.308388352394104,
      "learning_rate": 1.926353508907172e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 15350144,
      "step": 23520
    },
    {
      "epoch": 12.329664570230609,
      "grad_norm": 0.6444071531295776,
      "learning_rate": 1.9252406489283437e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 15352832,
      "step": 23525
    },
    {
      "epoch": 12.332285115303984,
      "grad_norm": 0.6118021011352539,
      "learning_rate": 1.9241279091826155e-05,
      "loss": 0.5087,
      "num_input_tokens_seen": 15355392,
      "step": 23530
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 0.6778067946434021,
      "learning_rate": 1.923015289902758e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 15357504,
      "step": 23535
    },
    {
      "epoch": 12.337526205450734,
      "grad_norm": 0.43443992733955383,
      "learning_rate": 1.9219027913215193e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 15360672,
      "step": 23540
    },
    {
      "epoch": 12.34014675052411,
      "grad_norm": 0.9688378572463989,
      "learning_rate": 1.9207904136716214e-05,
      "loss": 0.6325,
      "num_input_tokens_seen": 15364128,
      "step": 23545
    },
    {
      "epoch": 12.342767295597485,
      "grad_norm": 0.9348114728927612,
      "learning_rate": 1.9196781571857608e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 15367072,
      "step": 23550
    },
    {
      "epoch": 12.34538784067086,
      "grad_norm": 0.5876073837280273,
      "learning_rate": 1.9185660220966097e-05,
      "loss": 0.3458,
      "num_input_tokens_seen": 15370848,
      "step": 23555
    },
    {
      "epoch": 12.348008385744235,
      "grad_norm": 0.4418937861919403,
      "learning_rate": 1.9174540086368135e-05,
      "loss": 0.3918,
      "num_input_tokens_seen": 15375680,
      "step": 23560
    },
    {
      "epoch": 12.35062893081761,
      "grad_norm": 0.5601840615272522,
      "learning_rate": 1.9163421170389918e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 15378784,
      "step": 23565
    },
    {
      "epoch": 12.353249475890985,
      "grad_norm": 0.4012461304664612,
      "learning_rate": 1.9152303475357402e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 15381856,
      "step": 23570
    },
    {
      "epoch": 12.35587002096436,
      "grad_norm": 0.3646993041038513,
      "learning_rate": 1.9141187003596295e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 15385376,
      "step": 23575
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 0.5398817658424377,
      "learning_rate": 1.913007175743202e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 15389568,
      "step": 23580
    },
    {
      "epoch": 12.36111111111111,
      "grad_norm": 0.7126410007476807,
      "learning_rate": 1.9118957739189773e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 15392160,
      "step": 23585
    },
    {
      "epoch": 12.363731656184486,
      "grad_norm": 0.5800081491470337,
      "learning_rate": 1.910784495119447e-05,
      "loss": 0.5196,
      "num_input_tokens_seen": 15395520,
      "step": 23590
    },
    {
      "epoch": 12.366352201257861,
      "grad_norm": 0.4096889793872833,
      "learning_rate": 1.9096733395770778e-05,
      "loss": 0.5398,
      "num_input_tokens_seen": 15398592,
      "step": 23595
    },
    {
      "epoch": 12.368972746331236,
      "grad_norm": 0.3608287274837494,
      "learning_rate": 1.9085623075243114e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 15401824,
      "step": 23600
    },
    {
      "epoch": 12.371593291404611,
      "grad_norm": 0.49594685435295105,
      "learning_rate": 1.9074513991935623e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 15405536,
      "step": 23605
    },
    {
      "epoch": 12.374213836477987,
      "grad_norm": 0.4573974013328552,
      "learning_rate": 1.906340614817221e-05,
      "loss": 0.5123,
      "num_input_tokens_seen": 15408800,
      "step": 23610
    },
    {
      "epoch": 12.376834381551364,
      "grad_norm": 0.46974295377731323,
      "learning_rate": 1.9052299546276506e-05,
      "loss": 0.561,
      "num_input_tokens_seen": 15411392,
      "step": 23615
    },
    {
      "epoch": 12.379454926624739,
      "grad_norm": 0.6835116744041443,
      "learning_rate": 1.904119418857187e-05,
      "loss": 0.5789,
      "num_input_tokens_seen": 15414144,
      "step": 23620
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 0.40802881121635437,
      "learning_rate": 1.903009007738144e-05,
      "loss": 0.5391,
      "num_input_tokens_seen": 15417984,
      "step": 23625
    },
    {
      "epoch": 12.384696016771489,
      "grad_norm": 0.5005124807357788,
      "learning_rate": 1.9018987215028045e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 15421280,
      "step": 23630
    },
    {
      "epoch": 12.387316561844864,
      "grad_norm": 0.6346768140792847,
      "learning_rate": 1.9007885603834298e-05,
      "loss": 0.5417,
      "num_input_tokens_seen": 15423872,
      "step": 23635
    },
    {
      "epoch": 12.38993710691824,
      "grad_norm": 0.44961145520210266,
      "learning_rate": 1.8996785246122526e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 15426976,
      "step": 23640
    },
    {
      "epoch": 12.392557651991615,
      "grad_norm": 0.6823875904083252,
      "learning_rate": 1.898568614421478e-05,
      "loss": 0.3941,
      "num_input_tokens_seen": 15430240,
      "step": 23645
    },
    {
      "epoch": 12.39517819706499,
      "grad_norm": 0.360533207654953,
      "learning_rate": 1.897458830043288e-05,
      "loss": 0.3351,
      "num_input_tokens_seen": 15433664,
      "step": 23650
    },
    {
      "epoch": 12.397798742138365,
      "grad_norm": 0.5178430676460266,
      "learning_rate": 1.8963491717098372e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 15437088,
      "step": 23655
    },
    {
      "epoch": 12.40041928721174,
      "grad_norm": 0.5978736281394958,
      "learning_rate": 1.8952396396532522e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 15440096,
      "step": 23660
    },
    {
      "epoch": 12.403039832285115,
      "grad_norm": 0.3888499140739441,
      "learning_rate": 1.8941302341056356e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 15444352,
      "step": 23665
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 0.8830628395080566,
      "learning_rate": 1.8930209552990625e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 15448736,
      "step": 23670
    },
    {
      "epoch": 12.408280922431866,
      "grad_norm": 0.32729002833366394,
      "learning_rate": 1.8919118034655796e-05,
      "loss": 0.3925,
      "num_input_tokens_seen": 15451712,
      "step": 23675
    },
    {
      "epoch": 12.41090146750524,
      "grad_norm": 0.5478555560112,
      "learning_rate": 1.8908027788372107e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 15454272,
      "step": 23680
    },
    {
      "epoch": 12.413522012578616,
      "grad_norm": 0.3818920850753784,
      "learning_rate": 1.88969388164595e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 15457856,
      "step": 23685
    },
    {
      "epoch": 12.416142557651991,
      "grad_norm": 0.5702599287033081,
      "learning_rate": 1.8885851121237666e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 15461408,
      "step": 23690
    },
    {
      "epoch": 12.418763102725366,
      "grad_norm": 0.3687085509300232,
      "learning_rate": 1.8874764705026023e-05,
      "loss": 0.4993,
      "num_input_tokens_seen": 15464992,
      "step": 23695
    },
    {
      "epoch": 12.421383647798741,
      "grad_norm": 0.36162662506103516,
      "learning_rate": 1.8863679570143728e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 15468096,
      "step": 23700
    },
    {
      "epoch": 12.424004192872117,
      "grad_norm": 0.5981331467628479,
      "learning_rate": 1.885259571890966e-05,
      "loss": 0.3918,
      "num_input_tokens_seen": 15470752,
      "step": 23705
    },
    {
      "epoch": 12.426624737945493,
      "grad_norm": 0.49703818559646606,
      "learning_rate": 1.8841513153642427e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 15473952,
      "step": 23710
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 0.37165507674217224,
      "learning_rate": 1.8830431876660388e-05,
      "loss": 0.363,
      "num_input_tokens_seen": 15477760,
      "step": 23715
    },
    {
      "epoch": 12.431865828092244,
      "grad_norm": 0.5169023871421814,
      "learning_rate": 1.8819351890281607e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 15480384,
      "step": 23720
    },
    {
      "epoch": 12.434486373165619,
      "grad_norm": 0.664169430732727,
      "learning_rate": 1.880827319682391e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 15482976,
      "step": 23725
    },
    {
      "epoch": 12.437106918238994,
      "grad_norm": 0.3915455937385559,
      "learning_rate": 1.879719579860481e-05,
      "loss": 0.3469,
      "num_input_tokens_seen": 15486016,
      "step": 23730
    },
    {
      "epoch": 12.43972746331237,
      "grad_norm": 0.3788011372089386,
      "learning_rate": 1.878611969794159e-05,
      "loss": 0.3667,
      "num_input_tokens_seen": 15489472,
      "step": 23735
    },
    {
      "epoch": 12.442348008385745,
      "grad_norm": 0.5196203589439392,
      "learning_rate": 1.8775044897151232e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 15493120,
      "step": 23740
    },
    {
      "epoch": 12.44496855345912,
      "grad_norm": 0.3898351788520813,
      "learning_rate": 1.876397139855047e-05,
      "loss": 0.459,
      "num_input_tokens_seen": 15495968,
      "step": 23745
    },
    {
      "epoch": 12.447589098532495,
      "grad_norm": 0.6603909730911255,
      "learning_rate": 1.875289920445574e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 15498784,
      "step": 23750
    },
    {
      "epoch": 12.45020964360587,
      "grad_norm": 0.3657612204551697,
      "learning_rate": 1.8741828317183234e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 15502784,
      "step": 23755
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 0.46194958686828613,
      "learning_rate": 1.873075873904885e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 15505952,
      "step": 23760
    },
    {
      "epoch": 12.45545073375262,
      "grad_norm": 0.8274015784263611,
      "learning_rate": 1.87196904723682e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 15509344,
      "step": 23765
    },
    {
      "epoch": 12.458071278825996,
      "grad_norm": 0.4650874435901642,
      "learning_rate": 1.8708623519456665e-05,
      "loss": 0.5284,
      "num_input_tokens_seen": 15512544,
      "step": 23770
    },
    {
      "epoch": 12.46069182389937,
      "grad_norm": 0.4675927758216858,
      "learning_rate": 1.8697557882629308e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 15519104,
      "step": 23775
    },
    {
      "epoch": 12.463312368972746,
      "grad_norm": 0.5178950428962708,
      "learning_rate": 1.868649356420094e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 15522528,
      "step": 23780
    },
    {
      "epoch": 12.465932914046121,
      "grad_norm": 0.44952526688575745,
      "learning_rate": 1.86754305664861e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 15525440,
      "step": 23785
    },
    {
      "epoch": 12.468553459119496,
      "grad_norm": 0.39405739307403564,
      "learning_rate": 1.866436889179902e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 15528448,
      "step": 23790
    },
    {
      "epoch": 12.471174004192871,
      "grad_norm": 0.4163210093975067,
      "learning_rate": 1.865330854245369e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 15531392,
      "step": 23795
    },
    {
      "epoch": 12.473794549266247,
      "grad_norm": 0.6953192353248596,
      "learning_rate": 1.8642249520763804e-05,
      "loss": 0.313,
      "num_input_tokens_seen": 15534784,
      "step": 23800
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 0.32014232873916626,
      "learning_rate": 1.863119182904278e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 15538144,
      "step": 23805
    },
    {
      "epoch": 12.479035639412999,
      "grad_norm": 0.27093255519866943,
      "learning_rate": 1.8620135469603777e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 15541600,
      "step": 23810
    },
    {
      "epoch": 12.481656184486374,
      "grad_norm": 0.4740891456604004,
      "learning_rate": 1.8609080444759653e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 15545728,
      "step": 23815
    },
    {
      "epoch": 12.484276729559749,
      "grad_norm": 0.4204324781894684,
      "learning_rate": 1.8598026756822978e-05,
      "loss": 0.4227,
      "num_input_tokens_seen": 15548000,
      "step": 23820
    },
    {
      "epoch": 12.486897274633124,
      "grad_norm": 0.3734978139400482,
      "learning_rate": 1.8586974408106074e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 15550496,
      "step": 23825
    },
    {
      "epoch": 12.4895178197065,
      "grad_norm": 0.33672159910202026,
      "learning_rate": 1.8575923400920954e-05,
      "loss": 0.448,
      "num_input_tokens_seen": 15554304,
      "step": 23830
    },
    {
      "epoch": 12.492138364779874,
      "grad_norm": 0.3251475393772125,
      "learning_rate": 1.856487373757937e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 15557824,
      "step": 23835
    },
    {
      "epoch": 12.49475890985325,
      "grad_norm": 0.3954533338546753,
      "learning_rate": 1.8553825420392792e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 15560704,
      "step": 23840
    },
    {
      "epoch": 12.497379454926625,
      "grad_norm": 0.30266594886779785,
      "learning_rate": 1.8542778451672385e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 15564832,
      "step": 23845
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.47381696105003357,
      "learning_rate": 1.8531732833729063e-05,
      "loss": 0.3538,
      "num_input_tokens_seen": 15567904,
      "step": 23850
    },
    {
      "epoch": 12.502620545073375,
      "grad_norm": 0.45747366547584534,
      "learning_rate": 1.852068856887343e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 15571168,
      "step": 23855
    },
    {
      "epoch": 12.50524109014675,
      "grad_norm": 0.611757218837738,
      "learning_rate": 1.8509645659415837e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 15574144,
      "step": 23860
    },
    {
      "epoch": 12.507861635220126,
      "grad_norm": 0.5075322985649109,
      "learning_rate": 1.849860410766632e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 15576160,
      "step": 23865
    },
    {
      "epoch": 12.5104821802935,
      "grad_norm": 0.7022299766540527,
      "learning_rate": 1.8487563915934664e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 15579424,
      "step": 23870
    },
    {
      "epoch": 12.513102725366876,
      "grad_norm": 0.36086782813072205,
      "learning_rate": 1.847652508653033e-05,
      "loss": 0.2723,
      "num_input_tokens_seen": 15584032,
      "step": 23875
    },
    {
      "epoch": 12.515723270440251,
      "grad_norm": 1.264525055885315,
      "learning_rate": 1.846548762176252e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 15588544,
      "step": 23880
    },
    {
      "epoch": 12.518343815513626,
      "grad_norm": 0.4032195210456848,
      "learning_rate": 1.845445152394015e-05,
      "loss": 0.3656,
      "num_input_tokens_seen": 15594336,
      "step": 23885
    },
    {
      "epoch": 12.520964360587001,
      "grad_norm": 0.6523927450180054,
      "learning_rate": 1.8443416795371853e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 15597760,
      "step": 23890
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 0.49289724230766296,
      "learning_rate": 1.8432383438365957e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 15601824,
      "step": 23895
    },
    {
      "epoch": 12.526205450733752,
      "grad_norm": 0.5058708190917969,
      "learning_rate": 1.8421351455230528e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 15605824,
      "step": 23900
    },
    {
      "epoch": 12.528825995807129,
      "grad_norm": 0.6597735285758972,
      "learning_rate": 1.8410320848273315e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 15609312,
      "step": 23905
    },
    {
      "epoch": 12.531446540880504,
      "grad_norm": 0.3094722330570221,
      "learning_rate": 1.8399291619801794e-05,
      "loss": 0.5943,
      "num_input_tokens_seen": 15613216,
      "step": 23910
    },
    {
      "epoch": 12.534067085953879,
      "grad_norm": 0.6441361308097839,
      "learning_rate": 1.838826377212317e-05,
      "loss": 0.3356,
      "num_input_tokens_seen": 15616352,
      "step": 23915
    },
    {
      "epoch": 12.536687631027254,
      "grad_norm": 0.40377068519592285,
      "learning_rate": 1.8377237307544326e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 15619072,
      "step": 23920
    },
    {
      "epoch": 12.53930817610063,
      "grad_norm": 0.5294737815856934,
      "learning_rate": 1.8366212228371894e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 15622144,
      "step": 23925
    },
    {
      "epoch": 12.541928721174004,
      "grad_norm": 0.5511976480484009,
      "learning_rate": 1.8355188536912173e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 15625152,
      "step": 23930
    },
    {
      "epoch": 12.54454926624738,
      "grad_norm": 0.7102044224739075,
      "learning_rate": 1.83441662354712e-05,
      "loss": 0.4771,
      "num_input_tokens_seen": 15628128,
      "step": 23935
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 0.7155203819274902,
      "learning_rate": 1.8333145326354722e-05,
      "loss": 0.5175,
      "num_input_tokens_seen": 15631328,
      "step": 23940
    },
    {
      "epoch": 12.54979035639413,
      "grad_norm": 0.4899660050868988,
      "learning_rate": 1.8322125811868175e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 15634304,
      "step": 23945
    },
    {
      "epoch": 12.552410901467505,
      "grad_norm": 0.44686299562454224,
      "learning_rate": 1.831110769431673e-05,
      "loss": 0.3099,
      "num_input_tokens_seen": 15637984,
      "step": 23950
    },
    {
      "epoch": 12.55503144654088,
      "grad_norm": 0.7772340178489685,
      "learning_rate": 1.8300090976005247e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 15640672,
      "step": 23955
    },
    {
      "epoch": 12.557651991614255,
      "grad_norm": 0.618003249168396,
      "learning_rate": 1.8289075659238285e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 15644096,
      "step": 23960
    },
    {
      "epoch": 12.56027253668763,
      "grad_norm": 0.5703979134559631,
      "learning_rate": 1.8278061746320134e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 15647040,
      "step": 23965
    },
    {
      "epoch": 12.562893081761006,
      "grad_norm": 0.6926773190498352,
      "learning_rate": 1.8267049239554786e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 15649824,
      "step": 23970
    },
    {
      "epoch": 12.565513626834381,
      "grad_norm": 0.35886988043785095,
      "learning_rate": 1.8256038141245912e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 15652992,
      "step": 23975
    },
    {
      "epoch": 12.568134171907756,
      "grad_norm": 1.0853006839752197,
      "learning_rate": 1.824502845369693e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 15655680,
      "step": 23980
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 0.5399832725524902,
      "learning_rate": 1.8234020179210938e-05,
      "loss": 0.333,
      "num_input_tokens_seen": 15658496,
      "step": 23985
    },
    {
      "epoch": 12.573375262054507,
      "grad_norm": 0.5000417828559875,
      "learning_rate": 1.8223013320090715e-05,
      "loss": 0.5924,
      "num_input_tokens_seen": 15661248,
      "step": 23990
    },
    {
      "epoch": 12.575995807127882,
      "grad_norm": 0.42196762561798096,
      "learning_rate": 1.82120078786388e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 15663840,
      "step": 23995
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 0.5051140785217285,
      "learning_rate": 1.820100385715739e-05,
      "loss": 0.5092,
      "num_input_tokens_seen": 15667744,
      "step": 24000
    },
    {
      "epoch": 12.581236897274634,
      "grad_norm": 0.5577185153961182,
      "learning_rate": 1.8190001257948414e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 15670592,
      "step": 24005
    },
    {
      "epoch": 12.583857442348009,
      "grad_norm": 0.35052189230918884,
      "learning_rate": 1.8179000083313483e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 15674688,
      "step": 24010
    },
    {
      "epoch": 12.586477987421384,
      "grad_norm": 0.6711887121200562,
      "learning_rate": 1.8168000335553907e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 15677440,
      "step": 24015
    },
    {
      "epoch": 12.58909853249476,
      "grad_norm": 0.7367752194404602,
      "learning_rate": 1.8157002016970723e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 15680416,
      "step": 24020
    },
    {
      "epoch": 12.591719077568134,
      "grad_norm": 0.3445480167865753,
      "learning_rate": 1.8146005129864644e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 15684800,
      "step": 24025
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 0.4449228048324585,
      "learning_rate": 1.81350096765361e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 15687936,
      "step": 24030
    },
    {
      "epoch": 12.596960167714885,
      "grad_norm": 0.5168044567108154,
      "learning_rate": 1.812401565928521e-05,
      "loss": 0.3147,
      "num_input_tokens_seen": 15691104,
      "step": 24035
    },
    {
      "epoch": 12.59958071278826,
      "grad_norm": 0.5009704828262329,
      "learning_rate": 1.8113023080411806e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 15694432,
      "step": 24040
    },
    {
      "epoch": 12.602201257861635,
      "grad_norm": 0.3339747190475464,
      "learning_rate": 1.810203194221539e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 15697504,
      "step": 24045
    },
    {
      "epoch": 12.60482180293501,
      "grad_norm": 0.5981756448745728,
      "learning_rate": 1.8091042246995206e-05,
      "loss": 0.4926,
      "num_input_tokens_seen": 15701824,
      "step": 24050
    },
    {
      "epoch": 12.607442348008385,
      "grad_norm": 0.3876720666885376,
      "learning_rate": 1.8080053997050154e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 15704736,
      "step": 24055
    },
    {
      "epoch": 12.61006289308176,
      "grad_norm": 0.4923306703567505,
      "learning_rate": 1.8069067194678867e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 15707744,
      "step": 24060
    },
    {
      "epoch": 12.612683438155136,
      "grad_norm": 0.36037206649780273,
      "learning_rate": 1.8058081842179642e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 15710208,
      "step": 24065
    },
    {
      "epoch": 12.615303983228511,
      "grad_norm": 0.28525635600090027,
      "learning_rate": 1.804709794185051e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 15713792,
      "step": 24070
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 0.5709060430526733,
      "learning_rate": 1.8036115495989164e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 15717792,
      "step": 24075
    },
    {
      "epoch": 12.620545073375261,
      "grad_norm": 0.4040522575378418,
      "learning_rate": 1.8025134506892997e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 15720896,
      "step": 24080
    },
    {
      "epoch": 12.623165618448636,
      "grad_norm": 0.6227567195892334,
      "learning_rate": 1.8014154976859126e-05,
      "loss": 0.3096,
      "num_input_tokens_seen": 15723904,
      "step": 24085
    },
    {
      "epoch": 12.625786163522012,
      "grad_norm": 0.529623806476593,
      "learning_rate": 1.800317690818433e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 15727168,
      "step": 24090
    },
    {
      "epoch": 12.628406708595389,
      "grad_norm": 0.5349040031433105,
      "learning_rate": 1.799220030316511e-05,
      "loss": 0.6206,
      "num_input_tokens_seen": 15730688,
      "step": 24095
    },
    {
      "epoch": 12.631027253668764,
      "grad_norm": 0.40749162435531616,
      "learning_rate": 1.7981225164097633e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 15733664,
      "step": 24100
    },
    {
      "epoch": 12.633647798742139,
      "grad_norm": 0.3181414306163788,
      "learning_rate": 1.7970251493277774e-05,
      "loss": 0.5203,
      "num_input_tokens_seen": 15736352,
      "step": 24105
    },
    {
      "epoch": 12.636268343815514,
      "grad_norm": 0.46842682361602783,
      "learning_rate": 1.7959279293001102e-05,
      "loss": 0.3512,
      "num_input_tokens_seen": 15739648,
      "step": 24110
    },
    {
      "epoch": 12.63888888888889,
      "grad_norm": 0.5527225732803345,
      "learning_rate": 1.7948308565562876e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 15742464,
      "step": 24115
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 0.38912081718444824,
      "learning_rate": 1.793733931325805e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 15745984,
      "step": 24120
    },
    {
      "epoch": 12.64412997903564,
      "grad_norm": 0.5613442659378052,
      "learning_rate": 1.792637153838127e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 15748608,
      "step": 24125
    },
    {
      "epoch": 12.646750524109015,
      "grad_norm": 0.47603532671928406,
      "learning_rate": 1.7915405243226864e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 15751520,
      "step": 24130
    },
    {
      "epoch": 12.64937106918239,
      "grad_norm": 0.35369017720222473,
      "learning_rate": 1.7904440430088847e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 15754912,
      "step": 24135
    },
    {
      "epoch": 12.651991614255765,
      "grad_norm": 0.32724398374557495,
      "learning_rate": 1.7893477101260947e-05,
      "loss": 0.6197,
      "num_input_tokens_seen": 15758688,
      "step": 24140
    },
    {
      "epoch": 12.65461215932914,
      "grad_norm": 0.5159912109375,
      "learning_rate": 1.7882515259036558e-05,
      "loss": 0.3656,
      "num_input_tokens_seen": 15761248,
      "step": 24145
    },
    {
      "epoch": 12.657232704402515,
      "grad_norm": 0.468676894903183,
      "learning_rate": 1.787155490570878e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 15765088,
      "step": 24150
    },
    {
      "epoch": 12.65985324947589,
      "grad_norm": 0.658002495765686,
      "learning_rate": 1.7860596043570398e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 15768896,
      "step": 24155
    },
    {
      "epoch": 12.662473794549266,
      "grad_norm": 0.46523210406303406,
      "learning_rate": 1.7849638674913863e-05,
      "loss": 0.3892,
      "num_input_tokens_seen": 15771584,
      "step": 24160
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 0.5651682615280151,
      "learning_rate": 1.7838682802031344e-05,
      "loss": 0.5584,
      "num_input_tokens_seen": 15774080,
      "step": 24165
    },
    {
      "epoch": 12.667714884696016,
      "grad_norm": 0.3705746829509735,
      "learning_rate": 1.782772842721468e-05,
      "loss": 0.594,
      "num_input_tokens_seen": 15777280,
      "step": 24170
    },
    {
      "epoch": 12.670335429769391,
      "grad_norm": 0.5048642158508301,
      "learning_rate": 1.7816775552755406e-05,
      "loss": 0.5654,
      "num_input_tokens_seen": 15780544,
      "step": 24175
    },
    {
      "epoch": 12.672955974842766,
      "grad_norm": 0.7841337323188782,
      "learning_rate": 1.780582418094473e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 15783680,
      "step": 24180
    },
    {
      "epoch": 12.675576519916142,
      "grad_norm": 0.6657911539077759,
      "learning_rate": 1.7794874314073573e-05,
      "loss": 0.3985,
      "num_input_tokens_seen": 15786752,
      "step": 24185
    },
    {
      "epoch": 12.678197064989519,
      "grad_norm": 0.7456448674201965,
      "learning_rate": 1.7783925954432503e-05,
      "loss": 0.3818,
      "num_input_tokens_seen": 15790368,
      "step": 24190
    },
    {
      "epoch": 12.680817610062894,
      "grad_norm": 0.4154849052429199,
      "learning_rate": 1.777297910431179e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 15793120,
      "step": 24195
    },
    {
      "epoch": 12.683438155136269,
      "grad_norm": 0.5981258749961853,
      "learning_rate": 1.7762033766001402e-05,
      "loss": 0.3663,
      "num_input_tokens_seen": 15797600,
      "step": 24200
    },
    {
      "epoch": 12.686058700209644,
      "grad_norm": 0.35135647654533386,
      "learning_rate": 1.7751089941790977e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 15802304,
      "step": 24205
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 0.45717060565948486,
      "learning_rate": 1.7740147633969844e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 15804672,
      "step": 24210
    },
    {
      "epoch": 12.691299790356394,
      "grad_norm": 0.5522353649139404,
      "learning_rate": 1.772920684482699e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 15807168,
      "step": 24215
    },
    {
      "epoch": 12.69392033542977,
      "grad_norm": 0.6910194158554077,
      "learning_rate": 1.771826757665112e-05,
      "loss": 0.3267,
      "num_input_tokens_seen": 15809824,
      "step": 24220
    },
    {
      "epoch": 12.696540880503145,
      "grad_norm": 0.5108334422111511,
      "learning_rate": 1.770732983173059e-05,
      "loss": 0.5647,
      "num_input_tokens_seen": 15813408,
      "step": 24225
    },
    {
      "epoch": 12.69916142557652,
      "grad_norm": 0.5800829529762268,
      "learning_rate": 1.7696393612353467e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 15816096,
      "step": 24230
    },
    {
      "epoch": 12.701781970649895,
      "grad_norm": 0.33593878149986267,
      "learning_rate": 1.7685458920807474e-05,
      "loss": 0.3301,
      "num_input_tokens_seen": 15819328,
      "step": 24235
    },
    {
      "epoch": 12.70440251572327,
      "grad_norm": 1.1108622550964355,
      "learning_rate": 1.7674525759380033e-05,
      "loss": 0.4154,
      "num_input_tokens_seen": 15822176,
      "step": 24240
    },
    {
      "epoch": 12.707023060796645,
      "grad_norm": 0.539167582988739,
      "learning_rate": 1.766359413035823e-05,
      "loss": 0.649,
      "num_input_tokens_seen": 15824768,
      "step": 24245
    },
    {
      "epoch": 12.70964360587002,
      "grad_norm": 0.36335229873657227,
      "learning_rate": 1.7652664036028827e-05,
      "loss": 0.5378,
      "num_input_tokens_seen": 15828608,
      "step": 24250
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 0.7305319309234619,
      "learning_rate": 1.7641735478678288e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 15830880,
      "step": 24255
    },
    {
      "epoch": 12.714884696016771,
      "grad_norm": 0.4141513407230377,
      "learning_rate": 1.763080846059274e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 15834112,
      "step": 24260
    },
    {
      "epoch": 12.717505241090146,
      "grad_norm": 0.33686283230781555,
      "learning_rate": 1.7619882984057987e-05,
      "loss": 0.3325,
      "num_input_tokens_seen": 15836832,
      "step": 24265
    },
    {
      "epoch": 12.720125786163521,
      "grad_norm": 0.506130576133728,
      "learning_rate": 1.760895905135953e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 15839840,
      "step": 24270
    },
    {
      "epoch": 12.722746331236896,
      "grad_norm": 0.37776845693588257,
      "learning_rate": 1.7598036664782508e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 15842848,
      "step": 24275
    },
    {
      "epoch": 12.725366876310272,
      "grad_norm": 0.5777381062507629,
      "learning_rate": 1.7587115826611766e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 15846496,
      "step": 24280
    },
    {
      "epoch": 12.727987421383649,
      "grad_norm": 1.2266771793365479,
      "learning_rate": 1.7576196539131828e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 15852832,
      "step": 24285
    },
    {
      "epoch": 12.730607966457024,
      "grad_norm": 0.37167444825172424,
      "learning_rate": 1.7565278804626872e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 15856416,
      "step": 24290
    },
    {
      "epoch": 12.733228511530399,
      "grad_norm": 0.47226089239120483,
      "learning_rate": 1.755436262538078e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 15859104,
      "step": 24295
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 0.4274834990501404,
      "learning_rate": 1.7543448003677082e-05,
      "loss": 0.344,
      "num_input_tokens_seen": 15861792,
      "step": 24300
    },
    {
      "epoch": 12.73846960167715,
      "grad_norm": 0.3072866201400757,
      "learning_rate": 1.7532534941798985e-05,
      "loss": 0.5083,
      "num_input_tokens_seen": 15864320,
      "step": 24305
    },
    {
      "epoch": 12.741090146750524,
      "grad_norm": 0.5638278126716614,
      "learning_rate": 1.752162344202939e-05,
      "loss": 0.3065,
      "num_input_tokens_seen": 15867168,
      "step": 24310
    },
    {
      "epoch": 12.7437106918239,
      "grad_norm": 0.7127078175544739,
      "learning_rate": 1.751071350665085e-05,
      "loss": 0.418,
      "num_input_tokens_seen": 15870048,
      "step": 24315
    },
    {
      "epoch": 12.746331236897275,
      "grad_norm": 0.4929355978965759,
      "learning_rate": 1.7499805137945602e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 15873024,
      "step": 24320
    },
    {
      "epoch": 12.74895178197065,
      "grad_norm": 0.3952749967575073,
      "learning_rate": 1.7488898338195566e-05,
      "loss": 0.4518,
      "num_input_tokens_seen": 15875648,
      "step": 24325
    },
    {
      "epoch": 12.751572327044025,
      "grad_norm": 0.35362309217453003,
      "learning_rate": 1.7477993109682288e-05,
      "loss": 0.3348,
      "num_input_tokens_seen": 15878240,
      "step": 24330
    },
    {
      "epoch": 12.7541928721174,
      "grad_norm": 0.49629032611846924,
      "learning_rate": 1.7467089454687043e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 15881152,
      "step": 24335
    },
    {
      "epoch": 12.756813417190775,
      "grad_norm": 0.9924601316452026,
      "learning_rate": 1.745618737549074e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 15884384,
      "step": 24340
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 0.4626362919807434,
      "learning_rate": 1.7445286874373974e-05,
      "loss": 0.6055,
      "num_input_tokens_seen": 15887232,
      "step": 24345
    },
    {
      "epoch": 12.762054507337526,
      "grad_norm": 0.48701512813568115,
      "learning_rate": 1.743438795361701e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 15889344,
      "step": 24350
    },
    {
      "epoch": 12.764675052410901,
      "grad_norm": 0.5731306076049805,
      "learning_rate": 1.742349061549978e-05,
      "loss": 0.6389,
      "num_input_tokens_seen": 15892704,
      "step": 24355
    },
    {
      "epoch": 12.767295597484276,
      "grad_norm": 0.4454531669616699,
      "learning_rate": 1.741259486230186e-05,
      "loss": 0.6713,
      "num_input_tokens_seen": 15895840,
      "step": 24360
    },
    {
      "epoch": 12.769916142557651,
      "grad_norm": 0.488025426864624,
      "learning_rate": 1.7401700696302546e-05,
      "loss": 0.3982,
      "num_input_tokens_seen": 15898720,
      "step": 24365
    },
    {
      "epoch": 12.772536687631026,
      "grad_norm": 0.2860681414604187,
      "learning_rate": 1.739080811978075e-05,
      "loss": 0.3192,
      "num_input_tokens_seen": 15901408,
      "step": 24370
    },
    {
      "epoch": 12.775157232704402,
      "grad_norm": 0.7886847257614136,
      "learning_rate": 1.7379917135015093e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 15904160,
      "step": 24375
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 0.35697129368782043,
      "learning_rate": 1.736902774428384e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 15907488,
      "step": 24380
    },
    {
      "epoch": 12.780398322851154,
      "grad_norm": 0.3196607828140259,
      "learning_rate": 1.735813994986491e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 15910784,
      "step": 24385
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 0.43437260389328003,
      "learning_rate": 1.7347253754035933e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 15914016,
      "step": 24390
    },
    {
      "epoch": 12.785639412997904,
      "grad_norm": 0.2732278108596802,
      "learning_rate": 1.7336369159074154e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 15916672,
      "step": 24395
    },
    {
      "epoch": 12.78825995807128,
      "grad_norm": 0.4160202145576477,
      "learning_rate": 1.7325486167256522e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 15920288,
      "step": 24400
    },
    {
      "epoch": 12.790880503144654,
      "grad_norm": 0.36488375067710876,
      "learning_rate": 1.7314604780859623e-05,
      "loss": 0.5507,
      "num_input_tokens_seen": 15924736,
      "step": 24405
    },
    {
      "epoch": 12.79350104821803,
      "grad_norm": 0.4311424791812897,
      "learning_rate": 1.730372500215973e-05,
      "loss": 0.3475,
      "num_input_tokens_seen": 15928544,
      "step": 24410
    },
    {
      "epoch": 12.796121593291405,
      "grad_norm": 0.4405182898044586,
      "learning_rate": 1.7292846833432763e-05,
      "loss": 0.3148,
      "num_input_tokens_seen": 15932864,
      "step": 24415
    },
    {
      "epoch": 12.79874213836478,
      "grad_norm": 0.26425835490226746,
      "learning_rate": 1.7281970276954305e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 15936512,
      "step": 24420
    },
    {
      "epoch": 12.801362683438155,
      "grad_norm": 0.4314482510089874,
      "learning_rate": 1.727109533499961e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 15939392,
      "step": 24425
    },
    {
      "epoch": 12.80398322851153,
      "grad_norm": 0.4282304644584656,
      "learning_rate": 1.72602220098436e-05,
      "loss": 0.4074,
      "num_input_tokens_seen": 15942272,
      "step": 24430
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 0.6310799717903137,
      "learning_rate": 1.7249350303760846e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 15945280,
      "step": 24435
    },
    {
      "epoch": 12.80922431865828,
      "grad_norm": 0.9734005928039551,
      "learning_rate": 1.7238480219025593e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 15948704,
      "step": 24440
    },
    {
      "epoch": 12.811844863731656,
      "grad_norm": 0.3200117349624634,
      "learning_rate": 1.7227611757911723e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 15952320,
      "step": 24445
    },
    {
      "epoch": 12.814465408805031,
      "grad_norm": 0.42792508006095886,
      "learning_rate": 1.7216744922692794e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 15956576,
      "step": 24450
    },
    {
      "epoch": 12.817085953878406,
      "grad_norm": 0.31017711758613586,
      "learning_rate": 1.7205879715642037e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 15959232,
      "step": 24455
    },
    {
      "epoch": 12.819706498951781,
      "grad_norm": 0.31028345227241516,
      "learning_rate": 1.719501613903232e-05,
      "loss": 0.4017,
      "num_input_tokens_seen": 15962720,
      "step": 24460
    },
    {
      "epoch": 12.822327044025156,
      "grad_norm": 0.3722979724407196,
      "learning_rate": 1.7184154195136192e-05,
      "loss": 0.3801,
      "num_input_tokens_seen": 15966240,
      "step": 24465
    },
    {
      "epoch": 12.824947589098532,
      "grad_norm": 0.5010010004043579,
      "learning_rate": 1.7173293886225837e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 15968800,
      "step": 24470
    },
    {
      "epoch": 12.827568134171909,
      "grad_norm": 0.4230749011039734,
      "learning_rate": 1.71624352145731e-05,
      "loss": 0.4838,
      "num_input_tokens_seen": 15971680,
      "step": 24475
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 0.26994588971138,
      "learning_rate": 1.7151578182449508e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 15975552,
      "step": 24480
    },
    {
      "epoch": 12.832809224318659,
      "grad_norm": 0.6033501625061035,
      "learning_rate": 1.7140722792126212e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 15978656,
      "step": 24485
    },
    {
      "epoch": 12.835429769392034,
      "grad_norm": 0.5224382877349854,
      "learning_rate": 1.7129869045874053e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 15981856,
      "step": 24490
    },
    {
      "epoch": 12.83805031446541,
      "grad_norm": 0.912449300289154,
      "learning_rate": 1.7119016945963506e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 15984736,
      "step": 24495
    },
    {
      "epoch": 12.840670859538784,
      "grad_norm": 0.8330023884773254,
      "learning_rate": 1.7108166494664694e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 15987424,
      "step": 24500
    },
    {
      "epoch": 12.84329140461216,
      "grad_norm": 0.6009150147438049,
      "learning_rate": 1.7097317694247417e-05,
      "loss": 0.7015,
      "num_input_tokens_seen": 15991296,
      "step": 24505
    },
    {
      "epoch": 12.845911949685535,
      "grad_norm": 0.44275492429733276,
      "learning_rate": 1.7086470546981122e-05,
      "loss": 0.6209,
      "num_input_tokens_seen": 15993760,
      "step": 24510
    },
    {
      "epoch": 12.84853249475891,
      "grad_norm": 0.562827467918396,
      "learning_rate": 1.7075625055134904e-05,
      "loss": 0.3214,
      "num_input_tokens_seen": 15996288,
      "step": 24515
    },
    {
      "epoch": 12.851153039832285,
      "grad_norm": 0.44286468625068665,
      "learning_rate": 1.706478122097752e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 15999328,
      "step": 24520
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 0.5849445462226868,
      "learning_rate": 1.7053939046777384e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 16002336,
      "step": 24525
    },
    {
      "epoch": 12.856394129979035,
      "grad_norm": 0.4516699016094208,
      "learning_rate": 1.7043098534802533e-05,
      "loss": 0.3151,
      "num_input_tokens_seen": 16005440,
      "step": 24530
    },
    {
      "epoch": 12.85901467505241,
      "grad_norm": 0.45688751339912415,
      "learning_rate": 1.7032259687320695e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 16009376,
      "step": 24535
    },
    {
      "epoch": 12.861635220125786,
      "grad_norm": 0.46900489926338196,
      "learning_rate": 1.7021422506599228e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 16012768,
      "step": 24540
    },
    {
      "epoch": 12.864255765199161,
      "grad_norm": 0.381447970867157,
      "learning_rate": 1.7010586994905155e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 16015136,
      "step": 24545
    },
    {
      "epoch": 12.866876310272536,
      "grad_norm": 0.49661877751350403,
      "learning_rate": 1.6999753154505123e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 16018688,
      "step": 24550
    },
    {
      "epoch": 12.869496855345911,
      "grad_norm": 0.3827776312828064,
      "learning_rate": 1.6988920987665476e-05,
      "loss": 0.6529,
      "num_input_tokens_seen": 16021504,
      "step": 24555
    },
    {
      "epoch": 12.872117400419286,
      "grad_norm": 0.3896386921405792,
      "learning_rate": 1.6978090496652156e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 16024736,
      "step": 24560
    },
    {
      "epoch": 12.874737945492662,
      "grad_norm": 0.6928125023841858,
      "learning_rate": 1.6967261683730778e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 16027296,
      "step": 24565
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 0.2663097381591797,
      "learning_rate": 1.695643455116662e-05,
      "loss": 0.3522,
      "num_input_tokens_seen": 16029920,
      "step": 24570
    },
    {
      "epoch": 12.879979035639414,
      "grad_norm": 0.4855334758758545,
      "learning_rate": 1.694560910122458e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 16032672,
      "step": 24575
    },
    {
      "epoch": 12.882599580712789,
      "grad_norm": 1.1328405141830444,
      "learning_rate": 1.693478533616924e-05,
      "loss": 0.3335,
      "num_input_tokens_seen": 16035040,
      "step": 24580
    },
    {
      "epoch": 12.885220125786164,
      "grad_norm": 0.3716769218444824,
      "learning_rate": 1.692396325826479e-05,
      "loss": 0.3805,
      "num_input_tokens_seen": 16040064,
      "step": 24585
    },
    {
      "epoch": 12.88784067085954,
      "grad_norm": 0.36107710003852844,
      "learning_rate": 1.6913142869775088e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 16042784,
      "step": 24590
    },
    {
      "epoch": 12.890461215932914,
      "grad_norm": 0.28789761662483215,
      "learning_rate": 1.6902324172963635e-05,
      "loss": 0.3781,
      "num_input_tokens_seen": 16045952,
      "step": 24595
    },
    {
      "epoch": 12.89308176100629,
      "grad_norm": 0.6312145590782166,
      "learning_rate": 1.6891507170093586e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 16049248,
      "step": 24600
    },
    {
      "epoch": 12.895702306079665,
      "grad_norm": 0.4437845051288605,
      "learning_rate": 1.6880691863427725e-05,
      "loss": 0.503,
      "num_input_tokens_seen": 16052576,
      "step": 24605
    },
    {
      "epoch": 12.89832285115304,
      "grad_norm": 0.658717155456543,
      "learning_rate": 1.686987825522851e-05,
      "loss": 0.676,
      "num_input_tokens_seen": 16055936,
      "step": 24610
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 0.7346464991569519,
      "learning_rate": 1.6859066347758002e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 16058176,
      "step": 24615
    },
    {
      "epoch": 12.90356394129979,
      "grad_norm": 0.6078482270240784,
      "learning_rate": 1.6848256143277933e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 16061568,
      "step": 24620
    },
    {
      "epoch": 12.906184486373165,
      "grad_norm": 0.6066827178001404,
      "learning_rate": 1.683744764404968e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 16065536,
      "step": 24625
    },
    {
      "epoch": 12.90880503144654,
      "grad_norm": 0.36401885747909546,
      "learning_rate": 1.6826640852334257e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 16068192,
      "step": 24630
    },
    {
      "epoch": 12.911425576519916,
      "grad_norm": 0.7183124423027039,
      "learning_rate": 1.6815835770392324e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 16071008,
      "step": 24635
    },
    {
      "epoch": 12.914046121593291,
      "grad_norm": 0.35073697566986084,
      "learning_rate": 1.6805032400484184e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 16074688,
      "step": 24640
    },
    {
      "epoch": 12.916666666666666,
      "grad_norm": 0.6640740633010864,
      "learning_rate": 1.6794230744869767e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 16076928,
      "step": 24645
    },
    {
      "epoch": 12.919287211740041,
      "grad_norm": 0.6183056831359863,
      "learning_rate": 1.6783430805808658e-05,
      "loss": 0.5275,
      "num_input_tokens_seen": 16079904,
      "step": 24650
    },
    {
      "epoch": 12.921907756813416,
      "grad_norm": 0.23983658850193024,
      "learning_rate": 1.6772632585560084e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 16083168,
      "step": 24655
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 0.5965045094490051,
      "learning_rate": 1.6761836086382908e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 16086592,
      "step": 24660
    },
    {
      "epoch": 12.927148846960169,
      "grad_norm": 0.5166486501693726,
      "learning_rate": 1.6751041310535653e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 16089088,
      "step": 24665
    },
    {
      "epoch": 12.929769392033544,
      "grad_norm": 0.6076230406761169,
      "learning_rate": 1.6740248260276437e-05,
      "loss": 0.7862,
      "num_input_tokens_seen": 16091936,
      "step": 24670
    },
    {
      "epoch": 12.932389937106919,
      "grad_norm": 0.9542809724807739,
      "learning_rate": 1.6729456937863047e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 16094336,
      "step": 24675
    },
    {
      "epoch": 12.935010482180294,
      "grad_norm": 0.5198947787284851,
      "learning_rate": 1.6718667345552918e-05,
      "loss": 0.4287,
      "num_input_tokens_seen": 16098304,
      "step": 24680
    },
    {
      "epoch": 12.93763102725367,
      "grad_norm": 0.3680078387260437,
      "learning_rate": 1.6707879485603097e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 16101632,
      "step": 24685
    },
    {
      "epoch": 12.940251572327044,
      "grad_norm": 0.38716188073158264,
      "learning_rate": 1.669709336027029e-05,
      "loss": 0.4051,
      "num_input_tokens_seen": 16105024,
      "step": 24690
    },
    {
      "epoch": 12.94287211740042,
      "grad_norm": 0.3330269455909729,
      "learning_rate": 1.6686308971810833e-05,
      "loss": 0.3609,
      "num_input_tokens_seen": 16108096,
      "step": 24695
    },
    {
      "epoch": 12.945492662473795,
      "grad_norm": 0.45405736565589905,
      "learning_rate": 1.6675526322480677e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 16111648,
      "step": 24700
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 0.5118696689605713,
      "learning_rate": 1.6664745414535454e-05,
      "loss": 0.5015,
      "num_input_tokens_seen": 16114752,
      "step": 24705
    },
    {
      "epoch": 12.950733752620545,
      "grad_norm": 0.5541314482688904,
      "learning_rate": 1.665396625023039e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 16119616,
      "step": 24710
    },
    {
      "epoch": 12.95335429769392,
      "grad_norm": 0.4665435254573822,
      "learning_rate": 1.6643188831820375e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 16123232,
      "step": 24715
    },
    {
      "epoch": 12.955974842767295,
      "grad_norm": 0.31249526143074036,
      "learning_rate": 1.6632413161559912e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 16127552,
      "step": 24720
    },
    {
      "epoch": 12.95859538784067,
      "grad_norm": 0.5327035784721375,
      "learning_rate": 1.6621639241703162e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 16130368,
      "step": 24725
    },
    {
      "epoch": 12.961215932914046,
      "grad_norm": 0.6866206526756287,
      "learning_rate": 1.6610867074503893e-05,
      "loss": 0.402,
      "num_input_tokens_seen": 16134208,
      "step": 24730
    },
    {
      "epoch": 12.963836477987421,
      "grad_norm": 0.3520740866661072,
      "learning_rate": 1.660009666221552e-05,
      "loss": 0.4359,
      "num_input_tokens_seen": 16137120,
      "step": 24735
    },
    {
      "epoch": 12.966457023060796,
      "grad_norm": 0.37012624740600586,
      "learning_rate": 1.6589328007091093e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 16140576,
      "step": 24740
    },
    {
      "epoch": 12.969077568134171,
      "grad_norm": 0.8356857895851135,
      "learning_rate": 1.65785611113833e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 16143808,
      "step": 24745
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 0.45541030168533325,
      "learning_rate": 1.6567795977344447e-05,
      "loss": 0.528,
      "num_input_tokens_seen": 16147136,
      "step": 24750
    },
    {
      "epoch": 12.974318658280922,
      "grad_norm": 0.3446418046951294,
      "learning_rate": 1.6557032607226466e-05,
      "loss": 0.4226,
      "num_input_tokens_seen": 16150272,
      "step": 24755
    },
    {
      "epoch": 12.976939203354299,
      "grad_norm": 0.6657600402832031,
      "learning_rate": 1.654627100328095e-05,
      "loss": 0.5224,
      "num_input_tokens_seen": 16153920,
      "step": 24760
    },
    {
      "epoch": 12.979559748427674,
      "grad_norm": 0.28326788544654846,
      "learning_rate": 1.6535511167759085e-05,
      "loss": 0.367,
      "num_input_tokens_seen": 16157280,
      "step": 24765
    },
    {
      "epoch": 12.982180293501049,
      "grad_norm": 0.6506519317626953,
      "learning_rate": 1.652475310291172e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 16161056,
      "step": 24770
    },
    {
      "epoch": 12.984800838574424,
      "grad_norm": 0.4704277813434601,
      "learning_rate": 1.6513996810989317e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 16163776,
      "step": 24775
    },
    {
      "epoch": 12.9874213836478,
      "grad_norm": 0.5228295922279358,
      "learning_rate": 1.6503242294241973e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 16166176,
      "step": 24780
    },
    {
      "epoch": 12.990041928721174,
      "grad_norm": 0.4132659435272217,
      "learning_rate": 1.6492489554919404e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 16168704,
      "step": 24785
    },
    {
      "epoch": 12.99266247379455,
      "grad_norm": 0.42268478870391846,
      "learning_rate": 1.6481738595270953e-05,
      "loss": 0.4219,
      "num_input_tokens_seen": 16171584,
      "step": 24790
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 0.8967188000679016,
      "learning_rate": 1.6470989417545608e-05,
      "loss": 0.5487,
      "num_input_tokens_seen": 16174656,
      "step": 24795
    },
    {
      "epoch": 12.9979035639413,
      "grad_norm": 0.4042135179042816,
      "learning_rate": 1.646024202399197e-05,
      "loss": 0.6154,
      "num_input_tokens_seen": 16177344,
      "step": 24800
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.4534229338169098,
      "eval_runtime": 9.8141,
      "eval_samples_per_second": 86.406,
      "eval_steps_per_second": 21.601,
      "num_input_tokens_seen": 16179624,
      "step": 24804
    },
    {
      "epoch": 13.000524109014675,
      "grad_norm": 0.483938604593277,
      "learning_rate": 1.6449496416858284e-05,
      "loss": 0.5023,
      "num_input_tokens_seen": 16180040,
      "step": 24805
    },
    {
      "epoch": 13.00314465408805,
      "grad_norm": 0.7858051061630249,
      "learning_rate": 1.64387525983924e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 16183176,
      "step": 24810
    },
    {
      "epoch": 13.005765199161425,
      "grad_norm": 0.604370653629303,
      "learning_rate": 1.6428010570841788e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 16185928,
      "step": 24815
    },
    {
      "epoch": 13.0083857442348,
      "grad_norm": 0.3679589629173279,
      "learning_rate": 1.6417270336453573e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 16189160,
      "step": 24820
    },
    {
      "epoch": 13.011006289308176,
      "grad_norm": 0.3434964120388031,
      "learning_rate": 1.6406531897474492e-05,
      "loss": 0.3082,
      "num_input_tokens_seen": 16192392,
      "step": 24825
    },
    {
      "epoch": 13.01362683438155,
      "grad_norm": 0.31986331939697266,
      "learning_rate": 1.639579525615089e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 16195752,
      "step": 24830
    },
    {
      "epoch": 13.016247379454926,
      "grad_norm": 0.3516814708709717,
      "learning_rate": 1.638506041472877e-05,
      "loss": 0.3973,
      "num_input_tokens_seen": 16199176,
      "step": 24835
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 0.46920543909072876,
      "learning_rate": 1.6374327375453724e-05,
      "loss": 0.5647,
      "num_input_tokens_seen": 16201768,
      "step": 24840
    },
    {
      "epoch": 13.021488469601676,
      "grad_norm": 0.4778822660446167,
      "learning_rate": 1.6363596140570982e-05,
      "loss": 0.4077,
      "num_input_tokens_seen": 16204680,
      "step": 24845
    },
    {
      "epoch": 13.024109014675052,
      "grad_norm": 0.41058120131492615,
      "learning_rate": 1.6352866712325395e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 16208264,
      "step": 24850
    },
    {
      "epoch": 13.026729559748428,
      "grad_norm": 0.769529402256012,
      "learning_rate": 1.634213909296144e-05,
      "loss": 0.3759,
      "num_input_tokens_seen": 16210536,
      "step": 24855
    },
    {
      "epoch": 13.029350104821804,
      "grad_norm": 0.4597223699092865,
      "learning_rate": 1.6331413284723217e-05,
      "loss": 0.5701,
      "num_input_tokens_seen": 16214376,
      "step": 24860
    },
    {
      "epoch": 13.031970649895179,
      "grad_norm": 0.8276714086532593,
      "learning_rate": 1.6320689289854442e-05,
      "loss": 0.5593,
      "num_input_tokens_seen": 16217416,
      "step": 24865
    },
    {
      "epoch": 13.034591194968554,
      "grad_norm": 0.39989548921585083,
      "learning_rate": 1.6309967110598438e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 16220584,
      "step": 24870
    },
    {
      "epoch": 13.03721174004193,
      "grad_norm": 0.520903468132019,
      "learning_rate": 1.6299246749198183e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 16224040,
      "step": 24875
    },
    {
      "epoch": 13.039832285115304,
      "grad_norm": 0.9236353635787964,
      "learning_rate": 1.6288528207896236e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 16227016,
      "step": 24880
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 0.6970665454864502,
      "learning_rate": 1.627781148893481e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 16229768,
      "step": 24885
    },
    {
      "epoch": 13.045073375262055,
      "grad_norm": 0.2895418703556061,
      "learning_rate": 1.6267096594555703e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 16233224,
      "step": 24890
    },
    {
      "epoch": 13.04769392033543,
      "grad_norm": 0.5159406661987305,
      "learning_rate": 1.6256383527000372e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 16236264,
      "step": 24895
    },
    {
      "epoch": 13.050314465408805,
      "grad_norm": 0.304953396320343,
      "learning_rate": 1.6245672288509843e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 16238984,
      "step": 24900
    },
    {
      "epoch": 13.05293501048218,
      "grad_norm": 0.62250155210495,
      "learning_rate": 1.62349628813248e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 16241608,
      "step": 24905
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 0.580470085144043,
      "learning_rate": 1.6224255307685525e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 16244328,
      "step": 24910
    },
    {
      "epoch": 13.05817610062893,
      "grad_norm": 0.38348138332366943,
      "learning_rate": 1.6213549569831922e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 16247912,
      "step": 24915
    },
    {
      "epoch": 13.060796645702306,
      "grad_norm": 0.44520264863967896,
      "learning_rate": 1.62028456700035e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 16250920,
      "step": 24920
    },
    {
      "epoch": 13.06341719077568,
      "grad_norm": 0.4085645079612732,
      "learning_rate": 1.619214361043942e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 16254728,
      "step": 24925
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 0.4034603238105774,
      "learning_rate": 1.61814433933784e-05,
      "loss": 0.364,
      "num_input_tokens_seen": 16257640,
      "step": 24930
    },
    {
      "epoch": 13.068658280922431,
      "grad_norm": 0.8428234457969666,
      "learning_rate": 1.617074502105881e-05,
      "loss": 0.5298,
      "num_input_tokens_seen": 16260808,
      "step": 24935
    },
    {
      "epoch": 13.071278825995806,
      "grad_norm": 0.34938398003578186,
      "learning_rate": 1.616004849571864e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 16264264,
      "step": 24940
    },
    {
      "epoch": 13.073899371069182,
      "grad_norm": 0.27325138449668884,
      "learning_rate": 1.6149353819595463e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 16267528,
      "step": 24945
    },
    {
      "epoch": 13.076519916142558,
      "grad_norm": 0.5027593970298767,
      "learning_rate": 1.6138660994926508e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 16270088,
      "step": 24950
    },
    {
      "epoch": 13.079140461215934,
      "grad_norm": 0.6047419309616089,
      "learning_rate": 1.612797002394857e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 16273320,
      "step": 24955
    },
    {
      "epoch": 13.081761006289309,
      "grad_norm": 0.8206341862678528,
      "learning_rate": 1.611728090889808e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 16276232,
      "step": 24960
    },
    {
      "epoch": 13.084381551362684,
      "grad_norm": 0.5242393612861633,
      "learning_rate": 1.610659365201109e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 16279272,
      "step": 24965
    },
    {
      "epoch": 13.08700209643606,
      "grad_norm": 0.40081721544265747,
      "learning_rate": 1.609590825552324e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 16282024,
      "step": 24970
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 0.8257771134376526,
      "learning_rate": 1.6085224721669807e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 16284808,
      "step": 24975
    },
    {
      "epoch": 13.09224318658281,
      "grad_norm": 0.38392168283462524,
      "learning_rate": 1.607454305268566e-05,
      "loss": 0.6326,
      "num_input_tokens_seen": 16288200,
      "step": 24980
    },
    {
      "epoch": 13.094863731656185,
      "grad_norm": 0.2977435290813446,
      "learning_rate": 1.606386325080528e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 16291176,
      "step": 24985
    },
    {
      "epoch": 13.09748427672956,
      "grad_norm": 0.57732754945755,
      "learning_rate": 1.6053185318262746e-05,
      "loss": 0.5072,
      "num_input_tokens_seen": 16293896,
      "step": 24990
    },
    {
      "epoch": 13.100104821802935,
      "grad_norm": 0.5660539269447327,
      "learning_rate": 1.6042509257291783e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 16297192,
      "step": 24995
    },
    {
      "epoch": 13.10272536687631,
      "grad_norm": 0.35655367374420166,
      "learning_rate": 1.6031835070125685e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 16300680,
      "step": 25000
    },
    {
      "epoch": 13.105345911949685,
      "grad_norm": 0.7468746304512024,
      "learning_rate": 1.6021162758997382e-05,
      "loss": 0.253,
      "num_input_tokens_seen": 16303080,
      "step": 25005
    },
    {
      "epoch": 13.10796645702306,
      "grad_norm": 0.5868549942970276,
      "learning_rate": 1.6010492326139398e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 16305224,
      "step": 25010
    },
    {
      "epoch": 13.110587002096436,
      "grad_norm": 0.4886774718761444,
      "learning_rate": 1.599982377378385e-05,
      "loss": 0.5627,
      "num_input_tokens_seen": 16308168,
      "step": 25015
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 0.32925674319267273,
      "learning_rate": 1.5989157104162496e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 16312072,
      "step": 25020
    },
    {
      "epoch": 13.115828092243186,
      "grad_norm": 0.38404974341392517,
      "learning_rate": 1.5978492319506673e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 16315112,
      "step": 25025
    },
    {
      "epoch": 13.118448637316561,
      "grad_norm": 0.5211077332496643,
      "learning_rate": 1.596782942204733e-05,
      "loss": 0.5941,
      "num_input_tokens_seen": 16317704,
      "step": 25030
    },
    {
      "epoch": 13.121069182389936,
      "grad_norm": 0.5582291483879089,
      "learning_rate": 1.595716841401504e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 16321288,
      "step": 25035
    },
    {
      "epoch": 13.123689727463312,
      "grad_norm": 0.47015297412872314,
      "learning_rate": 1.594650929763994e-05,
      "loss": 0.3242,
      "num_input_tokens_seen": 16324360,
      "step": 25040
    },
    {
      "epoch": 13.126310272536688,
      "grad_norm": 0.2905695140361786,
      "learning_rate": 1.5935852075151812e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 16327784,
      "step": 25045
    },
    {
      "epoch": 13.128930817610064,
      "grad_norm": 1.0609148740768433,
      "learning_rate": 1.5925196748780016e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 16331848,
      "step": 25050
    },
    {
      "epoch": 13.131551362683439,
      "grad_norm": 0.6202066540718079,
      "learning_rate": 1.5914543320753528e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 16335592,
      "step": 25055
    },
    {
      "epoch": 13.134171907756814,
      "grad_norm": 0.2629740536212921,
      "learning_rate": 1.5903891793300932e-05,
      "loss": 0.5568,
      "num_input_tokens_seen": 16340200,
      "step": 25060
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 0.9414557814598083,
      "learning_rate": 1.5893242168650403e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 16343208,
      "step": 25065
    },
    {
      "epoch": 13.139412997903564,
      "grad_norm": 0.3517865240573883,
      "learning_rate": 1.5882594449029705e-05,
      "loss": 0.3522,
      "num_input_tokens_seen": 16346088,
      "step": 25070
    },
    {
      "epoch": 13.14203354297694,
      "grad_norm": 0.36063069105148315,
      "learning_rate": 1.587194863666624e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 16349768,
      "step": 25075
    },
    {
      "epoch": 13.144654088050315,
      "grad_norm": 0.5740429759025574,
      "learning_rate": 1.5861304733786974e-05,
      "loss": 0.4624,
      "num_input_tokens_seen": 16353960,
      "step": 25080
    },
    {
      "epoch": 13.14727463312369,
      "grad_norm": 0.4482402503490448,
      "learning_rate": 1.5850662742618505e-05,
      "loss": 0.5924,
      "num_input_tokens_seen": 16356904,
      "step": 25085
    },
    {
      "epoch": 13.149895178197065,
      "grad_norm": 0.4009535610675812,
      "learning_rate": 1.5840022665387004e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 16360040,
      "step": 25090
    },
    {
      "epoch": 13.15251572327044,
      "grad_norm": 1.9126839637756348,
      "learning_rate": 1.5829384504318268e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 16363496,
      "step": 25095
    },
    {
      "epoch": 13.155136268343815,
      "grad_norm": 0.7699492573738098,
      "learning_rate": 1.5818748261637662e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 16366376,
      "step": 25100
    },
    {
      "epoch": 13.15775681341719,
      "grad_norm": 0.3239366114139557,
      "learning_rate": 1.5808113939570173e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 16370184,
      "step": 25105
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 0.41080424189567566,
      "learning_rate": 1.5797481540340386e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 16374152,
      "step": 25110
    },
    {
      "epoch": 13.16299790356394,
      "grad_norm": 0.31809189915657043,
      "learning_rate": 1.5786851066172466e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 16376904,
      "step": 25115
    },
    {
      "epoch": 13.165618448637316,
      "grad_norm": 0.38795459270477295,
      "learning_rate": 1.5776222519290208e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 16379784,
      "step": 25120
    },
    {
      "epoch": 13.168238993710691,
      "grad_norm": 0.45472481846809387,
      "learning_rate": 1.576559590191696e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 16382888,
      "step": 25125
    },
    {
      "epoch": 13.170859538784066,
      "grad_norm": 0.6129626035690308,
      "learning_rate": 1.5754971216275692e-05,
      "loss": 0.3821,
      "num_input_tokens_seen": 16386472,
      "step": 25130
    },
    {
      "epoch": 13.173480083857442,
      "grad_norm": 1.0145490169525146,
      "learning_rate": 1.5744348464588975e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 16389608,
      "step": 25135
    },
    {
      "epoch": 13.176100628930818,
      "grad_norm": 0.5273874998092651,
      "learning_rate": 1.5733727649078968e-05,
      "loss": 0.3574,
      "num_input_tokens_seen": 16393032,
      "step": 25140
    },
    {
      "epoch": 13.178721174004194,
      "grad_norm": 0.2741822302341461,
      "learning_rate": 1.5723108771967414e-05,
      "loss": 0.3602,
      "num_input_tokens_seen": 16396712,
      "step": 25145
    },
    {
      "epoch": 13.181341719077569,
      "grad_norm": 1.601509928703308,
      "learning_rate": 1.5712491835475685e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 16399272,
      "step": 25150
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 0.4577283263206482,
      "learning_rate": 1.5701876841824698e-05,
      "loss": 0.3848,
      "num_input_tokens_seen": 16402088,
      "step": 25155
    },
    {
      "epoch": 13.18658280922432,
      "grad_norm": 0.38105347752571106,
      "learning_rate": 1.5691263793234993e-05,
      "loss": 0.6184,
      "num_input_tokens_seen": 16406600,
      "step": 25160
    },
    {
      "epoch": 13.189203354297694,
      "grad_norm": 0.40941283106803894,
      "learning_rate": 1.5680652691926706e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 16409224,
      "step": 25165
    },
    {
      "epoch": 13.19182389937107,
      "grad_norm": 0.5194299817085266,
      "learning_rate": 1.5670043540119553e-05,
      "loss": 0.7462,
      "num_input_tokens_seen": 16412136,
      "step": 25170
    },
    {
      "epoch": 13.194444444444445,
      "grad_norm": 0.4068811535835266,
      "learning_rate": 1.5659436340032855e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 16416392,
      "step": 25175
    },
    {
      "epoch": 13.19706498951782,
      "grad_norm": 0.430645614862442,
      "learning_rate": 1.5648831093885515e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 16419688,
      "step": 25180
    },
    {
      "epoch": 13.199685534591195,
      "grad_norm": 0.4042915403842926,
      "learning_rate": 1.5638227803896015e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 16422856,
      "step": 25185
    },
    {
      "epoch": 13.20230607966457,
      "grad_norm": 0.4205605983734131,
      "learning_rate": 1.562762647228246e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 16425608,
      "step": 25190
    },
    {
      "epoch": 13.204926624737945,
      "grad_norm": 0.5099202394485474,
      "learning_rate": 1.5617027101262515e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 16429288,
      "step": 25195
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 0.41014155745506287,
      "learning_rate": 1.5606429693053463e-05,
      "loss": 0.5016,
      "num_input_tokens_seen": 16432744,
      "step": 25200
    },
    {
      "epoch": 13.210167714884696,
      "grad_norm": 0.2659021019935608,
      "learning_rate": 1.5595834249872145e-05,
      "loss": 0.3221,
      "num_input_tokens_seen": 16435944,
      "step": 25205
    },
    {
      "epoch": 13.21278825995807,
      "grad_norm": 0.41320085525512695,
      "learning_rate": 1.558524077393502e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 16438760,
      "step": 25210
    },
    {
      "epoch": 13.215408805031446,
      "grad_norm": 0.3968907296657562,
      "learning_rate": 1.557464926745811e-05,
      "loss": 0.372,
      "num_input_tokens_seen": 16442280,
      "step": 25215
    },
    {
      "epoch": 13.218029350104821,
      "grad_norm": 0.8959255218505859,
      "learning_rate": 1.5564059732657045e-05,
      "loss": 0.3696,
      "num_input_tokens_seen": 16445384,
      "step": 25220
    },
    {
      "epoch": 13.220649895178196,
      "grad_norm": 0.5393037796020508,
      "learning_rate": 1.555347217174703e-05,
      "loss": 0.2971,
      "num_input_tokens_seen": 16448040,
      "step": 25225
    },
    {
      "epoch": 13.223270440251572,
      "grad_norm": 0.6175490617752075,
      "learning_rate": 1.554288658694287e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 16451432,
      "step": 25230
    },
    {
      "epoch": 13.225890985324948,
      "grad_norm": 0.3725316822528839,
      "learning_rate": 1.553230298045895e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 16455624,
      "step": 25235
    },
    {
      "epoch": 13.228511530398324,
      "grad_norm": 0.5684233903884888,
      "learning_rate": 1.5521721354509223e-05,
      "loss": 0.3451,
      "num_input_tokens_seen": 16459112,
      "step": 25240
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 0.35831740498542786,
      "learning_rate": 1.5511141711307258e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 16462312,
      "step": 25245
    },
    {
      "epoch": 13.233752620545074,
      "grad_norm": 1.0799431800842285,
      "learning_rate": 1.5500564053066184e-05,
      "loss": 0.6229,
      "num_input_tokens_seen": 16465128,
      "step": 25250
    },
    {
      "epoch": 13.23637316561845,
      "grad_norm": 0.27502527832984924,
      "learning_rate": 1.5489988381998743e-05,
      "loss": 0.3995,
      "num_input_tokens_seen": 16469320,
      "step": 25255
    },
    {
      "epoch": 13.238993710691824,
      "grad_norm": 0.471162348985672,
      "learning_rate": 1.5479414700317237e-05,
      "loss": 0.5753,
      "num_input_tokens_seen": 16472296,
      "step": 25260
    },
    {
      "epoch": 13.2416142557652,
      "grad_norm": 0.41983357071876526,
      "learning_rate": 1.546884301023356e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 16475240,
      "step": 25265
    },
    {
      "epoch": 13.244234800838575,
      "grad_norm": 0.8169326782226562,
      "learning_rate": 1.5458273313959186e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 16477736,
      "step": 25270
    },
    {
      "epoch": 13.24685534591195,
      "grad_norm": 0.5860379338264465,
      "learning_rate": 1.5447705613705177e-05,
      "loss": 0.4951,
      "num_input_tokens_seen": 16480776,
      "step": 25275
    },
    {
      "epoch": 13.249475890985325,
      "grad_norm": 0.737739622592926,
      "learning_rate": 1.5437139911682176e-05,
      "loss": 0.4172,
      "num_input_tokens_seen": 16484584,
      "step": 25280
    },
    {
      "epoch": 13.2520964360587,
      "grad_norm": 0.5632511973381042,
      "learning_rate": 1.54265762101004e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 16487400,
      "step": 25285
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 0.45749908685684204,
      "learning_rate": 1.5416014511169663e-05,
      "loss": 0.3981,
      "num_input_tokens_seen": 16490024,
      "step": 25290
    },
    {
      "epoch": 13.25733752620545,
      "grad_norm": 0.3756489157676697,
      "learning_rate": 1.5405454817099363e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 16493064,
      "step": 25295
    },
    {
      "epoch": 13.259958071278826,
      "grad_norm": 0.6566864848136902,
      "learning_rate": 1.5394897130098446e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 16496136,
      "step": 25300
    },
    {
      "epoch": 13.2625786163522,
      "grad_norm": 0.3668692111968994,
      "learning_rate": 1.5384341452375465e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 16500328,
      "step": 25305
    },
    {
      "epoch": 13.265199161425576,
      "grad_norm": 0.3973047435283661,
      "learning_rate": 1.5373787786138556e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 16503688,
      "step": 25310
    },
    {
      "epoch": 13.267819706498951,
      "grad_norm": 1.5061132907867432,
      "learning_rate": 1.5363236133595416e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 16507112,
      "step": 25315
    },
    {
      "epoch": 13.270440251572326,
      "grad_norm": 0.5650754570960999,
      "learning_rate": 1.5352686496953344e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 16510920,
      "step": 25320
    },
    {
      "epoch": 13.273060796645701,
      "grad_norm": 0.4709383249282837,
      "learning_rate": 1.534213887841919e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 16513736,
      "step": 25325
    },
    {
      "epoch": 13.275681341719078,
      "grad_norm": 0.31347763538360596,
      "learning_rate": 1.5331593280199395e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 16516744,
      "step": 25330
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 0.548573911190033,
      "learning_rate": 1.532104970449999e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 16519976,
      "step": 25335
    },
    {
      "epoch": 13.280922431865829,
      "grad_norm": 0.3745838403701782,
      "learning_rate": 1.5310508153526555e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 16523112,
      "step": 25340
    },
    {
      "epoch": 13.283542976939204,
      "grad_norm": 0.6844411492347717,
      "learning_rate": 1.5299968629484276e-05,
      "loss": 0.5402,
      "num_input_tokens_seen": 16526376,
      "step": 25345
    },
    {
      "epoch": 13.286163522012579,
      "grad_norm": 0.4633665978908539,
      "learning_rate": 1.5289431134577904e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 16533160,
      "step": 25350
    },
    {
      "epoch": 13.288784067085954,
      "grad_norm": 0.5243653655052185,
      "learning_rate": 1.527889567101174e-05,
      "loss": 0.3803,
      "num_input_tokens_seen": 16536680,
      "step": 25355
    },
    {
      "epoch": 13.29140461215933,
      "grad_norm": 0.5236226320266724,
      "learning_rate": 1.5268362240989707e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 16539656,
      "step": 25360
    },
    {
      "epoch": 13.294025157232705,
      "grad_norm": 0.5986313819885254,
      "learning_rate": 1.525783084671526e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 16542376,
      "step": 25365
    },
    {
      "epoch": 13.29664570230608,
      "grad_norm": 0.7087909579277039,
      "learning_rate": 1.5247301490391458e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 16545352,
      "step": 25370
    },
    {
      "epoch": 13.299266247379455,
      "grad_norm": 0.5612697601318359,
      "learning_rate": 1.5236774174220922e-05,
      "loss": 0.5138,
      "num_input_tokens_seen": 16548424,
      "step": 25375
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 0.42989253997802734,
      "learning_rate": 1.5226248900405852e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 16551592,
      "step": 25380
    },
    {
      "epoch": 13.304507337526205,
      "grad_norm": 0.3752943277359009,
      "learning_rate": 1.5215725671147996e-05,
      "loss": 0.3439,
      "num_input_tokens_seen": 16554568,
      "step": 25385
    },
    {
      "epoch": 13.30712788259958,
      "grad_norm": 0.38801825046539307,
      "learning_rate": 1.5205204488648709e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 16557416,
      "step": 25390
    },
    {
      "epoch": 13.309748427672956,
      "grad_norm": 0.9894915819168091,
      "learning_rate": 1.5194685355108893e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 16560424,
      "step": 25395
    },
    {
      "epoch": 13.31236897274633,
      "grad_norm": 0.25859731435775757,
      "learning_rate": 1.518416827272904e-05,
      "loss": 0.3792,
      "num_input_tokens_seen": 16563880,
      "step": 25400
    },
    {
      "epoch": 13.314989517819706,
      "grad_norm": 0.7203106880187988,
      "learning_rate": 1.5173653243709207e-05,
      "loss": 0.3717,
      "num_input_tokens_seen": 16567432,
      "step": 25405
    },
    {
      "epoch": 13.317610062893081,
      "grad_norm": 0.49173927307128906,
      "learning_rate": 1.5163140270249e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 16570888,
      "step": 25410
    },
    {
      "epoch": 13.320230607966456,
      "grad_norm": 0.9570600986480713,
      "learning_rate": 1.5152629354547626e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 16573384,
      "step": 25415
    },
    {
      "epoch": 13.322851153039831,
      "grad_norm": 0.5963931679725647,
      "learning_rate": 1.5142120498803842e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 16576584,
      "step": 25420
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 0.5372403264045715,
      "learning_rate": 1.5131613705215991e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 16580584,
      "step": 25425
    },
    {
      "epoch": 13.328092243186584,
      "grad_norm": 0.5910112261772156,
      "learning_rate": 1.5121108975981965e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 16583848,
      "step": 25430
    },
    {
      "epoch": 13.330712788259959,
      "grad_norm": 0.7025021910667419,
      "learning_rate": 1.5110606313299248e-05,
      "loss": 0.3344,
      "num_input_tokens_seen": 16587272,
      "step": 25435
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.5552392601966858,
      "learning_rate": 1.5100105719364855e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 16590088,
      "step": 25440
    },
    {
      "epoch": 13.335953878406709,
      "grad_norm": 1.0433907508850098,
      "learning_rate": 1.50896071963754e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 16593320,
      "step": 25445
    },
    {
      "epoch": 13.338574423480084,
      "grad_norm": 0.4425164461135864,
      "learning_rate": 1.5079110746527055e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 16596328,
      "step": 25450
    },
    {
      "epoch": 13.34119496855346,
      "grad_norm": 0.5553991198539734,
      "learning_rate": 1.5068616372015565e-05,
      "loss": 0.6044,
      "num_input_tokens_seen": 16599080,
      "step": 25455
    },
    {
      "epoch": 13.343815513626835,
      "grad_norm": 0.40042734146118164,
      "learning_rate": 1.5058124075036223e-05,
      "loss": 0.4008,
      "num_input_tokens_seen": 16602056,
      "step": 25460
    },
    {
      "epoch": 13.34643605870021,
      "grad_norm": 0.39006710052490234,
      "learning_rate": 1.5047633857783908e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 16606216,
      "step": 25465
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 0.44261133670806885,
      "learning_rate": 1.5037145722453044e-05,
      "loss": 0.3883,
      "num_input_tokens_seen": 16609064,
      "step": 25470
    },
    {
      "epoch": 13.35167714884696,
      "grad_norm": 0.7620410919189453,
      "learning_rate": 1.5026659671237624e-05,
      "loss": 0.5441,
      "num_input_tokens_seen": 16611560,
      "step": 25475
    },
    {
      "epoch": 13.354297693920335,
      "grad_norm": 0.32613417506217957,
      "learning_rate": 1.501617570633123e-05,
      "loss": 0.3691,
      "num_input_tokens_seen": 16614408,
      "step": 25480
    },
    {
      "epoch": 13.35691823899371,
      "grad_norm": 0.4532809555530548,
      "learning_rate": 1.500569382992697e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 16617256,
      "step": 25485
    },
    {
      "epoch": 13.359538784067086,
      "grad_norm": 0.37178900837898254,
      "learning_rate": 1.4995214044217547e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 16620232,
      "step": 25490
    },
    {
      "epoch": 13.36215932914046,
      "grad_norm": 0.5775800943374634,
      "learning_rate": 1.49847363513952e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 16623912,
      "step": 25495
    },
    {
      "epoch": 13.364779874213836,
      "grad_norm": 0.8015732765197754,
      "learning_rate": 1.4974260753651748e-05,
      "loss": 0.564,
      "num_input_tokens_seen": 16626440,
      "step": 25500
    },
    {
      "epoch": 13.367400419287211,
      "grad_norm": 0.7757188677787781,
      "learning_rate": 1.4963787253178568e-05,
      "loss": 0.5931,
      "num_input_tokens_seen": 16630440,
      "step": 25505
    },
    {
      "epoch": 13.370020964360586,
      "grad_norm": 0.40441930294036865,
      "learning_rate": 1.4953315852166594e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 16633352,
      "step": 25510
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 1.031402826309204,
      "learning_rate": 1.4942846552806322e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 16636392,
      "step": 25515
    },
    {
      "epoch": 13.375262054507338,
      "grad_norm": 0.6509634852409363,
      "learning_rate": 1.4932379357287829e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 16639688,
      "step": 25520
    },
    {
      "epoch": 13.377882599580714,
      "grad_norm": 0.47461697459220886,
      "learning_rate": 1.49219142678007e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 16643176,
      "step": 25525
    },
    {
      "epoch": 13.380503144654089,
      "grad_norm": 0.44728195667266846,
      "learning_rate": 1.4911451286534128e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 16646952,
      "step": 25530
    },
    {
      "epoch": 13.383123689727464,
      "grad_norm": 0.46908074617385864,
      "learning_rate": 1.490099041567686e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 16650792,
      "step": 25535
    },
    {
      "epoch": 13.385744234800839,
      "grad_norm": 0.4406309127807617,
      "learning_rate": 1.4890531657417176e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 16654568,
      "step": 25540
    },
    {
      "epoch": 13.388364779874214,
      "grad_norm": 0.6562884449958801,
      "learning_rate": 1.4880075013942938e-05,
      "loss": 0.5262,
      "num_input_tokens_seen": 16657608,
      "step": 25545
    },
    {
      "epoch": 13.39098532494759,
      "grad_norm": 0.4651445746421814,
      "learning_rate": 1.4869620487441558e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 16660360,
      "step": 25550
    },
    {
      "epoch": 13.393605870020965,
      "grad_norm": 0.3854161202907562,
      "learning_rate": 1.4859168080099988e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 16663048,
      "step": 25555
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 0.5518003702163696,
      "learning_rate": 1.484871779410477e-05,
      "loss": 0.3866,
      "num_input_tokens_seen": 16667080,
      "step": 25560
    },
    {
      "epoch": 13.398846960167715,
      "grad_norm": 0.7494410872459412,
      "learning_rate": 1.4838269631641971e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 16669608,
      "step": 25565
    },
    {
      "epoch": 13.40146750524109,
      "grad_norm": 0.26908573508262634,
      "learning_rate": 1.482782359489724e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 16676968,
      "step": 25570
    },
    {
      "epoch": 13.404088050314465,
      "grad_norm": 0.3627142608165741,
      "learning_rate": 1.4817379686055754e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 16680552,
      "step": 25575
    },
    {
      "epoch": 13.40670859538784,
      "grad_norm": 0.6105518937110901,
      "learning_rate": 1.4806937907302282e-05,
      "loss": 0.3957,
      "num_input_tokens_seen": 16683880,
      "step": 25580
    },
    {
      "epoch": 13.409329140461216,
      "grad_norm": 0.6245530247688293,
      "learning_rate": 1.4796498260821103e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 16686312,
      "step": 25585
    },
    {
      "epoch": 13.41194968553459,
      "grad_norm": 0.5335291624069214,
      "learning_rate": 1.4786060748796076e-05,
      "loss": 0.4539,
      "num_input_tokens_seen": 16688776,
      "step": 25590
    },
    {
      "epoch": 13.414570230607966,
      "grad_norm": 0.5355052947998047,
      "learning_rate": 1.4775625373410617e-05,
      "loss": 0.3213,
      "num_input_tokens_seen": 16692104,
      "step": 25595
    },
    {
      "epoch": 13.417190775681341,
      "grad_norm": 0.43517765402793884,
      "learning_rate": 1.4765192136847685e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 16694600,
      "step": 25600
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 0.505400538444519,
      "learning_rate": 1.47547610412898e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 16697384,
      "step": 25605
    },
    {
      "epoch": 13.422431865828091,
      "grad_norm": 0.5200650691986084,
      "learning_rate": 1.4744332088919011e-05,
      "loss": 0.4529,
      "num_input_tokens_seen": 16700776,
      "step": 25610
    },
    {
      "epoch": 13.425052410901468,
      "grad_norm": 0.42882663011550903,
      "learning_rate": 1.4733905281916954e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 16703272,
      "step": 25615
    },
    {
      "epoch": 13.427672955974844,
      "grad_norm": 0.3342805504798889,
      "learning_rate": 1.472348062246478e-05,
      "loss": 0.3374,
      "num_input_tokens_seen": 16706440,
      "step": 25620
    },
    {
      "epoch": 13.430293501048219,
      "grad_norm": 0.44186899065971375,
      "learning_rate": 1.4713058112743228e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 16709256,
      "step": 25625
    },
    {
      "epoch": 13.432914046121594,
      "grad_norm": 0.3382691740989685,
      "learning_rate": 1.4702637754932553e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 16712712,
      "step": 25630
    },
    {
      "epoch": 13.435534591194969,
      "grad_norm": 0.36530977487564087,
      "learning_rate": 1.4692219551212587e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 16715464,
      "step": 25635
    },
    {
      "epoch": 13.438155136268344,
      "grad_norm": 0.7327429056167603,
      "learning_rate": 1.4681803503762695e-05,
      "loss": 0.6308,
      "num_input_tokens_seen": 16719304,
      "step": 25640
    },
    {
      "epoch": 13.44077568134172,
      "grad_norm": 0.43512916564941406,
      "learning_rate": 1.4671389614761788e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 16722760,
      "step": 25645
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 0.5260469317436218,
      "learning_rate": 1.4660977886388339e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 16726152,
      "step": 25650
    },
    {
      "epoch": 13.44601677148847,
      "grad_norm": 0.8231216073036194,
      "learning_rate": 1.4650568320820362e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 16728808,
      "step": 25655
    },
    {
      "epoch": 13.448637316561845,
      "grad_norm": 0.6755306124687195,
      "learning_rate": 1.4640160920235423e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 16731240,
      "step": 25660
    },
    {
      "epoch": 13.45125786163522,
      "grad_norm": 0.4273129999637604,
      "learning_rate": 1.4629755686810631e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 16735112,
      "step": 25665
    },
    {
      "epoch": 13.453878406708595,
      "grad_norm": 0.4108061194419861,
      "learning_rate": 1.4619352622722632e-05,
      "loss": 0.4524,
      "num_input_tokens_seen": 16738568,
      "step": 25670
    },
    {
      "epoch": 13.45649895178197,
      "grad_norm": 0.7033745050430298,
      "learning_rate": 1.4608951730147635e-05,
      "loss": 0.42,
      "num_input_tokens_seen": 16742056,
      "step": 25675
    },
    {
      "epoch": 13.459119496855346,
      "grad_norm": 0.7011315226554871,
      "learning_rate": 1.4598553011261395e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 16744616,
      "step": 25680
    },
    {
      "epoch": 13.46174004192872,
      "grad_norm": 0.794048011302948,
      "learning_rate": 1.4588156468239192e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 16747464,
      "step": 25685
    },
    {
      "epoch": 13.464360587002096,
      "grad_norm": 0.28980183601379395,
      "learning_rate": 1.4577762103255882e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 16752104,
      "step": 25690
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 0.2253323644399643,
      "learning_rate": 1.4567369918485828e-05,
      "loss": 0.3518,
      "num_input_tokens_seen": 16755720,
      "step": 25695
    },
    {
      "epoch": 13.469601677148846,
      "grad_norm": 0.4151504635810852,
      "learning_rate": 1.4556979916102976e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 16758600,
      "step": 25700
    },
    {
      "epoch": 13.472222222222221,
      "grad_norm": 0.8128553032875061,
      "learning_rate": 1.454659209828077e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 16761352,
      "step": 25705
    },
    {
      "epoch": 13.474842767295598,
      "grad_norm": 0.26873043179512024,
      "learning_rate": 1.4536206467192243e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 16765512,
      "step": 25710
    },
    {
      "epoch": 13.477463312368974,
      "grad_norm": 0.6307510137557983,
      "learning_rate": 1.4525823025009944e-05,
      "loss": 0.5612,
      "num_input_tokens_seen": 16769064,
      "step": 25715
    },
    {
      "epoch": 13.480083857442349,
      "grad_norm": 0.4749489426612854,
      "learning_rate": 1.4515441773905983e-05,
      "loss": 0.5415,
      "num_input_tokens_seen": 16772232,
      "step": 25720
    },
    {
      "epoch": 13.482704402515724,
      "grad_norm": 0.45767760276794434,
      "learning_rate": 1.4505062716051988e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 16776616,
      "step": 25725
    },
    {
      "epoch": 13.485324947589099,
      "grad_norm": 0.5208693742752075,
      "learning_rate": 1.4494685853619133e-05,
      "loss": 0.4359,
      "num_input_tokens_seen": 16779496,
      "step": 25730
    },
    {
      "epoch": 13.487945492662474,
      "grad_norm": 0.46437889337539673,
      "learning_rate": 1.4484311188778143e-05,
      "loss": 0.3537,
      "num_input_tokens_seen": 16783368,
      "step": 25735
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 0.5413416624069214,
      "learning_rate": 1.4473938723699285e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 16785928,
      "step": 25740
    },
    {
      "epoch": 13.493186582809225,
      "grad_norm": 0.7054838538169861,
      "learning_rate": 1.4463568460552368e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 16789320,
      "step": 25745
    },
    {
      "epoch": 13.4958071278826,
      "grad_norm": 0.4816001057624817,
      "learning_rate": 1.4453200401506723e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 16793928,
      "step": 25750
    },
    {
      "epoch": 13.498427672955975,
      "grad_norm": 0.9227237105369568,
      "learning_rate": 1.4442834548731218e-05,
      "loss": 0.4902,
      "num_input_tokens_seen": 16796872,
      "step": 25755
    },
    {
      "epoch": 13.50104821802935,
      "grad_norm": 0.40334010124206543,
      "learning_rate": 1.4432470904394285e-05,
      "loss": 0.3812,
      "num_input_tokens_seen": 16800552,
      "step": 25760
    },
    {
      "epoch": 13.503668763102725,
      "grad_norm": 0.5650759339332581,
      "learning_rate": 1.4422109470663878e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 16803880,
      "step": 25765
    },
    {
      "epoch": 13.5062893081761,
      "grad_norm": 0.5992008447647095,
      "learning_rate": 1.441175024970749e-05,
      "loss": 0.4172,
      "num_input_tokens_seen": 16806600,
      "step": 25770
    },
    {
      "epoch": 13.508909853249476,
      "grad_norm": 0.43460336327552795,
      "learning_rate": 1.4401393243692163e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 16809576,
      "step": 25775
    },
    {
      "epoch": 13.51153039832285,
      "grad_norm": 0.48552176356315613,
      "learning_rate": 1.4391038454784455e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 16812776,
      "step": 25780
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 0.5956622362136841,
      "learning_rate": 1.4380685885150458e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 16815496,
      "step": 25785
    },
    {
      "epoch": 13.516771488469601,
      "grad_norm": 0.7759331464767456,
      "learning_rate": 1.437033553695582e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 16818984,
      "step": 25790
    },
    {
      "epoch": 13.519392033542976,
      "grad_norm": 0.40525999665260315,
      "learning_rate": 1.4359987412365721e-05,
      "loss": 0.3191,
      "num_input_tokens_seen": 16825096,
      "step": 25795
    },
    {
      "epoch": 13.522012578616351,
      "grad_norm": 0.7211340069770813,
      "learning_rate": 1.4349641513544876e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 16827880,
      "step": 25800
    },
    {
      "epoch": 13.524633123689728,
      "grad_norm": 0.5162200927734375,
      "learning_rate": 1.4339297842657523e-05,
      "loss": 0.5069,
      "num_input_tokens_seen": 16830504,
      "step": 25805
    },
    {
      "epoch": 13.527253668763104,
      "grad_norm": 0.7822476625442505,
      "learning_rate": 1.4328956401867427e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 16833128,
      "step": 25810
    },
    {
      "epoch": 13.529874213836479,
      "grad_norm": 0.47250664234161377,
      "learning_rate": 1.4318617193337908e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 16835592,
      "step": 25815
    },
    {
      "epoch": 13.532494758909854,
      "grad_norm": 0.7820215225219727,
      "learning_rate": 1.4308280219231812e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 16839752,
      "step": 25820
    },
    {
      "epoch": 13.535115303983229,
      "grad_norm": 0.4024304747581482,
      "learning_rate": 1.4297945481711527e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 16842568,
      "step": 25825
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 0.8130342960357666,
      "learning_rate": 1.4287612982938946e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 16845320,
      "step": 25830
    },
    {
      "epoch": 13.54035639412998,
      "grad_norm": 0.6680163145065308,
      "learning_rate": 1.4277282725075523e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 16848552,
      "step": 25835
    },
    {
      "epoch": 13.542976939203355,
      "grad_norm": 0.4107411503791809,
      "learning_rate": 1.426695471028221e-05,
      "loss": 0.5496,
      "num_input_tokens_seen": 16853608,
      "step": 25840
    },
    {
      "epoch": 13.54559748427673,
      "grad_norm": 0.41854676604270935,
      "learning_rate": 1.4256628940719524e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 16856648,
      "step": 25845
    },
    {
      "epoch": 13.548218029350105,
      "grad_norm": 0.3076823055744171,
      "learning_rate": 1.42463054185475e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 16861864,
      "step": 25850
    },
    {
      "epoch": 13.55083857442348,
      "grad_norm": 0.6141853332519531,
      "learning_rate": 1.4235984145925705e-05,
      "loss": 0.4167,
      "num_input_tokens_seen": 16865256,
      "step": 25855
    },
    {
      "epoch": 13.553459119496855,
      "grad_norm": 0.4320535659790039,
      "learning_rate": 1.4225665125013224e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 16868936,
      "step": 25860
    },
    {
      "epoch": 13.55607966457023,
      "grad_norm": 0.6503134965896606,
      "learning_rate": 1.4215348357968669e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 16872200,
      "step": 25865
    },
    {
      "epoch": 13.558700209643606,
      "grad_norm": 0.4424359202384949,
      "learning_rate": 1.42050338469502e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 16878824,
      "step": 25870
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 0.6568911671638489,
      "learning_rate": 1.4194721594115495e-05,
      "loss": 0.3777,
      "num_input_tokens_seen": 16883080,
      "step": 25875
    },
    {
      "epoch": 13.563941299790356,
      "grad_norm": 0.30357760190963745,
      "learning_rate": 1.4184411601621769e-05,
      "loss": 0.5477,
      "num_input_tokens_seen": 16887336,
      "step": 25880
    },
    {
      "epoch": 13.566561844863731,
      "grad_norm": 0.45144936442375183,
      "learning_rate": 1.4174103871625738e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 16890760,
      "step": 25885
    },
    {
      "epoch": 13.569182389937106,
      "grad_norm": 0.2900311350822449,
      "learning_rate": 1.4163798406283674e-05,
      "loss": 0.3748,
      "num_input_tokens_seen": 16893928,
      "step": 25890
    },
    {
      "epoch": 13.571802935010481,
      "grad_norm": 0.6249638795852661,
      "learning_rate": 1.415349520775135e-05,
      "loss": 0.3255,
      "num_input_tokens_seen": 16897032,
      "step": 25895
    },
    {
      "epoch": 13.574423480083858,
      "grad_norm": 0.8132590055465698,
      "learning_rate": 1.4143194278184086e-05,
      "loss": 0.365,
      "num_input_tokens_seen": 16900648,
      "step": 25900
    },
    {
      "epoch": 13.577044025157234,
      "grad_norm": 1.0391461849212646,
      "learning_rate": 1.4132895619736724e-05,
      "loss": 0.5578,
      "num_input_tokens_seen": 16903880,
      "step": 25905
    },
    {
      "epoch": 13.579664570230609,
      "grad_norm": 0.44456014037132263,
      "learning_rate": 1.4122599234563613e-05,
      "loss": 0.5081,
      "num_input_tokens_seen": 16907816,
      "step": 25910
    },
    {
      "epoch": 13.582285115303984,
      "grad_norm": 0.3267393112182617,
      "learning_rate": 1.4112305124818647e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 16910664,
      "step": 25915
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 0.33983972668647766,
      "learning_rate": 1.4102013292655242e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 16914408,
      "step": 25920
    },
    {
      "epoch": 13.587526205450734,
      "grad_norm": 0.4364328682422638,
      "learning_rate": 1.409172374022632e-05,
      "loss": 0.5877,
      "num_input_tokens_seen": 16918120,
      "step": 25925
    },
    {
      "epoch": 13.59014675052411,
      "grad_norm": 0.47913122177124023,
      "learning_rate": 1.4081436469684339e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 16920680,
      "step": 25930
    },
    {
      "epoch": 13.592767295597485,
      "grad_norm": 0.3126973509788513,
      "learning_rate": 1.4071151483181294e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 16923592,
      "step": 25935
    },
    {
      "epoch": 13.59538784067086,
      "grad_norm": 0.4046201705932617,
      "learning_rate": 1.4060868782868669e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 16926920,
      "step": 25940
    },
    {
      "epoch": 13.598008385744235,
      "grad_norm": 0.4797278940677643,
      "learning_rate": 1.405058837089749e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 16930056,
      "step": 25945
    },
    {
      "epoch": 13.60062893081761,
      "grad_norm": 0.785940408706665,
      "learning_rate": 1.404031024941832e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 16932424,
      "step": 25950
    },
    {
      "epoch": 13.603249475890985,
      "grad_norm": 0.5603991150856018,
      "learning_rate": 1.40300344205812e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 16936200,
      "step": 25955
    },
    {
      "epoch": 13.60587002096436,
      "grad_norm": 0.38892993330955505,
      "learning_rate": 1.4019760886535738e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 16939176,
      "step": 25960
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 0.2852663993835449,
      "learning_rate": 1.4009489649431018e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 16943048,
      "step": 25965
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.35146117210388184,
      "learning_rate": 1.3999220711415675e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 16946056,
      "step": 25970
    },
    {
      "epoch": 13.613731656184486,
      "grad_norm": 0.45963427424430847,
      "learning_rate": 1.3988954074637867e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 16948712,
      "step": 25975
    },
    {
      "epoch": 13.616352201257861,
      "grad_norm": 0.612196147441864,
      "learning_rate": 1.3978689741245238e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 16952168,
      "step": 25980
    },
    {
      "epoch": 13.618972746331236,
      "grad_norm": 0.4132981300354004,
      "learning_rate": 1.3968427713384986e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 16955496,
      "step": 25985
    },
    {
      "epoch": 13.621593291404611,
      "grad_norm": 0.4507858455181122,
      "learning_rate": 1.3958167993203791e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 16958472,
      "step": 25990
    },
    {
      "epoch": 13.624213836477988,
      "grad_norm": 0.237117201089859,
      "learning_rate": 1.3947910582847879e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 16961832,
      "step": 25995
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 0.7755764126777649,
      "learning_rate": 1.3937655484462985e-05,
      "loss": 0.3633,
      "num_input_tokens_seen": 16964552,
      "step": 26000
    },
    {
      "epoch": 13.629454926624739,
      "grad_norm": 0.4114496409893036,
      "learning_rate": 1.392740270019437e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 16967688,
      "step": 26005
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 0.5607171654701233,
      "learning_rate": 1.3917152232186776e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 16971176,
      "step": 26010
    },
    {
      "epoch": 13.634696016771489,
      "grad_norm": 0.45195263624191284,
      "learning_rate": 1.3906904082584504e-05,
      "loss": 0.3817,
      "num_input_tokens_seen": 16974568,
      "step": 26015
    },
    {
      "epoch": 13.637316561844864,
      "grad_norm": 0.6581764221191406,
      "learning_rate": 1.3896658253531334e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 16978728,
      "step": 26020
    },
    {
      "epoch": 13.63993710691824,
      "grad_norm": 0.44221168756484985,
      "learning_rate": 1.3886414747170581e-05,
      "loss": 0.3286,
      "num_input_tokens_seen": 16981928,
      "step": 26025
    },
    {
      "epoch": 13.642557651991615,
      "grad_norm": 0.7143174409866333,
      "learning_rate": 1.3876173565645072e-05,
      "loss": 0.4389,
      "num_input_tokens_seen": 16984104,
      "step": 26030
    },
    {
      "epoch": 13.64517819706499,
      "grad_norm": 0.5476198792457581,
      "learning_rate": 1.3865934711097156e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 16986952,
      "step": 26035
    },
    {
      "epoch": 13.647798742138365,
      "grad_norm": 0.30268287658691406,
      "learning_rate": 1.3855698185668675e-05,
      "loss": 0.3112,
      "num_input_tokens_seen": 16993160,
      "step": 26040
    },
    {
      "epoch": 13.65041928721174,
      "grad_norm": 0.5717089772224426,
      "learning_rate": 1.3845463991500985e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 16995560,
      "step": 26045
    },
    {
      "epoch": 13.653039832285115,
      "grad_norm": 0.5259094834327698,
      "learning_rate": 1.3835232130734968e-05,
      "loss": 0.4836,
      "num_input_tokens_seen": 16999016,
      "step": 26050
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 0.4200407564640045,
      "learning_rate": 1.3825002605511017e-05,
      "loss": 0.3449,
      "num_input_tokens_seen": 17001960,
      "step": 26055
    },
    {
      "epoch": 13.658280922431866,
      "grad_norm": 0.7178955674171448,
      "learning_rate": 1.3814775417969037e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 17004936,
      "step": 26060
    },
    {
      "epoch": 13.66090146750524,
      "grad_norm": 0.8378800749778748,
      "learning_rate": 1.3804550570248431e-05,
      "loss": 0.3564,
      "num_input_tokens_seen": 17008584,
      "step": 26065
    },
    {
      "epoch": 13.663522012578616,
      "grad_norm": 0.5705344080924988,
      "learning_rate": 1.3794328064488116e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 17011048,
      "step": 26070
    },
    {
      "epoch": 13.666142557651991,
      "grad_norm": 0.6474170684814453,
      "learning_rate": 1.3784107902826523e-05,
      "loss": 0.3657,
      "num_input_tokens_seen": 17013384,
      "step": 26075
    },
    {
      "epoch": 13.668763102725366,
      "grad_norm": 0.724386990070343,
      "learning_rate": 1.3773890087401603e-05,
      "loss": 0.4323,
      "num_input_tokens_seen": 17016584,
      "step": 26080
    },
    {
      "epoch": 13.671383647798741,
      "grad_norm": 0.6439083218574524,
      "learning_rate": 1.3763674620350802e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 17019368,
      "step": 26085
    },
    {
      "epoch": 13.674004192872118,
      "grad_norm": 0.5277048945426941,
      "learning_rate": 1.3753461503811089e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 17022184,
      "step": 26090
    },
    {
      "epoch": 13.676624737945493,
      "grad_norm": 0.39281103014945984,
      "learning_rate": 1.3743250739918922e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 17025288,
      "step": 26095
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 0.5787399411201477,
      "learning_rate": 1.3733042330810265e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 17028424,
      "step": 26100
    },
    {
      "epoch": 13.681865828092244,
      "grad_norm": 0.4817372262477875,
      "learning_rate": 1.3722836278620616e-05,
      "loss": 0.3272,
      "num_input_tokens_seen": 17031720,
      "step": 26105
    },
    {
      "epoch": 13.684486373165619,
      "grad_norm": 0.40907594561576843,
      "learning_rate": 1.3712632585484957e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 17034888,
      "step": 26110
    },
    {
      "epoch": 13.687106918238994,
      "grad_norm": 0.4773266613483429,
      "learning_rate": 1.37024312535378e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 17041576,
      "step": 26115
    },
    {
      "epoch": 13.68972746331237,
      "grad_norm": 0.5321083664894104,
      "learning_rate": 1.3692232284913133e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 17044424,
      "step": 26120
    },
    {
      "epoch": 13.692348008385745,
      "grad_norm": 0.36374685168266296,
      "learning_rate": 1.3682035681744459e-05,
      "loss": 0.3778,
      "num_input_tokens_seen": 17047976,
      "step": 26125
    },
    {
      "epoch": 13.69496855345912,
      "grad_norm": 0.5330352783203125,
      "learning_rate": 1.3671841446164795e-05,
      "loss": 0.6179,
      "num_input_tokens_seen": 17051208,
      "step": 26130
    },
    {
      "epoch": 13.697589098532495,
      "grad_norm": 0.4748504161834717,
      "learning_rate": 1.3661649580306669e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 17053640,
      "step": 26135
    },
    {
      "epoch": 13.70020964360587,
      "grad_norm": 1.060402750968933,
      "learning_rate": 1.3651460086302087e-05,
      "loss": 0.3689,
      "num_input_tokens_seen": 17056968,
      "step": 26140
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 0.6712900400161743,
      "learning_rate": 1.3641272966282609e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 17059688,
      "step": 26145
    },
    {
      "epoch": 13.70545073375262,
      "grad_norm": 0.7123062610626221,
      "learning_rate": 1.3631088222379218e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 17064712,
      "step": 26150
    },
    {
      "epoch": 13.708071278825996,
      "grad_norm": 0.4468853175640106,
      "learning_rate": 1.3620905856722468e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 17067528,
      "step": 26155
    },
    {
      "epoch": 13.71069182389937,
      "grad_norm": 0.38936901092529297,
      "learning_rate": 1.3610725871442392e-05,
      "loss": 0.3694,
      "num_input_tokens_seen": 17070952,
      "step": 26160
    },
    {
      "epoch": 13.713312368972746,
      "grad_norm": 0.45188501477241516,
      "learning_rate": 1.3600548268668532e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 17073992,
      "step": 26165
    },
    {
      "epoch": 13.715932914046121,
      "grad_norm": 0.456898957490921,
      "learning_rate": 1.3590373050529927e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 17078376,
      "step": 26170
    },
    {
      "epoch": 13.718553459119496,
      "grad_norm": 0.9831394553184509,
      "learning_rate": 1.3580200219155116e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 17082952,
      "step": 26175
    },
    {
      "epoch": 13.721174004192871,
      "grad_norm": 0.49215564131736755,
      "learning_rate": 1.3570029776672122e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 17085640,
      "step": 26180
    },
    {
      "epoch": 13.723794549266248,
      "grad_norm": 0.2919432520866394,
      "learning_rate": 1.35598617252085e-05,
      "loss": 0.3162,
      "num_input_tokens_seen": 17089096,
      "step": 26185
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 1.0305004119873047,
      "learning_rate": 1.354969606689129e-05,
      "loss": 0.6015,
      "num_input_tokens_seen": 17091752,
      "step": 26190
    },
    {
      "epoch": 13.729035639412999,
      "grad_norm": 0.29066455364227295,
      "learning_rate": 1.353953280384704e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 17094824,
      "step": 26195
    },
    {
      "epoch": 13.731656184486374,
      "grad_norm": 0.42951568961143494,
      "learning_rate": 1.352937193820177e-05,
      "loss": 0.5957,
      "num_input_tokens_seen": 17097992,
      "step": 26200
    },
    {
      "epoch": 13.734276729559749,
      "grad_norm": 0.5977164506912231,
      "learning_rate": 1.3519213472081038e-05,
      "loss": 0.5324,
      "num_input_tokens_seen": 17100968,
      "step": 26205
    },
    {
      "epoch": 13.736897274633124,
      "grad_norm": 0.7500107884407043,
      "learning_rate": 1.3509057407609854e-05,
      "loss": 0.4728,
      "num_input_tokens_seen": 17103496,
      "step": 26210
    },
    {
      "epoch": 13.7395178197065,
      "grad_norm": 0.3975032567977905,
      "learning_rate": 1.3498903746912769e-05,
      "loss": 0.4196,
      "num_input_tokens_seen": 17106920,
      "step": 26215
    },
    {
      "epoch": 13.742138364779874,
      "grad_norm": 0.7269766330718994,
      "learning_rate": 1.3488752492113807e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 17110472,
      "step": 26220
    },
    {
      "epoch": 13.74475890985325,
      "grad_norm": 0.3290219306945801,
      "learning_rate": 1.3478603645336508e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 17113640,
      "step": 26225
    },
    {
      "epoch": 13.747379454926625,
      "grad_norm": 0.5629852414131165,
      "learning_rate": 1.3468457208703877e-05,
      "loss": 0.5616,
      "num_input_tokens_seen": 17116808,
      "step": 26230
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.3537505567073822,
      "learning_rate": 1.3458313184338437e-05,
      "loss": 0.443,
      "num_input_tokens_seen": 17119144,
      "step": 26235
    },
    {
      "epoch": 13.752620545073375,
      "grad_norm": 0.47369033098220825,
      "learning_rate": 1.3448171574362201e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 17122056,
      "step": 26240
    },
    {
      "epoch": 13.75524109014675,
      "grad_norm": 0.4134799838066101,
      "learning_rate": 1.3438032380896681e-05,
      "loss": 0.3699,
      "num_input_tokens_seen": 17125352,
      "step": 26245
    },
    {
      "epoch": 13.757861635220126,
      "grad_norm": 0.8203812837600708,
      "learning_rate": 1.342789560606289e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 17128456,
      "step": 26250
    },
    {
      "epoch": 13.7604821802935,
      "grad_norm": 0.6157899498939514,
      "learning_rate": 1.341776125198131e-05,
      "loss": 0.3242,
      "num_input_tokens_seen": 17131496,
      "step": 26255
    },
    {
      "epoch": 13.763102725366876,
      "grad_norm": 0.5219053030014038,
      "learning_rate": 1.3407629320771948e-05,
      "loss": 0.3043,
      "num_input_tokens_seen": 17135528,
      "step": 26260
    },
    {
      "epoch": 13.765723270440251,
      "grad_norm": 0.4294511377811432,
      "learning_rate": 1.3397499814554266e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 17138728,
      "step": 26265
    },
    {
      "epoch": 13.768343815513626,
      "grad_norm": 0.6555507183074951,
      "learning_rate": 1.3387372735447256e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 17142600,
      "step": 26270
    },
    {
      "epoch": 13.770964360587001,
      "grad_norm": 0.8154400587081909,
      "learning_rate": 1.3377248085569399e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 17145448,
      "step": 26275
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 0.3850715160369873,
      "learning_rate": 1.336712586703863e-05,
      "loss": 0.3376,
      "num_input_tokens_seen": 17148712,
      "step": 26280
    },
    {
      "epoch": 13.776205450733752,
      "grad_norm": 0.6181591153144836,
      "learning_rate": 1.3357006081972414e-05,
      "loss": 0.3805,
      "num_input_tokens_seen": 17151144,
      "step": 26285
    },
    {
      "epoch": 13.778825995807129,
      "grad_norm": 0.49007681012153625,
      "learning_rate": 1.3346888732487706e-05,
      "loss": 0.3627,
      "num_input_tokens_seen": 17153864,
      "step": 26290
    },
    {
      "epoch": 13.781446540880504,
      "grad_norm": 0.631557822227478,
      "learning_rate": 1.3336773820700921e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 17156968,
      "step": 26295
    },
    {
      "epoch": 13.784067085953879,
      "grad_norm": 0.5464796423912048,
      "learning_rate": 1.3326661348727992e-05,
      "loss": 0.5093,
      "num_input_tokens_seen": 17160520,
      "step": 26300
    },
    {
      "epoch": 13.786687631027254,
      "grad_norm": 0.4879322648048401,
      "learning_rate": 1.3316551318684342e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 17163368,
      "step": 26305
    },
    {
      "epoch": 13.78930817610063,
      "grad_norm": 0.6345732808113098,
      "learning_rate": 1.3306443732684854e-05,
      "loss": 0.6187,
      "num_input_tokens_seen": 17166472,
      "step": 26310
    },
    {
      "epoch": 13.791928721174004,
      "grad_norm": 0.4372478127479553,
      "learning_rate": 1.3296338592843929e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 17170088,
      "step": 26315
    },
    {
      "epoch": 13.79454926624738,
      "grad_norm": 0.4159722924232483,
      "learning_rate": 1.3286235901275457e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 17172840,
      "step": 26320
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 0.3805217444896698,
      "learning_rate": 1.327613566009279e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 17175752,
      "step": 26325
    },
    {
      "epoch": 13.79979035639413,
      "grad_norm": 0.30582156777381897,
      "learning_rate": 1.3266037871408798e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 17178504,
      "step": 26330
    },
    {
      "epoch": 13.802410901467505,
      "grad_norm": 0.44186174869537354,
      "learning_rate": 1.3255942537335805e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 17180648,
      "step": 26335
    },
    {
      "epoch": 13.80503144654088,
      "grad_norm": 0.4543159604072571,
      "learning_rate": 1.3245849659985652e-05,
      "loss": 0.532,
      "num_input_tokens_seen": 17184648,
      "step": 26340
    },
    {
      "epoch": 13.807651991614255,
      "grad_norm": 0.7352887392044067,
      "learning_rate": 1.3235759241469658e-05,
      "loss": 0.5623,
      "num_input_tokens_seen": 17188264,
      "step": 26345
    },
    {
      "epoch": 13.81027253668763,
      "grad_norm": 0.3661147356033325,
      "learning_rate": 1.322567128389861e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 17191464,
      "step": 26350
    },
    {
      "epoch": 13.812893081761006,
      "grad_norm": 0.5697885751724243,
      "learning_rate": 1.3215585789382812e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 17194792,
      "step": 26355
    },
    {
      "epoch": 13.815513626834381,
      "grad_norm": 0.9185153245925903,
      "learning_rate": 1.320550276003201e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 17197864,
      "step": 26360
    },
    {
      "epoch": 13.818134171907756,
      "grad_norm": 0.707740843296051,
      "learning_rate": 1.3195422197955473e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 17200840,
      "step": 26365
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 0.4498235881328583,
      "learning_rate": 1.3185344105261937e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 17204296,
      "step": 26370
    },
    {
      "epoch": 13.823375262054507,
      "grad_norm": 0.33681780099868774,
      "learning_rate": 1.3175268484059639e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 17206920,
      "step": 26375
    },
    {
      "epoch": 13.825995807127882,
      "grad_norm": 0.5135623812675476,
      "learning_rate": 1.316519533645626e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 17209672,
      "step": 26380
    },
    {
      "epoch": 13.828616352201259,
      "grad_norm": 0.622076690196991,
      "learning_rate": 1.3155124664559004e-05,
      "loss": 0.6094,
      "num_input_tokens_seen": 17212296,
      "step": 26385
    },
    {
      "epoch": 13.831236897274634,
      "grad_norm": 0.4267202317714691,
      "learning_rate": 1.3145056470474525e-05,
      "loss": 0.5585,
      "num_input_tokens_seen": 17215880,
      "step": 26390
    },
    {
      "epoch": 13.833857442348009,
      "grad_norm": 0.5077509880065918,
      "learning_rate": 1.313499075630899e-05,
      "loss": 0.4854,
      "num_input_tokens_seen": 17218760,
      "step": 26395
    },
    {
      "epoch": 13.836477987421384,
      "grad_norm": 0.5987157225608826,
      "learning_rate": 1.312492752416802e-05,
      "loss": 0.5778,
      "num_input_tokens_seen": 17222408,
      "step": 26400
    },
    {
      "epoch": 13.83909853249476,
      "grad_norm": 0.37092992663383484,
      "learning_rate": 1.3114866776156748e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 17225800,
      "step": 26405
    },
    {
      "epoch": 13.841719077568134,
      "grad_norm": 0.3941640853881836,
      "learning_rate": 1.3104808514379748e-05,
      "loss": 0.3333,
      "num_input_tokens_seen": 17229256,
      "step": 26410
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 0.6779186129570007,
      "learning_rate": 1.3094752740941094e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 17232104,
      "step": 26415
    },
    {
      "epoch": 13.846960167714885,
      "grad_norm": 0.34696558117866516,
      "learning_rate": 1.3084699457944344e-05,
      "loss": 0.677,
      "num_input_tokens_seen": 17235528,
      "step": 26420
    },
    {
      "epoch": 13.84958071278826,
      "grad_norm": 0.6549261212348938,
      "learning_rate": 1.3074648667492528e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 17238088,
      "step": 26425
    },
    {
      "epoch": 13.852201257861635,
      "grad_norm": 0.32079729437828064,
      "learning_rate": 1.3064600371688171e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 17241288,
      "step": 26430
    },
    {
      "epoch": 13.85482180293501,
      "grad_norm": 0.2528837323188782,
      "learning_rate": 1.3054554572633249e-05,
      "loss": 0.3888,
      "num_input_tokens_seen": 17244840,
      "step": 26435
    },
    {
      "epoch": 13.857442348008385,
      "grad_norm": 0.32761162519454956,
      "learning_rate": 1.3044511272429222e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 17247688,
      "step": 26440
    },
    {
      "epoch": 13.86006289308176,
      "grad_norm": 0.7144145965576172,
      "learning_rate": 1.3034470473177038e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 17252360,
      "step": 26445
    },
    {
      "epoch": 13.862683438155136,
      "grad_norm": 0.765985369682312,
      "learning_rate": 1.3024432176977125e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 17255368,
      "step": 26450
    },
    {
      "epoch": 13.865303983228511,
      "grad_norm": 1.3459465503692627,
      "learning_rate": 1.3014396385929372e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 17257832,
      "step": 26455
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 0.4450909197330475,
      "learning_rate": 1.3004363102133165e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 17261480,
      "step": 26460
    },
    {
      "epoch": 13.870545073375261,
      "grad_norm": 0.3875693380832672,
      "learning_rate": 1.2994332327687348e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 17264360,
      "step": 26465
    },
    {
      "epoch": 13.873165618448636,
      "grad_norm": 0.4294387400150299,
      "learning_rate": 1.2984304064690228e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 17268104,
      "step": 26470
    },
    {
      "epoch": 13.875786163522012,
      "grad_norm": 0.7789985537528992,
      "learning_rate": 1.2974278315239619e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 17271048,
      "step": 26475
    },
    {
      "epoch": 13.878406708595389,
      "grad_norm": 0.5936257839202881,
      "learning_rate": 1.2964255081432786e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 17274056,
      "step": 26480
    },
    {
      "epoch": 13.881027253668764,
      "grad_norm": 0.8018804788589478,
      "learning_rate": 1.2954234365366488e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 17277192,
      "step": 26485
    },
    {
      "epoch": 13.883647798742139,
      "grad_norm": 0.6581233143806458,
      "learning_rate": 1.2944216169136936e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 17281128,
      "step": 26490
    },
    {
      "epoch": 13.886268343815514,
      "grad_norm": 0.3705587387084961,
      "learning_rate": 1.2934200494839815e-05,
      "loss": 0.384,
      "num_input_tokens_seen": 17283848,
      "step": 26495
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.5828413963317871,
      "learning_rate": 1.2924187344570293e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 17287208,
      "step": 26500
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 0.4386993944644928,
      "learning_rate": 1.2914176720423011e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 17290696,
      "step": 26505
    },
    {
      "epoch": 13.89412997903564,
      "grad_norm": 0.41144266724586487,
      "learning_rate": 1.2904168624492088e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 17294600,
      "step": 26510
    },
    {
      "epoch": 13.896750524109015,
      "grad_norm": 0.7647677063941956,
      "learning_rate": 1.2894163058871092e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 17297768,
      "step": 26515
    },
    {
      "epoch": 13.89937106918239,
      "grad_norm": 0.5313085913658142,
      "learning_rate": 1.2884160025653064e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 17300808,
      "step": 26520
    },
    {
      "epoch": 13.901991614255765,
      "grad_norm": 0.36933740973472595,
      "learning_rate": 1.2874159526930532e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 17303432,
      "step": 26525
    },
    {
      "epoch": 13.90461215932914,
      "grad_norm": 0.439666211605072,
      "learning_rate": 1.286416156479549e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 17306152,
      "step": 26530
    },
    {
      "epoch": 13.907232704402515,
      "grad_norm": 0.4482743740081787,
      "learning_rate": 1.2854166141339394e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 17308840,
      "step": 26535
    },
    {
      "epoch": 13.90985324947589,
      "grad_norm": 0.8368704319000244,
      "learning_rate": 1.2844173258653183e-05,
      "loss": 0.527,
      "num_input_tokens_seen": 17311944,
      "step": 26540
    },
    {
      "epoch": 13.912473794549266,
      "grad_norm": 0.5395371317863464,
      "learning_rate": 1.2834182918827247e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 17314888,
      "step": 26545
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 1.4697123765945435,
      "learning_rate": 1.282419512395144e-05,
      "loss": 0.5161,
      "num_input_tokens_seen": 17318248,
      "step": 26550
    },
    {
      "epoch": 13.917714884696016,
      "grad_norm": 0.32172825932502747,
      "learning_rate": 1.2814209876115102e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 17321320,
      "step": 26555
    },
    {
      "epoch": 13.920335429769391,
      "grad_norm": 0.39122167229652405,
      "learning_rate": 1.2804227177407038e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 17324680,
      "step": 26560
    },
    {
      "epoch": 13.922955974842766,
      "grad_norm": 0.34435051679611206,
      "learning_rate": 1.279424702991552e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 17328072,
      "step": 26565
    },
    {
      "epoch": 13.925576519916142,
      "grad_norm": 0.7126839756965637,
      "learning_rate": 1.2784269435728264e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 17331272,
      "step": 26570
    },
    {
      "epoch": 13.928197064989519,
      "grad_norm": 0.44600436091423035,
      "learning_rate": 1.2774294396932488e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 17334920,
      "step": 26575
    },
    {
      "epoch": 13.930817610062894,
      "grad_norm": 0.6411777138710022,
      "learning_rate": 1.2764321915614835e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 17338664,
      "step": 26580
    },
    {
      "epoch": 13.933438155136269,
      "grad_norm": 0.6319724917411804,
      "learning_rate": 1.2754351993861444e-05,
      "loss": 0.5458,
      "num_input_tokens_seen": 17341000,
      "step": 26585
    },
    {
      "epoch": 13.936058700209644,
      "grad_norm": 0.31361740827560425,
      "learning_rate": 1.2744384633757922e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 17344456,
      "step": 26590
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 0.6386637091636658,
      "learning_rate": 1.2734419837389306e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 17347976,
      "step": 26595
    },
    {
      "epoch": 13.941299790356394,
      "grad_norm": 0.49536627531051636,
      "learning_rate": 1.2724457606840123e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 17351336,
      "step": 26600
    },
    {
      "epoch": 13.94392033542977,
      "grad_norm": 0.2764771580696106,
      "learning_rate": 1.2714497944194376e-05,
      "loss": 0.402,
      "num_input_tokens_seen": 17356744,
      "step": 26605
    },
    {
      "epoch": 13.946540880503145,
      "grad_norm": 0.3238592743873596,
      "learning_rate": 1.2704540851535484e-05,
      "loss": 0.4592,
      "num_input_tokens_seen": 17359240,
      "step": 26610
    },
    {
      "epoch": 13.94916142557652,
      "grad_norm": 0.7713359594345093,
      "learning_rate": 1.2694586330946376e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 17363592,
      "step": 26615
    },
    {
      "epoch": 13.951781970649895,
      "grad_norm": 0.6582915186882019,
      "learning_rate": 1.2684634384509427e-05,
      "loss": 0.6403,
      "num_input_tokens_seen": 17365992,
      "step": 26620
    },
    {
      "epoch": 13.95440251572327,
      "grad_norm": 0.3467501997947693,
      "learning_rate": 1.2674685014306456e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 17368520,
      "step": 26625
    },
    {
      "epoch": 13.957023060796645,
      "grad_norm": 0.4093809425830841,
      "learning_rate": 1.2664738222418776e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 17371880,
      "step": 26630
    },
    {
      "epoch": 13.95964360587002,
      "grad_norm": 0.5336394906044006,
      "learning_rate": 1.2654794010927118e-05,
      "loss": 0.3843,
      "num_input_tokens_seen": 17374600,
      "step": 26635
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 0.4872286319732666,
      "learning_rate": 1.2644852381911715e-05,
      "loss": 0.358,
      "num_input_tokens_seen": 17378600,
      "step": 26640
    },
    {
      "epoch": 13.964884696016771,
      "grad_norm": 0.5715813636779785,
      "learning_rate": 1.2634913337452248e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 17381416,
      "step": 26645
    },
    {
      "epoch": 13.967505241090146,
      "grad_norm": 0.5769841074943542,
      "learning_rate": 1.2624976879627832e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 17384264,
      "step": 26650
    },
    {
      "epoch": 13.970125786163521,
      "grad_norm": 0.6675117611885071,
      "learning_rate": 1.2615043010517069e-05,
      "loss": 0.5313,
      "num_input_tokens_seen": 17387912,
      "step": 26655
    },
    {
      "epoch": 13.972746331236896,
      "grad_norm": 0.8053733110427856,
      "learning_rate": 1.2605111732198027e-05,
      "loss": 0.3438,
      "num_input_tokens_seen": 17390600,
      "step": 26660
    },
    {
      "epoch": 13.975366876310272,
      "grad_norm": 0.3502251207828522,
      "learning_rate": 1.2595183046748188e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 17393704,
      "step": 26665
    },
    {
      "epoch": 13.977987421383649,
      "grad_norm": 0.41726353764533997,
      "learning_rate": 1.2585256956244545e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 17396808,
      "step": 26670
    },
    {
      "epoch": 13.980607966457024,
      "grad_norm": 0.3908206522464752,
      "learning_rate": 1.2575333462763506e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 17400040,
      "step": 26675
    },
    {
      "epoch": 13.983228511530399,
      "grad_norm": 0.3458293378353119,
      "learning_rate": 1.2565412568380957e-05,
      "loss": 0.3425,
      "num_input_tokens_seen": 17403624,
      "step": 26680
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 0.4216941297054291,
      "learning_rate": 1.2555494275172237e-05,
      "loss": 0.6493,
      "num_input_tokens_seen": 17406760,
      "step": 26685
    },
    {
      "epoch": 13.98846960167715,
      "grad_norm": 0.336683452129364,
      "learning_rate": 1.2545578585212148e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 17409832,
      "step": 26690
    },
    {
      "epoch": 13.991090146750524,
      "grad_norm": 0.6003012657165527,
      "learning_rate": 1.2535665500574922e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 17413672,
      "step": 26695
    },
    {
      "epoch": 13.9937106918239,
      "grad_norm": 0.3471783995628357,
      "learning_rate": 1.2525755023334285e-05,
      "loss": 0.561,
      "num_input_tokens_seen": 17416936,
      "step": 26700
    },
    {
      "epoch": 13.996331236897275,
      "grad_norm": 0.3988579213619232,
      "learning_rate": 1.251584715556337e-05,
      "loss": 0.458,
      "num_input_tokens_seen": 17420520,
      "step": 26705
    },
    {
      "epoch": 13.99895178197065,
      "grad_norm": 0.34049737453460693,
      "learning_rate": 1.2505941899334805e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 17424840,
      "step": 26710
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.45284703373908997,
      "eval_runtime": 9.8974,
      "eval_samples_per_second": 85.679,
      "eval_steps_per_second": 21.42,
      "num_input_tokens_seen": 17425368,
      "step": 26712
    },
    {
      "epoch": 14.001572327044025,
      "grad_norm": 0.8855025768280029,
      "learning_rate": 1.2496039256720665e-05,
      "loss": 0.4327,
      "num_input_tokens_seen": 17427448,
      "step": 26715
    },
    {
      "epoch": 14.0041928721174,
      "grad_norm": 0.5433520674705505,
      "learning_rate": 1.2486139229792452e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 17430008,
      "step": 26720
    },
    {
      "epoch": 14.006813417190775,
      "grad_norm": 0.3068620264530182,
      "learning_rate": 1.2476241820621152e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 17433080,
      "step": 26725
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 0.4766952395439148,
      "learning_rate": 1.2466347031277176e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 17436088,
      "step": 26730
    },
    {
      "epoch": 14.012054507337526,
      "grad_norm": 0.5174327492713928,
      "learning_rate": 1.2456454863830411e-05,
      "loss": 0.3868,
      "num_input_tokens_seen": 17438680,
      "step": 26735
    },
    {
      "epoch": 14.014675052410901,
      "grad_norm": 0.4457821249961853,
      "learning_rate": 1.2446565320350185e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 17442424,
      "step": 26740
    },
    {
      "epoch": 14.017295597484276,
      "grad_norm": 0.886006236076355,
      "learning_rate": 1.2436678402905283e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 17445784,
      "step": 26745
    },
    {
      "epoch": 14.019916142557651,
      "grad_norm": 0.5058649778366089,
      "learning_rate": 1.2426794113563933e-05,
      "loss": 0.3164,
      "num_input_tokens_seen": 17448280,
      "step": 26750
    },
    {
      "epoch": 14.022536687631026,
      "grad_norm": 0.5832860469818115,
      "learning_rate": 1.24169124543938e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 17452248,
      "step": 26755
    },
    {
      "epoch": 14.025157232704403,
      "grad_norm": 0.43219420313835144,
      "learning_rate": 1.2407033427462026e-05,
      "loss": 0.3626,
      "num_input_tokens_seen": 17455672,
      "step": 26760
    },
    {
      "epoch": 14.027777777777779,
      "grad_norm": 0.3816930055618286,
      "learning_rate": 1.2397157034835192e-05,
      "loss": 0.3298,
      "num_input_tokens_seen": 17458328,
      "step": 26765
    },
    {
      "epoch": 14.030398322851154,
      "grad_norm": 0.5554039478302002,
      "learning_rate": 1.2387283278579325e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 17460856,
      "step": 26770
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 0.35359224677085876,
      "learning_rate": 1.237741216075991e-05,
      "loss": 0.3978,
      "num_input_tokens_seen": 17464440,
      "step": 26775
    },
    {
      "epoch": 14.035639412997904,
      "grad_norm": 0.41594788432121277,
      "learning_rate": 1.2367543683441868e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 17467512,
      "step": 26780
    },
    {
      "epoch": 14.03825995807128,
      "grad_norm": 0.688420832157135,
      "learning_rate": 1.2357677848689558e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 17471224,
      "step": 26785
    },
    {
      "epoch": 14.040880503144654,
      "grad_norm": 0.38054925203323364,
      "learning_rate": 1.2347814658566811e-05,
      "loss": 0.3285,
      "num_input_tokens_seen": 17474616,
      "step": 26790
    },
    {
      "epoch": 14.04350104821803,
      "grad_norm": 0.3521279990673065,
      "learning_rate": 1.2337954115136894e-05,
      "loss": 0.398,
      "num_input_tokens_seen": 17477880,
      "step": 26795
    },
    {
      "epoch": 14.046121593291405,
      "grad_norm": 0.3391832113265991,
      "learning_rate": 1.2328096220462529e-05,
      "loss": 0.5677,
      "num_input_tokens_seen": 17481016,
      "step": 26800
    },
    {
      "epoch": 14.04874213836478,
      "grad_norm": 0.32065457105636597,
      "learning_rate": 1.2318240976605865e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 17484568,
      "step": 26805
    },
    {
      "epoch": 14.051362683438155,
      "grad_norm": 0.5389261841773987,
      "learning_rate": 1.23083883856285e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 17491032,
      "step": 26810
    },
    {
      "epoch": 14.05398322851153,
      "grad_norm": 0.6474174857139587,
      "learning_rate": 1.229853844959149e-05,
      "loss": 0.3441,
      "num_input_tokens_seen": 17494136,
      "step": 26815
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 0.6524984240531921,
      "learning_rate": 1.228869117055533e-05,
      "loss": 0.5553,
      "num_input_tokens_seen": 17496920,
      "step": 26820
    },
    {
      "epoch": 14.05922431865828,
      "grad_norm": 0.4318070113658905,
      "learning_rate": 1.227884655057997e-05,
      "loss": 0.3084,
      "num_input_tokens_seen": 17500024,
      "step": 26825
    },
    {
      "epoch": 14.061844863731656,
      "grad_norm": 0.5172185301780701,
      "learning_rate": 1.2269004591724777e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 17503416,
      "step": 26830
    },
    {
      "epoch": 14.064465408805031,
      "grad_norm": 0.6385302543640137,
      "learning_rate": 1.2259165296048575e-05,
      "loss": 0.3476,
      "num_input_tokens_seen": 17506648,
      "step": 26835
    },
    {
      "epoch": 14.067085953878406,
      "grad_norm": 0.5415169596672058,
      "learning_rate": 1.2249328665609638e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 17509496,
      "step": 26840
    },
    {
      "epoch": 14.069706498951781,
      "grad_norm": 0.3988257050514221,
      "learning_rate": 1.2239494702465676e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 17513464,
      "step": 26845
    },
    {
      "epoch": 14.072327044025156,
      "grad_norm": 0.5011816024780273,
      "learning_rate": 1.2229663408673842e-05,
      "loss": 0.3087,
      "num_input_tokens_seen": 17516568,
      "step": 26850
    },
    {
      "epoch": 14.074947589098532,
      "grad_norm": 0.48250696063041687,
      "learning_rate": 1.221983478629074e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 17520440,
      "step": 26855
    },
    {
      "epoch": 14.077568134171909,
      "grad_norm": 0.46544623374938965,
      "learning_rate": 1.22100088373724e-05,
      "loss": 0.4314,
      "num_input_tokens_seen": 17523224,
      "step": 26860
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 0.2888934016227722,
      "learning_rate": 1.2200185563974283e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 17530008,
      "step": 26865
    },
    {
      "epoch": 14.082809224318659,
      "grad_norm": 0.48031195998191833,
      "learning_rate": 1.219036496815132e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 17533048,
      "step": 26870
    },
    {
      "epoch": 14.085429769392034,
      "grad_norm": 0.3362758755683899,
      "learning_rate": 1.2180547051957863e-05,
      "loss": 0.354,
      "num_input_tokens_seen": 17535608,
      "step": 26875
    },
    {
      "epoch": 14.08805031446541,
      "grad_norm": 0.36934104561805725,
      "learning_rate": 1.2170731817447723e-05,
      "loss": 0.3896,
      "num_input_tokens_seen": 17539096,
      "step": 26880
    },
    {
      "epoch": 14.090670859538784,
      "grad_norm": 0.4872642755508423,
      "learning_rate": 1.216091926667412e-05,
      "loss": 0.541,
      "num_input_tokens_seen": 17542936,
      "step": 26885
    },
    {
      "epoch": 14.09329140461216,
      "grad_norm": 0.49155595898628235,
      "learning_rate": 1.2151109401689723e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 17545688,
      "step": 26890
    },
    {
      "epoch": 14.095911949685535,
      "grad_norm": 0.29751256108283997,
      "learning_rate": 1.214130222454665e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 17548536,
      "step": 26895
    },
    {
      "epoch": 14.09853249475891,
      "grad_norm": 0.4329130947589874,
      "learning_rate": 1.2131497737296452e-05,
      "loss": 0.3841,
      "num_input_tokens_seen": 17552056,
      "step": 26900
    },
    {
      "epoch": 14.101153039832285,
      "grad_norm": 0.3174706995487213,
      "learning_rate": 1.2121695941990124e-05,
      "loss": 0.3563,
      "num_input_tokens_seen": 17555512,
      "step": 26905
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 0.8008637428283691,
      "learning_rate": 1.2111896840678075e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 17559320,
      "step": 26910
    },
    {
      "epoch": 14.106394129979035,
      "grad_norm": 0.3458684980869293,
      "learning_rate": 1.210210043541018e-05,
      "loss": 0.3174,
      "num_input_tokens_seen": 17562616,
      "step": 26915
    },
    {
      "epoch": 14.10901467505241,
      "grad_norm": 0.4061106741428375,
      "learning_rate": 1.2092306728235717e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 17566200,
      "step": 26920
    },
    {
      "epoch": 14.111635220125786,
      "grad_norm": 0.42161259055137634,
      "learning_rate": 1.2082515721203427e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 17569496,
      "step": 26925
    },
    {
      "epoch": 14.114255765199161,
      "grad_norm": 0.7314572334289551,
      "learning_rate": 1.2072727416361479e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 17571832,
      "step": 26930
    },
    {
      "epoch": 14.116876310272536,
      "grad_norm": 0.585176944732666,
      "learning_rate": 1.2062941815757484e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 17574744,
      "step": 26935
    },
    {
      "epoch": 14.119496855345911,
      "grad_norm": 0.38534337282180786,
      "learning_rate": 1.2053158921438454e-05,
      "loss": 0.4603,
      "num_input_tokens_seen": 17578232,
      "step": 26940
    },
    {
      "epoch": 14.122117400419286,
      "grad_norm": 0.36908572912216187,
      "learning_rate": 1.2043378735450885e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 17581688,
      "step": 26945
    },
    {
      "epoch": 14.124737945492662,
      "grad_norm": 0.669779360294342,
      "learning_rate": 1.2033601259840657e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 17585336,
      "step": 26950
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 0.4072117209434509,
      "learning_rate": 1.2023826496653118e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 17588120,
      "step": 26955
    },
    {
      "epoch": 14.129979035639414,
      "grad_norm": 0.41595712304115295,
      "learning_rate": 1.2014054447933044e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 17591928,
      "step": 26960
    },
    {
      "epoch": 14.132599580712789,
      "grad_norm": 0.31031084060668945,
      "learning_rate": 1.200428511572462e-05,
      "loss": 0.3276,
      "num_input_tokens_seen": 17596120,
      "step": 26965
    },
    {
      "epoch": 14.135220125786164,
      "grad_norm": 0.402108758687973,
      "learning_rate": 1.1994518502071484e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 17599416,
      "step": 26970
    },
    {
      "epoch": 14.13784067085954,
      "grad_norm": 0.41580548882484436,
      "learning_rate": 1.1984754609016716e-05,
      "loss": 0.4868,
      "num_input_tokens_seen": 17603192,
      "step": 26975
    },
    {
      "epoch": 14.140461215932914,
      "grad_norm": 0.6380018591880798,
      "learning_rate": 1.1974993438602785e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 17606520,
      "step": 26980
    },
    {
      "epoch": 14.14308176100629,
      "grad_norm": 0.5136463642120361,
      "learning_rate": 1.1965234992871641e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 17609112,
      "step": 26985
    },
    {
      "epoch": 14.145702306079665,
      "grad_norm": 0.5305076241493225,
      "learning_rate": 1.195547927386462e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 17612440,
      "step": 26990
    },
    {
      "epoch": 14.14832285115304,
      "grad_norm": 0.27001795172691345,
      "learning_rate": 1.1945726283622514e-05,
      "loss": 0.3798,
      "num_input_tokens_seen": 17615384,
      "step": 26995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 0.34126049280166626,
      "learning_rate": 1.1935976024185547e-05,
      "loss": 0.3514,
      "num_input_tokens_seen": 17618456,
      "step": 27000
    },
    {
      "epoch": 14.15356394129979,
      "grad_norm": 0.5234718322753906,
      "learning_rate": 1.1926228497593341e-05,
      "loss": 0.699,
      "num_input_tokens_seen": 17621720,
      "step": 27005
    },
    {
      "epoch": 14.156184486373165,
      "grad_norm": 0.4635133743286133,
      "learning_rate": 1.1916483705884985e-05,
      "loss": 0.5003,
      "num_input_tokens_seen": 17625112,
      "step": 27010
    },
    {
      "epoch": 14.15880503144654,
      "grad_norm": 0.7266581654548645,
      "learning_rate": 1.1906741651098983e-05,
      "loss": 0.5912,
      "num_input_tokens_seen": 17628824,
      "step": 27015
    },
    {
      "epoch": 14.161425576519916,
      "grad_norm": 1.5265554189682007,
      "learning_rate": 1.1897002335273239e-05,
      "loss": 0.3385,
      "num_input_tokens_seen": 17632568,
      "step": 27020
    },
    {
      "epoch": 14.164046121593291,
      "grad_norm": 0.30304768681526184,
      "learning_rate": 1.1887265760445121e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 17635928,
      "step": 27025
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.45233315229415894,
      "learning_rate": 1.1877531928651417e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 17639832,
      "step": 27030
    },
    {
      "epoch": 14.169287211740041,
      "grad_norm": 0.8383107781410217,
      "learning_rate": 1.1867800841928313e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 17643160,
      "step": 27035
    },
    {
      "epoch": 14.171907756813416,
      "grad_norm": 0.32418739795684814,
      "learning_rate": 1.1858072502311465e-05,
      "loss": 0.5369,
      "num_input_tokens_seen": 17646040,
      "step": 27040
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 0.6410925984382629,
      "learning_rate": 1.1848346911835908e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 17648600,
      "step": 27045
    },
    {
      "epoch": 14.177148846960169,
      "grad_norm": 0.887929379940033,
      "learning_rate": 1.1838624072536137e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 17652024,
      "step": 27050
    },
    {
      "epoch": 14.179769392033544,
      "grad_norm": 0.8105661869049072,
      "learning_rate": 1.1828903986446055e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 17655160,
      "step": 27055
    },
    {
      "epoch": 14.182389937106919,
      "grad_norm": 0.4993961453437805,
      "learning_rate": 1.1819186655599007e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 17658712,
      "step": 27060
    },
    {
      "epoch": 14.185010482180294,
      "grad_norm": 0.7880528569221497,
      "learning_rate": 1.1809472082027734e-05,
      "loss": 0.3374,
      "num_input_tokens_seen": 17661432,
      "step": 27065
    },
    {
      "epoch": 14.18763102725367,
      "grad_norm": 0.4143844246864319,
      "learning_rate": 1.1799760267764407e-05,
      "loss": 0.548,
      "num_input_tokens_seen": 17664408,
      "step": 27070
    },
    {
      "epoch": 14.190251572327044,
      "grad_norm": 0.6454504132270813,
      "learning_rate": 1.1790051214840634e-05,
      "loss": 0.3641,
      "num_input_tokens_seen": 17667608,
      "step": 27075
    },
    {
      "epoch": 14.19287211740042,
      "grad_norm": 0.5122662782669067,
      "learning_rate": 1.1780344925287443e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 17670872,
      "step": 27080
    },
    {
      "epoch": 14.195492662473795,
      "grad_norm": 0.3275659680366516,
      "learning_rate": 1.1770641401135282e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 17673368,
      "step": 27085
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 0.39264002442359924,
      "learning_rate": 1.1760940644414003e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 17675800,
      "step": 27090
    },
    {
      "epoch": 14.200733752620545,
      "grad_norm": 0.3793303668498993,
      "learning_rate": 1.1751242657152914e-05,
      "loss": 0.35,
      "num_input_tokens_seen": 17679288,
      "step": 27095
    },
    {
      "epoch": 14.20335429769392,
      "grad_norm": 0.6084994077682495,
      "learning_rate": 1.1741547441380698e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 17682392,
      "step": 27100
    },
    {
      "epoch": 14.205974842767295,
      "grad_norm": 0.8846498727798462,
      "learning_rate": 1.17318549991255e-05,
      "loss": 0.4837,
      "num_input_tokens_seen": 17685176,
      "step": 27105
    },
    {
      "epoch": 14.20859538784067,
      "grad_norm": 0.8409104347229004,
      "learning_rate": 1.1722165332414863e-05,
      "loss": 0.3786,
      "num_input_tokens_seen": 17687768,
      "step": 27110
    },
    {
      "epoch": 14.211215932914046,
      "grad_norm": 0.40813493728637695,
      "learning_rate": 1.1712478443275768e-05,
      "loss": 0.3259,
      "num_input_tokens_seen": 17690712,
      "step": 27115
    },
    {
      "epoch": 14.213836477987421,
      "grad_norm": 0.4016442596912384,
      "learning_rate": 1.170279433373459e-05,
      "loss": 0.6932,
      "num_input_tokens_seen": 17693944,
      "step": 27120
    },
    {
      "epoch": 14.216457023060796,
      "grad_norm": 0.3952736258506775,
      "learning_rate": 1.1693113005817125e-05,
      "loss": 0.3848,
      "num_input_tokens_seen": 17697528,
      "step": 27125
    },
    {
      "epoch": 14.219077568134171,
      "grad_norm": 0.5884512662887573,
      "learning_rate": 1.1683434461548606e-05,
      "loss": 0.4594,
      "num_input_tokens_seen": 17700472,
      "step": 27130
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 0.4722994863986969,
      "learning_rate": 1.1673758702953675e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 17704024,
      "step": 27135
    },
    {
      "epoch": 14.224318658280922,
      "grad_norm": 0.6101797819137573,
      "learning_rate": 1.1664085732056398e-05,
      "loss": 0.3832,
      "num_input_tokens_seen": 17706840,
      "step": 27140
    },
    {
      "epoch": 14.226939203354299,
      "grad_norm": 0.4719480872154236,
      "learning_rate": 1.1654415550880243e-05,
      "loss": 0.586,
      "num_input_tokens_seen": 17709592,
      "step": 27145
    },
    {
      "epoch": 14.229559748427674,
      "grad_norm": 0.3060236871242523,
      "learning_rate": 1.1644748161448088e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 17713112,
      "step": 27150
    },
    {
      "epoch": 14.232180293501049,
      "grad_norm": 0.808035671710968,
      "learning_rate": 1.1635083565782257e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 17716056,
      "step": 27155
    },
    {
      "epoch": 14.234800838574424,
      "grad_norm": 0.6658580899238586,
      "learning_rate": 1.1625421765904465e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 17719704,
      "step": 27160
    },
    {
      "epoch": 14.2374213836478,
      "grad_norm": 0.6715240478515625,
      "learning_rate": 1.1615762763835858e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 17722456,
      "step": 27165
    },
    {
      "epoch": 14.240041928721174,
      "grad_norm": 0.29919201135635376,
      "learning_rate": 1.1606106561596993e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 17725400,
      "step": 27170
    },
    {
      "epoch": 14.24266247379455,
      "grad_norm": 0.4209563136100769,
      "learning_rate": 1.1596453161207829e-05,
      "loss": 0.5759,
      "num_input_tokens_seen": 17728440,
      "step": 27175
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 1.8905696868896484,
      "learning_rate": 1.1586802564687741e-05,
      "loss": 0.3824,
      "num_input_tokens_seen": 17731192,
      "step": 27180
    },
    {
      "epoch": 14.2479035639413,
      "grad_norm": 0.332761287689209,
      "learning_rate": 1.1577154774055528e-05,
      "loss": 0.3367,
      "num_input_tokens_seen": 17735096,
      "step": 27185
    },
    {
      "epoch": 14.250524109014675,
      "grad_norm": 0.48899075388908386,
      "learning_rate": 1.1567509791329401e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 17739000,
      "step": 27190
    },
    {
      "epoch": 14.25314465408805,
      "grad_norm": 0.49358510971069336,
      "learning_rate": 1.155786761852699e-05,
      "loss": 0.389,
      "num_input_tokens_seen": 17741784,
      "step": 27195
    },
    {
      "epoch": 14.255765199161425,
      "grad_norm": 1.10570228099823,
      "learning_rate": 1.1548228257665317e-05,
      "loss": 0.5301,
      "num_input_tokens_seen": 17744440,
      "step": 27200
    },
    {
      "epoch": 14.2583857442348,
      "grad_norm": 0.47741225361824036,
      "learning_rate": 1.1538591710760815e-05,
      "loss": 0.6108,
      "num_input_tokens_seen": 17747352,
      "step": 27205
    },
    {
      "epoch": 14.261006289308176,
      "grad_norm": 0.4114724397659302,
      "learning_rate": 1.1528957979829355e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 17750104,
      "step": 27210
    },
    {
      "epoch": 14.26362683438155,
      "grad_norm": 0.4638059735298157,
      "learning_rate": 1.1519327066886198e-05,
      "loss": 0.3906,
      "num_input_tokens_seen": 17752728,
      "step": 27215
    },
    {
      "epoch": 14.266247379454926,
      "grad_norm": 0.3120374381542206,
      "learning_rate": 1.1509698973946032e-05,
      "loss": 0.3599,
      "num_input_tokens_seen": 17756248,
      "step": 27220
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 0.4361594617366791,
      "learning_rate": 1.1500073703022926e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 17759704,
      "step": 27225
    },
    {
      "epoch": 14.271488469601676,
      "grad_norm": 0.441344290971756,
      "learning_rate": 1.1490451256130394e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 17763064,
      "step": 27230
    },
    {
      "epoch": 14.274109014675052,
      "grad_norm": 0.6600255370140076,
      "learning_rate": 1.1480831635281328e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 17767384,
      "step": 27235
    },
    {
      "epoch": 14.276729559748428,
      "grad_norm": 0.3863745331764221,
      "learning_rate": 1.1471214842488046e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 17770328,
      "step": 27240
    },
    {
      "epoch": 14.279350104821804,
      "grad_norm": 0.8865317702293396,
      "learning_rate": 1.1461600879762277e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 17774072,
      "step": 27245
    },
    {
      "epoch": 14.281970649895179,
      "grad_norm": 0.8312270641326904,
      "learning_rate": 1.1451989749115157e-05,
      "loss": 0.3184,
      "num_input_tokens_seen": 17776344,
      "step": 27250
    },
    {
      "epoch": 14.284591194968554,
      "grad_norm": 0.45916301012039185,
      "learning_rate": 1.1442381452557219e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 17780536,
      "step": 27255
    },
    {
      "epoch": 14.28721174004193,
      "grad_norm": 0.2669757604598999,
      "learning_rate": 1.1432775992098397e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 17784024,
      "step": 27260
    },
    {
      "epoch": 14.289832285115304,
      "grad_norm": 0.45421600341796875,
      "learning_rate": 1.1423173369748056e-05,
      "loss": 0.5978,
      "num_input_tokens_seen": 17788248,
      "step": 27265
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 0.8383277058601379,
      "learning_rate": 1.1413573587514952e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 17790872,
      "step": 27270
    },
    {
      "epoch": 14.295073375262055,
      "grad_norm": 0.4390462636947632,
      "learning_rate": 1.140397664740726e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 17793912,
      "step": 27275
    },
    {
      "epoch": 14.29769392033543,
      "grad_norm": 0.44893205165863037,
      "learning_rate": 1.1394382551432534e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 17797336,
      "step": 27280
    },
    {
      "epoch": 14.300314465408805,
      "grad_norm": 0.7050559520721436,
      "learning_rate": 1.1384791301597767e-05,
      "loss": 0.5148,
      "num_input_tokens_seen": 17800728,
      "step": 27285
    },
    {
      "epoch": 14.30293501048218,
      "grad_norm": 0.5514563918113708,
      "learning_rate": 1.1375202899909317e-05,
      "loss": 0.6259,
      "num_input_tokens_seen": 17803448,
      "step": 27290
    },
    {
      "epoch": 14.305555555555555,
      "grad_norm": 0.3014252185821533,
      "learning_rate": 1.1365617348372984e-05,
      "loss": 0.3679,
      "num_input_tokens_seen": 17807896,
      "step": 27295
    },
    {
      "epoch": 14.30817610062893,
      "grad_norm": 0.4254356324672699,
      "learning_rate": 1.135603464899396e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 17811032,
      "step": 27300
    },
    {
      "epoch": 14.310796645702306,
      "grad_norm": 0.23366694152355194,
      "learning_rate": 1.1346454803776821e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 17815640,
      "step": 27305
    },
    {
      "epoch": 14.31341719077568,
      "grad_norm": 0.34418779611587524,
      "learning_rate": 1.1336877814725566e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 17818744,
      "step": 27310
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 0.5368207693099976,
      "learning_rate": 1.1327303683843607e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 17821432,
      "step": 27315
    },
    {
      "epoch": 14.318658280922431,
      "grad_norm": 0.6012990474700928,
      "learning_rate": 1.1317732413133723e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 17824312,
      "step": 27320
    },
    {
      "epoch": 14.321278825995806,
      "grad_norm": 0.38741132616996765,
      "learning_rate": 1.1308164004598118e-05,
      "loss": 0.3227,
      "num_input_tokens_seen": 17827288,
      "step": 27325
    },
    {
      "epoch": 14.323899371069182,
      "grad_norm": 0.5403540730476379,
      "learning_rate": 1.129859846023841e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 17830392,
      "step": 27330
    },
    {
      "epoch": 14.326519916142558,
      "grad_norm": 0.4447949528694153,
      "learning_rate": 1.1289035782055582e-05,
      "loss": 0.4304,
      "num_input_tokens_seen": 17833432,
      "step": 27335
    },
    {
      "epoch": 14.329140461215934,
      "grad_norm": 0.4714897572994232,
      "learning_rate": 1.1279475972050044e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 17837112,
      "step": 27340
    },
    {
      "epoch": 14.331761006289309,
      "grad_norm": 0.3544812500476837,
      "learning_rate": 1.1269919032221612e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 17840088,
      "step": 27345
    },
    {
      "epoch": 14.334381551362684,
      "grad_norm": 0.6511423587799072,
      "learning_rate": 1.1260364964569466e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 17843000,
      "step": 27350
    },
    {
      "epoch": 14.33700209643606,
      "grad_norm": 0.569607675075531,
      "learning_rate": 1.125081377109223e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 17846872,
      "step": 27355
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 0.4133397340774536,
      "learning_rate": 1.1241265453787882e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 17850232,
      "step": 27360
    },
    {
      "epoch": 14.34224318658281,
      "grad_norm": 0.40519431233406067,
      "learning_rate": 1.1231720014653837e-05,
      "loss": 0.6436,
      "num_input_tokens_seen": 17854360,
      "step": 27365
    },
    {
      "epoch": 14.344863731656185,
      "grad_norm": 0.5184076428413391,
      "learning_rate": 1.1222177455686896e-05,
      "loss": 0.4249,
      "num_input_tokens_seen": 17858040,
      "step": 27370
    },
    {
      "epoch": 14.34748427672956,
      "grad_norm": 0.8070068359375,
      "learning_rate": 1.1212637778883237e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 17861400,
      "step": 27375
    },
    {
      "epoch": 14.350104821802935,
      "grad_norm": 0.9169698357582092,
      "learning_rate": 1.1203100986238468e-05,
      "loss": 0.6467,
      "num_input_tokens_seen": 17865176,
      "step": 27380
    },
    {
      "epoch": 14.35272536687631,
      "grad_norm": 0.6256107091903687,
      "learning_rate": 1.1193567079747561e-05,
      "loss": 0.531,
      "num_input_tokens_seen": 17867544,
      "step": 27385
    },
    {
      "epoch": 14.355345911949685,
      "grad_norm": 0.7808119654655457,
      "learning_rate": 1.1184036061404907e-05,
      "loss": 0.3615,
      "num_input_tokens_seen": 17870584,
      "step": 27390
    },
    {
      "epoch": 14.35796645702306,
      "grad_norm": 0.48498108983039856,
      "learning_rate": 1.117450793320429e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 17873560,
      "step": 27395
    },
    {
      "epoch": 14.360587002096436,
      "grad_norm": 0.39404940605163574,
      "learning_rate": 1.1164982697138893e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 17876824,
      "step": 27400
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 0.9167343974113464,
      "learning_rate": 1.1155460355201267e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 17879928,
      "step": 27405
    },
    {
      "epoch": 14.365828092243186,
      "grad_norm": 0.5406451225280762,
      "learning_rate": 1.1145940909383399e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 17883672,
      "step": 27410
    },
    {
      "epoch": 14.368448637316561,
      "grad_norm": 0.5492587089538574,
      "learning_rate": 1.1136424361676626e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 17887032,
      "step": 27415
    },
    {
      "epoch": 14.371069182389936,
      "grad_norm": 0.7731547951698303,
      "learning_rate": 1.1126910714071711e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 17889816,
      "step": 27420
    },
    {
      "epoch": 14.373689727463312,
      "grad_norm": 0.646270751953125,
      "learning_rate": 1.1117399968558806e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 17893016,
      "step": 27425
    },
    {
      "epoch": 14.376310272536688,
      "grad_norm": 0.4644816815853119,
      "learning_rate": 1.1107892127127454e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 17896056,
      "step": 27430
    },
    {
      "epoch": 14.378930817610064,
      "grad_norm": 0.6471148133277893,
      "learning_rate": 1.109838719176658e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 17899224,
      "step": 27435
    },
    {
      "epoch": 14.381551362683439,
      "grad_norm": 0.4390222728252411,
      "learning_rate": 1.1088885164464497e-05,
      "loss": 0.5282,
      "num_input_tokens_seen": 17902840,
      "step": 27440
    },
    {
      "epoch": 14.384171907756814,
      "grad_norm": 0.8210170269012451,
      "learning_rate": 1.1079386047208932e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 17905272,
      "step": 27445
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 0.5647426843643188,
      "learning_rate": 1.1069889841986991e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 17907992,
      "step": 27450
    },
    {
      "epoch": 14.389412997903564,
      "grad_norm": 0.3335869610309601,
      "learning_rate": 1.1060396550785182e-05,
      "loss": 0.3786,
      "num_input_tokens_seen": 17910904,
      "step": 27455
    },
    {
      "epoch": 14.39203354297694,
      "grad_norm": 0.3898073434829712,
      "learning_rate": 1.1050906175589382e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 17914904,
      "step": 27460
    },
    {
      "epoch": 14.394654088050315,
      "grad_norm": 0.6678040623664856,
      "learning_rate": 1.1041418718384863e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 17918104,
      "step": 27465
    },
    {
      "epoch": 14.39727463312369,
      "grad_norm": 0.39694124460220337,
      "learning_rate": 1.1031934181156298e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 17921720,
      "step": 27470
    },
    {
      "epoch": 14.399895178197065,
      "grad_norm": 0.36473143100738525,
      "learning_rate": 1.1022452565887747e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 17924984,
      "step": 27475
    },
    {
      "epoch": 14.40251572327044,
      "grad_norm": 0.4733085334300995,
      "learning_rate": 1.1012973874562657e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 17927672,
      "step": 27480
    },
    {
      "epoch": 14.405136268343815,
      "grad_norm": 0.22987547516822815,
      "learning_rate": 1.1003498109163865e-05,
      "loss": 0.5558,
      "num_input_tokens_seen": 17931128,
      "step": 27485
    },
    {
      "epoch": 14.40775681341719,
      "grad_norm": 0.4761018455028534,
      "learning_rate": 1.0994025271673592e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 17934936,
      "step": 27490
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 0.5622671842575073,
      "learning_rate": 1.0984555364073431e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 17937432,
      "step": 27495
    },
    {
      "epoch": 14.41299790356394,
      "grad_norm": 0.7873948812484741,
      "learning_rate": 1.0975088388344393e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 17939800,
      "step": 27500
    },
    {
      "epoch": 14.415618448637316,
      "grad_norm": 0.43388116359710693,
      "learning_rate": 1.096562434646686e-05,
      "loss": 0.4645,
      "num_input_tokens_seen": 17942808,
      "step": 27505
    },
    {
      "epoch": 14.418238993710691,
      "grad_norm": 0.7587456107139587,
      "learning_rate": 1.0956163240420611e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 17945592,
      "step": 27510
    },
    {
      "epoch": 14.420859538784066,
      "grad_norm": 0.6490299105644226,
      "learning_rate": 1.094670507218479e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 17948536,
      "step": 27515
    },
    {
      "epoch": 14.423480083857442,
      "grad_norm": 0.47462886571884155,
      "learning_rate": 1.093724984373793e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 17952408,
      "step": 27520
    },
    {
      "epoch": 14.426100628930818,
      "grad_norm": 0.54812091588974,
      "learning_rate": 1.0927797557057968e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 17955896,
      "step": 27525
    },
    {
      "epoch": 14.428721174004194,
      "grad_norm": 0.7792067527770996,
      "learning_rate": 1.0918348214122209e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 17963128,
      "step": 27530
    },
    {
      "epoch": 14.431341719077569,
      "grad_norm": 0.35220304131507874,
      "learning_rate": 1.0908901816907366e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 17965688,
      "step": 27535
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 0.5586534738540649,
      "learning_rate": 1.08994583673895e-05,
      "loss": 0.3372,
      "num_input_tokens_seen": 17968888,
      "step": 27540
    },
    {
      "epoch": 14.43658280922432,
      "grad_norm": 0.4487116038799286,
      "learning_rate": 1.0890017867544068e-05,
      "loss": 0.4053,
      "num_input_tokens_seen": 17972120,
      "step": 27545
    },
    {
      "epoch": 14.439203354297694,
      "grad_norm": 0.7594118714332581,
      "learning_rate": 1.088058031934592e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 17975608,
      "step": 27550
    },
    {
      "epoch": 14.44182389937107,
      "grad_norm": 0.4160321354866028,
      "learning_rate": 1.087114572476929e-05,
      "loss": 0.5509,
      "num_input_tokens_seen": 17979128,
      "step": 27555
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.7781434655189514,
      "learning_rate": 1.0861714085787783e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 17982456,
      "step": 27560
    },
    {
      "epoch": 14.44706498951782,
      "grad_norm": 0.4111287295818329,
      "learning_rate": 1.0852285404374402e-05,
      "loss": 0.5181,
      "num_input_tokens_seen": 17985720,
      "step": 27565
    },
    {
      "epoch": 14.449685534591195,
      "grad_norm": 0.3564583659172058,
      "learning_rate": 1.0842859682501505e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 17989272,
      "step": 27570
    },
    {
      "epoch": 14.45230607966457,
      "grad_norm": 0.5234764218330383,
      "learning_rate": 1.0833436922140841e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 17992408,
      "step": 27575
    },
    {
      "epoch": 14.454926624737945,
      "grad_norm": 0.3193892538547516,
      "learning_rate": 1.0824017125263555e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 17996568,
      "step": 27580
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 0.4383573532104492,
      "learning_rate": 1.0814600293840154e-05,
      "loss": 0.413,
      "num_input_tokens_seen": 17999832,
      "step": 27585
    },
    {
      "epoch": 14.460167714884696,
      "grad_norm": 0.3668787479400635,
      "learning_rate": 1.080518642984055e-05,
      "loss": 0.5359,
      "num_input_tokens_seen": 18003704,
      "step": 27590
    },
    {
      "epoch": 14.46278825995807,
      "grad_norm": 0.903766930103302,
      "learning_rate": 1.0795775535233988e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 18006328,
      "step": 27595
    },
    {
      "epoch": 14.465408805031446,
      "grad_norm": 0.4505167603492737,
      "learning_rate": 1.0786367611989145e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 18010008,
      "step": 27600
    },
    {
      "epoch": 14.468029350104821,
      "grad_norm": 0.7459962964057922,
      "learning_rate": 1.0776962662074031e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 18013880,
      "step": 27605
    },
    {
      "epoch": 14.470649895178196,
      "grad_norm": 0.536212146282196,
      "learning_rate": 1.0767560687456062e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 18016824,
      "step": 27610
    },
    {
      "epoch": 14.473270440251572,
      "grad_norm": 0.6630700826644897,
      "learning_rate": 1.0758161690102036e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 18020472,
      "step": 27615
    },
    {
      "epoch": 14.475890985324948,
      "grad_norm": 0.5953046679496765,
      "learning_rate": 1.0748765671978095e-05,
      "loss": 0.364,
      "num_input_tokens_seen": 18023128,
      "step": 27620
    },
    {
      "epoch": 14.478511530398324,
      "grad_norm": 0.9538819193840027,
      "learning_rate": 1.0739372635049797e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 18026872,
      "step": 27625
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 0.5269328355789185,
      "learning_rate": 1.0729982581282042e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 18030200,
      "step": 27630
    },
    {
      "epoch": 14.483752620545074,
      "grad_norm": 0.5489579439163208,
      "learning_rate": 1.0720595512639129e-05,
      "loss": 0.3376,
      "num_input_tokens_seen": 18033528,
      "step": 27635
    },
    {
      "epoch": 14.48637316561845,
      "grad_norm": 0.7267798781394958,
      "learning_rate": 1.0711211431084728e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 18036088,
      "step": 27640
    },
    {
      "epoch": 14.488993710691824,
      "grad_norm": 0.2513332962989807,
      "learning_rate": 1.0701830338581887e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 18040408,
      "step": 27645
    },
    {
      "epoch": 14.4916142557652,
      "grad_norm": 0.581805408000946,
      "learning_rate": 1.069245223709301e-05,
      "loss": 0.3454,
      "num_input_tokens_seen": 18043096,
      "step": 27650
    },
    {
      "epoch": 14.494234800838575,
      "grad_norm": 0.355742871761322,
      "learning_rate": 1.0683077128579904e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 18046232,
      "step": 27655
    },
    {
      "epoch": 14.49685534591195,
      "grad_norm": 0.6951922178268433,
      "learning_rate": 1.0673705015003718e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 18048792,
      "step": 27660
    },
    {
      "epoch": 14.499475890985325,
      "grad_norm": 0.3905443847179413,
      "learning_rate": 1.0664335898325002e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 18052696,
      "step": 27665
    },
    {
      "epoch": 14.5020964360587,
      "grad_norm": 0.33356988430023193,
      "learning_rate": 1.065496978050367e-05,
      "loss": 0.334,
      "num_input_tokens_seen": 18055256,
      "step": 27670
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 0.35962775349617004,
      "learning_rate": 1.0645606663499e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 18057784,
      "step": 27675
    },
    {
      "epoch": 14.50733752620545,
      "grad_norm": 0.7942711710929871,
      "learning_rate": 1.0636246549269649e-05,
      "loss": 0.5668,
      "num_input_tokens_seen": 18060920,
      "step": 27680
    },
    {
      "epoch": 14.509958071278826,
      "grad_norm": 0.43169471621513367,
      "learning_rate": 1.062688943977366e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 18064184,
      "step": 27685
    },
    {
      "epoch": 14.5125786163522,
      "grad_norm": 0.49205446243286133,
      "learning_rate": 1.0617535336968413e-05,
      "loss": 0.5171,
      "num_input_tokens_seen": 18067672,
      "step": 27690
    },
    {
      "epoch": 14.515199161425576,
      "grad_norm": 0.5537139773368835,
      "learning_rate": 1.0608184242810698e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 18071032,
      "step": 27695
    },
    {
      "epoch": 14.517819706498951,
      "grad_norm": 0.6903154253959656,
      "learning_rate": 1.0598836159256642e-05,
      "loss": 0.6873,
      "num_input_tokens_seen": 18074200,
      "step": 27700
    },
    {
      "epoch": 14.520440251572326,
      "grad_norm": 0.4911239445209503,
      "learning_rate": 1.0589491088261763e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 18076824,
      "step": 27705
    },
    {
      "epoch": 14.523060796645701,
      "grad_norm": 1.1941879987716675,
      "learning_rate": 1.0580149031780945e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 18079512,
      "step": 27710
    },
    {
      "epoch": 14.525681341719078,
      "grad_norm": 0.48817798495292664,
      "learning_rate": 1.0570809991768447e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 18082424,
      "step": 27715
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 0.2998994290828705,
      "learning_rate": 1.0561473970177873e-05,
      "loss": 0.3887,
      "num_input_tokens_seen": 18085400,
      "step": 27720
    },
    {
      "epoch": 14.530922431865829,
      "grad_norm": 0.43778303265571594,
      "learning_rate": 1.0552140968962232e-05,
      "loss": 0.6103,
      "num_input_tokens_seen": 18088664,
      "step": 27725
    },
    {
      "epoch": 14.533542976939204,
      "grad_norm": 1.359464168548584,
      "learning_rate": 1.054281099007386e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 18091512,
      "step": 27730
    },
    {
      "epoch": 14.536163522012579,
      "grad_norm": 0.6890415549278259,
      "learning_rate": 1.0533484035464492e-05,
      "loss": 0.6207,
      "num_input_tokens_seen": 18093752,
      "step": 27735
    },
    {
      "epoch": 14.538784067085954,
      "grad_norm": 0.47603878378868103,
      "learning_rate": 1.0524160107085231e-05,
      "loss": 0.5132,
      "num_input_tokens_seen": 18096440,
      "step": 27740
    },
    {
      "epoch": 14.54140461215933,
      "grad_norm": 0.5311678647994995,
      "learning_rate": 1.0514839206886518e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 18099384,
      "step": 27745
    },
    {
      "epoch": 14.544025157232705,
      "grad_norm": 0.3568607568740845,
      "learning_rate": 1.0505521336818195e-05,
      "loss": 0.4211,
      "num_input_tokens_seen": 18102744,
      "step": 27750
    },
    {
      "epoch": 14.54664570230608,
      "grad_norm": 0.583606481552124,
      "learning_rate": 1.0496206498829438e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 18105720,
      "step": 27755
    },
    {
      "epoch": 14.549266247379455,
      "grad_norm": 0.5345736742019653,
      "learning_rate": 1.0486894694868814e-05,
      "loss": 0.3475,
      "num_input_tokens_seen": 18108536,
      "step": 27760
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 0.9786553382873535,
      "learning_rate": 1.0477585926884246e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 18110904,
      "step": 27765
    },
    {
      "epoch": 14.554507337526205,
      "grad_norm": 0.5160871744155884,
      "learning_rate": 1.046828019682303e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 18113944,
      "step": 27770
    },
    {
      "epoch": 14.55712788259958,
      "grad_norm": 0.39413273334503174,
      "learning_rate": 1.0458977506631808e-05,
      "loss": 0.535,
      "num_input_tokens_seen": 18117304,
      "step": 27775
    },
    {
      "epoch": 14.559748427672956,
      "grad_norm": 0.45086875557899475,
      "learning_rate": 1.0449677858256593e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 18119704,
      "step": 27780
    },
    {
      "epoch": 14.56236897274633,
      "grad_norm": 0.3652489483356476,
      "learning_rate": 1.0440381253642773e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 18124312,
      "step": 27785
    },
    {
      "epoch": 14.564989517819706,
      "grad_norm": 0.6620948910713196,
      "learning_rate": 1.0431087694735087e-05,
      "loss": 0.3921,
      "num_input_tokens_seen": 18128408,
      "step": 27790
    },
    {
      "epoch": 14.567610062893081,
      "grad_norm": 0.46348002552986145,
      "learning_rate": 1.0421797183477644e-05,
      "loss": 0.3752,
      "num_input_tokens_seen": 18131416,
      "step": 27795
    },
    {
      "epoch": 14.570230607966456,
      "grad_norm": 0.46146419644355774,
      "learning_rate": 1.041250972181392e-05,
      "loss": 0.5174,
      "num_input_tokens_seen": 18134872,
      "step": 27800
    },
    {
      "epoch": 14.572851153039831,
      "grad_norm": 0.28866180777549744,
      "learning_rate": 1.0403225311686743e-05,
      "loss": 0.3156,
      "num_input_tokens_seen": 18138776,
      "step": 27805
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 0.7852950096130371,
      "learning_rate": 1.0393943955038293e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 18141752,
      "step": 27810
    },
    {
      "epoch": 14.578092243186584,
      "grad_norm": 0.49494269490242004,
      "learning_rate": 1.038466565381013e-05,
      "loss": 0.3472,
      "num_input_tokens_seen": 18144152,
      "step": 27815
    },
    {
      "epoch": 14.580712788259959,
      "grad_norm": 0.6075924038887024,
      "learning_rate": 1.0375390409943175e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 18147192,
      "step": 27820
    },
    {
      "epoch": 14.583333333333334,
      "grad_norm": 0.28243663907051086,
      "learning_rate": 1.0366118225377708e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 18150456,
      "step": 27825
    },
    {
      "epoch": 14.585953878406709,
      "grad_norm": 0.38707560300827026,
      "learning_rate": 1.0356849102053353e-05,
      "loss": 0.4167,
      "num_input_tokens_seen": 18153464,
      "step": 27830
    },
    {
      "epoch": 14.588574423480084,
      "grad_norm": 0.39747029542922974,
      "learning_rate": 1.0347583041909104e-05,
      "loss": 0.5254,
      "num_input_tokens_seen": 18157336,
      "step": 27835
    },
    {
      "epoch": 14.59119496855346,
      "grad_norm": 0.39772090315818787,
      "learning_rate": 1.0338320046883315e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 18161016,
      "step": 27840
    },
    {
      "epoch": 14.593815513626835,
      "grad_norm": 0.41282954812049866,
      "learning_rate": 1.0329060118913705e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 18163480,
      "step": 27845
    },
    {
      "epoch": 14.59643605870021,
      "grad_norm": 0.5111899375915527,
      "learning_rate": 1.0319803259937353e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 18166584,
      "step": 27850
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 0.4570893943309784,
      "learning_rate": 1.0310549471890676e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 18169144,
      "step": 27855
    },
    {
      "epoch": 14.60167714884696,
      "grad_norm": 0.5423540472984314,
      "learning_rate": 1.0301298756709454e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 18172184,
      "step": 27860
    },
    {
      "epoch": 14.604297693920335,
      "grad_norm": 0.3651004433631897,
      "learning_rate": 1.0292051116328844e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 18174840,
      "step": 27865
    },
    {
      "epoch": 14.60691823899371,
      "grad_norm": 0.4726479649543762,
      "learning_rate": 1.028280655268334e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 18178584,
      "step": 27870
    },
    {
      "epoch": 14.609538784067086,
      "grad_norm": 0.5952588319778442,
      "learning_rate": 1.0273565067706806e-05,
      "loss": 0.415,
      "num_input_tokens_seen": 18184792,
      "step": 27875
    },
    {
      "epoch": 14.61215932914046,
      "grad_norm": 0.29688742756843567,
      "learning_rate": 1.0264326663332457e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 18188248,
      "step": 27880
    },
    {
      "epoch": 14.614779874213836,
      "grad_norm": 0.35524141788482666,
      "learning_rate": 1.0255091341492859e-05,
      "loss": 0.3582,
      "num_input_tokens_seen": 18191160,
      "step": 27885
    },
    {
      "epoch": 14.617400419287211,
      "grad_norm": 0.414098858833313,
      "learning_rate": 1.0245859104119923e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 18195000,
      "step": 27890
    },
    {
      "epoch": 14.620020964360586,
      "grad_norm": 0.8907266855239868,
      "learning_rate": 1.0236629953144938e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 18198232,
      "step": 27895
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 0.46912848949432373,
      "learning_rate": 1.0227403890498537e-05,
      "loss": 0.5731,
      "num_input_tokens_seen": 18200856,
      "step": 27900
    },
    {
      "epoch": 14.625262054507338,
      "grad_norm": 0.6567990183830261,
      "learning_rate": 1.021818091811072e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 18203256,
      "step": 27905
    },
    {
      "epoch": 14.627882599580714,
      "grad_norm": 1.0798312425613403,
      "learning_rate": 1.0208961037910813e-05,
      "loss": 0.5522,
      "num_input_tokens_seen": 18206008,
      "step": 27910
    },
    {
      "epoch": 14.630503144654089,
      "grad_norm": 0.4717761278152466,
      "learning_rate": 1.0199744251827503e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 18208984,
      "step": 27915
    },
    {
      "epoch": 14.633123689727464,
      "grad_norm": 0.5292760133743286,
      "learning_rate": 1.0190530561788845e-05,
      "loss": 0.3115,
      "num_input_tokens_seen": 18213848,
      "step": 27920
    },
    {
      "epoch": 14.635744234800839,
      "grad_norm": 0.49635571241378784,
      "learning_rate": 1.0181319969722236e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 18217496,
      "step": 27925
    },
    {
      "epoch": 14.638364779874214,
      "grad_norm": 0.7088976502418518,
      "learning_rate": 1.0172112477554441e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 18220536,
      "step": 27930
    },
    {
      "epoch": 14.64098532494759,
      "grad_norm": 0.8758652806282043,
      "learning_rate": 1.016290808721154e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 18224888,
      "step": 27935
    },
    {
      "epoch": 14.643605870020965,
      "grad_norm": 0.2714882493019104,
      "learning_rate": 1.0153706800619007e-05,
      "loss": 0.3088,
      "num_input_tokens_seen": 18229368,
      "step": 27940
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 0.5670976638793945,
      "learning_rate": 1.0144508619701623e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 18232760,
      "step": 27945
    },
    {
      "epoch": 14.648846960167715,
      "grad_norm": 0.44884514808654785,
      "learning_rate": 1.013531354638356e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 18235576,
      "step": 27950
    },
    {
      "epoch": 14.65146750524109,
      "grad_norm": 0.3930905759334564,
      "learning_rate": 1.0126121582588316e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 18239224,
      "step": 27955
    },
    {
      "epoch": 14.654088050314465,
      "grad_norm": 0.8853451609611511,
      "learning_rate": 1.0116932730238754e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 18241912,
      "step": 27960
    },
    {
      "epoch": 14.65670859538784,
      "grad_norm": 0.3709086775779724,
      "learning_rate": 1.0107746991257063e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 18244632,
      "step": 27965
    },
    {
      "epoch": 14.659329140461216,
      "grad_norm": 0.5087335705757141,
      "learning_rate": 1.0098564367564811e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 18247768,
      "step": 27970
    },
    {
      "epoch": 14.66194968553459,
      "grad_norm": 0.621911883354187,
      "learning_rate": 1.0089384861082876e-05,
      "loss": 0.381,
      "num_input_tokens_seen": 18250456,
      "step": 27975
    },
    {
      "epoch": 14.664570230607966,
      "grad_norm": 0.48834455013275146,
      "learning_rate": 1.0080208473731523e-05,
      "loss": 0.4376,
      "num_input_tokens_seen": 18253688,
      "step": 27980
    },
    {
      "epoch": 14.667190775681341,
      "grad_norm": 0.32048121094703674,
      "learning_rate": 1.0071035207430352e-05,
      "loss": 0.3669,
      "num_input_tokens_seen": 18256952,
      "step": 27985
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 0.45747944712638855,
      "learning_rate": 1.0061865064098285e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 18263064,
      "step": 27990
    },
    {
      "epoch": 14.672431865828091,
      "grad_norm": 0.49480143189430237,
      "learning_rate": 1.0052698045653625e-05,
      "loss": 0.3739,
      "num_input_tokens_seen": 18265848,
      "step": 27995
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.4090780019760132,
      "learning_rate": 1.0043534154014015e-05,
      "loss": 0.6014,
      "num_input_tokens_seen": 18269208,
      "step": 28000
    },
    {
      "epoch": 14.677672955974844,
      "grad_norm": 0.3662545382976532,
      "learning_rate": 1.003437339109642e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 18274136,
      "step": 28005
    },
    {
      "epoch": 14.680293501048219,
      "grad_norm": 0.30362027883529663,
      "learning_rate": 1.0025215758817183e-05,
      "loss": 0.3516,
      "num_input_tokens_seen": 18277016,
      "step": 28010
    },
    {
      "epoch": 14.682914046121594,
      "grad_norm": 0.6371168494224548,
      "learning_rate": 1.0016061259091958e-05,
      "loss": 0.3646,
      "num_input_tokens_seen": 18279640,
      "step": 28015
    },
    {
      "epoch": 14.685534591194969,
      "grad_norm": 0.8112293481826782,
      "learning_rate": 1.000690989383577e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 18281816,
      "step": 28020
    },
    {
      "epoch": 14.688155136268344,
      "grad_norm": 0.38889893889427185,
      "learning_rate": 9.997761664962993e-06,
      "loss": 0.5621,
      "num_input_tokens_seen": 18284568,
      "step": 28025
    },
    {
      "epoch": 14.69077568134172,
      "grad_norm": 0.37073957920074463,
      "learning_rate": 9.988616574387314e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 18288792,
      "step": 28030
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 0.5351162552833557,
      "learning_rate": 9.979474624021787e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 18292344,
      "step": 28035
    },
    {
      "epoch": 14.69601677148847,
      "grad_norm": 0.49681827425956726,
      "learning_rate": 9.970335815778815e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 18296184,
      "step": 28040
    },
    {
      "epoch": 14.698637316561845,
      "grad_norm": 0.38186749815940857,
      "learning_rate": 9.96120015157011e-06,
      "loss": 0.4862,
      "num_input_tokens_seen": 18299960,
      "step": 28045
    },
    {
      "epoch": 14.70125786163522,
      "grad_norm": 0.3474632203578949,
      "learning_rate": 9.952067633306767e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 18304024,
      "step": 28050
    },
    {
      "epoch": 14.703878406708595,
      "grad_norm": 0.4209977984428406,
      "learning_rate": 9.942938262899204e-06,
      "loss": 0.4697,
      "num_input_tokens_seen": 18306808,
      "step": 28055
    },
    {
      "epoch": 14.70649895178197,
      "grad_norm": 0.9834418296813965,
      "learning_rate": 9.93381204225717e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 18308984,
      "step": 28060
    },
    {
      "epoch": 14.709119496855346,
      "grad_norm": 0.5715864896774292,
      "learning_rate": 9.924688973289782e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 18311416,
      "step": 28065
    },
    {
      "epoch": 14.71174004192872,
      "grad_norm": 0.755233108997345,
      "learning_rate": 9.915569057905463e-06,
      "loss": 0.5413,
      "num_input_tokens_seen": 18314552,
      "step": 28070
    },
    {
      "epoch": 14.714360587002096,
      "grad_norm": 0.480559378862381,
      "learning_rate": 9.906452298012003e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 18317752,
      "step": 28075
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 0.339759886264801,
      "learning_rate": 9.897338695516529e-06,
      "loss": 0.4658,
      "num_input_tokens_seen": 18320888,
      "step": 28080
    },
    {
      "epoch": 14.719601677148846,
      "grad_norm": 0.5275718569755554,
      "learning_rate": 9.888228252325507e-06,
      "loss": 0.6034,
      "num_input_tokens_seen": 18324600,
      "step": 28085
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 0.7673593163490295,
      "learning_rate": 9.87912097034473e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 18328184,
      "step": 28090
    },
    {
      "epoch": 14.724842767295598,
      "grad_norm": 0.37272754311561584,
      "learning_rate": 9.870016851479327e-06,
      "loss": 0.4017,
      "num_input_tokens_seen": 18330680,
      "step": 28095
    },
    {
      "epoch": 14.727463312368974,
      "grad_norm": 0.5160172581672668,
      "learning_rate": 9.860915897633791e-06,
      "loss": 0.4826,
      "num_input_tokens_seen": 18334552,
      "step": 28100
    },
    {
      "epoch": 14.730083857442349,
      "grad_norm": 0.544396698474884,
      "learning_rate": 9.85181811071193e-06,
      "loss": 0.4412,
      "num_input_tokens_seen": 18338168,
      "step": 28105
    },
    {
      "epoch": 14.732704402515724,
      "grad_norm": 0.35150381922721863,
      "learning_rate": 9.842723492616913e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 18341528,
      "step": 28110
    },
    {
      "epoch": 14.735324947589099,
      "grad_norm": 0.41093137860298157,
      "learning_rate": 9.833632045251207e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 18344920,
      "step": 28115
    },
    {
      "epoch": 14.737945492662474,
      "grad_norm": 0.2629348635673523,
      "learning_rate": 9.824543770516656e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 18348312,
      "step": 28120
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 0.5390451550483704,
      "learning_rate": 9.815458670314412e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 18351160,
      "step": 28125
    },
    {
      "epoch": 14.743186582809225,
      "grad_norm": 1.1741244792938232,
      "learning_rate": 9.80637674654498e-06,
      "loss": 0.5164,
      "num_input_tokens_seen": 18354104,
      "step": 28130
    },
    {
      "epoch": 14.7458071278826,
      "grad_norm": 0.5449249744415283,
      "learning_rate": 9.797298001108193e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 18357784,
      "step": 28135
    },
    {
      "epoch": 14.748427672955975,
      "grad_norm": 0.4418110251426697,
      "learning_rate": 9.788222435903232e-06,
      "loss": 0.5371,
      "num_input_tokens_seen": 18360984,
      "step": 28140
    },
    {
      "epoch": 14.75104821802935,
      "grad_norm": 0.3629049062728882,
      "learning_rate": 9.779150052828596e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 18365080,
      "step": 28145
    },
    {
      "epoch": 14.753668763102725,
      "grad_norm": 0.5206747055053711,
      "learning_rate": 9.770080853782108e-06,
      "loss": 0.5138,
      "num_input_tokens_seen": 18367800,
      "step": 28150
    },
    {
      "epoch": 14.7562893081761,
      "grad_norm": 0.9713367223739624,
      "learning_rate": 9.76101484066095e-06,
      "loss": 0.6432,
      "num_input_tokens_seen": 18370392,
      "step": 28155
    },
    {
      "epoch": 14.758909853249476,
      "grad_norm": 0.3026851713657379,
      "learning_rate": 9.751952015361637e-06,
      "loss": 0.4692,
      "num_input_tokens_seen": 18373560,
      "step": 28160
    },
    {
      "epoch": 14.76153039832285,
      "grad_norm": 0.4716787040233612,
      "learning_rate": 9.742892379780003e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 18378296,
      "step": 28165
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 0.6320344805717468,
      "learning_rate": 9.733835935811237e-06,
      "loss": 0.4523,
      "num_input_tokens_seen": 18381016,
      "step": 28170
    },
    {
      "epoch": 14.766771488469601,
      "grad_norm": 0.4083341956138611,
      "learning_rate": 9.724782685349809e-06,
      "loss": 0.4939,
      "num_input_tokens_seen": 18384248,
      "step": 28175
    },
    {
      "epoch": 14.769392033542976,
      "grad_norm": 1.4514824151992798,
      "learning_rate": 9.715732630289576e-06,
      "loss": 0.5268,
      "num_input_tokens_seen": 18386648,
      "step": 28180
    },
    {
      "epoch": 14.772012578616351,
      "grad_norm": 0.4186674654483795,
      "learning_rate": 9.706685772523701e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 18390232,
      "step": 28185
    },
    {
      "epoch": 14.774633123689728,
      "grad_norm": 0.4999680817127228,
      "learning_rate": 9.697642113944692e-06,
      "loss": 0.4781,
      "num_input_tokens_seen": 18393624,
      "step": 28190
    },
    {
      "epoch": 14.777253668763104,
      "grad_norm": 0.48005592823028564,
      "learning_rate": 9.688601656444381e-06,
      "loss": 0.3596,
      "num_input_tokens_seen": 18397176,
      "step": 28195
    },
    {
      "epoch": 14.779874213836479,
      "grad_norm": 0.42248424887657166,
      "learning_rate": 9.679564401913921e-06,
      "loss": 0.4735,
      "num_input_tokens_seen": 18400472,
      "step": 28200
    },
    {
      "epoch": 14.782494758909854,
      "grad_norm": 0.49835270643234253,
      "learning_rate": 9.670530352243798e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 18403224,
      "step": 28205
    },
    {
      "epoch": 14.785115303983229,
      "grad_norm": 0.45914924144744873,
      "learning_rate": 9.661499509323835e-06,
      "loss": 0.332,
      "num_input_tokens_seen": 18406200,
      "step": 28210
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 0.5171177387237549,
      "learning_rate": 9.652471875043182e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 18409432,
      "step": 28215
    },
    {
      "epoch": 14.79035639412998,
      "grad_norm": 0.3740355968475342,
      "learning_rate": 9.64344745129033e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 18411736,
      "step": 28220
    },
    {
      "epoch": 14.792976939203355,
      "grad_norm": 0.4162824749946594,
      "learning_rate": 9.634426239953073e-06,
      "loss": 0.4526,
      "num_input_tokens_seen": 18415448,
      "step": 28225
    },
    {
      "epoch": 14.79559748427673,
      "grad_norm": 0.38272643089294434,
      "learning_rate": 9.62540824291854e-06,
      "loss": 0.5225,
      "num_input_tokens_seen": 18419032,
      "step": 28230
    },
    {
      "epoch": 14.798218029350105,
      "grad_norm": 0.47913751006126404,
      "learning_rate": 9.616393462073197e-06,
      "loss": 0.4778,
      "num_input_tokens_seen": 18422136,
      "step": 28235
    },
    {
      "epoch": 14.80083857442348,
      "grad_norm": 0.4448185861110687,
      "learning_rate": 9.607381899302839e-06,
      "loss": 0.6297,
      "num_input_tokens_seen": 18424696,
      "step": 28240
    },
    {
      "epoch": 14.803459119496855,
      "grad_norm": 0.6071577072143555,
      "learning_rate": 9.59837355649258e-06,
      "loss": 0.5171,
      "num_input_tokens_seen": 18427704,
      "step": 28245
    },
    {
      "epoch": 14.80607966457023,
      "grad_norm": 0.6538960933685303,
      "learning_rate": 9.589368435526868e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 18430328,
      "step": 28250
    },
    {
      "epoch": 14.808700209643606,
      "grad_norm": 0.6962639689445496,
      "learning_rate": 9.580366538289468e-06,
      "loss": 0.5065,
      "num_input_tokens_seen": 18432920,
      "step": 28255
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 0.7185736894607544,
      "learning_rate": 9.57136786666346e-06,
      "loss": 0.5409,
      "num_input_tokens_seen": 18435896,
      "step": 28260
    },
    {
      "epoch": 14.813941299790356,
      "grad_norm": 0.7335108518600464,
      "learning_rate": 9.562372422531276e-06,
      "loss": 0.4251,
      "num_input_tokens_seen": 18438712,
      "step": 28265
    },
    {
      "epoch": 14.816561844863731,
      "grad_norm": 0.5597232580184937,
      "learning_rate": 9.553380207774662e-06,
      "loss": 0.4526,
      "num_input_tokens_seen": 18441592,
      "step": 28270
    },
    {
      "epoch": 14.819182389937106,
      "grad_norm": 0.3735893666744232,
      "learning_rate": 9.54439122427469e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 18444120,
      "step": 28275
    },
    {
      "epoch": 14.821802935010481,
      "grad_norm": 0.5577102899551392,
      "learning_rate": 9.535405473911748e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 18447640,
      "step": 28280
    },
    {
      "epoch": 14.824423480083858,
      "grad_norm": 0.20533159375190735,
      "learning_rate": 9.526422958565546e-06,
      "loss": 0.4831,
      "num_input_tokens_seen": 18452312,
      "step": 28285
    },
    {
      "epoch": 14.827044025157234,
      "grad_norm": 0.417564332485199,
      "learning_rate": 9.517443680115124e-06,
      "loss": 0.434,
      "num_input_tokens_seen": 18455704,
      "step": 28290
    },
    {
      "epoch": 14.829664570230609,
      "grad_norm": 0.4072856605052948,
      "learning_rate": 9.50846764043885e-06,
      "loss": 0.3981,
      "num_input_tokens_seen": 18459384,
      "step": 28295
    },
    {
      "epoch": 14.832285115303984,
      "grad_norm": 0.5673676133155823,
      "learning_rate": 9.499494841414422e-06,
      "loss": 0.5585,
      "num_input_tokens_seen": 18463512,
      "step": 28300
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 0.6580725908279419,
      "learning_rate": 9.490525284918822e-06,
      "loss": 0.5419,
      "num_input_tokens_seen": 18466520,
      "step": 28305
    },
    {
      "epoch": 14.837526205450734,
      "grad_norm": 0.508817195892334,
      "learning_rate": 9.4815589728284e-06,
      "loss": 0.4012,
      "num_input_tokens_seen": 18469240,
      "step": 28310
    },
    {
      "epoch": 14.84014675052411,
      "grad_norm": 0.2492373138666153,
      "learning_rate": 9.472595907018788e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 18473272,
      "step": 28315
    },
    {
      "epoch": 14.842767295597485,
      "grad_norm": 0.25258079171180725,
      "learning_rate": 9.463636089364968e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 18478392,
      "step": 28320
    },
    {
      "epoch": 14.84538784067086,
      "grad_norm": 0.45152825117111206,
      "learning_rate": 9.454679521741239e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 18481048,
      "step": 28325
    },
    {
      "epoch": 14.848008385744235,
      "grad_norm": 0.36260107159614563,
      "learning_rate": 9.445726206021194e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 18484376,
      "step": 28330
    },
    {
      "epoch": 14.85062893081761,
      "grad_norm": 0.35340800881385803,
      "learning_rate": 9.436776144077772e-06,
      "loss": 0.3191,
      "num_input_tokens_seen": 18487928,
      "step": 28335
    },
    {
      "epoch": 14.853249475890985,
      "grad_norm": 0.4278166592121124,
      "learning_rate": 9.427829337783237e-06,
      "loss": 0.5688,
      "num_input_tokens_seen": 18491192,
      "step": 28340
    },
    {
      "epoch": 14.85587002096436,
      "grad_norm": 0.5764255523681641,
      "learning_rate": 9.418885789009138e-06,
      "loss": 0.3763,
      "num_input_tokens_seen": 18494744,
      "step": 28345
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 0.3479382395744324,
      "learning_rate": 9.409945499626374e-06,
      "loss": 0.4895,
      "num_input_tokens_seen": 18498584,
      "step": 28350
    },
    {
      "epoch": 14.86111111111111,
      "grad_norm": 0.5277671813964844,
      "learning_rate": 9.40100847150516e-06,
      "loss": 0.5177,
      "num_input_tokens_seen": 18501400,
      "step": 28355
    },
    {
      "epoch": 14.863731656184486,
      "grad_norm": 0.6476327776908875,
      "learning_rate": 9.392074706515003e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 18503864,
      "step": 28360
    },
    {
      "epoch": 14.866352201257861,
      "grad_norm": 0.3730814754962921,
      "learning_rate": 9.383144206524753e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 18506712,
      "step": 28365
    },
    {
      "epoch": 14.868972746331236,
      "grad_norm": 0.5190721750259399,
      "learning_rate": 9.374216973402575e-06,
      "loss": 0.4721,
      "num_input_tokens_seen": 18510360,
      "step": 28370
    },
    {
      "epoch": 14.871593291404611,
      "grad_norm": 0.43906930088996887,
      "learning_rate": 9.365293009015926e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 18513432,
      "step": 28375
    },
    {
      "epoch": 14.874213836477988,
      "grad_norm": 0.7476726770401001,
      "learning_rate": 9.356372315231623e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 18516632,
      "step": 28380
    },
    {
      "epoch": 14.876834381551364,
      "grad_norm": 0.3537196218967438,
      "learning_rate": 9.347454893915749e-06,
      "loss": 0.5173,
      "num_input_tokens_seen": 18521624,
      "step": 28385
    },
    {
      "epoch": 14.879454926624739,
      "grad_norm": 0.35662898421287537,
      "learning_rate": 9.338540746933741e-06,
      "loss": 0.4919,
      "num_input_tokens_seen": 18524760,
      "step": 28390
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 0.9257522821426392,
      "learning_rate": 9.329629876150339e-06,
      "loss": 0.3929,
      "num_input_tokens_seen": 18527416,
      "step": 28395
    },
    {
      "epoch": 14.884696016771489,
      "grad_norm": 0.44938185811042786,
      "learning_rate": 9.320722283429581e-06,
      "loss": 0.4866,
      "num_input_tokens_seen": 18531032,
      "step": 28400
    },
    {
      "epoch": 14.887316561844864,
      "grad_norm": 0.3077608048915863,
      "learning_rate": 9.311817970634854e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 18534744,
      "step": 28405
    },
    {
      "epoch": 14.88993710691824,
      "grad_norm": 0.4757938086986542,
      "learning_rate": 9.302916939628814e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 18538392,
      "step": 28410
    },
    {
      "epoch": 14.892557651991615,
      "grad_norm": 0.6558445692062378,
      "learning_rate": 9.29401919227347e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 18541400,
      "step": 28415
    },
    {
      "epoch": 14.89517819706499,
      "grad_norm": 1.3240702152252197,
      "learning_rate": 9.28512473043013e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 18544696,
      "step": 28420
    },
    {
      "epoch": 14.897798742138365,
      "grad_norm": 0.4022192656993866,
      "learning_rate": 9.276233555959419e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 18547864,
      "step": 28425
    },
    {
      "epoch": 14.90041928721174,
      "grad_norm": 0.5853098630905151,
      "learning_rate": 9.267345670721254e-06,
      "loss": 0.4248,
      "num_input_tokens_seen": 18550584,
      "step": 28430
    },
    {
      "epoch": 14.903039832285115,
      "grad_norm": 0.8444446325302124,
      "learning_rate": 9.258461076574895e-06,
      "loss": 0.5356,
      "num_input_tokens_seen": 18553624,
      "step": 28435
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 0.5271971225738525,
      "learning_rate": 9.24957977537888e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 18556472,
      "step": 28440
    },
    {
      "epoch": 14.908280922431866,
      "grad_norm": 0.48987558484077454,
      "learning_rate": 9.240701768991086e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 18558840,
      "step": 28445
    },
    {
      "epoch": 14.91090146750524,
      "grad_norm": 0.5672153830528259,
      "learning_rate": 9.231827059268692e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 18562008,
      "step": 28450
    },
    {
      "epoch": 14.913522012578616,
      "grad_norm": 0.3430073857307434,
      "learning_rate": 9.222955648068193e-06,
      "loss": 0.5232,
      "num_input_tokens_seen": 18564952,
      "step": 28455
    },
    {
      "epoch": 14.916142557651991,
      "grad_norm": 0.3944774866104126,
      "learning_rate": 9.214087537245378e-06,
      "loss": 0.4216,
      "num_input_tokens_seen": 18567800,
      "step": 28460
    },
    {
      "epoch": 14.918763102725366,
      "grad_norm": 0.5943798422813416,
      "learning_rate": 9.205222728655344e-06,
      "loss": 0.4944,
      "num_input_tokens_seen": 18571096,
      "step": 28465
    },
    {
      "epoch": 14.921383647798741,
      "grad_norm": 0.3694209158420563,
      "learning_rate": 9.196361224152522e-06,
      "loss": 0.3441,
      "num_input_tokens_seen": 18574136,
      "step": 28470
    },
    {
      "epoch": 14.924004192872118,
      "grad_norm": 1.2370924949645996,
      "learning_rate": 9.187503025590632e-06,
      "loss": 0.4634,
      "num_input_tokens_seen": 18576664,
      "step": 28475
    },
    {
      "epoch": 14.926624737945493,
      "grad_norm": 0.3998202979564667,
      "learning_rate": 9.17864813482272e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 18580824,
      "step": 28480
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 0.6529139876365662,
      "learning_rate": 9.169796553701122e-06,
      "loss": 0.3945,
      "num_input_tokens_seen": 18583864,
      "step": 28485
    },
    {
      "epoch": 14.931865828092244,
      "grad_norm": 0.6349070072174072,
      "learning_rate": 9.160948284077473e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 18586680,
      "step": 28490
    },
    {
      "epoch": 14.934486373165619,
      "grad_norm": 0.26279646158218384,
      "learning_rate": 9.152103327802738e-06,
      "loss": 0.3441,
      "num_input_tokens_seen": 18590776,
      "step": 28495
    },
    {
      "epoch": 14.937106918238994,
      "grad_norm": 0.5289042592048645,
      "learning_rate": 9.143261686727189e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 18594616,
      "step": 28500
    },
    {
      "epoch": 14.93972746331237,
      "grad_norm": 0.4664541482925415,
      "learning_rate": 9.134423362700392e-06,
      "loss": 0.5423,
      "num_input_tokens_seen": 18597528,
      "step": 28505
    },
    {
      "epoch": 14.942348008385745,
      "grad_norm": 0.775137186050415,
      "learning_rate": 9.12558835757123e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 18599992,
      "step": 28510
    },
    {
      "epoch": 14.94496855345912,
      "grad_norm": 1.0901044607162476,
      "learning_rate": 9.116756673187878e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 18602872,
      "step": 28515
    },
    {
      "epoch": 14.947589098532495,
      "grad_norm": 0.4416857361793518,
      "learning_rate": 9.107928311397817e-06,
      "loss": 0.4119,
      "num_input_tokens_seen": 18606360,
      "step": 28520
    },
    {
      "epoch": 14.95020964360587,
      "grad_norm": 0.485257089138031,
      "learning_rate": 9.09910327404785e-06,
      "loss": 0.3651,
      "num_input_tokens_seen": 18608824,
      "step": 28525
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.31645962595939636,
      "learning_rate": 9.090281562984068e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 18611864,
      "step": 28530
    },
    {
      "epoch": 14.95545073375262,
      "grad_norm": 0.2448282092809677,
      "learning_rate": 9.081463180051883e-06,
      "loss": 0.5282,
      "num_input_tokens_seen": 18615480,
      "step": 28535
    },
    {
      "epoch": 14.958071278825996,
      "grad_norm": 0.40688398480415344,
      "learning_rate": 9.072648127095996e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 18618392,
      "step": 28540
    },
    {
      "epoch": 14.96069182389937,
      "grad_norm": 0.6093529462814331,
      "learning_rate": 9.0638364059604e-06,
      "loss": 0.5238,
      "num_input_tokens_seen": 18622200,
      "step": 28545
    },
    {
      "epoch": 14.963312368972746,
      "grad_norm": 0.2649950385093689,
      "learning_rate": 9.055028018488417e-06,
      "loss": 0.4513,
      "num_input_tokens_seen": 18625080,
      "step": 28550
    },
    {
      "epoch": 14.965932914046121,
      "grad_norm": 0.40560251474380493,
      "learning_rate": 9.046222966522665e-06,
      "loss": 0.4837,
      "num_input_tokens_seen": 18627992,
      "step": 28555
    },
    {
      "epoch": 14.968553459119496,
      "grad_norm": 0.4406459629535675,
      "learning_rate": 9.037421251905059e-06,
      "loss": 0.4148,
      "num_input_tokens_seen": 18631192,
      "step": 28560
    },
    {
      "epoch": 14.971174004192871,
      "grad_norm": 0.4620476961135864,
      "learning_rate": 9.028622876476833e-06,
      "loss": 0.3821,
      "num_input_tokens_seen": 18634904,
      "step": 28565
    },
    {
      "epoch": 14.973794549266248,
      "grad_norm": 0.6035802364349365,
      "learning_rate": 9.019827842078474e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 18637784,
      "step": 28570
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 1.0014262199401855,
      "learning_rate": 9.011036150549817e-06,
      "loss": 0.438,
      "num_input_tokens_seen": 18640120,
      "step": 28575
    },
    {
      "epoch": 14.979035639412999,
      "grad_norm": 0.5800977945327759,
      "learning_rate": 9.002247803729987e-06,
      "loss": 0.5888,
      "num_input_tokens_seen": 18643672,
      "step": 28580
    },
    {
      "epoch": 14.981656184486374,
      "grad_norm": 0.5348493456840515,
      "learning_rate": 8.993462803457404e-06,
      "loss": 0.5408,
      "num_input_tokens_seen": 18646872,
      "step": 28585
    },
    {
      "epoch": 14.984276729559749,
      "grad_norm": 0.6892058849334717,
      "learning_rate": 8.984681151569799e-06,
      "loss": 0.3265,
      "num_input_tokens_seen": 18650744,
      "step": 28590
    },
    {
      "epoch": 14.986897274633124,
      "grad_norm": 0.5488628149032593,
      "learning_rate": 8.975902849904183e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 18654296,
      "step": 28595
    },
    {
      "epoch": 14.9895178197065,
      "grad_norm": 0.5651209354400635,
      "learning_rate": 8.967127900296874e-06,
      "loss": 0.4044,
      "num_input_tokens_seen": 18657560,
      "step": 28600
    },
    {
      "epoch": 14.992138364779874,
      "grad_norm": 0.42571306228637695,
      "learning_rate": 8.95835630458349e-06,
      "loss": 0.4292,
      "num_input_tokens_seen": 18660728,
      "step": 28605
    },
    {
      "epoch": 14.99475890985325,
      "grad_norm": 0.617886483669281,
      "learning_rate": 8.949588064598958e-06,
      "loss": 0.4207,
      "num_input_tokens_seen": 18663352,
      "step": 28610
    },
    {
      "epoch": 14.997379454926625,
      "grad_norm": 0.8657456636428833,
      "learning_rate": 8.940823182177499e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 18665816,
      "step": 28615
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.9788236021995544,
      "learning_rate": 8.932061659152607e-06,
      "loss": 0.3595,
      "num_input_tokens_seen": 18668536,
      "step": 28620
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.45286494493484497,
      "eval_runtime": 9.7193,
      "eval_samples_per_second": 87.249,
      "eval_steps_per_second": 21.812,
      "num_input_tokens_seen": 18668536,
      "step": 28620
    },
    {
      "epoch": 15.002620545073375,
      "grad_norm": 0.5765998959541321,
      "learning_rate": 8.92330349735711e-06,
      "loss": 0.7118,
      "num_input_tokens_seen": 18672024,
      "step": 28625
    },
    {
      "epoch": 15.00524109014675,
      "grad_norm": 0.42363113164901733,
      "learning_rate": 8.914548698623104e-06,
      "loss": 0.5433,
      "num_input_tokens_seen": 18674712,
      "step": 28630
    },
    {
      "epoch": 15.007861635220126,
      "grad_norm": 0.506928563117981,
      "learning_rate": 8.905797264781995e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 18682488,
      "step": 28635
    },
    {
      "epoch": 15.0104821802935,
      "grad_norm": 0.2685156464576721,
      "learning_rate": 8.89704919766449e-06,
      "loss": 0.4284,
      "num_input_tokens_seen": 18685816,
      "step": 28640
    },
    {
      "epoch": 15.013102725366876,
      "grad_norm": 0.5611968636512756,
      "learning_rate": 8.888304499100586e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 18688760,
      "step": 28645
    },
    {
      "epoch": 15.015723270440251,
      "grad_norm": 0.3407368063926697,
      "learning_rate": 8.879563170919566e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 18691160,
      "step": 28650
    },
    {
      "epoch": 15.018343815513626,
      "grad_norm": 0.5904470682144165,
      "learning_rate": 8.870825214950016e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 18694616,
      "step": 28655
    },
    {
      "epoch": 15.020964360587001,
      "grad_norm": 0.40709376335144043,
      "learning_rate": 8.862090633019818e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 18697656,
      "step": 28660
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 0.424934446811676,
      "learning_rate": 8.853359426956146e-06,
      "loss": 0.3738,
      "num_input_tokens_seen": 18700536,
      "step": 28665
    },
    {
      "epoch": 15.026205450733753,
      "grad_norm": 0.8022681474685669,
      "learning_rate": 8.844631598585482e-06,
      "loss": 0.549,
      "num_input_tokens_seen": 18703960,
      "step": 28670
    },
    {
      "epoch": 15.028825995807129,
      "grad_norm": 0.6118645668029785,
      "learning_rate": 8.835907149733569e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 18706744,
      "step": 28675
    },
    {
      "epoch": 15.031446540880504,
      "grad_norm": 0.3537578284740448,
      "learning_rate": 8.827186082225477e-06,
      "loss": 0.4361,
      "num_input_tokens_seen": 18710424,
      "step": 28680
    },
    {
      "epoch": 15.034067085953879,
      "grad_norm": 0.3764192461967468,
      "learning_rate": 8.81846839788554e-06,
      "loss": 0.5595,
      "num_input_tokens_seen": 18715640,
      "step": 28685
    },
    {
      "epoch": 15.036687631027254,
      "grad_norm": 0.26726481318473816,
      "learning_rate": 8.809754098537407e-06,
      "loss": 0.5373,
      "num_input_tokens_seen": 18718680,
      "step": 28690
    },
    {
      "epoch": 15.03930817610063,
      "grad_norm": 0.2772156000137329,
      "learning_rate": 8.801043186004016e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 18721688,
      "step": 28695
    },
    {
      "epoch": 15.041928721174004,
      "grad_norm": 0.4112829566001892,
      "learning_rate": 8.792335662107575e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 18724920,
      "step": 28700
    },
    {
      "epoch": 15.04454926624738,
      "grad_norm": 0.7725711464881897,
      "learning_rate": 8.783631528669605e-06,
      "loss": 0.585,
      "num_input_tokens_seen": 18728376,
      "step": 28705
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 0.8964946866035461,
      "learning_rate": 8.774930787510924e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 18731704,
      "step": 28710
    },
    {
      "epoch": 15.04979035639413,
      "grad_norm": 0.4319470524787903,
      "learning_rate": 8.766233440451608e-06,
      "loss": 0.5007,
      "num_input_tokens_seen": 18734968,
      "step": 28715
    },
    {
      "epoch": 15.052410901467505,
      "grad_norm": 0.5206273794174194,
      "learning_rate": 8.75753948931105e-06,
      "loss": 0.4907,
      "num_input_tokens_seen": 18738680,
      "step": 28720
    },
    {
      "epoch": 15.05503144654088,
      "grad_norm": 0.36139023303985596,
      "learning_rate": 8.748848935907936e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 18742136,
      "step": 28725
    },
    {
      "epoch": 15.057651991614255,
      "grad_norm": 0.33579355478286743,
      "learning_rate": 8.740161782060214e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 18745016,
      "step": 28730
    },
    {
      "epoch": 15.06027253668763,
      "grad_norm": 0.43236443400382996,
      "learning_rate": 8.731478029585144e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 18748920,
      "step": 28735
    },
    {
      "epoch": 15.062893081761006,
      "grad_norm": 0.7281990647315979,
      "learning_rate": 8.722797680299278e-06,
      "loss": 0.447,
      "num_input_tokens_seen": 18753560,
      "step": 28740
    },
    {
      "epoch": 15.065513626834381,
      "grad_norm": 0.3342204689979553,
      "learning_rate": 8.71412073601843e-06,
      "loss": 0.5419,
      "num_input_tokens_seen": 18756952,
      "step": 28745
    },
    {
      "epoch": 15.068134171907756,
      "grad_norm": 0.2868773639202118,
      "learning_rate": 8.705447198557731e-06,
      "loss": 0.4882,
      "num_input_tokens_seen": 18760472,
      "step": 28750
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 0.5061671733856201,
      "learning_rate": 8.696777069731574e-06,
      "loss": 0.4533,
      "num_input_tokens_seen": 18763928,
      "step": 28755
    },
    {
      "epoch": 15.073375262054507,
      "grad_norm": 0.7684540152549744,
      "learning_rate": 8.688110351353654e-06,
      "loss": 0.4615,
      "num_input_tokens_seen": 18767512,
      "step": 28760
    },
    {
      "epoch": 15.075995807127883,
      "grad_norm": 0.43927234411239624,
      "learning_rate": 8.679447045236962e-06,
      "loss": 0.5145,
      "num_input_tokens_seen": 18774040,
      "step": 28765
    },
    {
      "epoch": 15.078616352201259,
      "grad_norm": 0.9582570791244507,
      "learning_rate": 8.670787153193746e-06,
      "loss": 0.3898,
      "num_input_tokens_seen": 18777336,
      "step": 28770
    },
    {
      "epoch": 15.081236897274634,
      "grad_norm": 0.4824061095714569,
      "learning_rate": 8.662130677035574e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 18780504,
      "step": 28775
    },
    {
      "epoch": 15.083857442348009,
      "grad_norm": 0.33573389053344727,
      "learning_rate": 8.65347761857326e-06,
      "loss": 0.6225,
      "num_input_tokens_seen": 18783800,
      "step": 28780
    },
    {
      "epoch": 15.086477987421384,
      "grad_norm": 0.544387936592102,
      "learning_rate": 8.64482797961694e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 18786968,
      "step": 28785
    },
    {
      "epoch": 15.08909853249476,
      "grad_norm": 0.48012775182724,
      "learning_rate": 8.636181761976016e-06,
      "loss": 0.4135,
      "num_input_tokens_seen": 18789720,
      "step": 28790
    },
    {
      "epoch": 15.091719077568134,
      "grad_norm": 0.8642719388008118,
      "learning_rate": 8.62753896745919e-06,
      "loss": 0.3941,
      "num_input_tokens_seen": 18792952,
      "step": 28795
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 0.39602699875831604,
      "learning_rate": 8.618899597874413e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 18795480,
      "step": 28800
    },
    {
      "epoch": 15.096960167714885,
      "grad_norm": 0.7699505686759949,
      "learning_rate": 8.610263655028964e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 18798648,
      "step": 28805
    },
    {
      "epoch": 15.09958071278826,
      "grad_norm": 0.5375533699989319,
      "learning_rate": 8.601631140729366e-06,
      "loss": 0.5107,
      "num_input_tokens_seen": 18802680,
      "step": 28810
    },
    {
      "epoch": 15.102201257861635,
      "grad_norm": 0.6095139384269714,
      "learning_rate": 8.593002056781451e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 18805112,
      "step": 28815
    },
    {
      "epoch": 15.10482180293501,
      "grad_norm": 0.556999146938324,
      "learning_rate": 8.584376404990326e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 18809272,
      "step": 28820
    },
    {
      "epoch": 15.107442348008385,
      "grad_norm": 0.36899179220199585,
      "learning_rate": 8.575754187160384e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 18812440,
      "step": 28825
    },
    {
      "epoch": 15.11006289308176,
      "grad_norm": 0.4739876687526703,
      "learning_rate": 8.567135405095291e-06,
      "loss": 0.5726,
      "num_input_tokens_seen": 18815288,
      "step": 28830
    },
    {
      "epoch": 15.112683438155136,
      "grad_norm": 0.6785800457000732,
      "learning_rate": 8.558520060597985e-06,
      "loss": 0.5408,
      "num_input_tokens_seen": 18818424,
      "step": 28835
    },
    {
      "epoch": 15.115303983228511,
      "grad_norm": 0.4416884779930115,
      "learning_rate": 8.54990815547071e-06,
      "loss": 0.4981,
      "num_input_tokens_seen": 18821144,
      "step": 28840
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 0.39249372482299805,
      "learning_rate": 8.541299691514974e-06,
      "loss": 0.3656,
      "num_input_tokens_seen": 18824632,
      "step": 28845
    },
    {
      "epoch": 15.120545073375261,
      "grad_norm": 0.5058260560035706,
      "learning_rate": 8.53269467053158e-06,
      "loss": 0.4968,
      "num_input_tokens_seen": 18827224,
      "step": 28850
    },
    {
      "epoch": 15.123165618448636,
      "grad_norm": 0.39983487129211426,
      "learning_rate": 8.524093094320593e-06,
      "loss": 0.466,
      "num_input_tokens_seen": 18830072,
      "step": 28855
    },
    {
      "epoch": 15.125786163522013,
      "grad_norm": 0.5938878059387207,
      "learning_rate": 8.515494964681357e-06,
      "loss": 0.5299,
      "num_input_tokens_seen": 18832856,
      "step": 28860
    },
    {
      "epoch": 15.128406708595389,
      "grad_norm": 1.200000524520874,
      "learning_rate": 8.506900283412506e-06,
      "loss": 0.6321,
      "num_input_tokens_seen": 18835544,
      "step": 28865
    },
    {
      "epoch": 15.131027253668764,
      "grad_norm": 0.4279043674468994,
      "learning_rate": 8.498309052311953e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 18838360,
      "step": 28870
    },
    {
      "epoch": 15.133647798742139,
      "grad_norm": 0.32210472226142883,
      "learning_rate": 8.489721273176887e-06,
      "loss": 0.5767,
      "num_input_tokens_seen": 18842168,
      "step": 28875
    },
    {
      "epoch": 15.136268343815514,
      "grad_norm": 0.42276179790496826,
      "learning_rate": 8.481136947803777e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 18845496,
      "step": 28880
    },
    {
      "epoch": 15.13888888888889,
      "grad_norm": 0.4547508656978607,
      "learning_rate": 8.472556077988362e-06,
      "loss": 0.334,
      "num_input_tokens_seen": 18850232,
      "step": 28885
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 0.5930312275886536,
      "learning_rate": 8.46397866552565e-06,
      "loss": 0.4917,
      "num_input_tokens_seen": 18852536,
      "step": 28890
    },
    {
      "epoch": 15.14412997903564,
      "grad_norm": 0.3600548803806305,
      "learning_rate": 8.455404712209952e-06,
      "loss": 0.5046,
      "num_input_tokens_seen": 18855544,
      "step": 28895
    },
    {
      "epoch": 15.146750524109015,
      "grad_norm": 0.6118006110191345,
      "learning_rate": 8.446834219834836e-06,
      "loss": 0.5409,
      "num_input_tokens_seen": 18858584,
      "step": 28900
    },
    {
      "epoch": 15.14937106918239,
      "grad_norm": 0.3422732651233673,
      "learning_rate": 8.43826719019316e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 18861688,
      "step": 28905
    },
    {
      "epoch": 15.151991614255765,
      "grad_norm": 0.7075667977333069,
      "learning_rate": 8.429703625077043e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 18864056,
      "step": 28910
    },
    {
      "epoch": 15.15461215932914,
      "grad_norm": 0.4189656674861908,
      "learning_rate": 8.421143526277875e-06,
      "loss": 0.4122,
      "num_input_tokens_seen": 18867960,
      "step": 28915
    },
    {
      "epoch": 15.157232704402515,
      "grad_norm": 0.9753557443618774,
      "learning_rate": 8.41258689558634e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 18870328,
      "step": 28920
    },
    {
      "epoch": 15.15985324947589,
      "grad_norm": 0.3704627454280853,
      "learning_rate": 8.404033734792386e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 18873976,
      "step": 28925
    },
    {
      "epoch": 15.162473794549266,
      "grad_norm": 0.4158569574356079,
      "learning_rate": 8.395484045685248e-06,
      "loss": 0.3008,
      "num_input_tokens_seen": 18877112,
      "step": 28930
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 0.5929554104804993,
      "learning_rate": 8.386937830053412e-06,
      "loss": 0.4435,
      "num_input_tokens_seen": 18879480,
      "step": 28935
    },
    {
      "epoch": 15.167714884696016,
      "grad_norm": 0.5977626442909241,
      "learning_rate": 8.37839508968464e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 18882072,
      "step": 28940
    },
    {
      "epoch": 15.170335429769391,
      "grad_norm": 0.2542964518070221,
      "learning_rate": 8.369855826365988e-06,
      "loss": 0.5165,
      "num_input_tokens_seen": 18884760,
      "step": 28945
    },
    {
      "epoch": 15.172955974842766,
      "grad_norm": 0.5088359117507935,
      "learning_rate": 8.361320041883772e-06,
      "loss": 0.4249,
      "num_input_tokens_seen": 18887736,
      "step": 28950
    },
    {
      "epoch": 15.175576519916143,
      "grad_norm": 0.5924432873725891,
      "learning_rate": 8.352787738023576e-06,
      "loss": 0.3968,
      "num_input_tokens_seen": 18890584,
      "step": 28955
    },
    {
      "epoch": 15.178197064989519,
      "grad_norm": 0.5303402543067932,
      "learning_rate": 8.344258916570274e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 18893208,
      "step": 28960
    },
    {
      "epoch": 15.180817610062894,
      "grad_norm": 0.38402804732322693,
      "learning_rate": 8.335733579307988e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 18895864,
      "step": 28965
    },
    {
      "epoch": 15.183438155136269,
      "grad_norm": 0.5825015902519226,
      "learning_rate": 8.327211728020113e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 18898712,
      "step": 28970
    },
    {
      "epoch": 15.186058700209644,
      "grad_norm": 0.47289276123046875,
      "learning_rate": 8.31869336448933e-06,
      "loss": 0.4403,
      "num_input_tokens_seen": 18901624,
      "step": 28975
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 0.5159740447998047,
      "learning_rate": 8.310178490497586e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 18905176,
      "step": 28980
    },
    {
      "epoch": 15.191299790356394,
      "grad_norm": 0.665919840335846,
      "learning_rate": 8.301667107826103e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 18907864,
      "step": 28985
    },
    {
      "epoch": 15.19392033542977,
      "grad_norm": 0.3186282217502594,
      "learning_rate": 8.293159218255345e-06,
      "loss": 0.3489,
      "num_input_tokens_seen": 18911064,
      "step": 28990
    },
    {
      "epoch": 15.196540880503145,
      "grad_norm": 0.5040205717086792,
      "learning_rate": 8.284654823565088e-06,
      "loss": 0.5312,
      "num_input_tokens_seen": 18914136,
      "step": 28995
    },
    {
      "epoch": 15.19916142557652,
      "grad_norm": 0.3596234917640686,
      "learning_rate": 8.276153925534333e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 18917048,
      "step": 29000
    },
    {
      "epoch": 15.201781970649895,
      "grad_norm": 0.3226461410522461,
      "learning_rate": 8.267656525941383e-06,
      "loss": 0.5386,
      "num_input_tokens_seen": 18919864,
      "step": 29005
    },
    {
      "epoch": 15.20440251572327,
      "grad_norm": 0.4316904842853546,
      "learning_rate": 8.259162626563801e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 18922968,
      "step": 29010
    },
    {
      "epoch": 15.207023060796645,
      "grad_norm": 0.470796138048172,
      "learning_rate": 8.250672229178402e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 18926328,
      "step": 29015
    },
    {
      "epoch": 15.20964360587002,
      "grad_norm": 0.5597256422042847,
      "learning_rate": 8.242185335561295e-06,
      "loss": 0.4055,
      "num_input_tokens_seen": 18929368,
      "step": 29020
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 0.47480905055999756,
      "learning_rate": 8.23370194748782e-06,
      "loss": 0.4137,
      "num_input_tokens_seen": 18932056,
      "step": 29025
    },
    {
      "epoch": 15.214884696016771,
      "grad_norm": 0.5641597509384155,
      "learning_rate": 8.22522206673262e-06,
      "loss": 0.5534,
      "num_input_tokens_seen": 18935896,
      "step": 29030
    },
    {
      "epoch": 15.217505241090146,
      "grad_norm": 0.7560675144195557,
      "learning_rate": 8.216745695069589e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 18938968,
      "step": 29035
    },
    {
      "epoch": 15.220125786163521,
      "grad_norm": 0.24608513712882996,
      "learning_rate": 8.208272834271894e-06,
      "loss": 0.5088,
      "num_input_tokens_seen": 18943000,
      "step": 29040
    },
    {
      "epoch": 15.222746331236896,
      "grad_norm": 0.40689295530319214,
      "learning_rate": 8.19980348611194e-06,
      "loss": 0.4985,
      "num_input_tokens_seen": 18946104,
      "step": 29045
    },
    {
      "epoch": 15.225366876310272,
      "grad_norm": 0.5097924470901489,
      "learning_rate": 8.191337652361439e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 18950328,
      "step": 29050
    },
    {
      "epoch": 15.227987421383649,
      "grad_norm": 0.5582974553108215,
      "learning_rate": 8.182875334791332e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 18953144,
      "step": 29055
    },
    {
      "epoch": 15.230607966457024,
      "grad_norm": 0.4652482867240906,
      "learning_rate": 8.174416535171841e-06,
      "loss": 0.5127,
      "num_input_tokens_seen": 18960248,
      "step": 29060
    },
    {
      "epoch": 15.233228511530399,
      "grad_norm": 0.4657214879989624,
      "learning_rate": 8.165961255272467e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 18962744,
      "step": 29065
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 0.37981295585632324,
      "learning_rate": 8.157509496861935e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 18965816,
      "step": 29070
    },
    {
      "epoch": 15.23846960167715,
      "grad_norm": 0.8243053555488586,
      "learning_rate": 8.149061261708266e-06,
      "loss": 0.4267,
      "num_input_tokens_seen": 18969560,
      "step": 29075
    },
    {
      "epoch": 15.241090146750524,
      "grad_norm": 0.4661625623703003,
      "learning_rate": 8.140616551578745e-06,
      "loss": 0.2764,
      "num_input_tokens_seen": 18972792,
      "step": 29080
    },
    {
      "epoch": 15.2437106918239,
      "grad_norm": 0.5537683367729187,
      "learning_rate": 8.13217536823989e-06,
      "loss": 0.3327,
      "num_input_tokens_seen": 18975640,
      "step": 29085
    },
    {
      "epoch": 15.246331236897275,
      "grad_norm": 0.47190263867378235,
      "learning_rate": 8.12373771345752e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 18978840,
      "step": 29090
    },
    {
      "epoch": 15.24895178197065,
      "grad_norm": 0.7230726480484009,
      "learning_rate": 8.115303588996676e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 18981464,
      "step": 29095
    },
    {
      "epoch": 15.251572327044025,
      "grad_norm": 0.46103155612945557,
      "learning_rate": 8.10687299662169e-06,
      "loss": 0.3781,
      "num_input_tokens_seen": 18985368,
      "step": 29100
    },
    {
      "epoch": 15.2541928721174,
      "grad_norm": 0.4677373766899109,
      "learning_rate": 8.098445938096147e-06,
      "loss": 0.5038,
      "num_input_tokens_seen": 18988120,
      "step": 29105
    },
    {
      "epoch": 15.256813417190775,
      "grad_norm": 0.41060322523117065,
      "learning_rate": 8.090022415182898e-06,
      "loss": 0.5365,
      "num_input_tokens_seen": 18991640,
      "step": 29110
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 0.3737141191959381,
      "learning_rate": 8.081602429644033e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 18994552,
      "step": 29115
    },
    {
      "epoch": 15.262054507337526,
      "grad_norm": 0.5924021601676941,
      "learning_rate": 8.073185983240932e-06,
      "loss": 0.5631,
      "num_input_tokens_seen": 18999224,
      "step": 29120
    },
    {
      "epoch": 15.264675052410901,
      "grad_norm": 0.37093237042427063,
      "learning_rate": 8.064773077734206e-06,
      "loss": 0.6989,
      "num_input_tokens_seen": 19002648,
      "step": 29125
    },
    {
      "epoch": 15.267295597484276,
      "grad_norm": 0.520002543926239,
      "learning_rate": 8.05636371488374e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 19005976,
      "step": 29130
    },
    {
      "epoch": 15.269916142557651,
      "grad_norm": 0.4423578083515167,
      "learning_rate": 8.047957896448696e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 19009144,
      "step": 29135
    },
    {
      "epoch": 15.272536687631026,
      "grad_norm": 0.4071482717990875,
      "learning_rate": 8.039555624187451e-06,
      "loss": 0.363,
      "num_input_tokens_seen": 19011896,
      "step": 29140
    },
    {
      "epoch": 15.275157232704402,
      "grad_norm": 0.48089030385017395,
      "learning_rate": 8.031156899857681e-06,
      "loss": 0.5675,
      "num_input_tokens_seen": 19015256,
      "step": 29145
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.8373034596443176,
      "learning_rate": 8.022761725216288e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 19018360,
      "step": 29150
    },
    {
      "epoch": 15.280398322851154,
      "grad_norm": 0.3930700719356537,
      "learning_rate": 8.014370102019456e-06,
      "loss": 0.2971,
      "num_input_tokens_seen": 19022136,
      "step": 29155
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 0.7768833637237549,
      "learning_rate": 8.005982032022616e-06,
      "loss": 0.491,
      "num_input_tokens_seen": 19024920,
      "step": 29160
    },
    {
      "epoch": 15.285639412997904,
      "grad_norm": 0.38428786396980286,
      "learning_rate": 7.997597516980467e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 19029464,
      "step": 29165
    },
    {
      "epoch": 15.28825995807128,
      "grad_norm": 0.37864309549331665,
      "learning_rate": 7.989216558646942e-06,
      "loss": 0.479,
      "num_input_tokens_seen": 19033272,
      "step": 29170
    },
    {
      "epoch": 15.290880503144654,
      "grad_norm": 0.5114526748657227,
      "learning_rate": 7.980839158775235e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 19035736,
      "step": 29175
    },
    {
      "epoch": 15.29350104821803,
      "grad_norm": 0.5056114196777344,
      "learning_rate": 7.972465319117814e-06,
      "loss": 0.482,
      "num_input_tokens_seen": 19038904,
      "step": 29180
    },
    {
      "epoch": 15.296121593291405,
      "grad_norm": 0.35877084732055664,
      "learning_rate": 7.964095041426386e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 19042424,
      "step": 29185
    },
    {
      "epoch": 15.29874213836478,
      "grad_norm": 0.5117776393890381,
      "learning_rate": 7.95572832745192e-06,
      "loss": 0.313,
      "num_input_tokens_seen": 19045784,
      "step": 29190
    },
    {
      "epoch": 15.301362683438155,
      "grad_norm": 0.5079352259635925,
      "learning_rate": 7.947365178944643e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 19049016,
      "step": 29195
    },
    {
      "epoch": 15.30398322851153,
      "grad_norm": 0.6690245866775513,
      "learning_rate": 7.939005597654025e-06,
      "loss": 0.3662,
      "num_input_tokens_seen": 19052536,
      "step": 29200
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 0.57472163438797,
      "learning_rate": 7.930649585328787e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 19055480,
      "step": 29205
    },
    {
      "epoch": 15.30922431865828,
      "grad_norm": 0.3359847962856293,
      "learning_rate": 7.922297143716918e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 19059416,
      "step": 29210
    },
    {
      "epoch": 15.311844863731656,
      "grad_norm": 0.6988290548324585,
      "learning_rate": 7.913948274565652e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 19062360,
      "step": 29215
    },
    {
      "epoch": 15.314465408805031,
      "grad_norm": 0.34040990471839905,
      "learning_rate": 7.905602979621491e-06,
      "loss": 0.4019,
      "num_input_tokens_seen": 19065400,
      "step": 29220
    },
    {
      "epoch": 15.317085953878406,
      "grad_norm": 0.4010949730873108,
      "learning_rate": 7.897261260630159e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 19068376,
      "step": 29225
    },
    {
      "epoch": 15.319706498951781,
      "grad_norm": 0.7393447756767273,
      "learning_rate": 7.888923119336647e-06,
      "loss": 0.325,
      "num_input_tokens_seen": 19071032,
      "step": 29230
    },
    {
      "epoch": 15.322327044025156,
      "grad_norm": 0.5403968095779419,
      "learning_rate": 7.880588557485203e-06,
      "loss": 0.454,
      "num_input_tokens_seen": 19073944,
      "step": 29235
    },
    {
      "epoch": 15.324947589098532,
      "grad_norm": 0.31886664032936096,
      "learning_rate": 7.872257576819325e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 19076760,
      "step": 29240
    },
    {
      "epoch": 15.327568134171909,
      "grad_norm": 0.4677913784980774,
      "learning_rate": 7.863930179081769e-06,
      "loss": 0.5471,
      "num_input_tokens_seen": 19079480,
      "step": 29245
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 0.4710175395011902,
      "learning_rate": 7.855606366014517e-06,
      "loss": 0.5398,
      "num_input_tokens_seen": 19081944,
      "step": 29250
    },
    {
      "epoch": 15.332809224318659,
      "grad_norm": 0.42529964447021484,
      "learning_rate": 7.847286139358814e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 19085272,
      "step": 29255
    },
    {
      "epoch": 15.335429769392034,
      "grad_norm": 0.2647625207901001,
      "learning_rate": 7.838969500855162e-06,
      "loss": 0.4888,
      "num_input_tokens_seen": 19088152,
      "step": 29260
    },
    {
      "epoch": 15.33805031446541,
      "grad_norm": 0.561236560344696,
      "learning_rate": 7.830656452243307e-06,
      "loss": 0.5503,
      "num_input_tokens_seen": 19091576,
      "step": 29265
    },
    {
      "epoch": 15.340670859538784,
      "grad_norm": 0.34224456548690796,
      "learning_rate": 7.822346995262241e-06,
      "loss": 0.5054,
      "num_input_tokens_seen": 19095672,
      "step": 29270
    },
    {
      "epoch": 15.34329140461216,
      "grad_norm": 0.49450191855430603,
      "learning_rate": 7.814041131650223e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 19099416,
      "step": 29275
    },
    {
      "epoch": 15.345911949685535,
      "grad_norm": 0.6766583919525146,
      "learning_rate": 7.805738863144731e-06,
      "loss": 0.4278,
      "num_input_tokens_seen": 19102296,
      "step": 29280
    },
    {
      "epoch": 15.34853249475891,
      "grad_norm": 0.41316112875938416,
      "learning_rate": 7.797440191482496e-06,
      "loss": 0.4284,
      "num_input_tokens_seen": 19105432,
      "step": 29285
    },
    {
      "epoch": 15.351153039832285,
      "grad_norm": 1.279768943786621,
      "learning_rate": 7.789145118399518e-06,
      "loss": 0.3447,
      "num_input_tokens_seen": 19108632,
      "step": 29290
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 0.5507115721702576,
      "learning_rate": 7.780853645631032e-06,
      "loss": 0.5767,
      "num_input_tokens_seen": 19111832,
      "step": 29295
    },
    {
      "epoch": 15.356394129979035,
      "grad_norm": 0.37074556946754456,
      "learning_rate": 7.772565774911522e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 19115288,
      "step": 29300
    },
    {
      "epoch": 15.35901467505241,
      "grad_norm": 0.46862730383872986,
      "learning_rate": 7.76428150797471e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 19118456,
      "step": 29305
    },
    {
      "epoch": 15.361635220125786,
      "grad_norm": 0.4490462839603424,
      "learning_rate": 7.756000846553562e-06,
      "loss": 0.4022,
      "num_input_tokens_seen": 19121272,
      "step": 29310
    },
    {
      "epoch": 15.364255765199161,
      "grad_norm": 0.40739497542381287,
      "learning_rate": 7.74772379238031e-06,
      "loss": 0.3709,
      "num_input_tokens_seen": 19124440,
      "step": 29315
    },
    {
      "epoch": 15.366876310272536,
      "grad_norm": 0.4308534562587738,
      "learning_rate": 7.739450347186417e-06,
      "loss": 0.4694,
      "num_input_tokens_seen": 19127672,
      "step": 29320
    },
    {
      "epoch": 15.369496855345911,
      "grad_norm": 0.7621113657951355,
      "learning_rate": 7.731180512702599e-06,
      "loss": 0.4993,
      "num_input_tokens_seen": 19130776,
      "step": 29325
    },
    {
      "epoch": 15.372117400419286,
      "grad_norm": 0.804807186126709,
      "learning_rate": 7.722914290658795e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 19133720,
      "step": 29330
    },
    {
      "epoch": 15.374737945492662,
      "grad_norm": 0.9211879372596741,
      "learning_rate": 7.714651682784224e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 19137112,
      "step": 29335
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 0.5256362557411194,
      "learning_rate": 7.706392690807313e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 19140408,
      "step": 29340
    },
    {
      "epoch": 15.379979035639414,
      "grad_norm": 0.4047980308532715,
      "learning_rate": 7.69813731645575e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 19143672,
      "step": 29345
    },
    {
      "epoch": 15.382599580712789,
      "grad_norm": 0.4893946647644043,
      "learning_rate": 7.689885561456475e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 19147288,
      "step": 29350
    },
    {
      "epoch": 15.385220125786164,
      "grad_norm": 0.3573492467403412,
      "learning_rate": 7.681637427535663e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 19149688,
      "step": 29355
    },
    {
      "epoch": 15.38784067085954,
      "grad_norm": 0.27966222167015076,
      "learning_rate": 7.673392916418715e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 19153752,
      "step": 29360
    },
    {
      "epoch": 15.390461215932914,
      "grad_norm": 0.3881914019584656,
      "learning_rate": 7.665152029830303e-06,
      "loss": 0.534,
      "num_input_tokens_seen": 19157208,
      "step": 29365
    },
    {
      "epoch": 15.39308176100629,
      "grad_norm": 0.45761582255363464,
      "learning_rate": 7.656914769494314e-06,
      "loss": 0.5238,
      "num_input_tokens_seen": 19161656,
      "step": 29370
    },
    {
      "epoch": 15.395702306079665,
      "grad_norm": 0.3630979359149933,
      "learning_rate": 7.648681137133892e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 19164824,
      "step": 29375
    },
    {
      "epoch": 15.39832285115304,
      "grad_norm": 0.3243873715400696,
      "learning_rate": 7.640451134471432e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 19167480,
      "step": 29380
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 0.46797698736190796,
      "learning_rate": 7.632224763228538e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 19169752,
      "step": 29385
    },
    {
      "epoch": 15.40356394129979,
      "grad_norm": 0.647530734539032,
      "learning_rate": 7.6240020251260896e-06,
      "loss": 0.53,
      "num_input_tokens_seen": 19172696,
      "step": 29390
    },
    {
      "epoch": 15.406184486373165,
      "grad_norm": 0.432309091091156,
      "learning_rate": 7.615782921884174e-06,
      "loss": 0.4253,
      "num_input_tokens_seen": 19176600,
      "step": 29395
    },
    {
      "epoch": 15.40880503144654,
      "grad_norm": 0.6285560727119446,
      "learning_rate": 7.607567455222142e-06,
      "loss": 0.3196,
      "num_input_tokens_seen": 19179000,
      "step": 29400
    },
    {
      "epoch": 15.411425576519916,
      "grad_norm": 0.3916255235671997,
      "learning_rate": 7.599355626858582e-06,
      "loss": 0.4909,
      "num_input_tokens_seen": 19182264,
      "step": 29405
    },
    {
      "epoch": 15.414046121593291,
      "grad_norm": 0.5411303043365479,
      "learning_rate": 7.591147438511298e-06,
      "loss": 0.4227,
      "num_input_tokens_seen": 19185240,
      "step": 29410
    },
    {
      "epoch": 15.416666666666666,
      "grad_norm": 0.22440707683563232,
      "learning_rate": 7.5829428918973625e-06,
      "loss": 0.3784,
      "num_input_tokens_seen": 19188920,
      "step": 29415
    },
    {
      "epoch": 15.419287211740041,
      "grad_norm": 0.5170420408248901,
      "learning_rate": 7.574741988733075e-06,
      "loss": 0.4665,
      "num_input_tokens_seen": 19191672,
      "step": 29420
    },
    {
      "epoch": 15.421907756813416,
      "grad_norm": 0.673582136631012,
      "learning_rate": 7.56654473073396e-06,
      "loss": 0.3982,
      "num_input_tokens_seen": 19194744,
      "step": 29425
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 0.4811672568321228,
      "learning_rate": 7.558351119614796e-06,
      "loss": 0.5306,
      "num_input_tokens_seen": 19197848,
      "step": 29430
    },
    {
      "epoch": 15.427148846960169,
      "grad_norm": 0.4847262501716614,
      "learning_rate": 7.5501611570896e-06,
      "loss": 0.3024,
      "num_input_tokens_seen": 19201432,
      "step": 29435
    },
    {
      "epoch": 15.429769392033544,
      "grad_norm": 0.2722054421901703,
      "learning_rate": 7.541974844871602e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 19205304,
      "step": 29440
    },
    {
      "epoch": 15.432389937106919,
      "grad_norm": 0.912031888961792,
      "learning_rate": 7.5337921846732965e-06,
      "loss": 0.5033,
      "num_input_tokens_seen": 19208888,
      "step": 29445
    },
    {
      "epoch": 15.435010482180294,
      "grad_norm": 0.46537840366363525,
      "learning_rate": 7.525613178206409e-06,
      "loss": 0.4327,
      "num_input_tokens_seen": 19212216,
      "step": 29450
    },
    {
      "epoch": 15.43763102725367,
      "grad_norm": 0.5265535116195679,
      "learning_rate": 7.517437827181878e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 19215736,
      "step": 29455
    },
    {
      "epoch": 15.440251572327044,
      "grad_norm": 0.27801793813705444,
      "learning_rate": 7.509266133309908e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 19219224,
      "step": 29460
    },
    {
      "epoch": 15.44287211740042,
      "grad_norm": 0.39287734031677246,
      "learning_rate": 7.501098098299914e-06,
      "loss": 0.4901,
      "num_input_tokens_seen": 19222424,
      "step": 29465
    },
    {
      "epoch": 15.445492662473795,
      "grad_norm": 0.4905872941017151,
      "learning_rate": 7.492933723860557e-06,
      "loss": 0.5005,
      "num_input_tokens_seen": 19225784,
      "step": 29470
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 0.5864579677581787,
      "learning_rate": 7.4847730116997334e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 19228600,
      "step": 29475
    },
    {
      "epoch": 15.450733752620545,
      "grad_norm": 0.5962755084037781,
      "learning_rate": 7.476615963524583e-06,
      "loss": 0.4431,
      "num_input_tokens_seen": 19231864,
      "step": 29480
    },
    {
      "epoch": 15.45335429769392,
      "grad_norm": 0.5530815124511719,
      "learning_rate": 7.468462581041452e-06,
      "loss": 0.4993,
      "num_input_tokens_seen": 19235480,
      "step": 29485
    },
    {
      "epoch": 15.455974842767295,
      "grad_norm": 0.4673871397972107,
      "learning_rate": 7.460312865955932e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 19238904,
      "step": 29490
    },
    {
      "epoch": 15.45859538784067,
      "grad_norm": 0.5489235520362854,
      "learning_rate": 7.4521668199728584e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 19242040,
      "step": 29495
    },
    {
      "epoch": 15.461215932914046,
      "grad_norm": 0.2818138301372528,
      "learning_rate": 7.4440244447962884e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 19245336,
      "step": 29500
    },
    {
      "epoch": 15.463836477987421,
      "grad_norm": 0.28323256969451904,
      "learning_rate": 7.435885742129523e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 19250264,
      "step": 29505
    },
    {
      "epoch": 15.466457023060796,
      "grad_norm": 0.4807904362678528,
      "learning_rate": 7.427750713675071e-06,
      "loss": 0.7741,
      "num_input_tokens_seen": 19253592,
      "step": 29510
    },
    {
      "epoch": 15.469077568134171,
      "grad_norm": 0.4990747272968292,
      "learning_rate": 7.419619361134702e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 19256728,
      "step": 29515
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 0.6666510105133057,
      "learning_rate": 7.411491686209387e-06,
      "loss": 0.4497,
      "num_input_tokens_seen": 19259640,
      "step": 29520
    },
    {
      "epoch": 15.474318658280922,
      "grad_norm": 0.44352638721466064,
      "learning_rate": 7.4033676905993525e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 19263128,
      "step": 29525
    },
    {
      "epoch": 15.476939203354299,
      "grad_norm": 0.43344518542289734,
      "learning_rate": 7.3952473760040444e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 19267928,
      "step": 29530
    },
    {
      "epoch": 15.479559748427674,
      "grad_norm": 0.48265546560287476,
      "learning_rate": 7.387130744122148e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 19270552,
      "step": 29535
    },
    {
      "epoch": 15.482180293501049,
      "grad_norm": 0.3133746087551117,
      "learning_rate": 7.379017796651558e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 19274072,
      "step": 29540
    },
    {
      "epoch": 15.484800838574424,
      "grad_norm": 0.2850632965564728,
      "learning_rate": 7.3709085352894085e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 19277048,
      "step": 29545
    },
    {
      "epoch": 15.4874213836478,
      "grad_norm": 0.6435651779174805,
      "learning_rate": 7.362802961732071e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 19279576,
      "step": 29550
    },
    {
      "epoch": 15.490041928721174,
      "grad_norm": 0.8842888474464417,
      "learning_rate": 7.354701077675136e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 19282200,
      "step": 29555
    },
    {
      "epoch": 15.49266247379455,
      "grad_norm": 0.4111308157444,
      "learning_rate": 7.346602884813439e-06,
      "loss": 0.5451,
      "num_input_tokens_seen": 19285336,
      "step": 29560
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 0.4562310576438904,
      "learning_rate": 7.338508384841014e-06,
      "loss": 0.4688,
      "num_input_tokens_seen": 19288344,
      "step": 29565
    },
    {
      "epoch": 15.4979035639413,
      "grad_norm": 0.6896198987960815,
      "learning_rate": 7.3304175794511345e-06,
      "loss": 0.3758,
      "num_input_tokens_seen": 19291928,
      "step": 29570
    },
    {
      "epoch": 15.500524109014675,
      "grad_norm": 1.4307321310043335,
      "learning_rate": 7.3223304703363135e-06,
      "loss": 0.3662,
      "num_input_tokens_seen": 19295736,
      "step": 29575
    },
    {
      "epoch": 15.50314465408805,
      "grad_norm": 0.6882445216178894,
      "learning_rate": 7.314247059188281e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 19298744,
      "step": 29580
    },
    {
      "epoch": 15.505765199161425,
      "grad_norm": 0.601230800151825,
      "learning_rate": 7.306167347697992e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 19301368,
      "step": 29585
    },
    {
      "epoch": 15.5083857442348,
      "grad_norm": 0.38025081157684326,
      "learning_rate": 7.29809133755564e-06,
      "loss": 0.3573,
      "num_input_tokens_seen": 19304952,
      "step": 29590
    },
    {
      "epoch": 15.511006289308176,
      "grad_norm": 0.4186491072177887,
      "learning_rate": 7.290019030450629e-06,
      "loss": 0.5206,
      "num_input_tokens_seen": 19308088,
      "step": 29595
    },
    {
      "epoch": 15.51362683438155,
      "grad_norm": 0.6245154142379761,
      "learning_rate": 7.2819504280715794e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 19310840,
      "step": 29600
    },
    {
      "epoch": 15.516247379454926,
      "grad_norm": 0.7849913239479065,
      "learning_rate": 7.273885532106364e-06,
      "loss": 0.4762,
      "num_input_tokens_seen": 19313720,
      "step": 29605
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 0.583120584487915,
      "learning_rate": 7.265824344242064e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 19318424,
      "step": 29610
    },
    {
      "epoch": 15.521488469601676,
      "grad_norm": 0.480781614780426,
      "learning_rate": 7.257766866165e-06,
      "loss": 0.3147,
      "num_input_tokens_seen": 19322488,
      "step": 29615
    },
    {
      "epoch": 15.524109014675052,
      "grad_norm": 0.6111209988594055,
      "learning_rate": 7.24971309956069e-06,
      "loss": 0.6487,
      "num_input_tokens_seen": 19325592,
      "step": 29620
    },
    {
      "epoch": 15.526729559748428,
      "grad_norm": 0.5895760655403137,
      "learning_rate": 7.241663046113887e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 19328408,
      "step": 29625
    },
    {
      "epoch": 15.529350104821804,
      "grad_norm": 0.599341094493866,
      "learning_rate": 7.233616707508576e-06,
      "loss": 0.568,
      "num_input_tokens_seen": 19330776,
      "step": 29630
    },
    {
      "epoch": 15.531970649895179,
      "grad_norm": 0.7657185792922974,
      "learning_rate": 7.225574085427961e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 19334360,
      "step": 29635
    },
    {
      "epoch": 15.534591194968554,
      "grad_norm": 0.8616182208061218,
      "learning_rate": 7.217535181554474e-06,
      "loss": 0.3948,
      "num_input_tokens_seen": 19336664,
      "step": 29640
    },
    {
      "epoch": 15.53721174004193,
      "grad_norm": 0.36124035716056824,
      "learning_rate": 7.209499997569747e-06,
      "loss": 0.3904,
      "num_input_tokens_seen": 19339608,
      "step": 29645
    },
    {
      "epoch": 15.539832285115304,
      "grad_norm": 0.44696271419525146,
      "learning_rate": 7.201468535154663e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 19342712,
      "step": 29650
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 0.6865984201431274,
      "learning_rate": 7.193440795989295e-06,
      "loss": 0.3993,
      "num_input_tokens_seen": 19346296,
      "step": 29655
    },
    {
      "epoch": 15.545073375262055,
      "grad_norm": 0.37620219588279724,
      "learning_rate": 7.185416781752965e-06,
      "loss": 0.3179,
      "num_input_tokens_seen": 19349656,
      "step": 29660
    },
    {
      "epoch": 15.54769392033543,
      "grad_norm": 0.5747156143188477,
      "learning_rate": 7.177396494124206e-06,
      "loss": 0.4308,
      "num_input_tokens_seen": 19352280,
      "step": 29665
    },
    {
      "epoch": 15.550314465408805,
      "grad_norm": 0.4025892913341522,
      "learning_rate": 7.169379934780779e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 19355128,
      "step": 29670
    },
    {
      "epoch": 15.55293501048218,
      "grad_norm": 0.6157593131065369,
      "learning_rate": 7.161367105399644e-06,
      "loss": 0.5841,
      "num_input_tokens_seen": 19359096,
      "step": 29675
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.518398642539978,
      "learning_rate": 7.153358007656991e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 19361784,
      "step": 29680
    },
    {
      "epoch": 15.55817610062893,
      "grad_norm": 0.7583367824554443,
      "learning_rate": 7.145352643228237e-06,
      "loss": 0.3584,
      "num_input_tokens_seen": 19364824,
      "step": 29685
    },
    {
      "epoch": 15.560796645702306,
      "grad_norm": 0.644999086856842,
      "learning_rate": 7.137351013788013e-06,
      "loss": 0.5597,
      "num_input_tokens_seen": 19368056,
      "step": 29690
    },
    {
      "epoch": 15.56341719077568,
      "grad_norm": 0.32264527678489685,
      "learning_rate": 7.129353121010177e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 19371448,
      "step": 29695
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 0.4039873480796814,
      "learning_rate": 7.121358966567779e-06,
      "loss": 0.5702,
      "num_input_tokens_seen": 19375448,
      "step": 29700
    },
    {
      "epoch": 15.568658280922431,
      "grad_norm": 0.26246488094329834,
      "learning_rate": 7.113368552133126e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 19379256,
      "step": 29705
    },
    {
      "epoch": 15.571278825995806,
      "grad_norm": 0.5022045373916626,
      "learning_rate": 7.105381879377701e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 19383032,
      "step": 29710
    },
    {
      "epoch": 15.573899371069182,
      "grad_norm": 0.4104236662387848,
      "learning_rate": 7.097398949972234e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 19387640,
      "step": 29715
    },
    {
      "epoch": 15.576519916142558,
      "grad_norm": 0.8706347942352295,
      "learning_rate": 7.089419765586672e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 19389944,
      "step": 29720
    },
    {
      "epoch": 15.579140461215934,
      "grad_norm": 0.4528384208679199,
      "learning_rate": 7.081444327890152e-06,
      "loss": 0.4659,
      "num_input_tokens_seen": 19393464,
      "step": 29725
    },
    {
      "epoch": 15.581761006289309,
      "grad_norm": 0.737694263458252,
      "learning_rate": 7.073472638551054e-06,
      "loss": 0.3317,
      "num_input_tokens_seen": 19395864,
      "step": 29730
    },
    {
      "epoch": 15.584381551362684,
      "grad_norm": 0.7373142838478088,
      "learning_rate": 7.065504699236969e-06,
      "loss": 0.4677,
      "num_input_tokens_seen": 19398552,
      "step": 29735
    },
    {
      "epoch": 15.58700209643606,
      "grad_norm": 0.5420969128608704,
      "learning_rate": 7.057540511614688e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 19401688,
      "step": 29740
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 0.639394223690033,
      "learning_rate": 7.049580077350232e-06,
      "loss": 0.3922,
      "num_input_tokens_seen": 19404472,
      "step": 29745
    },
    {
      "epoch": 15.59224318658281,
      "grad_norm": 0.4824809432029724,
      "learning_rate": 7.041623398108843e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 19407544,
      "step": 29750
    },
    {
      "epoch": 15.594863731656185,
      "grad_norm": 0.4311022162437439,
      "learning_rate": 7.033670475554949e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 19410488,
      "step": 29755
    },
    {
      "epoch": 15.59748427672956,
      "grad_norm": 0.6420392394065857,
      "learning_rate": 7.025721311352224e-06,
      "loss": 0.6076,
      "num_input_tokens_seen": 19414360,
      "step": 29760
    },
    {
      "epoch": 15.600104821802935,
      "grad_norm": 0.4472962021827698,
      "learning_rate": 7.017775907163546e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 19418072,
      "step": 29765
    },
    {
      "epoch": 15.60272536687631,
      "grad_norm": 0.43704572319984436,
      "learning_rate": 7.009834264650989e-06,
      "loss": 0.4344,
      "num_input_tokens_seen": 19421016,
      "step": 29770
    },
    {
      "epoch": 15.605345911949685,
      "grad_norm": 0.25376713275909424,
      "learning_rate": 7.001896385475867e-06,
      "loss": 0.4873,
      "num_input_tokens_seen": 19424568,
      "step": 29775
    },
    {
      "epoch": 15.60796645702306,
      "grad_norm": 0.4612159729003906,
      "learning_rate": 6.993962271298682e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 19427352,
      "step": 29780
    },
    {
      "epoch": 15.610587002096436,
      "grad_norm": 0.6887850165367126,
      "learning_rate": 6.986031923779166e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 19430328,
      "step": 29785
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 0.33470624685287476,
      "learning_rate": 6.978105344576264e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 19433848,
      "step": 29790
    },
    {
      "epoch": 15.615828092243186,
      "grad_norm": 1.1080161333084106,
      "learning_rate": 6.970182535348111e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 19436888,
      "step": 29795
    },
    {
      "epoch": 15.618448637316561,
      "grad_norm": 0.8017227053642273,
      "learning_rate": 6.962263497752086e-06,
      "loss": 0.5498,
      "num_input_tokens_seen": 19441176,
      "step": 29800
    },
    {
      "epoch": 15.621069182389936,
      "grad_norm": 0.822776734828949,
      "learning_rate": 6.9543482334447436e-06,
      "loss": 0.4405,
      "num_input_tokens_seen": 19444664,
      "step": 29805
    },
    {
      "epoch": 15.623689727463312,
      "grad_norm": 0.5003507137298584,
      "learning_rate": 6.946436744081875e-06,
      "loss": 0.3978,
      "num_input_tokens_seen": 19447736,
      "step": 29810
    },
    {
      "epoch": 15.626310272536688,
      "grad_norm": 0.5442944169044495,
      "learning_rate": 6.938529031318472e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 19451384,
      "step": 29815
    },
    {
      "epoch": 15.628930817610064,
      "grad_norm": 0.6541792750358582,
      "learning_rate": 6.9306250968087485e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 19454552,
      "step": 29820
    },
    {
      "epoch": 15.631551362683439,
      "grad_norm": 0.5400993227958679,
      "learning_rate": 6.922724942206102e-06,
      "loss": 0.4111,
      "num_input_tokens_seen": 19457336,
      "step": 29825
    },
    {
      "epoch": 15.634171907756814,
      "grad_norm": 0.6310332417488098,
      "learning_rate": 6.914828569163167e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 19460344,
      "step": 29830
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 0.6805776357650757,
      "learning_rate": 6.906935979331763e-06,
      "loss": 0.5005,
      "num_input_tokens_seen": 19462744,
      "step": 29835
    },
    {
      "epoch": 15.639412997903564,
      "grad_norm": 0.5243613123893738,
      "learning_rate": 6.8990471743629356e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 19465880,
      "step": 29840
    },
    {
      "epoch": 15.64203354297694,
      "grad_norm": 0.5752059817314148,
      "learning_rate": 6.89116215590693e-06,
      "loss": 0.5014,
      "num_input_tokens_seen": 19468632,
      "step": 29845
    },
    {
      "epoch": 15.644654088050315,
      "grad_norm": 0.42781156301498413,
      "learning_rate": 6.8832809256132146e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 19471736,
      "step": 29850
    },
    {
      "epoch": 15.64727463312369,
      "grad_norm": 0.3466745913028717,
      "learning_rate": 6.875403485130444e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 19474424,
      "step": 29855
    },
    {
      "epoch": 15.649895178197065,
      "grad_norm": 0.35189583897590637,
      "learning_rate": 6.867529836106479e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 19477400,
      "step": 29860
    },
    {
      "epoch": 15.65251572327044,
      "grad_norm": 0.6287027597427368,
      "learning_rate": 6.8596599801884045e-06,
      "loss": 0.4135,
      "num_input_tokens_seen": 19480024,
      "step": 29865
    },
    {
      "epoch": 15.655136268343815,
      "grad_norm": 0.4955372214317322,
      "learning_rate": 6.851793919022509e-06,
      "loss": 0.4604,
      "num_input_tokens_seen": 19483256,
      "step": 29870
    },
    {
      "epoch": 15.65775681341719,
      "grad_norm": 0.4462738633155823,
      "learning_rate": 6.843931654254285e-06,
      "loss": 0.3849,
      "num_input_tokens_seen": 19486392,
      "step": 29875
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 0.3918355107307434,
      "learning_rate": 6.836073187528425e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 19489464,
      "step": 29880
    },
    {
      "epoch": 15.66299790356394,
      "grad_norm": 0.6936311721801758,
      "learning_rate": 6.828218520488821e-06,
      "loss": 0.4214,
      "num_input_tokens_seen": 19492952,
      "step": 29885
    },
    {
      "epoch": 15.665618448637316,
      "grad_norm": 0.4864693582057953,
      "learning_rate": 6.820367654778589e-06,
      "loss": 0.5356,
      "num_input_tokens_seen": 19495928,
      "step": 29890
    },
    {
      "epoch": 15.668238993710691,
      "grad_norm": 0.4054194688796997,
      "learning_rate": 6.812520592040039e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 19499800,
      "step": 29895
    },
    {
      "epoch": 15.670859538784066,
      "grad_norm": 0.5287234783172607,
      "learning_rate": 6.804677333914689e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 19504408,
      "step": 29900
    },
    {
      "epoch": 15.673480083857442,
      "grad_norm": 0.33822721242904663,
      "learning_rate": 6.796837882043261e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 19507576,
      "step": 29905
    },
    {
      "epoch": 15.676100628930818,
      "grad_norm": 0.8644176721572876,
      "learning_rate": 6.7890022380656785e-06,
      "loss": 0.6138,
      "num_input_tokens_seen": 19510712,
      "step": 29910
    },
    {
      "epoch": 15.678721174004194,
      "grad_norm": 0.4425170421600342,
      "learning_rate": 6.781170403621056e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 19514168,
      "step": 29915
    },
    {
      "epoch": 15.681341719077569,
      "grad_norm": 0.5472335815429688,
      "learning_rate": 6.773342380347736e-06,
      "loss": 0.3817,
      "num_input_tokens_seen": 19517304,
      "step": 29920
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 0.35324227809906006,
      "learning_rate": 6.765518169883248e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 19520504,
      "step": 29925
    },
    {
      "epoch": 15.68658280922432,
      "grad_norm": 0.2522197663784027,
      "learning_rate": 6.757697773864338e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 19527000,
      "step": 29930
    },
    {
      "epoch": 15.689203354297694,
      "grad_norm": 0.39610815048217773,
      "learning_rate": 6.749881193926932e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 19530008,
      "step": 29935
    },
    {
      "epoch": 15.69182389937107,
      "grad_norm": 0.44171392917633057,
      "learning_rate": 6.742068431706167e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 19533208,
      "step": 29940
    },
    {
      "epoch": 15.694444444444445,
      "grad_norm": 0.5250959396362305,
      "learning_rate": 6.734259488836386e-06,
      "loss": 0.6571,
      "num_input_tokens_seen": 19536248,
      "step": 29945
    },
    {
      "epoch": 15.69706498951782,
      "grad_norm": 0.268714964389801,
      "learning_rate": 6.726454366951135e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 19539480,
      "step": 29950
    },
    {
      "epoch": 15.699685534591195,
      "grad_norm": 0.6731241941452026,
      "learning_rate": 6.7186530676831616e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 19542584,
      "step": 29955
    },
    {
      "epoch": 15.70230607966457,
      "grad_norm": 0.4830050766468048,
      "learning_rate": 6.710855592664403e-06,
      "loss": 0.5711,
      "num_input_tokens_seen": 19546296,
      "step": 29960
    },
    {
      "epoch": 15.704926624737945,
      "grad_norm": 0.5074218511581421,
      "learning_rate": 6.703061943525993e-06,
      "loss": 0.4795,
      "num_input_tokens_seen": 19549400,
      "step": 29965
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 0.543550431728363,
      "learning_rate": 6.69527212189828e-06,
      "loss": 0.4819,
      "num_input_tokens_seen": 19552984,
      "step": 29970
    },
    {
      "epoch": 15.710167714884696,
      "grad_norm": 0.8553636074066162,
      "learning_rate": 6.687486129410811e-06,
      "loss": 0.5803,
      "num_input_tokens_seen": 19555448,
      "step": 29975
    },
    {
      "epoch": 15.71278825995807,
      "grad_norm": 0.5267041325569153,
      "learning_rate": 6.679703967692322e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 19558136,
      "step": 29980
    },
    {
      "epoch": 15.715408805031446,
      "grad_norm": 0.35377150774002075,
      "learning_rate": 6.67192563837076e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 19560824,
      "step": 29985
    },
    {
      "epoch": 15.718029350104821,
      "grad_norm": 0.39266327023506165,
      "learning_rate": 6.664151143073258e-06,
      "loss": 0.3334,
      "num_input_tokens_seen": 19564472,
      "step": 29990
    },
    {
      "epoch": 15.720649895178196,
      "grad_norm": 0.7316373586654663,
      "learning_rate": 6.656380483426141e-06,
      "loss": 0.563,
      "num_input_tokens_seen": 19567960,
      "step": 29995
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 0.6149336695671082,
      "learning_rate": 6.648613661054956e-06,
      "loss": 0.4245,
      "num_input_tokens_seen": 19570616,
      "step": 30000
    },
    {
      "epoch": 15.725890985324948,
      "grad_norm": 0.8420714139938354,
      "learning_rate": 6.6408506775844256e-06,
      "loss": 0.5777,
      "num_input_tokens_seen": 19573816,
      "step": 30005
    },
    {
      "epoch": 15.728511530398324,
      "grad_norm": 0.38807734847068787,
      "learning_rate": 6.633091534638492e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 19577272,
      "step": 30010
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 0.3850862681865692,
      "learning_rate": 6.625336233840257e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 19580408,
      "step": 30015
    },
    {
      "epoch": 15.733752620545074,
      "grad_norm": 0.4217720031738281,
      "learning_rate": 6.617584776812064e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 19583640,
      "step": 30020
    },
    {
      "epoch": 15.73637316561845,
      "grad_norm": 0.3482051193714142,
      "learning_rate": 6.6098371651754085e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 19587128,
      "step": 30025
    },
    {
      "epoch": 15.738993710691824,
      "grad_norm": 0.45594269037246704,
      "learning_rate": 6.602093400551012e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 19590584,
      "step": 30030
    },
    {
      "epoch": 15.7416142557652,
      "grad_norm": 0.5259227156639099,
      "learning_rate": 6.59435348455879e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 19594264,
      "step": 30035
    },
    {
      "epoch": 15.744234800838575,
      "grad_norm": 2.248137950897217,
      "learning_rate": 6.586617418817828e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 19597176,
      "step": 30040
    },
    {
      "epoch": 15.74685534591195,
      "grad_norm": 0.3279888927936554,
      "learning_rate": 6.578885204946439e-06,
      "loss": 0.5557,
      "num_input_tokens_seen": 19599896,
      "step": 30045
    },
    {
      "epoch": 15.749475890985325,
      "grad_norm": 0.4578329622745514,
      "learning_rate": 6.571156844562098e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 19602904,
      "step": 30050
    },
    {
      "epoch": 15.7520964360587,
      "grad_norm": 0.3039514124393463,
      "learning_rate": 6.563432339281497e-06,
      "loss": 0.4538,
      "num_input_tokens_seen": 19606456,
      "step": 30055
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 0.3072201609611511,
      "learning_rate": 6.555711690720517e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 19609304,
      "step": 30060
    },
    {
      "epoch": 15.75733752620545,
      "grad_norm": 0.34096086025238037,
      "learning_rate": 6.547994900494234e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 19613304,
      "step": 30065
    },
    {
      "epoch": 15.759958071278826,
      "grad_norm": 0.41437745094299316,
      "learning_rate": 6.540281970216899e-06,
      "loss": 0.4543,
      "num_input_tokens_seen": 19616952,
      "step": 30070
    },
    {
      "epoch": 15.7625786163522,
      "grad_norm": 0.6158509254455566,
      "learning_rate": 6.5325729015019845e-06,
      "loss": 0.4158,
      "num_input_tokens_seen": 19620056,
      "step": 30075
    },
    {
      "epoch": 15.765199161425576,
      "grad_norm": 0.5428687930107117,
      "learning_rate": 6.524867695962122e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 19623096,
      "step": 30080
    },
    {
      "epoch": 15.767819706498951,
      "grad_norm": 0.49721482396125793,
      "learning_rate": 6.517166355209165e-06,
      "loss": 0.6459,
      "num_input_tokens_seen": 19625624,
      "step": 30085
    },
    {
      "epoch": 15.770440251572326,
      "grad_norm": 0.5206719040870667,
      "learning_rate": 6.50946888085415e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 19628312,
      "step": 30090
    },
    {
      "epoch": 15.773060796645701,
      "grad_norm": 0.6795042157173157,
      "learning_rate": 6.5017752745072855e-06,
      "loss": 0.3779,
      "num_input_tokens_seen": 19632056,
      "step": 30095
    },
    {
      "epoch": 15.775681341719078,
      "grad_norm": 0.5699576735496521,
      "learning_rate": 6.494085537777994e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 19637560,
      "step": 30100
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 0.5155341029167175,
      "learning_rate": 6.48639967227489e-06,
      "loss": 0.451,
      "num_input_tokens_seen": 19640600,
      "step": 30105
    },
    {
      "epoch": 15.780922431865829,
      "grad_norm": 0.3484971821308136,
      "learning_rate": 6.47871767960575e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 19643608,
      "step": 30110
    },
    {
      "epoch": 15.783542976939204,
      "grad_norm": 0.5323516130447388,
      "learning_rate": 6.471039561377581e-06,
      "loss": 0.5214,
      "num_input_tokens_seen": 19646040,
      "step": 30115
    },
    {
      "epoch": 15.786163522012579,
      "grad_norm": 0.5385193824768066,
      "learning_rate": 6.463365319196538e-06,
      "loss": 0.5107,
      "num_input_tokens_seen": 19648600,
      "step": 30120
    },
    {
      "epoch": 15.788784067085954,
      "grad_norm": 0.4672623574733734,
      "learning_rate": 6.4556949546679905e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 19652280,
      "step": 30125
    },
    {
      "epoch": 15.79140461215933,
      "grad_norm": 0.48213061690330505,
      "learning_rate": 6.448028469396497e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 19655736,
      "step": 30130
    },
    {
      "epoch": 15.794025157232705,
      "grad_norm": 0.5174271464347839,
      "learning_rate": 6.440365864985801e-06,
      "loss": 0.4935,
      "num_input_tokens_seen": 19658456,
      "step": 30135
    },
    {
      "epoch": 15.79664570230608,
      "grad_norm": 0.43049976229667664,
      "learning_rate": 6.432707143038818e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 19661688,
      "step": 30140
    },
    {
      "epoch": 15.799266247379455,
      "grad_norm": 0.9127957224845886,
      "learning_rate": 6.4250523051576834e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 19665400,
      "step": 30145
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 0.38419222831726074,
      "learning_rate": 6.417401352943686e-06,
      "loss": 0.3971,
      "num_input_tokens_seen": 19668472,
      "step": 30150
    },
    {
      "epoch": 15.804507337526205,
      "grad_norm": 0.6568211317062378,
      "learning_rate": 6.409754287997322e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 19672088,
      "step": 30155
    },
    {
      "epoch": 15.80712788259958,
      "grad_norm": 0.320160835981369,
      "learning_rate": 6.40211111191828e-06,
      "loss": 0.425,
      "num_input_tokens_seen": 19675864,
      "step": 30160
    },
    {
      "epoch": 15.809748427672956,
      "grad_norm": 0.4384572207927704,
      "learning_rate": 6.394471826305409e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 19680216,
      "step": 30165
    },
    {
      "epoch": 15.81236897274633,
      "grad_norm": 0.6422553658485413,
      "learning_rate": 6.3868364327567795e-06,
      "loss": 0.5776,
      "num_input_tokens_seen": 19683480,
      "step": 30170
    },
    {
      "epoch": 15.814989517819706,
      "grad_norm": 0.6766823530197144,
      "learning_rate": 6.379204932869606e-06,
      "loss": 0.5391,
      "num_input_tokens_seen": 19687160,
      "step": 30175
    },
    {
      "epoch": 15.817610062893081,
      "grad_norm": 0.2675630450248718,
      "learning_rate": 6.371577328240327e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 19690424,
      "step": 30180
    },
    {
      "epoch": 15.820230607966456,
      "grad_norm": 0.33794477581977844,
      "learning_rate": 6.363953620464547e-06,
      "loss": 0.443,
      "num_input_tokens_seen": 19693656,
      "step": 30185
    },
    {
      "epoch": 15.822851153039831,
      "grad_norm": 0.5273880958557129,
      "learning_rate": 6.356333811137064e-06,
      "loss": 0.5788,
      "num_input_tokens_seen": 19696920,
      "step": 30190
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 0.4313899874687195,
      "learning_rate": 6.34871790185185e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 19700088,
      "step": 30195
    },
    {
      "epoch": 15.828092243186584,
      "grad_norm": 0.499671995639801,
      "learning_rate": 6.341105894202057e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 19703320,
      "step": 30200
    },
    {
      "epoch": 15.830712788259959,
      "grad_norm": 0.992438554763794,
      "learning_rate": 6.333497789780041e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 19706168,
      "step": 30205
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.7628089785575867,
      "learning_rate": 6.325893590177329e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 19708984,
      "step": 30210
    },
    {
      "epoch": 15.835953878406709,
      "grad_norm": 0.5103996396064758,
      "learning_rate": 6.318293296984631e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 19711928,
      "step": 30215
    },
    {
      "epoch": 15.838574423480084,
      "grad_norm": 0.4049149453639984,
      "learning_rate": 6.3106969117918495e-06,
      "loss": 0.4283,
      "num_input_tokens_seen": 19715640,
      "step": 30220
    },
    {
      "epoch": 15.84119496855346,
      "grad_norm": 0.2983691394329071,
      "learning_rate": 6.303104436188057e-06,
      "loss": 0.401,
      "num_input_tokens_seen": 19718616,
      "step": 30225
    },
    {
      "epoch": 15.843815513626835,
      "grad_norm": 0.547927975654602,
      "learning_rate": 6.2955158717615036e-06,
      "loss": 0.4143,
      "num_input_tokens_seen": 19721688,
      "step": 30230
    },
    {
      "epoch": 15.84643605870021,
      "grad_norm": 0.5311259627342224,
      "learning_rate": 6.287931220099638e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 19725272,
      "step": 30235
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 0.3960718810558319,
      "learning_rate": 6.280350482789082e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 19727960,
      "step": 30240
    },
    {
      "epoch": 15.85167714884696,
      "grad_norm": 0.656982421875,
      "learning_rate": 6.272773661415645e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 19730904,
      "step": 30245
    },
    {
      "epoch": 15.854297693920335,
      "grad_norm": 0.38986340165138245,
      "learning_rate": 6.26520075756431e-06,
      "loss": 0.5073,
      "num_input_tokens_seen": 19734552,
      "step": 30250
    },
    {
      "epoch": 15.85691823899371,
      "grad_norm": 0.3668936789035797,
      "learning_rate": 6.2576317728192304e-06,
      "loss": 0.3902,
      "num_input_tokens_seen": 19737400,
      "step": 30255
    },
    {
      "epoch": 15.859538784067086,
      "grad_norm": 0.3707769513130188,
      "learning_rate": 6.250066708763761e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 19740792,
      "step": 30260
    },
    {
      "epoch": 15.86215932914046,
      "grad_norm": 0.3108007311820984,
      "learning_rate": 6.242505566980422e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 19744248,
      "step": 30265
    },
    {
      "epoch": 15.864779874213836,
      "grad_norm": 0.2350563406944275,
      "learning_rate": 6.234948349050931e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 19746744,
      "step": 30270
    },
    {
      "epoch": 15.867400419287211,
      "grad_norm": 0.5718319416046143,
      "learning_rate": 6.227395056556162e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 19749688,
      "step": 30275
    },
    {
      "epoch": 15.870020964360586,
      "grad_norm": 0.46502256393432617,
      "learning_rate": 6.219845691076173e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 19753208,
      "step": 30280
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 0.45182666182518005,
      "learning_rate": 6.212300254190206e-06,
      "loss": 0.5166,
      "num_input_tokens_seen": 19756568,
      "step": 30285
    },
    {
      "epoch": 15.875262054507338,
      "grad_norm": 0.7001965045928955,
      "learning_rate": 6.204758747476688e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 19759576,
      "step": 30290
    },
    {
      "epoch": 15.877882599580714,
      "grad_norm": 0.5203524827957153,
      "learning_rate": 6.1972211725132095e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 19763608,
      "step": 30295
    },
    {
      "epoch": 15.880503144654089,
      "grad_norm": 0.35414204001426697,
      "learning_rate": 6.189687530876559e-06,
      "loss": 0.6229,
      "num_input_tokens_seen": 19766808,
      "step": 30300
    },
    {
      "epoch": 15.883123689727464,
      "grad_norm": 0.38046571612358093,
      "learning_rate": 6.182157824142676e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 19769752,
      "step": 30305
    },
    {
      "epoch": 15.885744234800839,
      "grad_norm": 0.6961907148361206,
      "learning_rate": 6.174632053886681e-06,
      "loss": 0.6197,
      "num_input_tokens_seen": 19772920,
      "step": 30310
    },
    {
      "epoch": 15.888364779874214,
      "grad_norm": 0.5718712210655212,
      "learning_rate": 6.167110221682893e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 19775512,
      "step": 30315
    },
    {
      "epoch": 15.89098532494759,
      "grad_norm": 0.3251066207885742,
      "learning_rate": 6.159592329104788e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 19778968,
      "step": 30320
    },
    {
      "epoch": 15.893605870020965,
      "grad_norm": 0.6737895607948303,
      "learning_rate": 6.152078377725032e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 19781656,
      "step": 30325
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 0.90559321641922,
      "learning_rate": 6.144568369115453e-06,
      "loss": 0.4933,
      "num_input_tokens_seen": 19785112,
      "step": 30330
    },
    {
      "epoch": 15.898846960167715,
      "grad_norm": 0.3875389099121094,
      "learning_rate": 6.137062304847046e-06,
      "loss": 0.4618,
      "num_input_tokens_seen": 19788088,
      "step": 30335
    },
    {
      "epoch": 15.90146750524109,
      "grad_norm": 0.49903565645217896,
      "learning_rate": 6.129560186490008e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 19791352,
      "step": 30340
    },
    {
      "epoch": 15.904088050314465,
      "grad_norm": 0.5437004566192627,
      "learning_rate": 6.122062015613694e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 19795288,
      "step": 30345
    },
    {
      "epoch": 15.90670859538784,
      "grad_norm": 0.4986122250556946,
      "learning_rate": 6.114567793786641e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 19798328,
      "step": 30350
    },
    {
      "epoch": 15.909329140461216,
      "grad_norm": 0.44881561398506165,
      "learning_rate": 6.107077522576543e-06,
      "loss": 0.3346,
      "num_input_tokens_seen": 19801432,
      "step": 30355
    },
    {
      "epoch": 15.91194968553459,
      "grad_norm": 0.7008413672447205,
      "learning_rate": 6.099591203550292e-06,
      "loss": 0.5231,
      "num_input_tokens_seen": 19804280,
      "step": 30360
    },
    {
      "epoch": 15.914570230607966,
      "grad_norm": 0.31214049458503723,
      "learning_rate": 6.092108838273927e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 19807768,
      "step": 30365
    },
    {
      "epoch": 15.917190775681341,
      "grad_norm": 0.31348928809165955,
      "learning_rate": 6.08463042831268e-06,
      "loss": 0.4223,
      "num_input_tokens_seen": 19810616,
      "step": 30370
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 0.48431214690208435,
      "learning_rate": 6.0771559752309496e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 19816472,
      "step": 30375
    },
    {
      "epoch": 15.922431865828091,
      "grad_norm": 0.4437524974346161,
      "learning_rate": 6.069685480592313e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 19819928,
      "step": 30380
    },
    {
      "epoch": 15.925052410901468,
      "grad_norm": 0.44424721598625183,
      "learning_rate": 6.062218945959497e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 19823704,
      "step": 30385
    },
    {
      "epoch": 15.927672955974844,
      "grad_norm": 0.7276006937026978,
      "learning_rate": 6.054756372894435e-06,
      "loss": 0.4829,
      "num_input_tokens_seen": 19826456,
      "step": 30390
    },
    {
      "epoch": 15.930293501048219,
      "grad_norm": 0.332974910736084,
      "learning_rate": 6.04729776295819e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 19829848,
      "step": 30395
    },
    {
      "epoch": 15.932914046121594,
      "grad_norm": 0.5385597348213196,
      "learning_rate": 6.0398431177110306e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 19832856,
      "step": 30400
    },
    {
      "epoch": 15.935534591194969,
      "grad_norm": 0.5024040937423706,
      "learning_rate": 6.032392438712389e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 19835704,
      "step": 30405
    },
    {
      "epoch": 15.938155136268344,
      "grad_norm": 0.6903315782546997,
      "learning_rate": 6.024945727520847e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 19839352,
      "step": 30410
    },
    {
      "epoch": 15.94077568134172,
      "grad_norm": 0.795326828956604,
      "learning_rate": 6.01750298569419e-06,
      "loss": 0.4332,
      "num_input_tokens_seen": 19842840,
      "step": 30415
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 0.3590032756328583,
      "learning_rate": 6.010064214789335e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 19846808,
      "step": 30420
    },
    {
      "epoch": 15.94601677148847,
      "grad_norm": 0.29034289717674255,
      "learning_rate": 6.002629416362399e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 19849816,
      "step": 30425
    },
    {
      "epoch": 15.948637316561845,
      "grad_norm": 0.4117405414581299,
      "learning_rate": 5.995198591968662e-06,
      "loss": 0.3283,
      "num_input_tokens_seen": 19852568,
      "step": 30430
    },
    {
      "epoch": 15.95125786163522,
      "grad_norm": 0.4805019795894623,
      "learning_rate": 5.987771743162554e-06,
      "loss": 0.3013,
      "num_input_tokens_seen": 19857976,
      "step": 30435
    },
    {
      "epoch": 15.953878406708595,
      "grad_norm": 0.5236485004425049,
      "learning_rate": 5.9803488714976955e-06,
      "loss": 0.4575,
      "num_input_tokens_seen": 19861432,
      "step": 30440
    },
    {
      "epoch": 15.95649895178197,
      "grad_norm": 0.3490901589393616,
      "learning_rate": 5.972929978526872e-06,
      "loss": 0.4897,
      "num_input_tokens_seen": 19865144,
      "step": 30445
    },
    {
      "epoch": 15.959119496855346,
      "grad_norm": 0.38609448075294495,
      "learning_rate": 5.965515065802019e-06,
      "loss": 0.3794,
      "num_input_tokens_seen": 19867928,
      "step": 30450
    },
    {
      "epoch": 15.96174004192872,
      "grad_norm": 0.3771040737628937,
      "learning_rate": 5.958104134874254e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 19870456,
      "step": 30455
    },
    {
      "epoch": 15.964360587002096,
      "grad_norm": 0.47746095061302185,
      "learning_rate": 5.950697187293872e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 19872824,
      "step": 30460
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 0.4702434241771698,
      "learning_rate": 5.943294224610305e-06,
      "loss": 0.363,
      "num_input_tokens_seen": 19875256,
      "step": 30465
    },
    {
      "epoch": 15.969601677148846,
      "grad_norm": 0.9999205470085144,
      "learning_rate": 5.935895248372175e-06,
      "loss": 0.5036,
      "num_input_tokens_seen": 19878328,
      "step": 30470
    },
    {
      "epoch": 15.972222222222221,
      "grad_norm": 0.7763075828552246,
      "learning_rate": 5.928500260127273e-06,
      "loss": 0.4985,
      "num_input_tokens_seen": 19881784,
      "step": 30475
    },
    {
      "epoch": 15.974842767295598,
      "grad_norm": 0.5115447044372559,
      "learning_rate": 5.921109261422531e-06,
      "loss": 0.4676,
      "num_input_tokens_seen": 19885144,
      "step": 30480
    },
    {
      "epoch": 15.977463312368974,
      "grad_norm": 0.5651865005493164,
      "learning_rate": 5.913722253804071e-06,
      "loss": 0.5061,
      "num_input_tokens_seen": 19888120,
      "step": 30485
    },
    {
      "epoch": 15.980083857442349,
      "grad_norm": 0.5495031476020813,
      "learning_rate": 5.906339238817163e-06,
      "loss": 0.5223,
      "num_input_tokens_seen": 19891352,
      "step": 30490
    },
    {
      "epoch": 15.982704402515724,
      "grad_norm": 0.6927979588508606,
      "learning_rate": 5.8989602180062554e-06,
      "loss": 0.5103,
      "num_input_tokens_seen": 19894872,
      "step": 30495
    },
    {
      "epoch": 15.985324947589099,
      "grad_norm": 0.25263410806655884,
      "learning_rate": 5.891585192914953e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 19898904,
      "step": 30500
    },
    {
      "epoch": 15.987945492662474,
      "grad_norm": 0.36811164021492004,
      "learning_rate": 5.884214165086036e-06,
      "loss": 0.4792,
      "num_input_tokens_seen": 19902840,
      "step": 30505
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 0.5855839848518372,
      "learning_rate": 5.876847136061428e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 19906168,
      "step": 30510
    },
    {
      "epoch": 15.993186582809225,
      "grad_norm": 0.6046812534332275,
      "learning_rate": 5.869484107382228e-06,
      "loss": 0.6571,
      "num_input_tokens_seen": 19908920,
      "step": 30515
    },
    {
      "epoch": 15.9958071278826,
      "grad_norm": 0.7280755043029785,
      "learning_rate": 5.862125080588696e-06,
      "loss": 0.529,
      "num_input_tokens_seen": 19912216,
      "step": 30520
    },
    {
      "epoch": 15.998427672955975,
      "grad_norm": 0.5279971361160278,
      "learning_rate": 5.854770057220263e-06,
      "loss": 0.4605,
      "num_input_tokens_seen": 19914904,
      "step": 30525
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.4528324604034424,
      "eval_runtime": 9.8847,
      "eval_samples_per_second": 85.79,
      "eval_steps_per_second": 21.447,
      "num_input_tokens_seen": 19916008,
      "step": 30528
    },
    {
      "epoch": 16.00104821802935,
      "grad_norm": 0.5590280294418335,
      "learning_rate": 5.8474190388155216e-06,
      "loss": 0.4745,
      "num_input_tokens_seen": 19917032,
      "step": 30530
    },
    {
      "epoch": 16.003668763102727,
      "grad_norm": 0.5062882900238037,
      "learning_rate": 5.840072026912205e-06,
      "loss": 0.5808,
      "num_input_tokens_seen": 19919976,
      "step": 30535
    },
    {
      "epoch": 16.0062893081761,
      "grad_norm": 0.7058495879173279,
      "learning_rate": 5.83272902304724e-06,
      "loss": 0.4708,
      "num_input_tokens_seen": 19922312,
      "step": 30540
    },
    {
      "epoch": 16.008909853249477,
      "grad_norm": 0.594010591506958,
      "learning_rate": 5.825390028756683e-06,
      "loss": 0.5721,
      "num_input_tokens_seen": 19924392,
      "step": 30545
    },
    {
      "epoch": 16.01153039832285,
      "grad_norm": 0.6770200133323669,
      "learning_rate": 5.818055045575776e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 19927528,
      "step": 30550
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 0.28073909878730774,
      "learning_rate": 5.810724075038917e-06,
      "loss": 0.4264,
      "num_input_tokens_seen": 19930920,
      "step": 30555
    },
    {
      "epoch": 16.0167714884696,
      "grad_norm": 0.43755415081977844,
      "learning_rate": 5.803397118679665e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 19934152,
      "step": 30560
    },
    {
      "epoch": 16.019392033542978,
      "grad_norm": 0.6768324375152588,
      "learning_rate": 5.796074178030727e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 19938216,
      "step": 30565
    },
    {
      "epoch": 16.02201257861635,
      "grad_norm": 0.31836602091789246,
      "learning_rate": 5.788755254623976e-06,
      "loss": 0.524,
      "num_input_tokens_seen": 19942088,
      "step": 30570
    },
    {
      "epoch": 16.02463312368973,
      "grad_norm": 0.5209043025970459,
      "learning_rate": 5.7814403499904525e-06,
      "loss": 0.7469,
      "num_input_tokens_seen": 19944744,
      "step": 30575
    },
    {
      "epoch": 16.0272536687631,
      "grad_norm": 0.33860841393470764,
      "learning_rate": 5.774129465660352e-06,
      "loss": 0.4314,
      "num_input_tokens_seen": 19948360,
      "step": 30580
    },
    {
      "epoch": 16.02987421383648,
      "grad_norm": 0.6149572134017944,
      "learning_rate": 5.7668226031630246e-06,
      "loss": 0.4623,
      "num_input_tokens_seen": 19951016,
      "step": 30585
    },
    {
      "epoch": 16.032494758909852,
      "grad_norm": 0.48534339666366577,
      "learning_rate": 5.759519764026999e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 19954376,
      "step": 30590
    },
    {
      "epoch": 16.03511530398323,
      "grad_norm": 0.6503050923347473,
      "learning_rate": 5.7522209497799185e-06,
      "loss": 0.4512,
      "num_input_tokens_seen": 19957864,
      "step": 30595
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 0.3555641770362854,
      "learning_rate": 5.7449261619486246e-06,
      "loss": 0.4004,
      "num_input_tokens_seen": 19961256,
      "step": 30600
    },
    {
      "epoch": 16.04035639412998,
      "grad_norm": 0.3608911335468292,
      "learning_rate": 5.737635402059105e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 19964744,
      "step": 30605
    },
    {
      "epoch": 16.042976939203353,
      "grad_norm": 0.24111449718475342,
      "learning_rate": 5.730348671636501e-06,
      "loss": 0.4581,
      "num_input_tokens_seen": 19968232,
      "step": 30610
    },
    {
      "epoch": 16.04559748427673,
      "grad_norm": 0.377582311630249,
      "learning_rate": 5.723065972205124e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 19971176,
      "step": 30615
    },
    {
      "epoch": 16.048218029350103,
      "grad_norm": 0.6094039082527161,
      "learning_rate": 5.715787305288423e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 19974888,
      "step": 30620
    },
    {
      "epoch": 16.05083857442348,
      "grad_norm": 0.3834526836872101,
      "learning_rate": 5.7085126724090024e-06,
      "loss": 0.3887,
      "num_input_tokens_seen": 19977864,
      "step": 30625
    },
    {
      "epoch": 16.053459119496857,
      "grad_norm": 0.5247219204902649,
      "learning_rate": 5.7012420750886445e-06,
      "loss": 0.5109,
      "num_input_tokens_seen": 19981640,
      "step": 30630
    },
    {
      "epoch": 16.05607966457023,
      "grad_norm": 0.41645336151123047,
      "learning_rate": 5.693975514848271e-06,
      "loss": 0.5119,
      "num_input_tokens_seen": 19985640,
      "step": 30635
    },
    {
      "epoch": 16.058700209643607,
      "grad_norm": 0.45882242918014526,
      "learning_rate": 5.686712993207971e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 19988776,
      "step": 30640
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 0.5353769659996033,
      "learning_rate": 5.679454511686977e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 19991976,
      "step": 30645
    },
    {
      "epoch": 16.063941299790358,
      "grad_norm": 0.9908624887466431,
      "learning_rate": 5.672200071803671e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 19995048,
      "step": 30650
    },
    {
      "epoch": 16.06656184486373,
      "grad_norm": 0.32540765404701233,
      "learning_rate": 5.664949675075607e-06,
      "loss": 0.4845,
      "num_input_tokens_seen": 19997992,
      "step": 30655
    },
    {
      "epoch": 16.069182389937108,
      "grad_norm": 0.5747808814048767,
      "learning_rate": 5.657703323019486e-06,
      "loss": 0.366,
      "num_input_tokens_seen": 20000520,
      "step": 30660
    },
    {
      "epoch": 16.07180293501048,
      "grad_norm": 0.511836051940918,
      "learning_rate": 5.65046101715116e-06,
      "loss": 0.603,
      "num_input_tokens_seen": 20003176,
      "step": 30665
    },
    {
      "epoch": 16.07442348008386,
      "grad_norm": 0.40365949273109436,
      "learning_rate": 5.643222758985644e-06,
      "loss": 0.5136,
      "num_input_tokens_seen": 20006664,
      "step": 30670
    },
    {
      "epoch": 16.07704402515723,
      "grad_norm": 0.45471683144569397,
      "learning_rate": 5.635988550037094e-06,
      "loss": 0.4576,
      "num_input_tokens_seen": 20009320,
      "step": 30675
    },
    {
      "epoch": 16.07966457023061,
      "grad_norm": 1.158348798751831,
      "learning_rate": 5.628758391818817e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 20011656,
      "step": 30680
    },
    {
      "epoch": 16.082285115303982,
      "grad_norm": 0.34822720289230347,
      "learning_rate": 5.621532285843287e-06,
      "loss": 0.438,
      "num_input_tokens_seen": 20015304,
      "step": 30685
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 0.4268607795238495,
      "learning_rate": 5.614310233622119e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 20018504,
      "step": 30690
    },
    {
      "epoch": 16.087526205450732,
      "grad_norm": 0.32341280579566956,
      "learning_rate": 5.607092236666093e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 20021096,
      "step": 30695
    },
    {
      "epoch": 16.09014675052411,
      "grad_norm": 0.5799938440322876,
      "learning_rate": 5.599878296485125e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 20023912,
      "step": 30700
    },
    {
      "epoch": 16.092767295597483,
      "grad_norm": 0.33715906739234924,
      "learning_rate": 5.592668414588284e-06,
      "loss": 0.467,
      "num_input_tokens_seen": 20026792,
      "step": 30705
    },
    {
      "epoch": 16.09538784067086,
      "grad_norm": 0.8565027117729187,
      "learning_rate": 5.585462592483801e-06,
      "loss": 0.4367,
      "num_input_tokens_seen": 20028840,
      "step": 30710
    },
    {
      "epoch": 16.098008385744233,
      "grad_norm": 0.30315670371055603,
      "learning_rate": 5.578260831679047e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 20032072,
      "step": 30715
    },
    {
      "epoch": 16.10062893081761,
      "grad_norm": 0.4351491332054138,
      "learning_rate": 5.571063133680562e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 20034824,
      "step": 30720
    },
    {
      "epoch": 16.103249475890987,
      "grad_norm": 0.2699405252933502,
      "learning_rate": 5.563869499994004e-06,
      "loss": 0.3258,
      "num_input_tokens_seen": 20038792,
      "step": 30725
    },
    {
      "epoch": 16.10587002096436,
      "grad_norm": 0.4025421440601349,
      "learning_rate": 5.556679932124215e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 20042024,
      "step": 30730
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 0.9720214605331421,
      "learning_rate": 5.549494431575159e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 20045576,
      "step": 30735
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.4419799745082855,
      "learning_rate": 5.542312999849961e-06,
      "loss": 0.3585,
      "num_input_tokens_seen": 20048872,
      "step": 30740
    },
    {
      "epoch": 16.113731656184488,
      "grad_norm": 0.4666426181793213,
      "learning_rate": 5.5351356384509e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 20053096,
      "step": 30745
    },
    {
      "epoch": 16.11635220125786,
      "grad_norm": 0.49286773800849915,
      "learning_rate": 5.527962348879401e-06,
      "loss": 0.4501,
      "num_input_tokens_seen": 20056296,
      "step": 30750
    },
    {
      "epoch": 16.118972746331238,
      "grad_norm": 0.6341814994812012,
      "learning_rate": 5.520793132636027e-06,
      "loss": 0.5278,
      "num_input_tokens_seen": 20059368,
      "step": 30755
    },
    {
      "epoch": 16.12159329140461,
      "grad_norm": 0.41372379660606384,
      "learning_rate": 5.513627991220502e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 20062472,
      "step": 30760
    },
    {
      "epoch": 16.12421383647799,
      "grad_norm": 0.4582538902759552,
      "learning_rate": 5.506466926131684e-06,
      "loss": 0.3637,
      "num_input_tokens_seen": 20066024,
      "step": 30765
    },
    {
      "epoch": 16.12683438155136,
      "grad_norm": 0.5145205855369568,
      "learning_rate": 5.4993099388675916e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 20068968,
      "step": 30770
    },
    {
      "epoch": 16.12945492662474,
      "grad_norm": 0.42707228660583496,
      "learning_rate": 5.49215703092539e-06,
      "loss": 0.434,
      "num_input_tokens_seen": 20071720,
      "step": 30775
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 0.4229941964149475,
      "learning_rate": 5.4850082038013755e-06,
      "loss": 0.6698,
      "num_input_tokens_seen": 20074984,
      "step": 30780
    },
    {
      "epoch": 16.13469601677149,
      "grad_norm": 0.29921871423721313,
      "learning_rate": 5.4778634589910115e-06,
      "loss": 0.5575,
      "num_input_tokens_seen": 20078760,
      "step": 30785
    },
    {
      "epoch": 16.137316561844862,
      "grad_norm": 0.6433526277542114,
      "learning_rate": 5.4707227979888834e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 20081480,
      "step": 30790
    },
    {
      "epoch": 16.13993710691824,
      "grad_norm": 0.5116033554077148,
      "learning_rate": 5.463586222288747e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 20084072,
      "step": 30795
    },
    {
      "epoch": 16.142557651991613,
      "grad_norm": 0.35748305916786194,
      "learning_rate": 5.456453733383493e-06,
      "loss": 0.5485,
      "num_input_tokens_seen": 20088488,
      "step": 30800
    },
    {
      "epoch": 16.14517819706499,
      "grad_norm": 0.3196178376674652,
      "learning_rate": 5.449325332765151e-06,
      "loss": 0.4236,
      "num_input_tokens_seen": 20092008,
      "step": 30805
    },
    {
      "epoch": 16.147798742138363,
      "grad_norm": 0.5631436705589294,
      "learning_rate": 5.4422010219249e-06,
      "loss": 0.5086,
      "num_input_tokens_seen": 20094664,
      "step": 30810
    },
    {
      "epoch": 16.15041928721174,
      "grad_norm": 0.6609331369400024,
      "learning_rate": 5.435080802353077e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 20101640,
      "step": 30815
    },
    {
      "epoch": 16.153039832285117,
      "grad_norm": 0.43386656045913696,
      "learning_rate": 5.4279646755391325e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 20104840,
      "step": 30820
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 0.3100167214870453,
      "learning_rate": 5.420852642971691e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 20108456,
      "step": 30825
    },
    {
      "epoch": 16.158280922431867,
      "grad_norm": 0.3281736969947815,
      "learning_rate": 5.41374470613851e-06,
      "loss": 0.5488,
      "num_input_tokens_seen": 20110792,
      "step": 30830
    },
    {
      "epoch": 16.16090146750524,
      "grad_norm": 0.8348004817962646,
      "learning_rate": 5.40664086652648e-06,
      "loss": 0.3337,
      "num_input_tokens_seen": 20113256,
      "step": 30835
    },
    {
      "epoch": 16.163522012578618,
      "grad_norm": 0.6051434278488159,
      "learning_rate": 5.399541125621646e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 20116424,
      "step": 30840
    },
    {
      "epoch": 16.16614255765199,
      "grad_norm": 0.5453125834465027,
      "learning_rate": 5.392445484909203e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 20120200,
      "step": 30845
    },
    {
      "epoch": 16.168763102725368,
      "grad_norm": 0.3995283842086792,
      "learning_rate": 5.385353945873464e-06,
      "loss": 0.4825,
      "num_input_tokens_seen": 20122664,
      "step": 30850
    },
    {
      "epoch": 16.17138364779874,
      "grad_norm": 0.4638817310333252,
      "learning_rate": 5.378266509997909e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 20125224,
      "step": 30855
    },
    {
      "epoch": 16.17400419287212,
      "grad_norm": 0.3827146291732788,
      "learning_rate": 5.371183178765135e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 20128968,
      "step": 30860
    },
    {
      "epoch": 16.17662473794549,
      "grad_norm": 1.0926812887191772,
      "learning_rate": 5.364103953656904e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 20131656,
      "step": 30865
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 0.7518755197525024,
      "learning_rate": 5.35702883615411e-06,
      "loss": 0.4695,
      "num_input_tokens_seen": 20134696,
      "step": 30870
    },
    {
      "epoch": 16.181865828092242,
      "grad_norm": 0.6114441752433777,
      "learning_rate": 5.34995782773679e-06,
      "loss": 0.5106,
      "num_input_tokens_seen": 20137192,
      "step": 30875
    },
    {
      "epoch": 16.18448637316562,
      "grad_norm": 0.4543515145778656,
      "learning_rate": 5.3428909298841115e-06,
      "loss": 0.3218,
      "num_input_tokens_seen": 20140712,
      "step": 30880
    },
    {
      "epoch": 16.187106918238992,
      "grad_norm": 0.6027383208274841,
      "learning_rate": 5.335828144074384e-06,
      "loss": 0.5586,
      "num_input_tokens_seen": 20143496,
      "step": 30885
    },
    {
      "epoch": 16.18972746331237,
      "grad_norm": 0.3509059548377991,
      "learning_rate": 5.3287694717850695e-06,
      "loss": 0.3582,
      "num_input_tokens_seen": 20148296,
      "step": 30890
    },
    {
      "epoch": 16.192348008385743,
      "grad_norm": 0.5010210871696472,
      "learning_rate": 5.321714914492759e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 20152168,
      "step": 30895
    },
    {
      "epoch": 16.19496855345912,
      "grad_norm": 0.3550039529800415,
      "learning_rate": 5.314664473673195e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 20155784,
      "step": 30900
    },
    {
      "epoch": 16.197589098532493,
      "grad_norm": 0.40698572993278503,
      "learning_rate": 5.307618150801239e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 20159112,
      "step": 30905
    },
    {
      "epoch": 16.20020964360587,
      "grad_norm": 0.5540254712104797,
      "learning_rate": 5.300575947350897e-06,
      "loss": 0.3947,
      "num_input_tokens_seen": 20162152,
      "step": 30910
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 0.5336487889289856,
      "learning_rate": 5.293537864795325e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 20165032,
      "step": 30915
    },
    {
      "epoch": 16.20545073375262,
      "grad_norm": 0.6787645220756531,
      "learning_rate": 5.2865039046068075e-06,
      "loss": 0.4509,
      "num_input_tokens_seen": 20168136,
      "step": 30920
    },
    {
      "epoch": 16.208071278825997,
      "grad_norm": 0.44483643770217896,
      "learning_rate": 5.279474068256767e-06,
      "loss": 0.3637,
      "num_input_tokens_seen": 20171368,
      "step": 30925
    },
    {
      "epoch": 16.21069182389937,
      "grad_norm": 0.43562808632850647,
      "learning_rate": 5.272448357215779e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 20174664,
      "step": 30930
    },
    {
      "epoch": 16.213312368972748,
      "grad_norm": 0.5950190424919128,
      "learning_rate": 5.26542677295353e-06,
      "loss": 0.631,
      "num_input_tokens_seen": 20177224,
      "step": 30935
    },
    {
      "epoch": 16.21593291404612,
      "grad_norm": 0.5118923783302307,
      "learning_rate": 5.258409316938848e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 20180200,
      "step": 30940
    },
    {
      "epoch": 16.218553459119498,
      "grad_norm": 0.5544978380203247,
      "learning_rate": 5.251395990639713e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 20183592,
      "step": 30945
    },
    {
      "epoch": 16.22117400419287,
      "grad_norm": 0.7804810404777527,
      "learning_rate": 5.244386795523235e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 20186760,
      "step": 30950
    },
    {
      "epoch": 16.22379454926625,
      "grad_norm": 0.42631080746650696,
      "learning_rate": 5.23738173305566e-06,
      "loss": 0.4058,
      "num_input_tokens_seen": 20189864,
      "step": 30955
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 0.3241351544857025,
      "learning_rate": 5.230380804702362e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 20193096,
      "step": 30960
    },
    {
      "epoch": 16.229035639413,
      "grad_norm": 0.5164960622787476,
      "learning_rate": 5.223384011927851e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 20196136,
      "step": 30965
    },
    {
      "epoch": 16.231656184486372,
      "grad_norm": 0.5088607668876648,
      "learning_rate": 5.216391356195779e-06,
      "loss": 0.4566,
      "num_input_tokens_seen": 20199880,
      "step": 30970
    },
    {
      "epoch": 16.23427672955975,
      "grad_norm": 0.27237457036972046,
      "learning_rate": 5.209402838968935e-06,
      "loss": 0.4404,
      "num_input_tokens_seen": 20203752,
      "step": 30975
    },
    {
      "epoch": 16.236897274633122,
      "grad_norm": 0.8806478381156921,
      "learning_rate": 5.202418461709233e-06,
      "loss": 0.362,
      "num_input_tokens_seen": 20206760,
      "step": 30980
    },
    {
      "epoch": 16.2395178197065,
      "grad_norm": 0.2960238754749298,
      "learning_rate": 5.195438225877738e-06,
      "loss": 0.4731,
      "num_input_tokens_seen": 20209096,
      "step": 30985
    },
    {
      "epoch": 16.242138364779873,
      "grad_norm": 0.4471920132637024,
      "learning_rate": 5.18846213293461e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 20212520,
      "step": 30990
    },
    {
      "epoch": 16.24475890985325,
      "grad_norm": 0.5776588320732117,
      "learning_rate": 5.181490184339185e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 20215720,
      "step": 30995
    },
    {
      "epoch": 16.247379454926623,
      "grad_norm": 0.48465341329574585,
      "learning_rate": 5.17452238154991e-06,
      "loss": 0.3155,
      "num_input_tokens_seen": 20219048,
      "step": 31000
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.4989895820617676,
      "learning_rate": 5.1675587260243715e-06,
      "loss": 0.5985,
      "num_input_tokens_seen": 20222440,
      "step": 31005
    },
    {
      "epoch": 16.252620545073377,
      "grad_norm": 0.3751395344734192,
      "learning_rate": 5.160599219219295e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 20225224,
      "step": 31010
    },
    {
      "epoch": 16.25524109014675,
      "grad_norm": 0.5638219714164734,
      "learning_rate": 5.1536438625905185e-06,
      "loss": 0.4803,
      "num_input_tokens_seen": 20228680,
      "step": 31015
    },
    {
      "epoch": 16.257861635220127,
      "grad_norm": 0.6327576637268066,
      "learning_rate": 5.146692657593025e-06,
      "loss": 0.6103,
      "num_input_tokens_seen": 20231176,
      "step": 31020
    },
    {
      "epoch": 16.2604821802935,
      "grad_norm": 0.6669365167617798,
      "learning_rate": 5.139745605680926e-06,
      "loss": 0.4737,
      "num_input_tokens_seen": 20233672,
      "step": 31025
    },
    {
      "epoch": 16.263102725366878,
      "grad_norm": 0.8267830014228821,
      "learning_rate": 5.132802708307469e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 20236424,
      "step": 31030
    },
    {
      "epoch": 16.26572327044025,
      "grad_norm": 0.531225860118866,
      "learning_rate": 5.125863966925035e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 20239336,
      "step": 31035
    },
    {
      "epoch": 16.268343815513628,
      "grad_norm": 0.34785252809524536,
      "learning_rate": 5.118929382985116e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 20242056,
      "step": 31040
    },
    {
      "epoch": 16.270964360587,
      "grad_norm": 0.4264688193798065,
      "learning_rate": 5.111998957938363e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 20245384,
      "step": 31045
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 0.45479950308799744,
      "learning_rate": 5.105072693234525e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 20248936,
      "step": 31050
    },
    {
      "epoch": 16.27620545073375,
      "grad_norm": 0.3982606530189514,
      "learning_rate": 5.098150590322506e-06,
      "loss": 0.3878,
      "num_input_tokens_seen": 20253416,
      "step": 31055
    },
    {
      "epoch": 16.27882599580713,
      "grad_norm": 0.33390235900878906,
      "learning_rate": 5.091232650650335e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 20256968,
      "step": 31060
    },
    {
      "epoch": 16.281446540880502,
      "grad_norm": 1.1475342512130737,
      "learning_rate": 5.0843188756651645e-06,
      "loss": 0.5861,
      "num_input_tokens_seen": 20259752,
      "step": 31065
    },
    {
      "epoch": 16.28406708595388,
      "grad_norm": 0.420341819524765,
      "learning_rate": 5.077409266813276e-06,
      "loss": 0.482,
      "num_input_tokens_seen": 20262600,
      "step": 31070
    },
    {
      "epoch": 16.286687631027252,
      "grad_norm": 0.31664589047431946,
      "learning_rate": 5.0705038255400765e-06,
      "loss": 0.3817,
      "num_input_tokens_seen": 20265896,
      "step": 31075
    },
    {
      "epoch": 16.28930817610063,
      "grad_norm": 0.37988653779029846,
      "learning_rate": 5.063602553290106e-06,
      "loss": 0.518,
      "num_input_tokens_seen": 20269224,
      "step": 31080
    },
    {
      "epoch": 16.291928721174003,
      "grad_norm": 0.3700060248374939,
      "learning_rate": 5.056705451507035e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 20272072,
      "step": 31085
    },
    {
      "epoch": 16.29454926624738,
      "grad_norm": 1.0495984554290771,
      "learning_rate": 5.049812521633668e-06,
      "loss": 0.4559,
      "num_input_tokens_seen": 20275400,
      "step": 31090
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 0.5219842195510864,
      "learning_rate": 5.042923765111907e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 20279560,
      "step": 31095
    },
    {
      "epoch": 16.29979035639413,
      "grad_norm": 0.5708805322647095,
      "learning_rate": 5.036039183382821e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 20282728,
      "step": 31100
    },
    {
      "epoch": 16.302410901467507,
      "grad_norm": 0.4078352451324463,
      "learning_rate": 5.02915877788657e-06,
      "loss": 0.5405,
      "num_input_tokens_seen": 20285224,
      "step": 31105
    },
    {
      "epoch": 16.30503144654088,
      "grad_norm": 0.4188995659351349,
      "learning_rate": 5.0222825500624625e-06,
      "loss": 0.3748,
      "num_input_tokens_seen": 20288552,
      "step": 31110
    },
    {
      "epoch": 16.307651991614257,
      "grad_norm": 0.5270510315895081,
      "learning_rate": 5.015410501348935e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 20292776,
      "step": 31115
    },
    {
      "epoch": 16.31027253668763,
      "grad_norm": 0.5526954531669617,
      "learning_rate": 5.008542633183527e-06,
      "loss": 0.5956,
      "num_input_tokens_seen": 20295208,
      "step": 31120
    },
    {
      "epoch": 16.312893081761008,
      "grad_norm": 0.49696487188339233,
      "learning_rate": 5.001678947002927e-06,
      "loss": 0.4401,
      "num_input_tokens_seen": 20297960,
      "step": 31125
    },
    {
      "epoch": 16.31551362683438,
      "grad_norm": 0.5625665783882141,
      "learning_rate": 4.994819444242943e-06,
      "loss": 0.5532,
      "num_input_tokens_seen": 20300968,
      "step": 31130
    },
    {
      "epoch": 16.318134171907758,
      "grad_norm": 0.7859742045402527,
      "learning_rate": 4.9879641263384956e-06,
      "loss": 0.5019,
      "num_input_tokens_seen": 20304648,
      "step": 31135
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 0.6932491660118103,
      "learning_rate": 4.981112994723644e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 20308488,
      "step": 31140
    },
    {
      "epoch": 16.32337526205451,
      "grad_norm": 0.33608120679855347,
      "learning_rate": 4.974266050831572e-06,
      "loss": 0.4809,
      "num_input_tokens_seen": 20311560,
      "step": 31145
    },
    {
      "epoch": 16.32599580712788,
      "grad_norm": 0.365516722202301,
      "learning_rate": 4.967423296094567e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 20315592,
      "step": 31150
    },
    {
      "epoch": 16.32861635220126,
      "grad_norm": 0.48520103096961975,
      "learning_rate": 4.960584731944076e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 20319304,
      "step": 31155
    },
    {
      "epoch": 16.331236897274632,
      "grad_norm": 0.3082974851131439,
      "learning_rate": 4.953750359810627e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 20324264,
      "step": 31160
    },
    {
      "epoch": 16.33385744234801,
      "grad_norm": 0.7010086178779602,
      "learning_rate": 4.946920181123904e-06,
      "loss": 0.3468,
      "num_input_tokens_seen": 20326696,
      "step": 31165
    },
    {
      "epoch": 16.336477987421382,
      "grad_norm": 0.5224611163139343,
      "learning_rate": 4.940094197312709e-06,
      "loss": 0.5202,
      "num_input_tokens_seen": 20330248,
      "step": 31170
    },
    {
      "epoch": 16.33909853249476,
      "grad_norm": 0.5301392674446106,
      "learning_rate": 4.933272409804946e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 20332552,
      "step": 31175
    },
    {
      "epoch": 16.341719077568133,
      "grad_norm": 0.39866116642951965,
      "learning_rate": 4.926454820027663e-06,
      "loss": 0.5036,
      "num_input_tokens_seen": 20335304,
      "step": 31180
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 0.6188821792602539,
      "learning_rate": 4.919641429407024e-06,
      "loss": 0.5291,
      "num_input_tokens_seen": 20337992,
      "step": 31185
    },
    {
      "epoch": 16.346960167714883,
      "grad_norm": 0.6109158992767334,
      "learning_rate": 4.912832239368304e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 20340680,
      "step": 31190
    },
    {
      "epoch": 16.34958071278826,
      "grad_norm": 0.5444687008857727,
      "learning_rate": 4.906027251335918e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 20343848,
      "step": 31195
    },
    {
      "epoch": 16.352201257861637,
      "grad_norm": 0.47605258226394653,
      "learning_rate": 4.899226466733381e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 20346792,
      "step": 31200
    },
    {
      "epoch": 16.35482180293501,
      "grad_norm": 0.4181668758392334,
      "learning_rate": 4.892429886983346e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 20349544,
      "step": 31205
    },
    {
      "epoch": 16.357442348008387,
      "grad_norm": 0.5515347719192505,
      "learning_rate": 4.88563751350758e-06,
      "loss": 0.4775,
      "num_input_tokens_seen": 20352904,
      "step": 31210
    },
    {
      "epoch": 16.36006289308176,
      "grad_norm": 0.53714919090271,
      "learning_rate": 4.878849347726977e-06,
      "loss": 0.3307,
      "num_input_tokens_seen": 20356424,
      "step": 31215
    },
    {
      "epoch": 16.362683438155138,
      "grad_norm": 0.3858570158481598,
      "learning_rate": 4.872065391061526e-06,
      "loss": 0.4107,
      "num_input_tokens_seen": 20360264,
      "step": 31220
    },
    {
      "epoch": 16.36530398322851,
      "grad_norm": 0.4485689401626587,
      "learning_rate": 4.865285644930373e-06,
      "loss": 0.4361,
      "num_input_tokens_seen": 20363848,
      "step": 31225
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 0.512252151966095,
      "learning_rate": 4.858510110751746e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 20366280,
      "step": 31230
    },
    {
      "epoch": 16.37054507337526,
      "grad_norm": 0.27239498496055603,
      "learning_rate": 4.851738789943019e-06,
      "loss": 0.2849,
      "num_input_tokens_seen": 20370568,
      "step": 31235
    },
    {
      "epoch": 16.37316561844864,
      "grad_norm": 0.4721827805042267,
      "learning_rate": 4.844971683920674e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 20373928,
      "step": 31240
    },
    {
      "epoch": 16.37578616352201,
      "grad_norm": 0.591084897518158,
      "learning_rate": 4.838208794100316e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 20376488,
      "step": 31245
    },
    {
      "epoch": 16.37840670859539,
      "grad_norm": 0.6133202910423279,
      "learning_rate": 4.831450121896661e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 20379176,
      "step": 31250
    },
    {
      "epoch": 16.381027253668762,
      "grad_norm": 0.7524320483207703,
      "learning_rate": 4.824695668723542e-06,
      "loss": 0.4915,
      "num_input_tokens_seen": 20381736,
      "step": 31255
    },
    {
      "epoch": 16.38364779874214,
      "grad_norm": 0.4052830934524536,
      "learning_rate": 4.817945435993915e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 20385064,
      "step": 31260
    },
    {
      "epoch": 16.386268343815512,
      "grad_norm": 0.29583391547203064,
      "learning_rate": 4.811199425119853e-06,
      "loss": 0.4725,
      "num_input_tokens_seen": 20388072,
      "step": 31265
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.955905020236969,
      "learning_rate": 4.804457637512552e-06,
      "loss": 0.5457,
      "num_input_tokens_seen": 20391080,
      "step": 31270
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 1.0036671161651611,
      "learning_rate": 4.79772007458231e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 20394024,
      "step": 31275
    },
    {
      "epoch": 16.39412997903564,
      "grad_norm": 0.21621665358543396,
      "learning_rate": 4.7909867377385424e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 20397672,
      "step": 31280
    },
    {
      "epoch": 16.396750524109013,
      "grad_norm": 0.30327674746513367,
      "learning_rate": 4.784257628389794e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 20401352,
      "step": 31285
    },
    {
      "epoch": 16.39937106918239,
      "grad_norm": 0.4654964208602905,
      "learning_rate": 4.7775327479437166e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 20404008,
      "step": 31290
    },
    {
      "epoch": 16.401991614255767,
      "grad_norm": 0.3761501610279083,
      "learning_rate": 4.770812097807076e-06,
      "loss": 0.6065,
      "num_input_tokens_seen": 20407496,
      "step": 31295
    },
    {
      "epoch": 16.40461215932914,
      "grad_norm": 0.7468111515045166,
      "learning_rate": 4.764095679385766e-06,
      "loss": 0.4215,
      "num_input_tokens_seen": 20410696,
      "step": 31300
    },
    {
      "epoch": 16.407232704402517,
      "grad_norm": 0.8434670567512512,
      "learning_rate": 4.7573834940847745e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 20413736,
      "step": 31305
    },
    {
      "epoch": 16.40985324947589,
      "grad_norm": 0.8876402974128723,
      "learning_rate": 4.750675543308212e-06,
      "loss": 0.4105,
      "num_input_tokens_seen": 20417160,
      "step": 31310
    },
    {
      "epoch": 16.412473794549268,
      "grad_norm": 0.29259607195854187,
      "learning_rate": 4.743971828459307e-06,
      "loss": 0.436,
      "num_input_tokens_seen": 20420520,
      "step": 31315
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 0.43541190028190613,
      "learning_rate": 4.7372723509404025e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 20423272,
      "step": 31320
    },
    {
      "epoch": 16.417714884696018,
      "grad_norm": 0.7143450975418091,
      "learning_rate": 4.7305771121529615e-06,
      "loss": 0.4116,
      "num_input_tokens_seen": 20426216,
      "step": 31325
    },
    {
      "epoch": 16.42033542976939,
      "grad_norm": 0.5469868183135986,
      "learning_rate": 4.723886113497542e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 20429512,
      "step": 31330
    },
    {
      "epoch": 16.42295597484277,
      "grad_norm": 0.27262985706329346,
      "learning_rate": 4.7171993563738195e-06,
      "loss": 0.5061,
      "num_input_tokens_seen": 20433064,
      "step": 31335
    },
    {
      "epoch": 16.42557651991614,
      "grad_norm": 1.1054240465164185,
      "learning_rate": 4.7105168421805924e-06,
      "loss": 0.5116,
      "num_input_tokens_seen": 20436232,
      "step": 31340
    },
    {
      "epoch": 16.42819706498952,
      "grad_norm": 0.46651124954223633,
      "learning_rate": 4.703838572315767e-06,
      "loss": 0.4817,
      "num_input_tokens_seen": 20439688,
      "step": 31345
    },
    {
      "epoch": 16.430817610062892,
      "grad_norm": 0.5360879898071289,
      "learning_rate": 4.697164548176367e-06,
      "loss": 0.6498,
      "num_input_tokens_seen": 20442888,
      "step": 31350
    },
    {
      "epoch": 16.43343815513627,
      "grad_norm": 1.2025073766708374,
      "learning_rate": 4.690494771158519e-06,
      "loss": 0.5012,
      "num_input_tokens_seen": 20445928,
      "step": 31355
    },
    {
      "epoch": 16.436058700209642,
      "grad_norm": 0.5107671618461609,
      "learning_rate": 4.683829242657453e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 20448552,
      "step": 31360
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 0.4986889064311981,
      "learning_rate": 4.677167964067533e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 20451304,
      "step": 31365
    },
    {
      "epoch": 16.441299790356393,
      "grad_norm": 0.6244668960571289,
      "learning_rate": 4.670510936782216e-06,
      "loss": 0.4661,
      "num_input_tokens_seen": 20454408,
      "step": 31370
    },
    {
      "epoch": 16.44392033542977,
      "grad_norm": 0.4184506833553314,
      "learning_rate": 4.6638581621940815e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 20457384,
      "step": 31375
    },
    {
      "epoch": 16.446540880503143,
      "grad_norm": 0.4168553054332733,
      "learning_rate": 4.657209641694821e-06,
      "loss": 0.422,
      "num_input_tokens_seen": 20460456,
      "step": 31380
    },
    {
      "epoch": 16.44916142557652,
      "grad_norm": 0.41415032744407654,
      "learning_rate": 4.65056537667522e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 20463336,
      "step": 31385
    },
    {
      "epoch": 16.451781970649897,
      "grad_norm": 0.9774859547615051,
      "learning_rate": 4.643925368525176e-06,
      "loss": 0.4876,
      "num_input_tokens_seen": 20465928,
      "step": 31390
    },
    {
      "epoch": 16.45440251572327,
      "grad_norm": 0.5209803581237793,
      "learning_rate": 4.637289618633711e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 20468584,
      "step": 31395
    },
    {
      "epoch": 16.457023060796647,
      "grad_norm": 0.550237774848938,
      "learning_rate": 4.630658128388948e-06,
      "loss": 0.4685,
      "num_input_tokens_seen": 20471336,
      "step": 31400
    },
    {
      "epoch": 16.45964360587002,
      "grad_norm": 0.7663949728012085,
      "learning_rate": 4.624030899178125e-06,
      "loss": 0.396,
      "num_input_tokens_seen": 20474696,
      "step": 31405
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 0.495883971452713,
      "learning_rate": 4.617407932387569e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 20477832,
      "step": 31410
    },
    {
      "epoch": 16.46488469601677,
      "grad_norm": 0.5130205750465393,
      "learning_rate": 4.610789229402743e-06,
      "loss": 0.5271,
      "num_input_tokens_seen": 20481160,
      "step": 31415
    },
    {
      "epoch": 16.467505241090148,
      "grad_norm": 0.4176497161388397,
      "learning_rate": 4.604174791608193e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 20484648,
      "step": 31420
    },
    {
      "epoch": 16.47012578616352,
      "grad_norm": 0.8837645649909973,
      "learning_rate": 4.597564620387587e-06,
      "loss": 0.2745,
      "num_input_tokens_seen": 20486696,
      "step": 31425
    },
    {
      "epoch": 16.4727463312369,
      "grad_norm": 0.5385043025016785,
      "learning_rate": 4.590958717123708e-06,
      "loss": 0.4316,
      "num_input_tokens_seen": 20489640,
      "step": 31430
    },
    {
      "epoch": 16.47536687631027,
      "grad_norm": 0.38821759819984436,
      "learning_rate": 4.5843570831984174e-06,
      "loss": 0.5505,
      "num_input_tokens_seen": 20492392,
      "step": 31435
    },
    {
      "epoch": 16.47798742138365,
      "grad_norm": 0.5635508894920349,
      "learning_rate": 4.577759719992719e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 20494984,
      "step": 31440
    },
    {
      "epoch": 16.480607966457022,
      "grad_norm": 0.6416059732437134,
      "learning_rate": 4.571166628886692e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 20497832,
      "step": 31445
    },
    {
      "epoch": 16.4832285115304,
      "grad_norm": 0.4074351191520691,
      "learning_rate": 4.564577811259541e-06,
      "loss": 0.3752,
      "num_input_tokens_seen": 20500488,
      "step": 31450
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 0.45696666836738586,
      "learning_rate": 4.557993268489574e-06,
      "loss": 0.4708,
      "num_input_tokens_seen": 20503688,
      "step": 31455
    },
    {
      "epoch": 16.48846960167715,
      "grad_norm": 0.4014734923839569,
      "learning_rate": 4.551413001954205e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 20507816,
      "step": 31460
    },
    {
      "epoch": 16.491090146750523,
      "grad_norm": 0.663882315158844,
      "learning_rate": 4.54483701302994e-06,
      "loss": 0.5617,
      "num_input_tokens_seen": 20510920,
      "step": 31465
    },
    {
      "epoch": 16.4937106918239,
      "grad_norm": 0.4321327209472656,
      "learning_rate": 4.538265303092415e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 20514408,
      "step": 31470
    },
    {
      "epoch": 16.496331236897273,
      "grad_norm": 0.6247777342796326,
      "learning_rate": 4.531697873516338e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 20518472,
      "step": 31475
    },
    {
      "epoch": 16.49895178197065,
      "grad_norm": 0.47849658131599426,
      "learning_rate": 4.525134725675556e-06,
      "loss": 0.5705,
      "num_input_tokens_seen": 20521064,
      "step": 31480
    },
    {
      "epoch": 16.501572327044027,
      "grad_norm": 1.0221699476242065,
      "learning_rate": 4.518575860943003e-06,
      "loss": 0.5,
      "num_input_tokens_seen": 20523784,
      "step": 31485
    },
    {
      "epoch": 16.5041928721174,
      "grad_norm": 0.2964824438095093,
      "learning_rate": 4.512021280690712e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 20527976,
      "step": 31490
    },
    {
      "epoch": 16.506813417190777,
      "grad_norm": 0.49394917488098145,
      "learning_rate": 4.505470986289828e-06,
      "loss": 0.5408,
      "num_input_tokens_seen": 20531784,
      "step": 31495
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 0.806463897228241,
      "learning_rate": 4.498924979110605e-06,
      "loss": 0.4934,
      "num_input_tokens_seen": 20535048,
      "step": 31500
    },
    {
      "epoch": 16.512054507337528,
      "grad_norm": 0.4510241746902466,
      "learning_rate": 4.492383260522384e-06,
      "loss": 0.5468,
      "num_input_tokens_seen": 20537672,
      "step": 31505
    },
    {
      "epoch": 16.5146750524109,
      "grad_norm": 0.5960895419120789,
      "learning_rate": 4.485845831893626e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 20540264,
      "step": 31510
    },
    {
      "epoch": 16.517295597484278,
      "grad_norm": 0.4014683961868286,
      "learning_rate": 4.479312694591878e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 20543208,
      "step": 31515
    },
    {
      "epoch": 16.51991614255765,
      "grad_norm": 0.6494557857513428,
      "learning_rate": 4.4727838499838e-06,
      "loss": 0.405,
      "num_input_tokens_seen": 20546632,
      "step": 31520
    },
    {
      "epoch": 16.52253668763103,
      "grad_norm": 0.5472298860549927,
      "learning_rate": 4.4662592994351545e-06,
      "loss": 0.5545,
      "num_input_tokens_seen": 20549256,
      "step": 31525
    },
    {
      "epoch": 16.5251572327044,
      "grad_norm": 0.369203120470047,
      "learning_rate": 4.4597390443108065e-06,
      "loss": 0.5466,
      "num_input_tokens_seen": 20551880,
      "step": 31530
    },
    {
      "epoch": 16.52777777777778,
      "grad_norm": 0.5061458349227905,
      "learning_rate": 4.453223085974712e-06,
      "loss": 0.4147,
      "num_input_tokens_seen": 20555368,
      "step": 31535
    },
    {
      "epoch": 16.530398322851152,
      "grad_norm": 0.3589203655719757,
      "learning_rate": 4.4467114257899425e-06,
      "loss": 0.323,
      "num_input_tokens_seen": 20557864,
      "step": 31540
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 0.34070518612861633,
      "learning_rate": 4.44020406511865e-06,
      "loss": 0.3953,
      "num_input_tokens_seen": 20561800,
      "step": 31545
    },
    {
      "epoch": 16.535639412997902,
      "grad_norm": 0.35560476779937744,
      "learning_rate": 4.43370100532211e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 20565128,
      "step": 31550
    },
    {
      "epoch": 16.53825995807128,
      "grad_norm": 0.7484133839607239,
      "learning_rate": 4.427202247760695e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 20568648,
      "step": 31555
    },
    {
      "epoch": 16.540880503144653,
      "grad_norm": 0.3927772641181946,
      "learning_rate": 4.420707793793854e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 20571816,
      "step": 31560
    },
    {
      "epoch": 16.54350104821803,
      "grad_norm": 0.44547367095947266,
      "learning_rate": 4.414217644780172e-06,
      "loss": 0.5253,
      "num_input_tokens_seen": 20576136,
      "step": 31565
    },
    {
      "epoch": 16.546121593291403,
      "grad_norm": 0.5914384722709656,
      "learning_rate": 4.407731802077294e-06,
      "loss": 0.5025,
      "num_input_tokens_seen": 20578472,
      "step": 31570
    },
    {
      "epoch": 16.54874213836478,
      "grad_norm": 0.34498751163482666,
      "learning_rate": 4.401250267041996e-06,
      "loss": 0.5232,
      "num_input_tokens_seen": 20581800,
      "step": 31575
    },
    {
      "epoch": 16.551362683438157,
      "grad_norm": 0.47004902362823486,
      "learning_rate": 4.394773041030137e-06,
      "loss": 0.6365,
      "num_input_tokens_seen": 20584296,
      "step": 31580
    },
    {
      "epoch": 16.55398322851153,
      "grad_norm": 0.6239703893661499,
      "learning_rate": 4.388300125396685e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 20587368,
      "step": 31585
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 0.6053426265716553,
      "learning_rate": 4.381831521495699e-06,
      "loss": 0.535,
      "num_input_tokens_seen": 20590504,
      "step": 31590
    },
    {
      "epoch": 16.55922431865828,
      "grad_norm": 0.3399604558944702,
      "learning_rate": 4.37536723068033e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 20593448,
      "step": 31595
    },
    {
      "epoch": 16.561844863731658,
      "grad_norm": 1.6320246458053589,
      "learning_rate": 4.368907254302837e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 20596904,
      "step": 31600
    },
    {
      "epoch": 16.56446540880503,
      "grad_norm": 0.6842033863067627,
      "learning_rate": 4.3624515937145725e-06,
      "loss": 0.4447,
      "num_input_tokens_seen": 20600040,
      "step": 31605
    },
    {
      "epoch": 16.567085953878408,
      "grad_norm": 1.2335925102233887,
      "learning_rate": 4.3560002502659905e-06,
      "loss": 0.4941,
      "num_input_tokens_seen": 20603112,
      "step": 31610
    },
    {
      "epoch": 16.56970649895178,
      "grad_norm": 0.4814818501472473,
      "learning_rate": 4.349553225306644e-06,
      "loss": 0.447,
      "num_input_tokens_seen": 20606760,
      "step": 31615
    },
    {
      "epoch": 16.572327044025158,
      "grad_norm": 0.8102845549583435,
      "learning_rate": 4.343110520185167e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 20609896,
      "step": 31620
    },
    {
      "epoch": 16.57494758909853,
      "grad_norm": 1.7019517421722412,
      "learning_rate": 4.336672136249298e-06,
      "loss": 0.4966,
      "num_input_tokens_seen": 20612712,
      "step": 31625
    },
    {
      "epoch": 16.57756813417191,
      "grad_norm": 0.4854390323162079,
      "learning_rate": 4.33023807484588e-06,
      "loss": 0.7015,
      "num_input_tokens_seen": 20616488,
      "step": 31630
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 0.693790853023529,
      "learning_rate": 4.323808337320842e-06,
      "loss": 0.5534,
      "num_input_tokens_seen": 20619944,
      "step": 31635
    },
    {
      "epoch": 16.58280922431866,
      "grad_norm": 0.4685388505458832,
      "learning_rate": 4.31738292501922e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 20622952,
      "step": 31640
    },
    {
      "epoch": 16.585429769392032,
      "grad_norm": 0.4195692241191864,
      "learning_rate": 4.31096183928513e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 20626248,
      "step": 31645
    },
    {
      "epoch": 16.58805031446541,
      "grad_norm": 0.39890629053115845,
      "learning_rate": 4.304545081461783e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 20630088,
      "step": 31650
    },
    {
      "epoch": 16.590670859538783,
      "grad_norm": 0.8101566433906555,
      "learning_rate": 4.298132652891501e-06,
      "loss": 0.4648,
      "num_input_tokens_seen": 20632840,
      "step": 31655
    },
    {
      "epoch": 16.59329140461216,
      "grad_norm": 0.4153977334499359,
      "learning_rate": 4.291724554915688e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 20635720,
      "step": 31660
    },
    {
      "epoch": 16.595911949685533,
      "grad_norm": 1.028037190437317,
      "learning_rate": 4.285320788874853e-06,
      "loss": 0.3565,
      "num_input_tokens_seen": 20639400,
      "step": 31665
    },
    {
      "epoch": 16.59853249475891,
      "grad_norm": 0.3797186315059662,
      "learning_rate": 4.278921356108581e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 20643656,
      "step": 31670
    },
    {
      "epoch": 16.601153039832283,
      "grad_norm": 0.43504202365875244,
      "learning_rate": 4.272526257955556e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 20646504,
      "step": 31675
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 0.7556583285331726,
      "learning_rate": 4.266135495753567e-06,
      "loss": 0.5273,
      "num_input_tokens_seen": 20650120,
      "step": 31680
    },
    {
      "epoch": 16.606394129979037,
      "grad_norm": 0.7427923083305359,
      "learning_rate": 4.259749070839491e-06,
      "loss": 0.3943,
      "num_input_tokens_seen": 20653064,
      "step": 31685
    },
    {
      "epoch": 16.60901467505241,
      "grad_norm": 0.2896905839443207,
      "learning_rate": 4.253366984549289e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 20657160,
      "step": 31690
    },
    {
      "epoch": 16.611635220125788,
      "grad_norm": 0.5825565457344055,
      "learning_rate": 4.24698923821803e-06,
      "loss": 0.3403,
      "num_input_tokens_seen": 20659912,
      "step": 31695
    },
    {
      "epoch": 16.61425576519916,
      "grad_norm": 0.4440615773200989,
      "learning_rate": 4.240615833179861e-06,
      "loss": 0.3439,
      "num_input_tokens_seen": 20662312,
      "step": 31700
    },
    {
      "epoch": 16.616876310272538,
      "grad_norm": 0.7463188171386719,
      "learning_rate": 4.234246770768016e-06,
      "loss": 0.4158,
      "num_input_tokens_seen": 20666248,
      "step": 31705
    },
    {
      "epoch": 16.61949685534591,
      "grad_norm": 0.5782614350318909,
      "learning_rate": 4.227882052314841e-06,
      "loss": 0.4122,
      "num_input_tokens_seen": 20670888,
      "step": 31710
    },
    {
      "epoch": 16.622117400419288,
      "grad_norm": 0.7724469304084778,
      "learning_rate": 4.2215216791517575e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 20673320,
      "step": 31715
    },
    {
      "epoch": 16.62473794549266,
      "grad_norm": 0.447590708732605,
      "learning_rate": 4.215165652609293e-06,
      "loss": 0.3255,
      "num_input_tokens_seen": 20676552,
      "step": 31720
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 0.34846773743629456,
      "learning_rate": 4.208813974017048e-06,
      "loss": 0.4335,
      "num_input_tokens_seen": 20678792,
      "step": 31725
    },
    {
      "epoch": 16.629979035639412,
      "grad_norm": 0.5123885273933411,
      "learning_rate": 4.202466644703712e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 20683240,
      "step": 31730
    },
    {
      "epoch": 16.63259958071279,
      "grad_norm": 0.4443627893924713,
      "learning_rate": 4.196123665997087e-06,
      "loss": 0.5388,
      "num_input_tokens_seen": 20686536,
      "step": 31735
    },
    {
      "epoch": 16.635220125786162,
      "grad_norm": 0.3809460401535034,
      "learning_rate": 4.189785039224045e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 20689832,
      "step": 31740
    },
    {
      "epoch": 16.63784067085954,
      "grad_norm": 0.568185567855835,
      "learning_rate": 4.183450765710567e-06,
      "loss": 0.4968,
      "num_input_tokens_seen": 20692648,
      "step": 31745
    },
    {
      "epoch": 16.640461215932913,
      "grad_norm": 0.7726665139198303,
      "learning_rate": 4.177120846781693e-06,
      "loss": 0.4626,
      "num_input_tokens_seen": 20695048,
      "step": 31750
    },
    {
      "epoch": 16.64308176100629,
      "grad_norm": 0.41180533170700073,
      "learning_rate": 4.170795283761583e-06,
      "loss": 0.3957,
      "num_input_tokens_seen": 20698216,
      "step": 31755
    },
    {
      "epoch": 16.645702306079663,
      "grad_norm": 0.23435325920581818,
      "learning_rate": 4.164474077973462e-06,
      "loss": 0.4819,
      "num_input_tokens_seen": 20702024,
      "step": 31760
    },
    {
      "epoch": 16.64832285115304,
      "grad_norm": 0.4454030692577362,
      "learning_rate": 4.158157230739659e-06,
      "loss": 0.3327,
      "num_input_tokens_seen": 20705032,
      "step": 31765
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 0.2972162067890167,
      "learning_rate": 4.151844743381589e-06,
      "loss": 0.3595,
      "num_input_tokens_seen": 20708072,
      "step": 31770
    },
    {
      "epoch": 16.65356394129979,
      "grad_norm": 0.54209965467453,
      "learning_rate": 4.145536617219756e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 20711432,
      "step": 31775
    },
    {
      "epoch": 16.656184486373167,
      "grad_norm": 0.7790505290031433,
      "learning_rate": 4.139232853573735e-06,
      "loss": 0.5015,
      "num_input_tokens_seen": 20715112,
      "step": 31780
    },
    {
      "epoch": 16.65880503144654,
      "grad_norm": 1.3607159852981567,
      "learning_rate": 4.132933453762214e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 20718024,
      "step": 31785
    },
    {
      "epoch": 16.661425576519918,
      "grad_norm": 0.6125066876411438,
      "learning_rate": 4.126638419102946e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 20721192,
      "step": 31790
    },
    {
      "epoch": 16.66404612159329,
      "grad_norm": 0.796346127986908,
      "learning_rate": 4.120347750912784e-06,
      "loss": 0.3738,
      "num_input_tokens_seen": 20723656,
      "step": 31795
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.4392852783203125,
      "learning_rate": 4.1140614505076724e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 20726088,
      "step": 31800
    },
    {
      "epoch": 16.66928721174004,
      "grad_norm": 0.39947670698165894,
      "learning_rate": 4.10777951920262e-06,
      "loss": 0.3523,
      "num_input_tokens_seen": 20728840,
      "step": 31805
    },
    {
      "epoch": 16.671907756813418,
      "grad_norm": 0.4488472044467926,
      "learning_rate": 4.1015019583117455e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 20732424,
      "step": 31810
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 0.3063036799430847,
      "learning_rate": 4.095228769148232e-06,
      "loss": 0.3315,
      "num_input_tokens_seen": 20738184,
      "step": 31815
    },
    {
      "epoch": 16.67714884696017,
      "grad_norm": 0.2170371264219284,
      "learning_rate": 4.088959953024368e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 20743464,
      "step": 31820
    },
    {
      "epoch": 16.679769392033542,
      "grad_norm": 0.4124796688556671,
      "learning_rate": 4.082695511251522e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 20746536,
      "step": 31825
    },
    {
      "epoch": 16.68238993710692,
      "grad_norm": 0.8376900553703308,
      "learning_rate": 4.076435445140131e-06,
      "loss": 0.5242,
      "num_input_tokens_seen": 20749896,
      "step": 31830
    },
    {
      "epoch": 16.685010482180292,
      "grad_norm": 0.4961246848106384,
      "learning_rate": 4.070179755999737e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 20753448,
      "step": 31835
    },
    {
      "epoch": 16.68763102725367,
      "grad_norm": 0.5700386166572571,
      "learning_rate": 4.063928445138962e-06,
      "loss": 0.4725,
      "num_input_tokens_seen": 20757032,
      "step": 31840
    },
    {
      "epoch": 16.690251572327043,
      "grad_norm": 0.324314683675766,
      "learning_rate": 4.057681513865499e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 20760136,
      "step": 31845
    },
    {
      "epoch": 16.69287211740042,
      "grad_norm": 0.5968024730682373,
      "learning_rate": 4.0514389634861426e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 20763144,
      "step": 31850
    },
    {
      "epoch": 16.695492662473793,
      "grad_norm": 0.5920572280883789,
      "learning_rate": 4.045200795306767e-06,
      "loss": 0.4829,
      "num_input_tokens_seen": 20766248,
      "step": 31855
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 0.7591057419776917,
      "learning_rate": 4.038967010632314e-06,
      "loss": 0.489,
      "num_input_tokens_seen": 20769160,
      "step": 31860
    },
    {
      "epoch": 16.700733752620543,
      "grad_norm": 0.43104565143585205,
      "learning_rate": 4.032737610766826e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 20773384,
      "step": 31865
    },
    {
      "epoch": 16.70335429769392,
      "grad_norm": 0.5477659106254578,
      "learning_rate": 4.026512597013429e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 20777416,
      "step": 31870
    },
    {
      "epoch": 16.705974842767297,
      "grad_norm": 0.6329734325408936,
      "learning_rate": 4.020291970674312e-06,
      "loss": 0.5054,
      "num_input_tokens_seen": 20780808,
      "step": 31875
    },
    {
      "epoch": 16.70859538784067,
      "grad_norm": 0.45595598220825195,
      "learning_rate": 4.0140757330507705e-06,
      "loss": 0.3052,
      "num_input_tokens_seen": 20784040,
      "step": 31880
    },
    {
      "epoch": 16.711215932914047,
      "grad_norm": 0.35527244210243225,
      "learning_rate": 4.007863885443161e-06,
      "loss": 0.4294,
      "num_input_tokens_seen": 20787144,
      "step": 31885
    },
    {
      "epoch": 16.71383647798742,
      "grad_norm": 0.2843407690525055,
      "learning_rate": 4.001656429150935e-06,
      "loss": 0.4233,
      "num_input_tokens_seen": 20790408,
      "step": 31890
    },
    {
      "epoch": 16.716457023060798,
      "grad_norm": 0.501812756061554,
      "learning_rate": 3.9954533654726235e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 20793768,
      "step": 31895
    },
    {
      "epoch": 16.71907756813417,
      "grad_norm": 0.25991126894950867,
      "learning_rate": 3.989254695705841e-06,
      "loss": 0.4004,
      "num_input_tokens_seen": 20799624,
      "step": 31900
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 0.5884345173835754,
      "learning_rate": 3.983060421147272e-06,
      "loss": 0.4784,
      "num_input_tokens_seen": 20802152,
      "step": 31905
    },
    {
      "epoch": 16.72431865828092,
      "grad_norm": 0.4422594904899597,
      "learning_rate": 3.976870543092684e-06,
      "loss": 0.4531,
      "num_input_tokens_seen": 20804776,
      "step": 31910
    },
    {
      "epoch": 16.7269392033543,
      "grad_norm": 0.5799729228019714,
      "learning_rate": 3.970685062836932e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 20807880,
      "step": 31915
    },
    {
      "epoch": 16.729559748427672,
      "grad_norm": 0.5831668972969055,
      "learning_rate": 3.96450398167395e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 20810568,
      "step": 31920
    },
    {
      "epoch": 16.73218029350105,
      "grad_norm": 0.2695061266422272,
      "learning_rate": 3.958327300896758e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 20815368,
      "step": 31925
    },
    {
      "epoch": 16.734800838574422,
      "grad_norm": 0.4588421881198883,
      "learning_rate": 3.952155021797429e-06,
      "loss": 0.5382,
      "num_input_tokens_seen": 20818472,
      "step": 31930
    },
    {
      "epoch": 16.7374213836478,
      "grad_norm": 0.44994938373565674,
      "learning_rate": 3.945987145667151e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 20822120,
      "step": 31935
    },
    {
      "epoch": 16.740041928721173,
      "grad_norm": 0.5474562048912048,
      "learning_rate": 3.939823673796162e-06,
      "loss": 0.4391,
      "num_input_tokens_seen": 20825160,
      "step": 31940
    },
    {
      "epoch": 16.74266247379455,
      "grad_norm": 0.32536837458610535,
      "learning_rate": 3.933664607473789e-06,
      "loss": 0.6182,
      "num_input_tokens_seen": 20830056,
      "step": 31945
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 0.37334588170051575,
      "learning_rate": 3.927509947988445e-06,
      "loss": 0.4895,
      "num_input_tokens_seen": 20833032,
      "step": 31950
    },
    {
      "epoch": 16.7479035639413,
      "grad_norm": 0.33285483717918396,
      "learning_rate": 3.9213596966276164e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 20836200,
      "step": 31955
    },
    {
      "epoch": 16.750524109014677,
      "grad_norm": 0.4449104964733124,
      "learning_rate": 3.9152138546778625e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 20839784,
      "step": 31960
    },
    {
      "epoch": 16.75314465408805,
      "grad_norm": 0.3639489710330963,
      "learning_rate": 3.9090724234248175e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 20842952,
      "step": 31965
    },
    {
      "epoch": 16.755765199161427,
      "grad_norm": 0.3778412342071533,
      "learning_rate": 3.902935404153202e-06,
      "loss": 0.4949,
      "num_input_tokens_seen": 20846344,
      "step": 31970
    },
    {
      "epoch": 16.7583857442348,
      "grad_norm": 0.7759495377540588,
      "learning_rate": 3.896802798146809e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 20849512,
      "step": 31975
    },
    {
      "epoch": 16.761006289308177,
      "grad_norm": 0.7010807394981384,
      "learning_rate": 3.890674606688524e-06,
      "loss": 0.5847,
      "num_input_tokens_seen": 20852744,
      "step": 31980
    },
    {
      "epoch": 16.76362683438155,
      "grad_norm": 0.4894975423812866,
      "learning_rate": 3.88455083106028e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 20856392,
      "step": 31985
    },
    {
      "epoch": 16.766247379454928,
      "grad_norm": 0.7589858174324036,
      "learning_rate": 3.878431472543098e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 20859048,
      "step": 31990
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 0.5071627497673035,
      "learning_rate": 3.872316532417086e-06,
      "loss": 0.4215,
      "num_input_tokens_seen": 20861992,
      "step": 31995
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 0.613289475440979,
      "learning_rate": 3.866206011961418e-06,
      "loss": 0.4738,
      "num_input_tokens_seen": 20864904,
      "step": 32000
    },
    {
      "epoch": 16.77410901467505,
      "grad_norm": 0.4172918498516083,
      "learning_rate": 3.860099912454346e-06,
      "loss": 0.4519,
      "num_input_tokens_seen": 20867976,
      "step": 32005
    },
    {
      "epoch": 16.77672955974843,
      "grad_norm": 0.4674493968486786,
      "learning_rate": 3.853998235173201e-06,
      "loss": 0.4446,
      "num_input_tokens_seen": 20871464,
      "step": 32010
    },
    {
      "epoch": 16.779350104821802,
      "grad_norm": 0.46679946780204773,
      "learning_rate": 3.847900981394379e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 20874984,
      "step": 32015
    },
    {
      "epoch": 16.78197064989518,
      "grad_norm": 0.41592174768447876,
      "learning_rate": 3.841808152393348e-06,
      "loss": 0.3472,
      "num_input_tokens_seen": 20877832,
      "step": 32020
    },
    {
      "epoch": 16.784591194968552,
      "grad_norm": 0.4755140542984009,
      "learning_rate": 3.835719749444666e-06,
      "loss": 0.5736,
      "num_input_tokens_seen": 20881096,
      "step": 32025
    },
    {
      "epoch": 16.78721174004193,
      "grad_norm": 0.4487319588661194,
      "learning_rate": 3.8296357738219606e-06,
      "loss": 0.4736,
      "num_input_tokens_seen": 20884680,
      "step": 32030
    },
    {
      "epoch": 16.789832285115303,
      "grad_norm": 0.5725118517875671,
      "learning_rate": 3.823556226797931e-06,
      "loss": 0.5669,
      "num_input_tokens_seen": 20888008,
      "step": 32035
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 0.3368745446205139,
      "learning_rate": 3.817481109644344e-06,
      "loss": 0.4098,
      "num_input_tokens_seen": 20891240,
      "step": 32040
    },
    {
      "epoch": 16.795073375262053,
      "grad_norm": 0.5217442512512207,
      "learning_rate": 3.811410423632042e-06,
      "loss": 0.3553,
      "num_input_tokens_seen": 20894664,
      "step": 32045
    },
    {
      "epoch": 16.79769392033543,
      "grad_norm": 0.45046383142471313,
      "learning_rate": 3.8053441700309473e-06,
      "loss": 0.4767,
      "num_input_tokens_seen": 20897608,
      "step": 32050
    },
    {
      "epoch": 16.800314465408803,
      "grad_norm": 0.45033958554267883,
      "learning_rate": 3.7992823501100507e-06,
      "loss": 0.4114,
      "num_input_tokens_seen": 20901096,
      "step": 32055
    },
    {
      "epoch": 16.80293501048218,
      "grad_norm": 0.5661202669143677,
      "learning_rate": 3.7932249651374207e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 20904520,
      "step": 32060
    },
    {
      "epoch": 16.805555555555557,
      "grad_norm": 0.4171622395515442,
      "learning_rate": 3.787172016380183e-06,
      "loss": 0.3898,
      "num_input_tokens_seen": 20907592,
      "step": 32065
    },
    {
      "epoch": 16.80817610062893,
      "grad_norm": 0.4307806193828583,
      "learning_rate": 3.7811235051045558e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 20911752,
      "step": 32070
    },
    {
      "epoch": 16.810796645702307,
      "grad_norm": 0.4464666247367859,
      "learning_rate": 3.775079432575809e-06,
      "loss": 0.4519,
      "num_input_tokens_seen": 20914728,
      "step": 32075
    },
    {
      "epoch": 16.81341719077568,
      "grad_norm": 0.6146876811981201,
      "learning_rate": 3.7690398000582964e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 20917704,
      "step": 32080
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 0.5023220181465149,
      "learning_rate": 3.763004608815443e-06,
      "loss": 0.4637,
      "num_input_tokens_seen": 20920808,
      "step": 32085
    },
    {
      "epoch": 16.81865828092243,
      "grad_norm": 0.5065116286277771,
      "learning_rate": 3.756973860109747e-06,
      "loss": 0.4348,
      "num_input_tokens_seen": 20923528,
      "step": 32090
    },
    {
      "epoch": 16.821278825995808,
      "grad_norm": 0.4040040075778961,
      "learning_rate": 3.7509475552027663e-06,
      "loss": 0.4091,
      "num_input_tokens_seen": 20926376,
      "step": 32095
    },
    {
      "epoch": 16.82389937106918,
      "grad_norm": 0.3612760901451111,
      "learning_rate": 3.744925695355131e-06,
      "loss": 0.3817,
      "num_input_tokens_seen": 20929672,
      "step": 32100
    },
    {
      "epoch": 16.82651991614256,
      "grad_norm": 0.5036816000938416,
      "learning_rate": 3.7389082818265476e-06,
      "loss": 0.5128,
      "num_input_tokens_seen": 20932616,
      "step": 32105
    },
    {
      "epoch": 16.829140461215932,
      "grad_norm": 1.1791945695877075,
      "learning_rate": 3.732895315875792e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 20935400,
      "step": 32110
    },
    {
      "epoch": 16.83176100628931,
      "grad_norm": 0.33845219016075134,
      "learning_rate": 3.726886798760715e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 20938824,
      "step": 32115
    },
    {
      "epoch": 16.834381551362682,
      "grad_norm": 0.5261827111244202,
      "learning_rate": 3.720882731738215e-06,
      "loss": 0.5765,
      "num_input_tokens_seen": 20941960,
      "step": 32120
    },
    {
      "epoch": 16.83700209643606,
      "grad_norm": 0.6685577630996704,
      "learning_rate": 3.714883116064291e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 20948712,
      "step": 32125
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 0.4798477292060852,
      "learning_rate": 3.708887952993978e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 20951592,
      "step": 32130
    },
    {
      "epoch": 16.84224318658281,
      "grad_norm": 0.45379406213760376,
      "learning_rate": 3.7028972437814057e-06,
      "loss": 0.53,
      "num_input_tokens_seen": 20955880,
      "step": 32135
    },
    {
      "epoch": 16.844863731656183,
      "grad_norm": 0.4386698007583618,
      "learning_rate": 3.696910989679764e-06,
      "loss": 0.41,
      "num_input_tokens_seen": 20959112,
      "step": 32140
    },
    {
      "epoch": 16.84748427672956,
      "grad_norm": 0.37516868114471436,
      "learning_rate": 3.6909291919413003e-06,
      "loss": 0.3235,
      "num_input_tokens_seen": 20961992,
      "step": 32145
    },
    {
      "epoch": 16.850104821802937,
      "grad_norm": 0.40331509709358215,
      "learning_rate": 3.684951851817342e-06,
      "loss": 0.3462,
      "num_input_tokens_seen": 20964776,
      "step": 32150
    },
    {
      "epoch": 16.85272536687631,
      "grad_norm": 0.4615892767906189,
      "learning_rate": 3.678978970558286e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 20967944,
      "step": 32155
    },
    {
      "epoch": 16.855345911949687,
      "grad_norm": 0.42440760135650635,
      "learning_rate": 3.673010549413583e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 20971976,
      "step": 32160
    },
    {
      "epoch": 16.85796645702306,
      "grad_norm": 0.3796471059322357,
      "learning_rate": 3.6670465896317592e-06,
      "loss": 0.3978,
      "num_input_tokens_seen": 20975208,
      "step": 32165
    },
    {
      "epoch": 16.860587002096437,
      "grad_norm": 0.8946788311004639,
      "learning_rate": 3.661087092460419e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 20977736,
      "step": 32170
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 0.4529218077659607,
      "learning_rate": 3.655132059146202e-06,
      "loss": 0.4376,
      "num_input_tokens_seen": 20981704,
      "step": 32175
    },
    {
      "epoch": 16.865828092243188,
      "grad_norm": 0.41341274976730347,
      "learning_rate": 3.6491814909348526e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 20985992,
      "step": 32180
    },
    {
      "epoch": 16.86844863731656,
      "grad_norm": 0.49700212478637695,
      "learning_rate": 3.6432353890711424e-06,
      "loss": 0.5155,
      "num_input_tokens_seen": 20989544,
      "step": 32185
    },
    {
      "epoch": 16.871069182389938,
      "grad_norm": 0.7067729830741882,
      "learning_rate": 3.6372937547989425e-06,
      "loss": 0.4878,
      "num_input_tokens_seen": 20996392,
      "step": 32190
    },
    {
      "epoch": 16.87368972746331,
      "grad_norm": 0.5116971731185913,
      "learning_rate": 3.631356589361173e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 20999400,
      "step": 32195
    },
    {
      "epoch": 16.87631027253669,
      "grad_norm": 0.4202554225921631,
      "learning_rate": 3.625423893999813e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 21001704,
      "step": 32200
    },
    {
      "epoch": 16.878930817610062,
      "grad_norm": 0.5570971369743347,
      "learning_rate": 3.6194956699559217e-06,
      "loss": 0.4887,
      "num_input_tokens_seen": 21004872,
      "step": 32205
    },
    {
      "epoch": 16.88155136268344,
      "grad_norm": 0.417581707239151,
      "learning_rate": 3.6135719184696188e-06,
      "loss": 0.499,
      "num_input_tokens_seen": 21011656,
      "step": 32210
    },
    {
      "epoch": 16.884171907756812,
      "grad_norm": 0.44718053936958313,
      "learning_rate": 3.6076526407800766e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 21014472,
      "step": 32215
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 0.9049344658851624,
      "learning_rate": 3.601737838125552e-06,
      "loss": 0.3995,
      "num_input_tokens_seen": 21017960,
      "step": 32220
    },
    {
      "epoch": 16.889412997903563,
      "grad_norm": 0.4514579772949219,
      "learning_rate": 3.595827511743341e-06,
      "loss": 0.4613,
      "num_input_tokens_seen": 21021288,
      "step": 32225
    },
    {
      "epoch": 16.89203354297694,
      "grad_norm": 0.4058665931224823,
      "learning_rate": 3.5899216628698217e-06,
      "loss": 0.4624,
      "num_input_tokens_seen": 21024872,
      "step": 32230
    },
    {
      "epoch": 16.894654088050313,
      "grad_norm": 0.7809538245201111,
      "learning_rate": 3.584020292740431e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 21028008,
      "step": 32235
    },
    {
      "epoch": 16.89727463312369,
      "grad_norm": 0.3853861391544342,
      "learning_rate": 3.578123402589675e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 21030920,
      "step": 32240
    },
    {
      "epoch": 16.899895178197063,
      "grad_norm": 0.3509085178375244,
      "learning_rate": 3.5722309936511044e-06,
      "loss": 0.5301,
      "num_input_tokens_seen": 21034248,
      "step": 32245
    },
    {
      "epoch": 16.90251572327044,
      "grad_norm": 0.45182785391807556,
      "learning_rate": 3.5663430671573572e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 21037672,
      "step": 32250
    },
    {
      "epoch": 16.905136268343817,
      "grad_norm": 0.6054587364196777,
      "learning_rate": 3.5604596243401038e-06,
      "loss": 0.481,
      "num_input_tokens_seen": 21041192,
      "step": 32255
    },
    {
      "epoch": 16.90775681341719,
      "grad_norm": 0.4382011294364929,
      "learning_rate": 3.554580666430102e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 21044488,
      "step": 32260
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 0.6514379382133484,
      "learning_rate": 3.5487061946571614e-06,
      "loss": 0.5178,
      "num_input_tokens_seen": 21047752,
      "step": 32265
    },
    {
      "epoch": 16.91299790356394,
      "grad_norm": 0.2612212002277374,
      "learning_rate": 3.542836210250161e-06,
      "loss": 0.4007,
      "num_input_tokens_seen": 21050344,
      "step": 32270
    },
    {
      "epoch": 16.915618448637318,
      "grad_norm": 0.39611002802848816,
      "learning_rate": 3.536970714437032e-06,
      "loss": 0.4765,
      "num_input_tokens_seen": 21053288,
      "step": 32275
    },
    {
      "epoch": 16.91823899371069,
      "grad_norm": 0.6931784152984619,
      "learning_rate": 3.5311097084447575e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 21056904,
      "step": 32280
    },
    {
      "epoch": 16.920859538784068,
      "grad_norm": 0.41538938879966736,
      "learning_rate": 3.5252531934994033e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 21059848,
      "step": 32285
    },
    {
      "epoch": 16.92348008385744,
      "grad_norm": 0.4285498857498169,
      "learning_rate": 3.5194011708260866e-06,
      "loss": 0.4116,
      "num_input_tokens_seen": 21062472,
      "step": 32290
    },
    {
      "epoch": 16.92610062893082,
      "grad_norm": 0.14703959226608276,
      "learning_rate": 3.513553641648984e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 21067752,
      "step": 32295
    },
    {
      "epoch": 16.928721174004192,
      "grad_norm": 1.4269434213638306,
      "learning_rate": 3.5077106071913307e-06,
      "loss": 0.5116,
      "num_input_tokens_seen": 21070632,
      "step": 32300
    },
    {
      "epoch": 16.93134171907757,
      "grad_norm": 0.4254535138607025,
      "learning_rate": 3.5018720686754133e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 21074568,
      "step": 32305
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 0.35577747225761414,
      "learning_rate": 3.4960380273225985e-06,
      "loss": 0.3854,
      "num_input_tokens_seen": 21077288,
      "step": 32310
    },
    {
      "epoch": 16.93658280922432,
      "grad_norm": 0.41759803891181946,
      "learning_rate": 3.490208484353297e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 21079912,
      "step": 32315
    },
    {
      "epoch": 16.939203354297693,
      "grad_norm": 1.1205804347991943,
      "learning_rate": 3.4843834409869824e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 21084040,
      "step": 32320
    },
    {
      "epoch": 16.94182389937107,
      "grad_norm": 0.5834856629371643,
      "learning_rate": 3.4785628984421977e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 21087528,
      "step": 32325
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.5348899960517883,
      "learning_rate": 3.4727468579365235e-06,
      "loss": 0.3953,
      "num_input_tokens_seen": 21090888,
      "step": 32330
    },
    {
      "epoch": 16.94706498951782,
      "grad_norm": 0.5834195017814636,
      "learning_rate": 3.466935320686604e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 21094792,
      "step": 32335
    },
    {
      "epoch": 16.949685534591197,
      "grad_norm": 0.5467596054077148,
      "learning_rate": 3.4611282879081553e-06,
      "loss": 0.401,
      "num_input_tokens_seen": 21097736,
      "step": 32340
    },
    {
      "epoch": 16.95230607966457,
      "grad_norm": 0.42953354120254517,
      "learning_rate": 3.455325760815939e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 21100776,
      "step": 32345
    },
    {
      "epoch": 16.954926624737947,
      "grad_norm": 0.4702867567539215,
      "learning_rate": 3.4495277406237843e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 21104040,
      "step": 32350
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 0.45788732171058655,
      "learning_rate": 3.443734228544565e-06,
      "loss": 0.3671,
      "num_input_tokens_seen": 21106856,
      "step": 32355
    },
    {
      "epoch": 16.960167714884697,
      "grad_norm": 1.0676242113113403,
      "learning_rate": 3.437945225790215e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 21109768,
      "step": 32360
    },
    {
      "epoch": 16.96278825995807,
      "grad_norm": 0.4152146577835083,
      "learning_rate": 3.432160733571729e-06,
      "loss": 0.4284,
      "num_input_tokens_seen": 21113512,
      "step": 32365
    },
    {
      "epoch": 16.965408805031448,
      "grad_norm": 0.607083797454834,
      "learning_rate": 3.4263807530991595e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 21117544,
      "step": 32370
    },
    {
      "epoch": 16.96802935010482,
      "grad_norm": 0.3475773334503174,
      "learning_rate": 3.4206052855816155e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 21121864,
      "step": 32375
    },
    {
      "epoch": 16.970649895178198,
      "grad_norm": 0.4714374840259552,
      "learning_rate": 3.4148343322272567e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 21124328,
      "step": 32380
    },
    {
      "epoch": 16.97327044025157,
      "grad_norm": 0.42447879910469055,
      "learning_rate": 3.4090678942432923e-06,
      "loss": 0.3397,
      "num_input_tokens_seen": 21127336,
      "step": 32385
    },
    {
      "epoch": 16.97589098532495,
      "grad_norm": 0.4055270850658417,
      "learning_rate": 3.4033059728360018e-06,
      "loss": 0.3453,
      "num_input_tokens_seen": 21129896,
      "step": 32390
    },
    {
      "epoch": 16.978511530398322,
      "grad_norm": 0.7546952366828918,
      "learning_rate": 3.397548569210715e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 21133512,
      "step": 32395
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 0.4056323170661926,
      "learning_rate": 3.3917956845718136e-06,
      "loss": 0.4234,
      "num_input_tokens_seen": 21135816,
      "step": 32400
    },
    {
      "epoch": 16.983752620545072,
      "grad_norm": 0.4176761507987976,
      "learning_rate": 3.38604732012274e-06,
      "loss": 0.3108,
      "num_input_tokens_seen": 21138536,
      "step": 32405
    },
    {
      "epoch": 16.98637316561845,
      "grad_norm": 0.5285802483558655,
      "learning_rate": 3.3803034770659825e-06,
      "loss": 0.4631,
      "num_input_tokens_seen": 21141160,
      "step": 32410
    },
    {
      "epoch": 16.988993710691823,
      "grad_norm": 0.48148828744888306,
      "learning_rate": 3.3745641566030826e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 21143624,
      "step": 32415
    },
    {
      "epoch": 16.9916142557652,
      "grad_norm": 0.4011727571487427,
      "learning_rate": 3.3688293599346442e-06,
      "loss": 0.5013,
      "num_input_tokens_seen": 21147528,
      "step": 32420
    },
    {
      "epoch": 16.994234800838573,
      "grad_norm": 0.3499581515789032,
      "learning_rate": 3.3630990882603213e-06,
      "loss": 0.4257,
      "num_input_tokens_seen": 21150376,
      "step": 32425
    },
    {
      "epoch": 16.99685534591195,
      "grad_norm": 0.7231867909431458,
      "learning_rate": 3.35737334277883e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 21155016,
      "step": 32430
    },
    {
      "epoch": 16.999475890985323,
      "grad_norm": 0.36498382687568665,
      "learning_rate": 3.351652124687918e-06,
      "loss": 0.2781,
      "num_input_tokens_seen": 21157992,
      "step": 32435
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.45275527238845825,
      "eval_runtime": 9.9021,
      "eval_samples_per_second": 85.638,
      "eval_steps_per_second": 21.41,
      "num_input_tokens_seen": 21158120,
      "step": 32436
    },
    {
      "epoch": 17.0020964360587,
      "grad_norm": 0.2743113040924072,
      "learning_rate": 3.345935435184408e-06,
      "loss": 0.3188,
      "num_input_tokens_seen": 21160712,
      "step": 32440
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 0.4948602616786957,
      "learning_rate": 3.3402232754641576e-06,
      "loss": 0.4207,
      "num_input_tokens_seen": 21163528,
      "step": 32445
    },
    {
      "epoch": 17.00733752620545,
      "grad_norm": 0.6249662637710571,
      "learning_rate": 3.334515646722089e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 21167304,
      "step": 32450
    },
    {
      "epoch": 17.009958071278827,
      "grad_norm": 0.48653918504714966,
      "learning_rate": 3.32881255015218e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 21170792,
      "step": 32455
    },
    {
      "epoch": 17.0125786163522,
      "grad_norm": 0.4728476107120514,
      "learning_rate": 3.3231139869474433e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 21174408,
      "step": 32460
    },
    {
      "epoch": 17.015199161425578,
      "grad_norm": 0.31199902296066284,
      "learning_rate": 3.317419958299961e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 21178248,
      "step": 32465
    },
    {
      "epoch": 17.01781970649895,
      "grad_norm": 0.5971257090568542,
      "learning_rate": 3.31173046540085e-06,
      "loss": 0.3225,
      "num_input_tokens_seen": 21181064,
      "step": 32470
    },
    {
      "epoch": 17.020440251572328,
      "grad_norm": 0.4594717025756836,
      "learning_rate": 3.306045509440295e-06,
      "loss": 0.4854,
      "num_input_tokens_seen": 21184296,
      "step": 32475
    },
    {
      "epoch": 17.0230607966457,
      "grad_norm": 1.3904085159301758,
      "learning_rate": 3.300365091607521e-06,
      "loss": 0.5244,
      "num_input_tokens_seen": 21187176,
      "step": 32480
    },
    {
      "epoch": 17.02568134171908,
      "grad_norm": 1.0121864080429077,
      "learning_rate": 3.2946892130908127e-06,
      "loss": 0.3604,
      "num_input_tokens_seen": 21189800,
      "step": 32485
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 0.4558420777320862,
      "learning_rate": 3.289017875077488e-06,
      "loss": 0.5177,
      "num_input_tokens_seen": 21196328,
      "step": 32490
    },
    {
      "epoch": 17.03092243186583,
      "grad_norm": 0.34336891770362854,
      "learning_rate": 3.283351078753938e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 21199400,
      "step": 32495
    },
    {
      "epoch": 17.033542976939202,
      "grad_norm": 0.28304871916770935,
      "learning_rate": 3.2776888253055777e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 21202248,
      "step": 32500
    },
    {
      "epoch": 17.03616352201258,
      "grad_norm": 0.7011508941650391,
      "learning_rate": 3.272031115916896e-06,
      "loss": 0.5229,
      "num_input_tokens_seen": 21205448,
      "step": 32505
    },
    {
      "epoch": 17.038784067085953,
      "grad_norm": 0.33498746156692505,
      "learning_rate": 3.2663779517714238e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 21209576,
      "step": 32510
    },
    {
      "epoch": 17.04140461215933,
      "grad_norm": 0.532198965549469,
      "learning_rate": 3.260729334051729e-06,
      "loss": 0.4894,
      "num_input_tokens_seen": 21212904,
      "step": 32515
    },
    {
      "epoch": 17.044025157232703,
      "grad_norm": 0.5620784759521484,
      "learning_rate": 3.2550852639394446e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 21216104,
      "step": 32520
    },
    {
      "epoch": 17.04664570230608,
      "grad_norm": 0.5757774710655212,
      "learning_rate": 3.2494457426152452e-06,
      "loss": 0.5036,
      "num_input_tokens_seen": 21220104,
      "step": 32525
    },
    {
      "epoch": 17.049266247379453,
      "grad_norm": 0.7357773184776306,
      "learning_rate": 3.243810771258851e-06,
      "loss": 0.511,
      "num_input_tokens_seen": 21222920,
      "step": 32530
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 0.4117731750011444,
      "learning_rate": 3.2381803510490387e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 21226408,
      "step": 32535
    },
    {
      "epoch": 17.054507337526207,
      "grad_norm": 0.7339961528778076,
      "learning_rate": 3.2325544831636217e-06,
      "loss": 0.5314,
      "num_input_tokens_seen": 21230184,
      "step": 32540
    },
    {
      "epoch": 17.05712788259958,
      "grad_norm": 0.720086395740509,
      "learning_rate": 3.22693316877947e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 21233288,
      "step": 32545
    },
    {
      "epoch": 17.059748427672957,
      "grad_norm": 0.5826966762542725,
      "learning_rate": 3.2213164090725057e-06,
      "loss": 0.4365,
      "num_input_tokens_seen": 21235720,
      "step": 32550
    },
    {
      "epoch": 17.06236897274633,
      "grad_norm": 0.42083120346069336,
      "learning_rate": 3.2157042052176763e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 21238408,
      "step": 32555
    },
    {
      "epoch": 17.064989517819708,
      "grad_norm": 0.3168710470199585,
      "learning_rate": 3.2100965583890003e-06,
      "loss": 0.3868,
      "num_input_tokens_seen": 21242632,
      "step": 32560
    },
    {
      "epoch": 17.06761006289308,
      "grad_norm": 0.3123473823070526,
      "learning_rate": 3.204493469759537e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 21246056,
      "step": 32565
    },
    {
      "epoch": 17.070230607966458,
      "grad_norm": 0.9582141041755676,
      "learning_rate": 3.1988949405013764e-06,
      "loss": 0.5438,
      "num_input_tokens_seen": 21249032,
      "step": 32570
    },
    {
      "epoch": 17.07285115303983,
      "grad_norm": 0.6135792136192322,
      "learning_rate": 3.193300971785676e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 21251816,
      "step": 32575
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 0.33285951614379883,
      "learning_rate": 3.1877115647826335e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 21256200,
      "step": 32580
    },
    {
      "epoch": 17.078092243186582,
      "grad_norm": 0.5944743752479553,
      "learning_rate": 3.1821267206614797e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 21260392,
      "step": 32585
    },
    {
      "epoch": 17.08071278825996,
      "grad_norm": 0.377162367105484,
      "learning_rate": 3.176546440590508e-06,
      "loss": 0.4743,
      "num_input_tokens_seen": 21264296,
      "step": 32590
    },
    {
      "epoch": 17.083333333333332,
      "grad_norm": 0.36711686849594116,
      "learning_rate": 3.170970725737041e-06,
      "loss": 0.4771,
      "num_input_tokens_seen": 21267848,
      "step": 32595
    },
    {
      "epoch": 17.08595387840671,
      "grad_norm": 0.4375876188278198,
      "learning_rate": 3.165399577267461e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 21271560,
      "step": 32600
    },
    {
      "epoch": 17.088574423480082,
      "grad_norm": 0.8662404417991638,
      "learning_rate": 3.159832996347184e-06,
      "loss": 0.4634,
      "num_input_tokens_seen": 21274760,
      "step": 32605
    },
    {
      "epoch": 17.09119496855346,
      "grad_norm": 0.488116592168808,
      "learning_rate": 3.1542709841406865e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 21277352,
      "step": 32610
    },
    {
      "epoch": 17.093815513626833,
      "grad_norm": 0.7430114150047302,
      "learning_rate": 3.148713541811468e-06,
      "loss": 0.5955,
      "num_input_tokens_seen": 21280168,
      "step": 32615
    },
    {
      "epoch": 17.09643605870021,
      "grad_norm": 0.5133503675460815,
      "learning_rate": 3.1431606705220803e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 21283688,
      "step": 32620
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 0.7039333581924438,
      "learning_rate": 3.1376123714341234e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 21286920,
      "step": 32625
    },
    {
      "epoch": 17.10167714884696,
      "grad_norm": 0.5730239152908325,
      "learning_rate": 3.1320686457082387e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 21290600,
      "step": 32630
    },
    {
      "epoch": 17.104297693920337,
      "grad_norm": 0.49318066239356995,
      "learning_rate": 3.126529494504113e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 21293032,
      "step": 32635
    },
    {
      "epoch": 17.10691823899371,
      "grad_norm": 0.5917571783065796,
      "learning_rate": 3.1209949189804744e-06,
      "loss": 0.472,
      "num_input_tokens_seen": 21295976,
      "step": 32640
    },
    {
      "epoch": 17.109538784067087,
      "grad_norm": 0.2997184097766876,
      "learning_rate": 3.115464920295089e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 21299816,
      "step": 32645
    },
    {
      "epoch": 17.11215932914046,
      "grad_norm": 0.759816586971283,
      "learning_rate": 3.1099394996047666e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 21302696,
      "step": 32650
    },
    {
      "epoch": 17.114779874213838,
      "grad_norm": 0.5111802816390991,
      "learning_rate": 3.104418658065367e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 21305288,
      "step": 32655
    },
    {
      "epoch": 17.11740041928721,
      "grad_norm": 0.8056383728981018,
      "learning_rate": 3.098902396831782e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 21308808,
      "step": 32660
    },
    {
      "epoch": 17.120020964360588,
      "grad_norm": 0.6235367655754089,
      "learning_rate": 3.0933907170579646e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 21312008,
      "step": 32665
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 0.7032302618026733,
      "learning_rate": 3.0878836198968835e-06,
      "loss": 0.4076,
      "num_input_tokens_seen": 21314280,
      "step": 32670
    },
    {
      "epoch": 17.12526205450734,
      "grad_norm": 0.6720888614654541,
      "learning_rate": 3.0823811065005564e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 21316808,
      "step": 32675
    },
    {
      "epoch": 17.127882599580712,
      "grad_norm": 0.7164550423622131,
      "learning_rate": 3.0768831780200542e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 21320712,
      "step": 32680
    },
    {
      "epoch": 17.13050314465409,
      "grad_norm": 0.5511178970336914,
      "learning_rate": 3.071389835605476e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 21324968,
      "step": 32685
    },
    {
      "epoch": 17.133123689727462,
      "grad_norm": 0.49193209409713745,
      "learning_rate": 3.065901080405978e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 21328360,
      "step": 32690
    },
    {
      "epoch": 17.13574423480084,
      "grad_norm": 0.6899136304855347,
      "learning_rate": 3.060416913569733e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 21331336,
      "step": 32695
    },
    {
      "epoch": 17.138364779874212,
      "grad_norm": 0.4145905673503876,
      "learning_rate": 3.054937336243968e-06,
      "loss": 0.5662,
      "num_input_tokens_seen": 21334344,
      "step": 32700
    },
    {
      "epoch": 17.14098532494759,
      "grad_norm": 0.32795071601867676,
      "learning_rate": 3.049462349574947e-06,
      "loss": 0.4778,
      "num_input_tokens_seen": 21340360,
      "step": 32705
    },
    {
      "epoch": 17.143605870020963,
      "grad_norm": 0.7270508408546448,
      "learning_rate": 3.043991954707981e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 21343176,
      "step": 32710
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 0.4254271388053894,
      "learning_rate": 3.0385261527874075e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 21346024,
      "step": 32715
    },
    {
      "epoch": 17.148846960167713,
      "grad_norm": 0.29187485575675964,
      "learning_rate": 3.0330649449566205e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 21350120,
      "step": 32720
    },
    {
      "epoch": 17.15146750524109,
      "grad_norm": 0.35437414050102234,
      "learning_rate": 3.027608332358034e-06,
      "loss": 0.5574,
      "num_input_tokens_seen": 21353608,
      "step": 32725
    },
    {
      "epoch": 17.154088050314467,
      "grad_norm": 0.4407079517841339,
      "learning_rate": 3.022156316133104e-06,
      "loss": 0.4208,
      "num_input_tokens_seen": 21358376,
      "step": 32730
    },
    {
      "epoch": 17.15670859538784,
      "grad_norm": 0.6637725234031677,
      "learning_rate": 3.016708897422335e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 21361128,
      "step": 32735
    },
    {
      "epoch": 17.159329140461217,
      "grad_norm": 0.6820401549339294,
      "learning_rate": 3.0112660773652685e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 21364584,
      "step": 32740
    },
    {
      "epoch": 17.16194968553459,
      "grad_norm": 0.3958576023578644,
      "learning_rate": 3.005827857100482e-06,
      "loss": 0.4096,
      "num_input_tokens_seen": 21368168,
      "step": 32745
    },
    {
      "epoch": 17.164570230607968,
      "grad_norm": 0.2658218443393707,
      "learning_rate": 3.000394237765586e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 21371752,
      "step": 32750
    },
    {
      "epoch": 17.16719077568134,
      "grad_norm": 0.5135958194732666,
      "learning_rate": 2.9949652204972254e-06,
      "loss": 0.5325,
      "num_input_tokens_seen": 21374632,
      "step": 32755
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 0.32554763555526733,
      "learning_rate": 2.989540806431093e-06,
      "loss": 0.4472,
      "num_input_tokens_seen": 21377608,
      "step": 32760
    },
    {
      "epoch": 17.17243186582809,
      "grad_norm": 0.4280675947666168,
      "learning_rate": 2.984120996701917e-06,
      "loss": 0.508,
      "num_input_tokens_seen": 21381192,
      "step": 32765
    },
    {
      "epoch": 17.17505241090147,
      "grad_norm": 0.6888055801391602,
      "learning_rate": 2.9787057924434618e-06,
      "loss": 0.6135,
      "num_input_tokens_seen": 21384168,
      "step": 32770
    },
    {
      "epoch": 17.177672955974842,
      "grad_norm": 0.7403122186660767,
      "learning_rate": 2.9732951947885166e-06,
      "loss": 0.4382,
      "num_input_tokens_seen": 21387240,
      "step": 32775
    },
    {
      "epoch": 17.18029350104822,
      "grad_norm": 0.3584553599357605,
      "learning_rate": 2.967889204868929e-06,
      "loss": 0.374,
      "num_input_tokens_seen": 21390216,
      "step": 32780
    },
    {
      "epoch": 17.182914046121592,
      "grad_norm": 0.6181259155273438,
      "learning_rate": 2.96248782381556e-06,
      "loss": 0.4548,
      "num_input_tokens_seen": 21393544,
      "step": 32785
    },
    {
      "epoch": 17.18553459119497,
      "grad_norm": 0.45591285824775696,
      "learning_rate": 2.957091052758321e-06,
      "loss": 0.4602,
      "num_input_tokens_seen": 21397288,
      "step": 32790
    },
    {
      "epoch": 17.188155136268342,
      "grad_norm": 0.3354123532772064,
      "learning_rate": 2.9516988928261517e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 21399912,
      "step": 32795
    },
    {
      "epoch": 17.19077568134172,
      "grad_norm": 0.5829130411148071,
      "learning_rate": 2.9463113451470403e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 21402472,
      "step": 32800
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 0.3594891428947449,
      "learning_rate": 2.94092841084799e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 21405352,
      "step": 32805
    },
    {
      "epoch": 17.19601677148847,
      "grad_norm": 0.43738794326782227,
      "learning_rate": 2.9355500910550586e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 21408008,
      "step": 32810
    },
    {
      "epoch": 17.198637316561843,
      "grad_norm": 0.597459614276886,
      "learning_rate": 2.9301763868933157e-06,
      "loss": 0.5409,
      "num_input_tokens_seen": 21411880,
      "step": 32815
    },
    {
      "epoch": 17.20125786163522,
      "grad_norm": 0.7161991596221924,
      "learning_rate": 2.9248072994868896e-06,
      "loss": 0.4326,
      "num_input_tokens_seen": 21414504,
      "step": 32820
    },
    {
      "epoch": 17.203878406708597,
      "grad_norm": 0.7491389513015747,
      "learning_rate": 2.919442829958932e-06,
      "loss": 0.4062,
      "num_input_tokens_seen": 21417608,
      "step": 32825
    },
    {
      "epoch": 17.20649895178197,
      "grad_norm": 0.5131235718727112,
      "learning_rate": 2.9140829794316234e-06,
      "loss": 0.4899,
      "num_input_tokens_seen": 21420712,
      "step": 32830
    },
    {
      "epoch": 17.209119496855347,
      "grad_norm": 0.4975450932979584,
      "learning_rate": 2.908727749026191e-06,
      "loss": 0.63,
      "num_input_tokens_seen": 21423400,
      "step": 32835
    },
    {
      "epoch": 17.21174004192872,
      "grad_norm": 0.7881034016609192,
      "learning_rate": 2.9033771398628756e-06,
      "loss": 0.5082,
      "num_input_tokens_seen": 21426888,
      "step": 32840
    },
    {
      "epoch": 17.214360587002098,
      "grad_norm": 0.4533248245716095,
      "learning_rate": 2.898031153060973e-06,
      "loss": 0.4687,
      "num_input_tokens_seen": 21429736,
      "step": 32845
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 0.5802170634269714,
      "learning_rate": 2.8926897897388033e-06,
      "loss": 0.4491,
      "num_input_tokens_seen": 21433320,
      "step": 32850
    },
    {
      "epoch": 17.219601677148848,
      "grad_norm": 0.7219507694244385,
      "learning_rate": 2.8873530510137136e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 21436584,
      "step": 32855
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.37790578603744507,
      "learning_rate": 2.8820209380020892e-06,
      "loss": 0.5546,
      "num_input_tokens_seen": 21439528,
      "step": 32860
    },
    {
      "epoch": 17.2248427672956,
      "grad_norm": 0.6222763657569885,
      "learning_rate": 2.8766934518193566e-06,
      "loss": 0.3986,
      "num_input_tokens_seen": 21442824,
      "step": 32865
    },
    {
      "epoch": 17.22746331236897,
      "grad_norm": 0.67730313539505,
      "learning_rate": 2.87137059357995e-06,
      "loss": 0.497,
      "num_input_tokens_seen": 21446024,
      "step": 32870
    },
    {
      "epoch": 17.23008385744235,
      "grad_norm": 0.44388118386268616,
      "learning_rate": 2.8660523643973618e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 21448744,
      "step": 32875
    },
    {
      "epoch": 17.232704402515722,
      "grad_norm": 0.5815220475196838,
      "learning_rate": 2.860738765384105e-06,
      "loss": 0.4559,
      "num_input_tokens_seen": 21451688,
      "step": 32880
    },
    {
      "epoch": 17.2353249475891,
      "grad_norm": 0.5561701655387878,
      "learning_rate": 2.8554297976517154e-06,
      "loss": 0.4445,
      "num_input_tokens_seen": 21453864,
      "step": 32885
    },
    {
      "epoch": 17.237945492662472,
      "grad_norm": 0.5648148655891418,
      "learning_rate": 2.8501254623107707e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 21456936,
      "step": 32890
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 0.7112119197845459,
      "learning_rate": 2.84482576047089e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 21460424,
      "step": 32895
    },
    {
      "epoch": 17.243186582809223,
      "grad_norm": 0.5020313858985901,
      "learning_rate": 2.839530693240694e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 21463016,
      "step": 32900
    },
    {
      "epoch": 17.2458071278826,
      "grad_norm": 0.30597546696662903,
      "learning_rate": 2.8342402617278606e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 21466632,
      "step": 32905
    },
    {
      "epoch": 17.248427672955973,
      "grad_norm": 0.4673578441143036,
      "learning_rate": 2.82895446703908e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 21469544,
      "step": 32910
    },
    {
      "epoch": 17.25104821802935,
      "grad_norm": 0.5540447235107422,
      "learning_rate": 2.823673310280084e-06,
      "loss": 0.3714,
      "num_input_tokens_seen": 21472040,
      "step": 32915
    },
    {
      "epoch": 17.253668763102727,
      "grad_norm": 0.4095008969306946,
      "learning_rate": 2.8183967925556305e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 21475784,
      "step": 32920
    },
    {
      "epoch": 17.2562893081761,
      "grad_norm": 0.2844364643096924,
      "learning_rate": 2.8131249149695115e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 21479656,
      "step": 32925
    },
    {
      "epoch": 17.258909853249477,
      "grad_norm": 0.4115624725818634,
      "learning_rate": 2.8078576786245397e-06,
      "loss": 0.3943,
      "num_input_tokens_seen": 21482824,
      "step": 32930
    },
    {
      "epoch": 17.26153039832285,
      "grad_norm": 0.7465366125106812,
      "learning_rate": 2.8025950846225557e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 21485672,
      "step": 32935
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 0.6110915541648865,
      "learning_rate": 2.7973371340644373e-06,
      "loss": 0.4006,
      "num_input_tokens_seen": 21488616,
      "step": 32940
    },
    {
      "epoch": 17.2667714884696,
      "grad_norm": 0.4325870871543884,
      "learning_rate": 2.7920838280500907e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 21490952,
      "step": 32945
    },
    {
      "epoch": 17.269392033542978,
      "grad_norm": 0.39020660519599915,
      "learning_rate": 2.7868351676784504e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 21493832,
      "step": 32950
    },
    {
      "epoch": 17.27201257861635,
      "grad_norm": 0.7295225262641907,
      "learning_rate": 2.7815911540474665e-06,
      "loss": 0.4197,
      "num_input_tokens_seen": 21496136,
      "step": 32955
    },
    {
      "epoch": 17.27463312368973,
      "grad_norm": 0.3983314633369446,
      "learning_rate": 2.7763517882541364e-06,
      "loss": 0.4685,
      "num_input_tokens_seen": 21499432,
      "step": 32960
    },
    {
      "epoch": 17.2772536687631,
      "grad_norm": 0.2980658710002899,
      "learning_rate": 2.771117071394466e-06,
      "loss": 0.5727,
      "num_input_tokens_seen": 21502472,
      "step": 32965
    },
    {
      "epoch": 17.27987421383648,
      "grad_norm": 0.3359861671924591,
      "learning_rate": 2.7658870045635055e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 21506088,
      "step": 32970
    },
    {
      "epoch": 17.282494758909852,
      "grad_norm": 0.3953231871128082,
      "learning_rate": 2.7606615888553234e-06,
      "loss": 0.4023,
      "num_input_tokens_seen": 21508904,
      "step": 32975
    },
    {
      "epoch": 17.28511530398323,
      "grad_norm": 0.32274243235588074,
      "learning_rate": 2.755440825363023e-06,
      "loss": 0.3741,
      "num_input_tokens_seen": 21511624,
      "step": 32980
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 0.3642297387123108,
      "learning_rate": 2.75022471517872e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 21514504,
      "step": 32985
    },
    {
      "epoch": 17.29035639412998,
      "grad_norm": 0.6494759917259216,
      "learning_rate": 2.745013259393564e-06,
      "loss": 0.5368,
      "num_input_tokens_seen": 21517256,
      "step": 32990
    },
    {
      "epoch": 17.292976939203353,
      "grad_norm": 0.36520662903785706,
      "learning_rate": 2.739806459097735e-06,
      "loss": 0.4041,
      "num_input_tokens_seen": 21520648,
      "step": 32995
    },
    {
      "epoch": 17.29559748427673,
      "grad_norm": 0.396038681268692,
      "learning_rate": 2.7346043153804403e-06,
      "loss": 0.3952,
      "num_input_tokens_seen": 21523688,
      "step": 33000
    },
    {
      "epoch": 17.298218029350103,
      "grad_norm": 0.432074636220932,
      "learning_rate": 2.729406829329903e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 21527016,
      "step": 33005
    },
    {
      "epoch": 17.30083857442348,
      "grad_norm": 0.4663503170013428,
      "learning_rate": 2.72421400203339e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 21530408,
      "step": 33010
    },
    {
      "epoch": 17.303459119496857,
      "grad_norm": 0.4450157880783081,
      "learning_rate": 2.719025834577163e-06,
      "loss": 0.5145,
      "num_input_tokens_seen": 21533288,
      "step": 33015
    },
    {
      "epoch": 17.30607966457023,
      "grad_norm": 0.6010509133338928,
      "learning_rate": 2.713842328046537e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 21536520,
      "step": 33020
    },
    {
      "epoch": 17.308700209643607,
      "grad_norm": 0.32745468616485596,
      "learning_rate": 2.7086634835258416e-06,
      "loss": 0.4958,
      "num_input_tokens_seen": 21539304,
      "step": 33025
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 0.7341368794441223,
      "learning_rate": 2.7034893020984333e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 21542472,
      "step": 33030
    },
    {
      "epoch": 17.313941299790358,
      "grad_norm": 0.39335402846336365,
      "learning_rate": 2.6983197848466952e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 21545736,
      "step": 33035
    },
    {
      "epoch": 17.31656184486373,
      "grad_norm": 0.49384039640426636,
      "learning_rate": 2.6931549328520265e-06,
      "loss": 0.5396,
      "num_input_tokens_seen": 21548744,
      "step": 33040
    },
    {
      "epoch": 17.319182389937108,
      "grad_norm": 0.5735076069831848,
      "learning_rate": 2.6879947471948523e-06,
      "loss": 0.5604,
      "num_input_tokens_seen": 21551592,
      "step": 33045
    },
    {
      "epoch": 17.32180293501048,
      "grad_norm": 0.49290478229522705,
      "learning_rate": 2.6828392289546285e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 21554216,
      "step": 33050
    },
    {
      "epoch": 17.32442348008386,
      "grad_norm": 0.4255881607532501,
      "learning_rate": 2.677688379209828e-06,
      "loss": 0.3224,
      "num_input_tokens_seen": 21556776,
      "step": 33055
    },
    {
      "epoch": 17.32704402515723,
      "grad_norm": 0.296811580657959,
      "learning_rate": 2.6725421990379586e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 21559976,
      "step": 33060
    },
    {
      "epoch": 17.32966457023061,
      "grad_norm": 0.43507546186447144,
      "learning_rate": 2.667400689515537e-06,
      "loss": 0.451,
      "num_input_tokens_seen": 21562792,
      "step": 33065
    },
    {
      "epoch": 17.332285115303982,
      "grad_norm": 0.4299238920211792,
      "learning_rate": 2.6622638517181e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 21565736,
      "step": 33070
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 0.39244988560676575,
      "learning_rate": 2.657131686720224e-06,
      "loss": 0.3572,
      "num_input_tokens_seen": 21569864,
      "step": 33075
    },
    {
      "epoch": 17.337526205450732,
      "grad_norm": 0.4372183382511139,
      "learning_rate": 2.652004195595498e-06,
      "loss": 0.2766,
      "num_input_tokens_seen": 21572776,
      "step": 33080
    },
    {
      "epoch": 17.34014675052411,
      "grad_norm": 0.5247169733047485,
      "learning_rate": 2.6468813794165356e-06,
      "loss": 0.5989,
      "num_input_tokens_seen": 21576584,
      "step": 33085
    },
    {
      "epoch": 17.342767295597483,
      "grad_norm": 0.5060827732086182,
      "learning_rate": 2.641763239254977e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 21579496,
      "step": 33090
    },
    {
      "epoch": 17.34538784067086,
      "grad_norm": 0.3457559645175934,
      "learning_rate": 2.6366497761814713e-06,
      "loss": 0.5155,
      "num_input_tokens_seen": 21582120,
      "step": 33095
    },
    {
      "epoch": 17.348008385744233,
      "grad_norm": 0.42166197299957275,
      "learning_rate": 2.631540991265691e-06,
      "loss": 0.4984,
      "num_input_tokens_seen": 21585224,
      "step": 33100
    },
    {
      "epoch": 17.35062893081761,
      "grad_norm": 0.7844440937042236,
      "learning_rate": 2.6264368855763443e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 21588264,
      "step": 33105
    },
    {
      "epoch": 17.353249475890987,
      "grad_norm": 0.614430844783783,
      "learning_rate": 2.6213374601811504e-06,
      "loss": 0.4977,
      "num_input_tokens_seen": 21591464,
      "step": 33110
    },
    {
      "epoch": 17.35587002096436,
      "grad_norm": 0.7404043078422546,
      "learning_rate": 2.616242716146855e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 21594024,
      "step": 33115
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 0.6344650387763977,
      "learning_rate": 2.6111526545392145e-06,
      "loss": 0.5143,
      "num_input_tokens_seen": 21596776,
      "step": 33120
    },
    {
      "epoch": 17.36111111111111,
      "grad_norm": 0.5691174864768982,
      "learning_rate": 2.6060672764230106e-06,
      "loss": 0.2996,
      "num_input_tokens_seen": 21600392,
      "step": 33125
    },
    {
      "epoch": 17.363731656184488,
      "grad_norm": 0.7355707883834839,
      "learning_rate": 2.600986582862047e-06,
      "loss": 0.6123,
      "num_input_tokens_seen": 21603880,
      "step": 33130
    },
    {
      "epoch": 17.36635220125786,
      "grad_norm": 0.44954705238342285,
      "learning_rate": 2.5959105749191498e-06,
      "loss": 0.4449,
      "num_input_tokens_seen": 21606600,
      "step": 33135
    },
    {
      "epoch": 17.368972746331238,
      "grad_norm": 0.28756123781204224,
      "learning_rate": 2.590839253656166e-06,
      "loss": 0.5266,
      "num_input_tokens_seen": 21610856,
      "step": 33140
    },
    {
      "epoch": 17.37159329140461,
      "grad_norm": 0.45045992732048035,
      "learning_rate": 2.5857726201339473e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 21615016,
      "step": 33145
    },
    {
      "epoch": 17.37421383647799,
      "grad_norm": 0.7505216598510742,
      "learning_rate": 2.5807106754123873e-06,
      "loss": 0.4724,
      "num_input_tokens_seen": 21618920,
      "step": 33150
    },
    {
      "epoch": 17.37683438155136,
      "grad_norm": 0.4487265944480896,
      "learning_rate": 2.575653420550378e-06,
      "loss": 0.3403,
      "num_input_tokens_seen": 21621960,
      "step": 33155
    },
    {
      "epoch": 17.37945492662474,
      "grad_norm": 0.7126368284225464,
      "learning_rate": 2.57060085660584e-06,
      "loss": 0.4377,
      "num_input_tokens_seen": 21624584,
      "step": 33160
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 0.4674534201622009,
      "learning_rate": 2.565552984635716e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 21628200,
      "step": 33165
    },
    {
      "epoch": 17.38469601677149,
      "grad_norm": 0.39824941754341125,
      "learning_rate": 2.5605098056959652e-06,
      "loss": 0.3381,
      "num_input_tokens_seen": 21630792,
      "step": 33170
    },
    {
      "epoch": 17.387316561844862,
      "grad_norm": 0.5995372533798218,
      "learning_rate": 2.555471320841557e-06,
      "loss": 0.5229,
      "num_input_tokens_seen": 21633480,
      "step": 33175
    },
    {
      "epoch": 17.38993710691824,
      "grad_norm": 0.49992263317108154,
      "learning_rate": 2.55043753112649e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 21636584,
      "step": 33180
    },
    {
      "epoch": 17.392557651991613,
      "grad_norm": 0.4200281500816345,
      "learning_rate": 2.5454084376037717e-06,
      "loss": 0.4521,
      "num_input_tokens_seen": 21639624,
      "step": 33185
    },
    {
      "epoch": 17.39517819706499,
      "grad_norm": 0.9247296452522278,
      "learning_rate": 2.5403840413254304e-06,
      "loss": 0.565,
      "num_input_tokens_seen": 21642248,
      "step": 33190
    },
    {
      "epoch": 17.397798742138363,
      "grad_norm": 1.218247652053833,
      "learning_rate": 2.5353643433425196e-06,
      "loss": 0.571,
      "num_input_tokens_seen": 21645576,
      "step": 33195
    },
    {
      "epoch": 17.40041928721174,
      "grad_norm": 0.4986112117767334,
      "learning_rate": 2.5303493447050965e-06,
      "loss": 0.5075,
      "num_input_tokens_seen": 21648712,
      "step": 33200
    },
    {
      "epoch": 17.403039832285117,
      "grad_norm": 0.6502639651298523,
      "learning_rate": 2.525339046462244e-06,
      "loss": 0.4804,
      "num_input_tokens_seen": 21651624,
      "step": 33205
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 0.6230336427688599,
      "learning_rate": 2.5203334496620544e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 21655624,
      "step": 33210
    },
    {
      "epoch": 17.408280922431867,
      "grad_norm": 0.5212849378585815,
      "learning_rate": 2.515332555351646e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 21658632,
      "step": 33215
    },
    {
      "epoch": 17.41090146750524,
      "grad_norm": 0.38138216733932495,
      "learning_rate": 2.510336364577154e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 21661736,
      "step": 33220
    },
    {
      "epoch": 17.413522012578618,
      "grad_norm": 0.54705810546875,
      "learning_rate": 2.5053448783837143e-06,
      "loss": 0.5014,
      "num_input_tokens_seen": 21664776,
      "step": 33225
    },
    {
      "epoch": 17.41614255765199,
      "grad_norm": 0.357619047164917,
      "learning_rate": 2.5003580978154917e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 21667752,
      "step": 33230
    },
    {
      "epoch": 17.418763102725368,
      "grad_norm": 0.25070706009864807,
      "learning_rate": 2.4953760239156716e-06,
      "loss": 0.5519,
      "num_input_tokens_seen": 21672168,
      "step": 33235
    },
    {
      "epoch": 17.42138364779874,
      "grad_norm": 0.4159049987792969,
      "learning_rate": 2.4903986577264366e-06,
      "loss": 0.5152,
      "num_input_tokens_seen": 21674472,
      "step": 33240
    },
    {
      "epoch": 17.42400419287212,
      "grad_norm": 0.7394114136695862,
      "learning_rate": 2.4854260002889986e-06,
      "loss": 0.6434,
      "num_input_tokens_seen": 21677608,
      "step": 33245
    },
    {
      "epoch": 17.42662473794549,
      "grad_norm": 0.49488815665245056,
      "learning_rate": 2.480458052643586e-06,
      "loss": 0.4375,
      "num_input_tokens_seen": 21680552,
      "step": 33250
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 0.44012877345085144,
      "learning_rate": 2.4754948158294293e-06,
      "loss": 0.5042,
      "num_input_tokens_seen": 21682696,
      "step": 33255
    },
    {
      "epoch": 17.431865828092242,
      "grad_norm": 0.4526166021823883,
      "learning_rate": 2.4705362908847814e-06,
      "loss": 0.5037,
      "num_input_tokens_seen": 21686920,
      "step": 33260
    },
    {
      "epoch": 17.43448637316562,
      "grad_norm": 0.6665984988212585,
      "learning_rate": 2.4655824788469172e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 21689416,
      "step": 33265
    },
    {
      "epoch": 17.437106918238992,
      "grad_norm": 0.3376377820968628,
      "learning_rate": 2.4606333807521087e-06,
      "loss": 0.2883,
      "num_input_tokens_seen": 21693480,
      "step": 33270
    },
    {
      "epoch": 17.43972746331237,
      "grad_norm": 1.022185206413269,
      "learning_rate": 2.455688997635658e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 21696296,
      "step": 33275
    },
    {
      "epoch": 17.442348008385743,
      "grad_norm": 0.4118252098560333,
      "learning_rate": 2.4507493305318652e-06,
      "loss": 0.3536,
      "num_input_tokens_seen": 21702088,
      "step": 33280
    },
    {
      "epoch": 17.44496855345912,
      "grad_norm": 0.6295987963676453,
      "learning_rate": 2.445814380474057e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 21705160,
      "step": 33285
    },
    {
      "epoch": 17.447589098532493,
      "grad_norm": 0.3877316415309906,
      "learning_rate": 2.4408841484945684e-06,
      "loss": 0.498,
      "num_input_tokens_seen": 21708328,
      "step": 33290
    },
    {
      "epoch": 17.45020964360587,
      "grad_norm": 0.3501114845275879,
      "learning_rate": 2.435958635624755e-06,
      "loss": 0.5605,
      "num_input_tokens_seen": 21713768,
      "step": 33295
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 0.8472925424575806,
      "learning_rate": 2.4310378428949678e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 21717192,
      "step": 33300
    },
    {
      "epoch": 17.45545073375262,
      "grad_norm": 0.3371170163154602,
      "learning_rate": 2.4261217713345804e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 21721384,
      "step": 33305
    },
    {
      "epoch": 17.458071278825997,
      "grad_norm": 0.41010546684265137,
      "learning_rate": 2.421210421971981e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 21724392,
      "step": 33310
    },
    {
      "epoch": 17.46069182389937,
      "grad_norm": 0.7298856377601624,
      "learning_rate": 2.416303795834571e-06,
      "loss": 0.5177,
      "num_input_tokens_seen": 21727208,
      "step": 33315
    },
    {
      "epoch": 17.463312368972748,
      "grad_norm": 0.45552799105644226,
      "learning_rate": 2.4114018939487615e-06,
      "loss": 0.3604,
      "num_input_tokens_seen": 21729960,
      "step": 33320
    },
    {
      "epoch": 17.46593291404612,
      "grad_norm": 0.3425939679145813,
      "learning_rate": 2.4065047173399662e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 21734216,
      "step": 33325
    },
    {
      "epoch": 17.468553459119498,
      "grad_norm": 0.7880104780197144,
      "learning_rate": 2.4016122670326295e-06,
      "loss": 0.5659,
      "num_input_tokens_seen": 21738152,
      "step": 33330
    },
    {
      "epoch": 17.47117400419287,
      "grad_norm": 0.514617383480072,
      "learning_rate": 2.396724544050186e-06,
      "loss": 0.4949,
      "num_input_tokens_seen": 21741576,
      "step": 33335
    },
    {
      "epoch": 17.47379454926625,
      "grad_norm": 0.3449259400367737,
      "learning_rate": 2.3918415494150944e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 21745224,
      "step": 33340
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 0.322870135307312,
      "learning_rate": 2.3869632841488245e-06,
      "loss": 0.5464,
      "num_input_tokens_seen": 21748424,
      "step": 33345
    },
    {
      "epoch": 17.479035639413,
      "grad_norm": 0.833107054233551,
      "learning_rate": 2.382089749271857e-06,
      "loss": 0.4546,
      "num_input_tokens_seen": 21751336,
      "step": 33350
    },
    {
      "epoch": 17.481656184486372,
      "grad_norm": 0.3476419150829315,
      "learning_rate": 2.3772209458036737e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 21754568,
      "step": 33355
    },
    {
      "epoch": 17.48427672955975,
      "grad_norm": 0.5157994627952576,
      "learning_rate": 2.372356874762771e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 21757832,
      "step": 33360
    },
    {
      "epoch": 17.486897274633122,
      "grad_norm": 0.4370240569114685,
      "learning_rate": 2.36749753716666e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 21760968,
      "step": 33365
    },
    {
      "epoch": 17.4895178197065,
      "grad_norm": 0.3917359709739685,
      "learning_rate": 2.3626429340318613e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 21765160,
      "step": 33370
    },
    {
      "epoch": 17.492138364779873,
      "grad_norm": 0.47879552841186523,
      "learning_rate": 2.357793066373901e-06,
      "loss": 0.6106,
      "num_input_tokens_seen": 21768584,
      "step": 33375
    },
    {
      "epoch": 17.49475890985325,
      "grad_norm": 0.6090266108512878,
      "learning_rate": 2.352947935207317e-06,
      "loss": 0.4319,
      "num_input_tokens_seen": 21771304,
      "step": 33380
    },
    {
      "epoch": 17.497379454926623,
      "grad_norm": 0.4930512309074402,
      "learning_rate": 2.3481075415456495e-06,
      "loss": 0.5423,
      "num_input_tokens_seen": 21774568,
      "step": 33385
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.4942540228366852,
      "learning_rate": 2.3432718864014576e-06,
      "loss": 0.5517,
      "num_input_tokens_seen": 21777608,
      "step": 33390
    },
    {
      "epoch": 17.502620545073377,
      "grad_norm": 0.45059067010879517,
      "learning_rate": 2.338440970786304e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 21781288,
      "step": 33395
    },
    {
      "epoch": 17.50524109014675,
      "grad_norm": 0.5429723858833313,
      "learning_rate": 2.3336147957107644e-06,
      "loss": 0.5229,
      "num_input_tokens_seen": 21784168,
      "step": 33400
    },
    {
      "epoch": 17.507861635220127,
      "grad_norm": 0.6041876077651978,
      "learning_rate": 2.3287933621844227e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 21786760,
      "step": 33405
    },
    {
      "epoch": 17.5104821802935,
      "grad_norm": 0.4630086421966553,
      "learning_rate": 2.3239766712158633e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 21789352,
      "step": 33410
    },
    {
      "epoch": 17.513102725366878,
      "grad_norm": 0.5894788503646851,
      "learning_rate": 2.319164723812675e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 21792296,
      "step": 33415
    },
    {
      "epoch": 17.51572327044025,
      "grad_norm": 0.5500302314758301,
      "learning_rate": 2.314357520981472e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 21794824,
      "step": 33420
    },
    {
      "epoch": 17.518343815513628,
      "grad_norm": 0.43999600410461426,
      "learning_rate": 2.309555063727864e-06,
      "loss": 0.4424,
      "num_input_tokens_seen": 21797928,
      "step": 33425
    },
    {
      "epoch": 17.520964360587,
      "grad_norm": 0.4980275630950928,
      "learning_rate": 2.304757353056475e-06,
      "loss": 0.6329,
      "num_input_tokens_seen": 21801352,
      "step": 33430
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 0.6343475580215454,
      "learning_rate": 2.2999643899709267e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 21805768,
      "step": 33435
    },
    {
      "epoch": 17.52620545073375,
      "grad_norm": 0.5391068458557129,
      "learning_rate": 2.2951761754738456e-06,
      "loss": 0.3517,
      "num_input_tokens_seen": 21808648,
      "step": 33440
    },
    {
      "epoch": 17.52882599580713,
      "grad_norm": 0.8317370414733887,
      "learning_rate": 2.29039271056688e-06,
      "loss": 0.4929,
      "num_input_tokens_seen": 21811080,
      "step": 33445
    },
    {
      "epoch": 17.531446540880502,
      "grad_norm": 0.5866454243659973,
      "learning_rate": 2.285613996250674e-06,
      "loss": 0.3664,
      "num_input_tokens_seen": 21814568,
      "step": 33450
    },
    {
      "epoch": 17.53406708595388,
      "grad_norm": 0.7155541181564331,
      "learning_rate": 2.2808400335248865e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 21817416,
      "step": 33455
    },
    {
      "epoch": 17.536687631027252,
      "grad_norm": 0.7494574189186096,
      "learning_rate": 2.2760708233881654e-06,
      "loss": 0.428,
      "num_input_tokens_seen": 21820648,
      "step": 33460
    },
    {
      "epoch": 17.53930817610063,
      "grad_norm": 0.8956611752510071,
      "learning_rate": 2.271306366838183e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 21823560,
      "step": 33465
    },
    {
      "epoch": 17.541928721174003,
      "grad_norm": 0.49931442737579346,
      "learning_rate": 2.266546664871602e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 21827208,
      "step": 33470
    },
    {
      "epoch": 17.54454926624738,
      "grad_norm": 0.6710150837898254,
      "learning_rate": 2.261791718484102e-06,
      "loss": 0.5287,
      "num_input_tokens_seen": 21830120,
      "step": 33475
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 0.50983726978302,
      "learning_rate": 2.257041528670362e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 21833064,
      "step": 33480
    },
    {
      "epoch": 17.54979035639413,
      "grad_norm": 0.7281284332275391,
      "learning_rate": 2.2522960964240753e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 21835752,
      "step": 33485
    },
    {
      "epoch": 17.552410901467507,
      "grad_norm": 0.4764060974121094,
      "learning_rate": 2.247555422737924e-06,
      "loss": 0.5682,
      "num_input_tokens_seen": 21838664,
      "step": 33490
    },
    {
      "epoch": 17.55503144654088,
      "grad_norm": 0.5753706097602844,
      "learning_rate": 2.2428195086036024e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 21842184,
      "step": 33495
    },
    {
      "epoch": 17.557651991614257,
      "grad_norm": 0.47039929032325745,
      "learning_rate": 2.238088355011811e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 21845224,
      "step": 33500
    },
    {
      "epoch": 17.56027253668763,
      "grad_norm": 0.6942668557167053,
      "learning_rate": 2.2333619629522518e-06,
      "loss": 0.6619,
      "num_input_tokens_seen": 21848552,
      "step": 33505
    },
    {
      "epoch": 17.562893081761008,
      "grad_norm": 0.5516318678855896,
      "learning_rate": 2.2286403334136424e-06,
      "loss": 0.4742,
      "num_input_tokens_seen": 21850856,
      "step": 33510
    },
    {
      "epoch": 17.56551362683438,
      "grad_norm": 0.6859117150306702,
      "learning_rate": 2.223923467383679e-06,
      "loss": 0.3844,
      "num_input_tokens_seen": 21856104,
      "step": 33515
    },
    {
      "epoch": 17.568134171907758,
      "grad_norm": 0.5414899587631226,
      "learning_rate": 2.2192113658490886e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 21858440,
      "step": 33520
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 0.5291438102722168,
      "learning_rate": 2.2145040297955776e-06,
      "loss": 0.7232,
      "num_input_tokens_seen": 21861064,
      "step": 33525
    },
    {
      "epoch": 17.57337526205451,
      "grad_norm": 0.8061471581459045,
      "learning_rate": 2.209801460207875e-06,
      "loss": 0.267,
      "num_input_tokens_seen": 21863848,
      "step": 33530
    },
    {
      "epoch": 17.57599580712788,
      "grad_norm": 0.426688015460968,
      "learning_rate": 2.2051036580697042e-06,
      "loss": 0.419,
      "num_input_tokens_seen": 21867208,
      "step": 33535
    },
    {
      "epoch": 17.57861635220126,
      "grad_norm": 0.4883347451686859,
      "learning_rate": 2.2004106243637885e-06,
      "loss": 0.3938,
      "num_input_tokens_seen": 21871944,
      "step": 33540
    },
    {
      "epoch": 17.581236897274632,
      "grad_norm": 0.42778486013412476,
      "learning_rate": 2.1957223600718587e-06,
      "loss": 0.2534,
      "num_input_tokens_seen": 21874728,
      "step": 33545
    },
    {
      "epoch": 17.58385744234801,
      "grad_norm": 0.9102915525436401,
      "learning_rate": 2.1910388661746493e-06,
      "loss": 0.3975,
      "num_input_tokens_seen": 21877320,
      "step": 33550
    },
    {
      "epoch": 17.586477987421382,
      "grad_norm": 0.5013969540596008,
      "learning_rate": 2.186360143651886e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 21880488,
      "step": 33555
    },
    {
      "epoch": 17.58909853249476,
      "grad_norm": 0.2553679347038269,
      "learning_rate": 2.1816861934823103e-06,
      "loss": 0.5064,
      "num_input_tokens_seen": 21883080,
      "step": 33560
    },
    {
      "epoch": 17.591719077568133,
      "grad_norm": 0.4633180499076843,
      "learning_rate": 2.1770170166436595e-06,
      "loss": 0.5346,
      "num_input_tokens_seen": 21886536,
      "step": 33565
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 0.3002517521381378,
      "learning_rate": 2.172352614112669e-06,
      "loss": 0.445,
      "num_input_tokens_seen": 21889896,
      "step": 33570
    },
    {
      "epoch": 17.596960167714883,
      "grad_norm": 0.3834555447101593,
      "learning_rate": 2.167692986865083e-06,
      "loss": 0.4252,
      "num_input_tokens_seen": 21893256,
      "step": 33575
    },
    {
      "epoch": 17.59958071278826,
      "grad_norm": 0.5939236283302307,
      "learning_rate": 2.1630381358756362e-06,
      "loss": 0.4488,
      "num_input_tokens_seen": 21896328,
      "step": 33580
    },
    {
      "epoch": 17.602201257861637,
      "grad_norm": 0.6759885549545288,
      "learning_rate": 2.158388062118072e-06,
      "loss": 0.5169,
      "num_input_tokens_seen": 21898856,
      "step": 33585
    },
    {
      "epoch": 17.60482180293501,
      "grad_norm": 0.6449843645095825,
      "learning_rate": 2.1537427665651393e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 21901992,
      "step": 33590
    },
    {
      "epoch": 17.607442348008387,
      "grad_norm": 0.6739323139190674,
      "learning_rate": 2.1491022501885726e-06,
      "loss": 0.3542,
      "num_input_tokens_seen": 21904904,
      "step": 33595
    },
    {
      "epoch": 17.61006289308176,
      "grad_norm": 0.36839908361434937,
      "learning_rate": 2.1444665139591174e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 21908808,
      "step": 33600
    },
    {
      "epoch": 17.612683438155138,
      "grad_norm": 0.4455377459526062,
      "learning_rate": 2.139835558846523e-06,
      "loss": 0.4633,
      "num_input_tokens_seen": 21911720,
      "step": 33605
    },
    {
      "epoch": 17.61530398322851,
      "grad_norm": 0.48865070939064026,
      "learning_rate": 2.1352093858195225e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 21914024,
      "step": 33610
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 0.3654732406139374,
      "learning_rate": 2.1305879958458676e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 21916648,
      "step": 33615
    },
    {
      "epoch": 17.62054507337526,
      "grad_norm": 0.2984830141067505,
      "learning_rate": 2.1259713898922894e-06,
      "loss": 0.4426,
      "num_input_tokens_seen": 21920136,
      "step": 33620
    },
    {
      "epoch": 17.62316561844864,
      "grad_norm": 0.43742072582244873,
      "learning_rate": 2.1213595689245386e-06,
      "loss": 0.3958,
      "num_input_tokens_seen": 21924168,
      "step": 33625
    },
    {
      "epoch": 17.62578616352201,
      "grad_norm": 1.239521861076355,
      "learning_rate": 2.1167525339073513e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 21931656,
      "step": 33630
    },
    {
      "epoch": 17.62840670859539,
      "grad_norm": 0.46043750643730164,
      "learning_rate": 2.112150285804473e-06,
      "loss": 0.409,
      "num_input_tokens_seen": 21934184,
      "step": 33635
    },
    {
      "epoch": 17.631027253668762,
      "grad_norm": 0.33860406279563904,
      "learning_rate": 2.107552825578632e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 21938856,
      "step": 33640
    },
    {
      "epoch": 17.63364779874214,
      "grad_norm": 0.2819269299507141,
      "learning_rate": 2.102960154191572e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 21941736,
      "step": 33645
    },
    {
      "epoch": 17.636268343815512,
      "grad_norm": 0.41597238183021545,
      "learning_rate": 2.098372272604021e-06,
      "loss": 0.4998,
      "num_input_tokens_seen": 21945256,
      "step": 33650
    },
    {
      "epoch": 17.63888888888889,
      "grad_norm": 0.8232507109642029,
      "learning_rate": 2.0937891817757155e-06,
      "loss": 0.4744,
      "num_input_tokens_seen": 21947432,
      "step": 33655
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 0.5127698183059692,
      "learning_rate": 2.089210882665385e-06,
      "loss": 0.5266,
      "num_input_tokens_seen": 21950664,
      "step": 33660
    },
    {
      "epoch": 17.64412997903564,
      "grad_norm": 0.5849958062171936,
      "learning_rate": 2.0846373762307635e-06,
      "loss": 0.4914,
      "num_input_tokens_seen": 21953704,
      "step": 33665
    },
    {
      "epoch": 17.646750524109013,
      "grad_norm": 0.27314943075180054,
      "learning_rate": 2.080068663428569e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 21956936,
      "step": 33670
    },
    {
      "epoch": 17.64937106918239,
      "grad_norm": 0.4773518443107605,
      "learning_rate": 2.075504745214521e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 21960200,
      "step": 33675
    },
    {
      "epoch": 17.651991614255767,
      "grad_norm": 0.4047268033027649,
      "learning_rate": 2.0709456225433437e-06,
      "loss": 0.4019,
      "num_input_tokens_seen": 21962824,
      "step": 33680
    },
    {
      "epoch": 17.65461215932914,
      "grad_norm": 0.5893195867538452,
      "learning_rate": 2.0663912963687554e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 21966504,
      "step": 33685
    },
    {
      "epoch": 17.657232704402517,
      "grad_norm": 0.5199195742607117,
      "learning_rate": 2.0618417676434713e-06,
      "loss": 0.4023,
      "num_input_tokens_seen": 21970664,
      "step": 33690
    },
    {
      "epoch": 17.65985324947589,
      "grad_norm": 0.29842454195022583,
      "learning_rate": 2.0572970373191975e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 21974312,
      "step": 33695
    },
    {
      "epoch": 17.662473794549268,
      "grad_norm": 0.43079134821891785,
      "learning_rate": 2.052757106346634e-06,
      "loss": 0.5128,
      "num_input_tokens_seen": 21977288,
      "step": 33700
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 0.8669826984405518,
      "learning_rate": 2.048221975675488e-06,
      "loss": 0.4765,
      "num_input_tokens_seen": 21979720,
      "step": 33705
    },
    {
      "epoch": 17.667714884696018,
      "grad_norm": 0.3238977789878845,
      "learning_rate": 2.043691646254456e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 21982600,
      "step": 33710
    },
    {
      "epoch": 17.67033542976939,
      "grad_norm": 0.5239455103874207,
      "learning_rate": 2.039166119031233e-06,
      "loss": 0.454,
      "num_input_tokens_seen": 21986184,
      "step": 33715
    },
    {
      "epoch": 17.67295597484277,
      "grad_norm": 0.43722787499427795,
      "learning_rate": 2.034645394952511e-06,
      "loss": 0.4998,
      "num_input_tokens_seen": 21989352,
      "step": 33720
    },
    {
      "epoch": 17.67557651991614,
      "grad_norm": 0.41047465801239014,
      "learning_rate": 2.030129474963971e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 21992072,
      "step": 33725
    },
    {
      "epoch": 17.67819706498952,
      "grad_norm": 0.6236319541931152,
      "learning_rate": 2.025618360010284e-06,
      "loss": 0.4859,
      "num_input_tokens_seen": 21996200,
      "step": 33730
    },
    {
      "epoch": 17.680817610062892,
      "grad_norm": 0.5507522821426392,
      "learning_rate": 2.0211120510351317e-06,
      "loss": 0.4427,
      "num_input_tokens_seen": 21998728,
      "step": 33735
    },
    {
      "epoch": 17.68343815513627,
      "grad_norm": 0.43319520354270935,
      "learning_rate": 2.0166105489811822e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 22001480,
      "step": 33740
    },
    {
      "epoch": 17.686058700209642,
      "grad_norm": 0.49084216356277466,
      "learning_rate": 2.0121138547901016e-06,
      "loss": 0.5557,
      "num_input_tokens_seen": 22005800,
      "step": 33745
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 0.6563897132873535,
      "learning_rate": 2.0076219694025427e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 22009096,
      "step": 33750
    },
    {
      "epoch": 17.691299790356393,
      "grad_norm": 0.36699366569519043,
      "learning_rate": 2.003134893758152e-06,
      "loss": 0.4265,
      "num_input_tokens_seen": 22012072,
      "step": 33755
    },
    {
      "epoch": 17.69392033542977,
      "grad_norm": 0.42149144411087036,
      "learning_rate": 1.9986526287955805e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 22015304,
      "step": 33760
    },
    {
      "epoch": 17.696540880503143,
      "grad_norm": 0.293273389339447,
      "learning_rate": 1.9941751754524644e-06,
      "loss": 0.3922,
      "num_input_tokens_seen": 22019016,
      "step": 33765
    },
    {
      "epoch": 17.69916142557652,
      "grad_norm": 0.3715484142303467,
      "learning_rate": 1.9897025346654412e-06,
      "loss": 0.4882,
      "num_input_tokens_seen": 22021992,
      "step": 33770
    },
    {
      "epoch": 17.701781970649897,
      "grad_norm": 1.0254333019256592,
      "learning_rate": 1.985234707370134e-06,
      "loss": 0.3609,
      "num_input_tokens_seen": 22024808,
      "step": 33775
    },
    {
      "epoch": 17.70440251572327,
      "grad_norm": 0.6872379779815674,
      "learning_rate": 1.9807716945011544e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 22027336,
      "step": 33780
    },
    {
      "epoch": 17.707023060796647,
      "grad_norm": 0.7203341126441956,
      "learning_rate": 1.9763134969921215e-06,
      "loss": 0.4124,
      "num_input_tokens_seen": 22030440,
      "step": 33785
    },
    {
      "epoch": 17.70964360587002,
      "grad_norm": 0.562602162361145,
      "learning_rate": 1.971860115775634e-06,
      "loss": 0.5949,
      "num_input_tokens_seen": 22033160,
      "step": 33790
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 0.34936389327049255,
      "learning_rate": 1.967411551783294e-06,
      "loss": 0.2769,
      "num_input_tokens_seen": 22035720,
      "step": 33795
    },
    {
      "epoch": 17.71488469601677,
      "grad_norm": 0.7498680949211121,
      "learning_rate": 1.9629678059456895e-06,
      "loss": 0.529,
      "num_input_tokens_seen": 22038664,
      "step": 33800
    },
    {
      "epoch": 17.717505241090148,
      "grad_norm": 0.47266191244125366,
      "learning_rate": 1.9585288791924004e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 22041512,
      "step": 33805
    },
    {
      "epoch": 17.72012578616352,
      "grad_norm": 0.4643246531486511,
      "learning_rate": 1.9540947724519947e-06,
      "loss": 0.4129,
      "num_input_tokens_seen": 22045320,
      "step": 33810
    },
    {
      "epoch": 17.7227463312369,
      "grad_norm": 0.654042661190033,
      "learning_rate": 1.9496654866520414e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 22048232,
      "step": 33815
    },
    {
      "epoch": 17.72536687631027,
      "grad_norm": 0.39572304487228394,
      "learning_rate": 1.9452410227190937e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 22051528,
      "step": 33820
    },
    {
      "epoch": 17.72798742138365,
      "grad_norm": 0.5136584639549255,
      "learning_rate": 1.9408213815787087e-06,
      "loss": 0.4839,
      "num_input_tokens_seen": 22054248,
      "step": 33825
    },
    {
      "epoch": 17.730607966457022,
      "grad_norm": 0.4960099756717682,
      "learning_rate": 1.936406564155413e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 22058600,
      "step": 33830
    },
    {
      "epoch": 17.7332285115304,
      "grad_norm": 0.5480375289916992,
      "learning_rate": 1.9319965713727435e-06,
      "loss": 0.3945,
      "num_input_tokens_seen": 22061896,
      "step": 33835
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 0.44767236709594727,
      "learning_rate": 1.9275914041532144e-06,
      "loss": 0.5505,
      "num_input_tokens_seen": 22065192,
      "step": 33840
    },
    {
      "epoch": 17.73846960167715,
      "grad_norm": 0.49956250190734863,
      "learning_rate": 1.9231910634183394e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 22068296,
      "step": 33845
    },
    {
      "epoch": 17.741090146750523,
      "grad_norm": 0.4243514835834503,
      "learning_rate": 1.918795550088623e-06,
      "loss": 0.504,
      "num_input_tokens_seen": 22071976,
      "step": 33850
    },
    {
      "epoch": 17.7437106918239,
      "grad_norm": 0.5587326884269714,
      "learning_rate": 1.9144048650835496e-06,
      "loss": 0.5438,
      "num_input_tokens_seen": 22075656,
      "step": 33855
    },
    {
      "epoch": 17.746331236897273,
      "grad_norm": 0.44918301701545715,
      "learning_rate": 1.9100190093216093e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 22079816,
      "step": 33860
    },
    {
      "epoch": 17.74895178197065,
      "grad_norm": 0.7546128630638123,
      "learning_rate": 1.9056379837202627e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 22082472,
      "step": 33865
    },
    {
      "epoch": 17.751572327044027,
      "grad_norm": 0.4093237817287445,
      "learning_rate": 1.9012617891959766e-06,
      "loss": 0.463,
      "num_input_tokens_seen": 22086024,
      "step": 33870
    },
    {
      "epoch": 17.7541928721174,
      "grad_norm": 0.42950332164764404,
      "learning_rate": 1.8968904266641996e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 22090024,
      "step": 33875
    },
    {
      "epoch": 17.756813417190777,
      "grad_norm": 0.4994814395904541,
      "learning_rate": 1.892523897039372e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 22092680,
      "step": 33880
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 1.1000500917434692,
      "learning_rate": 1.888162201234922e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 22095112,
      "step": 33885
    },
    {
      "epoch": 17.762054507337528,
      "grad_norm": 0.4239879846572876,
      "learning_rate": 1.8838053401632666e-06,
      "loss": 0.4742,
      "num_input_tokens_seen": 22098696,
      "step": 33890
    },
    {
      "epoch": 17.7646750524109,
      "grad_norm": 0.6724371910095215,
      "learning_rate": 1.8794533147358074e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 22102024,
      "step": 33895
    },
    {
      "epoch": 17.767295597484278,
      "grad_norm": 0.6057414412498474,
      "learning_rate": 1.875106125862941e-06,
      "loss": 0.414,
      "num_input_tokens_seen": 22105064,
      "step": 33900
    },
    {
      "epoch": 17.76991614255765,
      "grad_norm": 0.5031081438064575,
      "learning_rate": 1.870763774454057e-06,
      "loss": 0.5703,
      "num_input_tokens_seen": 22108968,
      "step": 33905
    },
    {
      "epoch": 17.77253668763103,
      "grad_norm": 0.32448336482048035,
      "learning_rate": 1.8664262614175115e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 22113288,
      "step": 33910
    },
    {
      "epoch": 17.7751572327044,
      "grad_norm": 0.3490169942378998,
      "learning_rate": 1.862093587660671e-06,
      "loss": 0.4684,
      "num_input_tokens_seen": 22117768,
      "step": 33915
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.4831335246562958,
      "learning_rate": 1.8577657540898851e-06,
      "loss": 0.3953,
      "num_input_tokens_seen": 22120936,
      "step": 33920
    },
    {
      "epoch": 17.780398322851152,
      "grad_norm": 0.346133828163147,
      "learning_rate": 1.8534427616104798e-06,
      "loss": 0.3501,
      "num_input_tokens_seen": 22124616,
      "step": 33925
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 0.7020349502563477,
      "learning_rate": 1.8491246111267846e-06,
      "loss": 0.4112,
      "num_input_tokens_seen": 22128072,
      "step": 33930
    },
    {
      "epoch": 17.785639412997902,
      "grad_norm": 0.6604073643684387,
      "learning_rate": 1.8448113035420934e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 22131272,
      "step": 33935
    },
    {
      "epoch": 17.78825995807128,
      "grad_norm": 0.38618090748786926,
      "learning_rate": 1.840502839758712e-06,
      "loss": 0.3474,
      "num_input_tokens_seen": 22134888,
      "step": 33940
    },
    {
      "epoch": 17.790880503144653,
      "grad_norm": 0.42278262972831726,
      "learning_rate": 1.8361992206779226e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 22137480,
      "step": 33945
    },
    {
      "epoch": 17.79350104821803,
      "grad_norm": 0.6522396206855774,
      "learning_rate": 1.8319004471999856e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 22141000,
      "step": 33950
    },
    {
      "epoch": 17.796121593291403,
      "grad_norm": 0.37627461552619934,
      "learning_rate": 1.8276065202241592e-06,
      "loss": 0.325,
      "num_input_tokens_seen": 22143752,
      "step": 33955
    },
    {
      "epoch": 17.79874213836478,
      "grad_norm": 0.42163652181625366,
      "learning_rate": 1.8233174406486863e-06,
      "loss": 0.4196,
      "num_input_tokens_seen": 22147272,
      "step": 33960
    },
    {
      "epoch": 17.801362683438157,
      "grad_norm": 0.301672101020813,
      "learning_rate": 1.8190332093707907e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 22151080,
      "step": 33965
    },
    {
      "epoch": 17.80398322851153,
      "grad_norm": 2.366849422454834,
      "learning_rate": 1.8147538272866833e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 22154664,
      "step": 33970
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 1.0383819341659546,
      "learning_rate": 1.8104792952915672e-06,
      "loss": 0.4264,
      "num_input_tokens_seen": 22157608,
      "step": 33975
    },
    {
      "epoch": 17.80922431865828,
      "grad_norm": 0.9738420248031616,
      "learning_rate": 1.8062096142796192e-06,
      "loss": 0.5095,
      "num_input_tokens_seen": 22160136,
      "step": 33980
    },
    {
      "epoch": 17.811844863731658,
      "grad_norm": 0.60719233751297,
      "learning_rate": 1.8019447851440163e-06,
      "loss": 0.3957,
      "num_input_tokens_seen": 22162376,
      "step": 33985
    },
    {
      "epoch": 17.81446540880503,
      "grad_norm": 0.46878987550735474,
      "learning_rate": 1.7976848087769032e-06,
      "loss": 0.621,
      "num_input_tokens_seen": 22166376,
      "step": 33990
    },
    {
      "epoch": 17.817085953878408,
      "grad_norm": 0.44272580742836,
      "learning_rate": 1.79342968606942e-06,
      "loss": 0.4044,
      "num_input_tokens_seen": 22169128,
      "step": 33995
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.5321689248085022,
      "learning_rate": 1.7891794179116933e-06,
      "loss": 0.4978,
      "num_input_tokens_seen": 22173544,
      "step": 34000
    },
    {
      "epoch": 17.822327044025158,
      "grad_norm": 0.30775341391563416,
      "learning_rate": 1.7849340051928316e-06,
      "loss": 0.4037,
      "num_input_tokens_seen": 22176712,
      "step": 34005
    },
    {
      "epoch": 17.82494758909853,
      "grad_norm": 0.6161625385284424,
      "learning_rate": 1.780693448800927e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 22179816,
      "step": 34010
    },
    {
      "epoch": 17.82756813417191,
      "grad_norm": 0.5431587100028992,
      "learning_rate": 1.7764577496230478e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 22182888,
      "step": 34015
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 0.41610780358314514,
      "learning_rate": 1.7722269085452602e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 22186472,
      "step": 34020
    },
    {
      "epoch": 17.83280922431866,
      "grad_norm": 0.4929746985435486,
      "learning_rate": 1.7680009264526038e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 22189320,
      "step": 34025
    },
    {
      "epoch": 17.835429769392032,
      "grad_norm": 0.6084170341491699,
      "learning_rate": 1.7637798042291126e-06,
      "loss": 0.4482,
      "num_input_tokens_seen": 22192200,
      "step": 34030
    },
    {
      "epoch": 17.83805031446541,
      "grad_norm": 0.5383867621421814,
      "learning_rate": 1.7595635427577944e-06,
      "loss": 0.3633,
      "num_input_tokens_seen": 22194696,
      "step": 34035
    },
    {
      "epoch": 17.840670859538783,
      "grad_norm": 0.4439990818500519,
      "learning_rate": 1.7553521429206465e-06,
      "loss": 0.3825,
      "num_input_tokens_seen": 22197192,
      "step": 34040
    },
    {
      "epoch": 17.84329140461216,
      "grad_norm": 0.42650702595710754,
      "learning_rate": 1.7511456055986335e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 22199976,
      "step": 34045
    },
    {
      "epoch": 17.845911949685533,
      "grad_norm": 0.646995484828949,
      "learning_rate": 1.746943931671724e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 22203176,
      "step": 34050
    },
    {
      "epoch": 17.84853249475891,
      "grad_norm": 0.2638079524040222,
      "learning_rate": 1.742747122018859e-06,
      "loss": 0.3158,
      "num_input_tokens_seen": 22207368,
      "step": 34055
    },
    {
      "epoch": 17.851153039832283,
      "grad_norm": 0.6413215398788452,
      "learning_rate": 1.738555177517967e-06,
      "loss": 0.4736,
      "num_input_tokens_seen": 22210376,
      "step": 34060
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 0.25924932956695557,
      "learning_rate": 1.734368099045952e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 22214856,
      "step": 34065
    },
    {
      "epoch": 17.856394129979037,
      "grad_norm": 0.4068414866924286,
      "learning_rate": 1.730185887478697e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 22218504,
      "step": 34070
    },
    {
      "epoch": 17.85901467505241,
      "grad_norm": 0.7301166653633118,
      "learning_rate": 1.7260085436910794e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 22221384,
      "step": 34075
    },
    {
      "epoch": 17.861635220125788,
      "grad_norm": 0.6781858205795288,
      "learning_rate": 1.7218360685569506e-06,
      "loss": 0.5274,
      "num_input_tokens_seen": 22225480,
      "step": 34080
    },
    {
      "epoch": 17.86425576519916,
      "grad_norm": 0.5699490308761597,
      "learning_rate": 1.7176684629491513e-06,
      "loss": 0.513,
      "num_input_tokens_seen": 22228488,
      "step": 34085
    },
    {
      "epoch": 17.866876310272538,
      "grad_norm": 0.3816727101802826,
      "learning_rate": 1.7135057277394868e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 22231912,
      "step": 34090
    },
    {
      "epoch": 17.86949685534591,
      "grad_norm": 0.616074800491333,
      "learning_rate": 1.7093478637987548e-06,
      "loss": 0.4701,
      "num_input_tokens_seen": 22235144,
      "step": 34095
    },
    {
      "epoch": 17.872117400419288,
      "grad_norm": 0.6629840135574341,
      "learning_rate": 1.7051948719967375e-06,
      "loss": 0.5658,
      "num_input_tokens_seen": 22238088,
      "step": 34100
    },
    {
      "epoch": 17.87473794549266,
      "grad_norm": 0.30856719613075256,
      "learning_rate": 1.7010467532021901e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 22241800,
      "step": 34105
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 0.4184569716453552,
      "learning_rate": 1.6969035082828544e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 22245512,
      "step": 34110
    },
    {
      "epoch": 17.879979035639412,
      "grad_norm": 0.5637756586074829,
      "learning_rate": 1.692765138105451e-06,
      "loss": 0.5113,
      "num_input_tokens_seen": 22248328,
      "step": 34115
    },
    {
      "epoch": 17.88259958071279,
      "grad_norm": 0.4349548816680908,
      "learning_rate": 1.6886316435356792e-06,
      "loss": 0.4932,
      "num_input_tokens_seen": 22252744,
      "step": 34120
    },
    {
      "epoch": 17.885220125786162,
      "grad_norm": 0.5367198586463928,
      "learning_rate": 1.6845030254382138e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 22257544,
      "step": 34125
    },
    {
      "epoch": 17.88784067085954,
      "grad_norm": 0.7345085144042969,
      "learning_rate": 1.6803792846767197e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 22260456,
      "step": 34130
    },
    {
      "epoch": 17.890461215932913,
      "grad_norm": 0.44902798533439636,
      "learning_rate": 1.6762604221138316e-06,
      "loss": 0.4619,
      "num_input_tokens_seen": 22264328,
      "step": 34135
    },
    {
      "epoch": 17.89308176100629,
      "grad_norm": 0.6539924740791321,
      "learning_rate": 1.6721464386111774e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 22267080,
      "step": 34140
    },
    {
      "epoch": 17.895702306079663,
      "grad_norm": 0.41912731528282166,
      "learning_rate": 1.6680373350293516e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 22270248,
      "step": 34145
    },
    {
      "epoch": 17.89832285115304,
      "grad_norm": 0.3652501702308655,
      "learning_rate": 1.6639331122279223e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 22273960,
      "step": 34150
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 1.1331146955490112,
      "learning_rate": 1.6598337710654583e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 22277256,
      "step": 34155
    },
    {
      "epoch": 17.90356394129979,
      "grad_norm": 0.3246707618236542,
      "learning_rate": 1.6557393123994874e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 22280424,
      "step": 34160
    },
    {
      "epoch": 17.906184486373167,
      "grad_norm": 0.45128414034843445,
      "learning_rate": 1.651649737086533e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 22284808,
      "step": 34165
    },
    {
      "epoch": 17.90880503144654,
      "grad_norm": 0.5211299657821655,
      "learning_rate": 1.6475650459820774e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 22287912,
      "step": 34170
    },
    {
      "epoch": 17.911425576519918,
      "grad_norm": 0.4525025188922882,
      "learning_rate": 1.643485239940598e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 22291112,
      "step": 34175
    },
    {
      "epoch": 17.91404612159329,
      "grad_norm": 0.3695257008075714,
      "learning_rate": 1.6394103198155375e-06,
      "loss": 0.4724,
      "num_input_tokens_seen": 22294280,
      "step": 34180
    },
    {
      "epoch": 17.916666666666668,
      "grad_norm": 0.3750275671482086,
      "learning_rate": 1.6353402864593276e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 22297640,
      "step": 34185
    },
    {
      "epoch": 17.91928721174004,
      "grad_norm": 0.624716579914093,
      "learning_rate": 1.6312751407233706e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 22300712,
      "step": 34190
    },
    {
      "epoch": 17.921907756813418,
      "grad_norm": 0.2848202884197235,
      "learning_rate": 1.6272148834580559e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 22304072,
      "step": 34195
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 0.5480794310569763,
      "learning_rate": 1.6231595155127317e-06,
      "loss": 0.5928,
      "num_input_tokens_seen": 22307336,
      "step": 34200
    },
    {
      "epoch": 17.92714884696017,
      "grad_norm": 0.5489487648010254,
      "learning_rate": 1.6191090377357414e-06,
      "loss": 0.4089,
      "num_input_tokens_seen": 22310216,
      "step": 34205
    },
    {
      "epoch": 17.929769392033542,
      "grad_norm": 0.5808227062225342,
      "learning_rate": 1.6150634509743939e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 22314216,
      "step": 34210
    },
    {
      "epoch": 17.93238993710692,
      "grad_norm": 1.2141022682189941,
      "learning_rate": 1.611022756074984e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 22317416,
      "step": 34215
    },
    {
      "epoch": 17.935010482180292,
      "grad_norm": 0.7777513861656189,
      "learning_rate": 1.606986953882783e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 22320456,
      "step": 34220
    },
    {
      "epoch": 17.93763102725367,
      "grad_norm": 0.3718630075454712,
      "learning_rate": 1.6029560452420238e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 22323048,
      "step": 34225
    },
    {
      "epoch": 17.940251572327043,
      "grad_norm": 0.5310676693916321,
      "learning_rate": 1.5989300309959377e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 22326568,
      "step": 34230
    },
    {
      "epoch": 17.94287211740042,
      "grad_norm": 0.4224800169467926,
      "learning_rate": 1.5949089119867116e-06,
      "loss": 0.4543,
      "num_input_tokens_seen": 22329864,
      "step": 34235
    },
    {
      "epoch": 17.945492662473793,
      "grad_norm": 0.9031092524528503,
      "learning_rate": 1.5908926890555204e-06,
      "loss": 0.633,
      "num_input_tokens_seen": 22333000,
      "step": 34240
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 0.5138852000236511,
      "learning_rate": 1.5868813630425167e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 22336328,
      "step": 34245
    },
    {
      "epoch": 17.950733752620543,
      "grad_norm": 0.5346965193748474,
      "learning_rate": 1.5828749347868183e-06,
      "loss": 0.3373,
      "num_input_tokens_seen": 22338920,
      "step": 34250
    },
    {
      "epoch": 17.95335429769392,
      "grad_norm": 0.6353639960289001,
      "learning_rate": 1.5788734051265268e-06,
      "loss": 0.6671,
      "num_input_tokens_seen": 22342568,
      "step": 34255
    },
    {
      "epoch": 17.955974842767297,
      "grad_norm": 0.6439529061317444,
      "learning_rate": 1.5748767748987231e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 22345512,
      "step": 34260
    },
    {
      "epoch": 17.95859538784067,
      "grad_norm": 0.8989706635475159,
      "learning_rate": 1.5708850449394436e-06,
      "loss": 0.5311,
      "num_input_tokens_seen": 22348488,
      "step": 34265
    },
    {
      "epoch": 17.961215932914047,
      "grad_norm": 0.44774454832077026,
      "learning_rate": 1.5668982160837176e-06,
      "loss": 0.3702,
      "num_input_tokens_seen": 22351848,
      "step": 34270
    },
    {
      "epoch": 17.96383647798742,
      "grad_norm": 1.3602508306503296,
      "learning_rate": 1.562916289165553e-06,
      "loss": 0.408,
      "num_input_tokens_seen": 22354856,
      "step": 34275
    },
    {
      "epoch": 17.966457023060798,
      "grad_norm": 0.3313102722167969,
      "learning_rate": 1.5589392650179114e-06,
      "loss": 0.2902,
      "num_input_tokens_seen": 22358152,
      "step": 34280
    },
    {
      "epoch": 17.96907756813417,
      "grad_norm": 0.551128089427948,
      "learning_rate": 1.5549671444727437e-06,
      "loss": 0.5007,
      "num_input_tokens_seen": 22361704,
      "step": 34285
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 0.38994520902633667,
      "learning_rate": 1.55099992836098e-06,
      "loss": 0.5476,
      "num_input_tokens_seen": 22364680,
      "step": 34290
    },
    {
      "epoch": 17.97431865828092,
      "grad_norm": 0.7517342567443848,
      "learning_rate": 1.5470376175125061e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 22367752,
      "step": 34295
    },
    {
      "epoch": 17.9769392033543,
      "grad_norm": 0.5482284426689148,
      "learning_rate": 1.5430802127562038e-06,
      "loss": 0.4228,
      "num_input_tokens_seen": 22370632,
      "step": 34300
    },
    {
      "epoch": 17.979559748427672,
      "grad_norm": 0.4307979941368103,
      "learning_rate": 1.5391277149199023e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 22373224,
      "step": 34305
    },
    {
      "epoch": 17.98218029350105,
      "grad_norm": 0.526813805103302,
      "learning_rate": 1.5351801248304292e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 22379816,
      "step": 34310
    },
    {
      "epoch": 17.984800838574422,
      "grad_norm": 0.4810706675052643,
      "learning_rate": 1.5312374433135767e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 22383112,
      "step": 34315
    },
    {
      "epoch": 17.9874213836478,
      "grad_norm": 0.6247139573097229,
      "learning_rate": 1.5272996711941018e-06,
      "loss": 0.533,
      "num_input_tokens_seen": 22386472,
      "step": 34320
    },
    {
      "epoch": 17.990041928721173,
      "grad_norm": 0.6078426837921143,
      "learning_rate": 1.5233668092957482e-06,
      "loss": 0.5322,
      "num_input_tokens_seen": 22388968,
      "step": 34325
    },
    {
      "epoch": 17.99266247379455,
      "grad_norm": 0.5207186937332153,
      "learning_rate": 1.5194388584412217e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 22392008,
      "step": 34330
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 0.8785868883132935,
      "learning_rate": 1.5155158194522013e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 22394984,
      "step": 34335
    },
    {
      "epoch": 17.9979035639413,
      "grad_norm": 0.4595358669757843,
      "learning_rate": 1.5115976931493498e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 22398376,
      "step": 34340
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.45284271240234375,
      "eval_runtime": 9.8279,
      "eval_samples_per_second": 86.285,
      "eval_steps_per_second": 21.571,
      "num_input_tokens_seen": 22400368,
      "step": 34344
    },
    {
      "epoch": 18.000524109014677,
      "grad_norm": 0.7216786742210388,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.662,
      "num_input_tokens_seen": 22400816,
      "step": 34345
    },
    {
      "epoch": 18.00314465408805,
      "grad_norm": 0.26473820209503174,
      "learning_rate": 1.5037761818796236e-06,
      "loss": 0.4424,
      "num_input_tokens_seen": 22403888,
      "step": 34350
    },
    {
      "epoch": 18.005765199161427,
      "grad_norm": 0.7240591645240784,
      "learning_rate": 1.4998727985489235e-06,
      "loss": 0.4987,
      "num_input_tokens_seen": 22408208,
      "step": 34355
    },
    {
      "epoch": 18.0083857442348,
      "grad_norm": 0.40903592109680176,
      "learning_rate": 1.4959743311767249e-06,
      "loss": 0.4228,
      "num_input_tokens_seen": 22411856,
      "step": 34360
    },
    {
      "epoch": 18.011006289308177,
      "grad_norm": 0.42144259810447693,
      "learning_rate": 1.4920807805785502e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 22415248,
      "step": 34365
    },
    {
      "epoch": 18.01362683438155,
      "grad_norm": 0.37621310353279114,
      "learning_rate": 1.4881921475688815e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 22418416,
      "step": 34370
    },
    {
      "epoch": 18.016247379454928,
      "grad_norm": 0.6350898742675781,
      "learning_rate": 1.4843084329611844e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 22421456,
      "step": 34375
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 0.6267496943473816,
      "learning_rate": 1.480429637567879e-06,
      "loss": 0.3241,
      "num_input_tokens_seen": 22424464,
      "step": 34380
    },
    {
      "epoch": 18.021488469601678,
      "grad_norm": 0.45678192377090454,
      "learning_rate": 1.4765557622003684e-06,
      "loss": 0.3714,
      "num_input_tokens_seen": 22427152,
      "step": 34385
    },
    {
      "epoch": 18.02410901467505,
      "grad_norm": 0.44883373379707336,
      "learning_rate": 1.4726868076690186e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 22431696,
      "step": 34390
    },
    {
      "epoch": 18.02672955974843,
      "grad_norm": 0.6164387464523315,
      "learning_rate": 1.4688227747831767e-06,
      "loss": 0.5909,
      "num_input_tokens_seen": 22434704,
      "step": 34395
    },
    {
      "epoch": 18.029350104821802,
      "grad_norm": 0.8542937636375427,
      "learning_rate": 1.4649636643511573e-06,
      "loss": 0.4932,
      "num_input_tokens_seen": 22437776,
      "step": 34400
    },
    {
      "epoch": 18.03197064989518,
      "grad_norm": 0.514518141746521,
      "learning_rate": 1.4611094771802364e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 22440912,
      "step": 34405
    },
    {
      "epoch": 18.034591194968552,
      "grad_norm": 0.6036940813064575,
      "learning_rate": 1.457260214076664e-06,
      "loss": 0.4697,
      "num_input_tokens_seen": 22444112,
      "step": 34410
    },
    {
      "epoch": 18.03721174004193,
      "grad_norm": 0.4097764194011688,
      "learning_rate": 1.4534158758456623e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 22448080,
      "step": 34415
    },
    {
      "epoch": 18.039832285115303,
      "grad_norm": 2.5019314289093018,
      "learning_rate": 1.4495764632914271e-06,
      "loss": 0.4757,
      "num_input_tokens_seen": 22450800,
      "step": 34420
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 0.3754488229751587,
      "learning_rate": 1.4457419772171215e-06,
      "loss": 0.5222,
      "num_input_tokens_seen": 22453648,
      "step": 34425
    },
    {
      "epoch": 18.045073375262053,
      "grad_norm": 0.7491334676742554,
      "learning_rate": 1.4419124184248733e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 22458480,
      "step": 34430
    },
    {
      "epoch": 18.04769392033543,
      "grad_norm": 0.30482378602027893,
      "learning_rate": 1.4380877877157834e-06,
      "loss": 0.4516,
      "num_input_tokens_seen": 22461840,
      "step": 34435
    },
    {
      "epoch": 18.050314465408807,
      "grad_norm": 0.5402563810348511,
      "learning_rate": 1.4342680858899143e-06,
      "loss": 0.3065,
      "num_input_tokens_seen": 22465104,
      "step": 34440
    },
    {
      "epoch": 18.05293501048218,
      "grad_norm": 0.42311182618141174,
      "learning_rate": 1.4304533137463134e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 22468976,
      "step": 34445
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.642067015171051,
      "learning_rate": 1.4266434720829808e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 22471888,
      "step": 34450
    },
    {
      "epoch": 18.05817610062893,
      "grad_norm": 0.42694759368896484,
      "learning_rate": 1.4228385616968987e-06,
      "loss": 0.4236,
      "num_input_tokens_seen": 22474640,
      "step": 34455
    },
    {
      "epoch": 18.060796645702307,
      "grad_norm": 0.4440756142139435,
      "learning_rate": 1.419038583384008e-06,
      "loss": 0.5661,
      "num_input_tokens_seen": 22477168,
      "step": 34460
    },
    {
      "epoch": 18.06341719077568,
      "grad_norm": 0.503850519657135,
      "learning_rate": 1.4152435379392148e-06,
      "loss": 0.551,
      "num_input_tokens_seen": 22480240,
      "step": 34465
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 0.5291957855224609,
      "learning_rate": 1.411453426156406e-06,
      "loss": 0.4592,
      "num_input_tokens_seen": 22483344,
      "step": 34470
    },
    {
      "epoch": 18.06865828092243,
      "grad_norm": 1.3661142587661743,
      "learning_rate": 1.4076682488284282e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 22485840,
      "step": 34475
    },
    {
      "epoch": 18.071278825995808,
      "grad_norm": 0.5179477334022522,
      "learning_rate": 1.4038880067470978e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 22489392,
      "step": 34480
    },
    {
      "epoch": 18.07389937106918,
      "grad_norm": 0.5452614426612854,
      "learning_rate": 1.400112700703196e-06,
      "loss": 0.3683,
      "num_input_tokens_seen": 22491696,
      "step": 34485
    },
    {
      "epoch": 18.07651991614256,
      "grad_norm": 0.7602025270462036,
      "learning_rate": 1.3963423314864805e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 22494448,
      "step": 34490
    },
    {
      "epoch": 18.079140461215932,
      "grad_norm": 0.7158231139183044,
      "learning_rate": 1.3925768998856587e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 22497264,
      "step": 34495
    },
    {
      "epoch": 18.08176100628931,
      "grad_norm": 0.3956576883792877,
      "learning_rate": 1.388816406688423e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 22500080,
      "step": 34500
    },
    {
      "epoch": 18.084381551362682,
      "grad_norm": 0.5123652815818787,
      "learning_rate": 1.3850608526814246e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 22503152,
      "step": 34505
    },
    {
      "epoch": 18.08700209643606,
      "grad_norm": 0.6657505631446838,
      "learning_rate": 1.3813102386502853e-06,
      "loss": 0.5023,
      "num_input_tokens_seen": 22506288,
      "step": 34510
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 0.7452619075775146,
      "learning_rate": 1.3775645653795881e-06,
      "loss": 0.526,
      "num_input_tokens_seen": 22508496,
      "step": 34515
    },
    {
      "epoch": 18.09224318658281,
      "grad_norm": 0.5094922780990601,
      "learning_rate": 1.373823833652882e-06,
      "loss": 0.4629,
      "num_input_tokens_seen": 22512496,
      "step": 34520
    },
    {
      "epoch": 18.094863731656183,
      "grad_norm": 0.23108632862567902,
      "learning_rate": 1.3700880442526876e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 22515888,
      "step": 34525
    },
    {
      "epoch": 18.09748427672956,
      "grad_norm": 0.666943371295929,
      "learning_rate": 1.3663571979604884e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 22518928,
      "step": 34530
    },
    {
      "epoch": 18.100104821802937,
      "grad_norm": 0.4355166256427765,
      "learning_rate": 1.3626312955567406e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 22523120,
      "step": 34535
    },
    {
      "epoch": 18.10272536687631,
      "grad_norm": 0.5510913133621216,
      "learning_rate": 1.358910337820854e-06,
      "loss": 0.3817,
      "num_input_tokens_seen": 22525392,
      "step": 34540
    },
    {
      "epoch": 18.105345911949687,
      "grad_norm": 0.7062119841575623,
      "learning_rate": 1.3551943255312172e-06,
      "loss": 0.4073,
      "num_input_tokens_seen": 22527920,
      "step": 34545
    },
    {
      "epoch": 18.10796645702306,
      "grad_norm": 0.33245548605918884,
      "learning_rate": 1.3514832594651694e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 22531472,
      "step": 34550
    },
    {
      "epoch": 18.110587002096437,
      "grad_norm": 0.5468137264251709,
      "learning_rate": 1.347777140399026e-06,
      "loss": 0.5015,
      "num_input_tokens_seen": 22534672,
      "step": 34555
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 0.3574463725090027,
      "learning_rate": 1.344075969108069e-06,
      "loss": 0.4775,
      "num_input_tokens_seen": 22537424,
      "step": 34560
    },
    {
      "epoch": 18.115828092243188,
      "grad_norm": 0.3794313073158264,
      "learning_rate": 1.3403797463665352e-06,
      "loss": 0.5288,
      "num_input_tokens_seen": 22540944,
      "step": 34565
    },
    {
      "epoch": 18.11844863731656,
      "grad_norm": 0.3875862956047058,
      "learning_rate": 1.3366884729476365e-06,
      "loss": 0.3767,
      "num_input_tokens_seen": 22544368,
      "step": 34570
    },
    {
      "epoch": 18.121069182389938,
      "grad_norm": 0.4470454156398773,
      "learning_rate": 1.3330021496235467e-06,
      "loss": 0.4388,
      "num_input_tokens_seen": 22547568,
      "step": 34575
    },
    {
      "epoch": 18.12368972746331,
      "grad_norm": 0.5269919633865356,
      "learning_rate": 1.3293207771653965e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 22551504,
      "step": 34580
    },
    {
      "epoch": 18.12631027253669,
      "grad_norm": 0.37407681345939636,
      "learning_rate": 1.3256443563432891e-06,
      "loss": 0.4441,
      "num_input_tokens_seen": 22554416,
      "step": 34585
    },
    {
      "epoch": 18.128930817610062,
      "grad_norm": 0.592533528804779,
      "learning_rate": 1.3219728879262956e-06,
      "loss": 0.298,
      "num_input_tokens_seen": 22557232,
      "step": 34590
    },
    {
      "epoch": 18.13155136268344,
      "grad_norm": 0.30981335043907166,
      "learning_rate": 1.3183063726824375e-06,
      "loss": 0.38,
      "num_input_tokens_seen": 22560976,
      "step": 34595
    },
    {
      "epoch": 18.134171907756812,
      "grad_norm": 0.5568313002586365,
      "learning_rate": 1.3146448113787151e-06,
      "loss": 0.6013,
      "num_input_tokens_seen": 22564272,
      "step": 34600
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 0.5603847503662109,
      "learning_rate": 1.3109882047810773e-06,
      "loss": 0.7153,
      "num_input_tokens_seen": 22567280,
      "step": 34605
    },
    {
      "epoch": 18.139412997903563,
      "grad_norm": 0.4474177658557892,
      "learning_rate": 1.307336553654448e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 22571120,
      "step": 34610
    },
    {
      "epoch": 18.14203354297694,
      "grad_norm": 0.3671526312828064,
      "learning_rate": 1.303689858762716e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 22574672,
      "step": 34615
    },
    {
      "epoch": 18.144654088050313,
      "grad_norm": 0.5424829721450806,
      "learning_rate": 1.3000481208687188e-06,
      "loss": 0.4807,
      "num_input_tokens_seen": 22577616,
      "step": 34620
    },
    {
      "epoch": 18.14727463312369,
      "grad_norm": 1.1069504022598267,
      "learning_rate": 1.2964113407342743e-06,
      "loss": 0.5441,
      "num_input_tokens_seen": 22579888,
      "step": 34625
    },
    {
      "epoch": 18.149895178197063,
      "grad_norm": 0.9612701535224915,
      "learning_rate": 1.292779519120152e-06,
      "loss": 0.4747,
      "num_input_tokens_seen": 22582896,
      "step": 34630
    },
    {
      "epoch": 18.15251572327044,
      "grad_norm": 0.6656397581100464,
      "learning_rate": 1.2891526567860856e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 22585616,
      "step": 34635
    },
    {
      "epoch": 18.155136268343817,
      "grad_norm": 0.6356598138809204,
      "learning_rate": 1.2855307544907825e-06,
      "loss": 0.5522,
      "num_input_tokens_seen": 22588496,
      "step": 34640
    },
    {
      "epoch": 18.15775681341719,
      "grad_norm": 0.5698803067207336,
      "learning_rate": 1.281913812991889e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 22590896,
      "step": 34645
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 1.215293288230896,
      "learning_rate": 1.2783018330460389e-06,
      "loss": 0.7417,
      "num_input_tokens_seen": 22593104,
      "step": 34650
    },
    {
      "epoch": 18.16299790356394,
      "grad_norm": 0.35973817110061646,
      "learning_rate": 1.2746948154088111e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 22596656,
      "step": 34655
    },
    {
      "epoch": 18.165618448637318,
      "grad_norm": 0.28895601630210876,
      "learning_rate": 1.2710927608347605e-06,
      "loss": 0.4315,
      "num_input_tokens_seen": 22599312,
      "step": 34660
    },
    {
      "epoch": 18.16823899371069,
      "grad_norm": 0.27390068769454956,
      "learning_rate": 1.2674956700773872e-06,
      "loss": 0.3621,
      "num_input_tokens_seen": 22603440,
      "step": 34665
    },
    {
      "epoch": 18.170859538784068,
      "grad_norm": 0.6790630221366882,
      "learning_rate": 1.2639035438891666e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 22606512,
      "step": 34670
    },
    {
      "epoch": 18.17348008385744,
      "grad_norm": 0.2941032350063324,
      "learning_rate": 1.260316383021526e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 22609584,
      "step": 34675
    },
    {
      "epoch": 18.17610062893082,
      "grad_norm": 0.4726249575614929,
      "learning_rate": 1.2567341882248645e-06,
      "loss": 0.4794,
      "num_input_tokens_seen": 22613712,
      "step": 34680
    },
    {
      "epoch": 18.178721174004192,
      "grad_norm": 0.3796800971031189,
      "learning_rate": 1.2531569602485305e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 22617008,
      "step": 34685
    },
    {
      "epoch": 18.18134171907757,
      "grad_norm": 0.8150178790092468,
      "learning_rate": 1.2495846998408445e-06,
      "loss": 0.2935,
      "num_input_tokens_seen": 22620112,
      "step": 34690
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 0.49267175793647766,
      "learning_rate": 1.246017407749084e-06,
      "loss": 0.3504,
      "num_input_tokens_seen": 22622608,
      "step": 34695
    },
    {
      "epoch": 18.18658280922432,
      "grad_norm": 0.819338321685791,
      "learning_rate": 1.2424550847194739e-06,
      "loss": 0.5327,
      "num_input_tokens_seen": 22626416,
      "step": 34700
    },
    {
      "epoch": 18.189203354297693,
      "grad_norm": 0.5445960760116577,
      "learning_rate": 1.2388977314972238e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 22629328,
      "step": 34705
    },
    {
      "epoch": 18.19182389937107,
      "grad_norm": 0.9072891473770142,
      "learning_rate": 1.235345348826486e-06,
      "loss": 0.5283,
      "num_input_tokens_seen": 22632144,
      "step": 34710
    },
    {
      "epoch": 18.194444444444443,
      "grad_norm": 0.4106820523738861,
      "learning_rate": 1.2317979374503825e-06,
      "loss": 0.4248,
      "num_input_tokens_seen": 22634992,
      "step": 34715
    },
    {
      "epoch": 18.19706498951782,
      "grad_norm": 0.281177818775177,
      "learning_rate": 1.228255498110989e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 22638256,
      "step": 34720
    },
    {
      "epoch": 18.199685534591197,
      "grad_norm": 0.3857070207595825,
      "learning_rate": 1.224718031549338e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 22641712,
      "step": 34725
    },
    {
      "epoch": 18.20230607966457,
      "grad_norm": 0.35756075382232666,
      "learning_rate": 1.2211855385054344e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 22645904,
      "step": 34730
    },
    {
      "epoch": 18.204926624737947,
      "grad_norm": 0.6718102097511292,
      "learning_rate": 1.217658019718232e-06,
      "loss": 0.4791,
      "num_input_tokens_seen": 22648944,
      "step": 34735
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 0.6002209782600403,
      "learning_rate": 1.2141354759256484e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 22651888,
      "step": 34740
    },
    {
      "epoch": 18.210167714884697,
      "grad_norm": 0.9285193681716919,
      "learning_rate": 1.2106179078645641e-06,
      "loss": 0.49,
      "num_input_tokens_seen": 22655120,
      "step": 34745
    },
    {
      "epoch": 18.21278825995807,
      "grad_norm": 0.836797833442688,
      "learning_rate": 1.2071053162708096e-06,
      "loss": 0.7596,
      "num_input_tokens_seen": 22658928,
      "step": 34750
    },
    {
      "epoch": 18.215408805031448,
      "grad_norm": 0.6614350080490112,
      "learning_rate": 1.203597701879175e-06,
      "loss": 0.4406,
      "num_input_tokens_seen": 22662288,
      "step": 34755
    },
    {
      "epoch": 18.21802935010482,
      "grad_norm": 1.221771478652954,
      "learning_rate": 1.2000950654234177e-06,
      "loss": 0.4314,
      "num_input_tokens_seen": 22664848,
      "step": 34760
    },
    {
      "epoch": 18.220649895178198,
      "grad_norm": 0.3748328387737274,
      "learning_rate": 1.1965974076362491e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 22667376,
      "step": 34765
    },
    {
      "epoch": 18.22327044025157,
      "grad_norm": 0.5683348774909973,
      "learning_rate": 1.193104729249342e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 22671440,
      "step": 34770
    },
    {
      "epoch": 18.22589098532495,
      "grad_norm": 0.3490750193595886,
      "learning_rate": 1.189617030993323e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 22675408,
      "step": 34775
    },
    {
      "epoch": 18.228511530398322,
      "grad_norm": 0.474846214056015,
      "learning_rate": 1.1861343135977753e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 22678224,
      "step": 34780
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 0.5456288456916809,
      "learning_rate": 1.182656577791244e-06,
      "loss": 0.3566,
      "num_input_tokens_seen": 22681712,
      "step": 34785
    },
    {
      "epoch": 18.233752620545072,
      "grad_norm": 0.42453351616859436,
      "learning_rate": 1.1791838243012359e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 22685008,
      "step": 34790
    },
    {
      "epoch": 18.23637316561845,
      "grad_norm": 0.6223408579826355,
      "learning_rate": 1.1757160538542117e-06,
      "loss": 0.4968,
      "num_input_tokens_seen": 22688368,
      "step": 34795
    },
    {
      "epoch": 18.238993710691823,
      "grad_norm": 0.6370689272880554,
      "learning_rate": 1.1722532671755882e-06,
      "loss": 0.4597,
      "num_input_tokens_seen": 22691184,
      "step": 34800
    },
    {
      "epoch": 18.2416142557652,
      "grad_norm": 0.38927096128463745,
      "learning_rate": 1.168795464989736e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 22695024,
      "step": 34805
    },
    {
      "epoch": 18.244234800838573,
      "grad_norm": 0.9250155687332153,
      "learning_rate": 1.165342648019993e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 22699088,
      "step": 34810
    },
    {
      "epoch": 18.24685534591195,
      "grad_norm": 0.5059627294540405,
      "learning_rate": 1.1618948169886457e-06,
      "loss": 0.3954,
      "num_input_tokens_seen": 22702480,
      "step": 34815
    },
    {
      "epoch": 18.249475890985323,
      "grad_norm": 0.6837442517280579,
      "learning_rate": 1.1584519726169446e-06,
      "loss": 0.3005,
      "num_input_tokens_seen": 22705840,
      "step": 34820
    },
    {
      "epoch": 18.2520964360587,
      "grad_norm": 0.6688764691352844,
      "learning_rate": 1.1550141156250942e-06,
      "loss": 0.4523,
      "num_input_tokens_seen": 22708752,
      "step": 34825
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 0.9774647355079651,
      "learning_rate": 1.1515812467322556e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 22711504,
      "step": 34830
    },
    {
      "epoch": 18.25733752620545,
      "grad_norm": 0.6082244515419006,
      "learning_rate": 1.1481533666565375e-06,
      "loss": 0.608,
      "num_input_tokens_seen": 22714384,
      "step": 34835
    },
    {
      "epoch": 18.259958071278827,
      "grad_norm": 0.7821647524833679,
      "learning_rate": 1.144730476115019e-06,
      "loss": 0.4531,
      "num_input_tokens_seen": 22718192,
      "step": 34840
    },
    {
      "epoch": 18.2625786163522,
      "grad_norm": 0.8449783325195312,
      "learning_rate": 1.1413125758237304e-06,
      "loss": 0.4955,
      "num_input_tokens_seen": 22721616,
      "step": 34845
    },
    {
      "epoch": 18.265199161425578,
      "grad_norm": 0.4197094738483429,
      "learning_rate": 1.1378996664976577e-06,
      "loss": 0.519,
      "num_input_tokens_seen": 22724624,
      "step": 34850
    },
    {
      "epoch": 18.26781970649895,
      "grad_norm": 0.41397637128829956,
      "learning_rate": 1.134491748850741e-06,
      "loss": 0.5046,
      "num_input_tokens_seen": 22727600,
      "step": 34855
    },
    {
      "epoch": 18.270440251572328,
      "grad_norm": 0.32598480582237244,
      "learning_rate": 1.13108882359588e-06,
      "loss": 0.3114,
      "num_input_tokens_seen": 22730608,
      "step": 34860
    },
    {
      "epoch": 18.2730607966457,
      "grad_norm": 0.510888397693634,
      "learning_rate": 1.127690891444924e-06,
      "loss": 0.4374,
      "num_input_tokens_seen": 22733424,
      "step": 34865
    },
    {
      "epoch": 18.27568134171908,
      "grad_norm": 0.3282552659511566,
      "learning_rate": 1.1242979531086829e-06,
      "loss": 0.4192,
      "num_input_tokens_seen": 22737008,
      "step": 34870
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 0.6902869343757629,
      "learning_rate": 1.1209100092969244e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 22740272,
      "step": 34875
    },
    {
      "epoch": 18.28092243186583,
      "grad_norm": 0.2842594087123871,
      "learning_rate": 1.1175270607183625e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 22744464,
      "step": 34880
    },
    {
      "epoch": 18.283542976939202,
      "grad_norm": 0.45631903409957886,
      "learning_rate": 1.114149108080678e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 22747152,
      "step": 34885
    },
    {
      "epoch": 18.28616352201258,
      "grad_norm": 0.5504592657089233,
      "learning_rate": 1.1107761520904925e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 22751408,
      "step": 34890
    },
    {
      "epoch": 18.288784067085953,
      "grad_norm": 0.4239780902862549,
      "learning_rate": 1.1074081934533908e-06,
      "loss": 0.5181,
      "num_input_tokens_seen": 22755568,
      "step": 34895
    },
    {
      "epoch": 18.29140461215933,
      "grad_norm": 0.41071146726608276,
      "learning_rate": 1.1040452328739153e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 22758576,
      "step": 34900
    },
    {
      "epoch": 18.294025157232703,
      "grad_norm": 0.4278794229030609,
      "learning_rate": 1.1006872710555615e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 22762128,
      "step": 34905
    },
    {
      "epoch": 18.29664570230608,
      "grad_norm": 0.4316238760948181,
      "learning_rate": 1.0973343087007703e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 22764784,
      "step": 34910
    },
    {
      "epoch": 18.299266247379457,
      "grad_norm": 0.6257055997848511,
      "learning_rate": 1.0939863465109473e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 22768304,
      "step": 34915
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 0.37178167700767517,
      "learning_rate": 1.0906433851864433e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 22771632,
      "step": 34920
    },
    {
      "epoch": 18.304507337526207,
      "grad_norm": 0.507583737373352,
      "learning_rate": 1.0873054254265713e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 22774928,
      "step": 34925
    },
    {
      "epoch": 18.30712788259958,
      "grad_norm": 0.4367640018463135,
      "learning_rate": 1.0839724679295976e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 22778640,
      "step": 34930
    },
    {
      "epoch": 18.309748427672957,
      "grad_norm": 0.3968718647956848,
      "learning_rate": 1.0806445133927311e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 22781936,
      "step": 34935
    },
    {
      "epoch": 18.31236897274633,
      "grad_norm": 0.45239850878715515,
      "learning_rate": 1.0773215625121457e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 22784528,
      "step": 34940
    },
    {
      "epoch": 18.314989517819708,
      "grad_norm": 0.5630651712417603,
      "learning_rate": 1.0740036159829715e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 22790704,
      "step": 34945
    },
    {
      "epoch": 18.31761006289308,
      "grad_norm": 0.3356512486934662,
      "learning_rate": 1.0706906744992724e-06,
      "loss": 0.4665,
      "num_input_tokens_seen": 22794320,
      "step": 34950
    },
    {
      "epoch": 18.320230607966458,
      "grad_norm": 0.27729296684265137,
      "learning_rate": 1.0673827387540857e-06,
      "loss": 0.4784,
      "num_input_tokens_seen": 22798544,
      "step": 34955
    },
    {
      "epoch": 18.32285115303983,
      "grad_norm": 0.5310739278793335,
      "learning_rate": 1.0640798094393916e-06,
      "loss": 0.4224,
      "num_input_tokens_seen": 22801392,
      "step": 34960
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 0.32858923077583313,
      "learning_rate": 1.060781887246126e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 22804880,
      "step": 34965
    },
    {
      "epoch": 18.328092243186582,
      "grad_norm": 0.3721480667591095,
      "learning_rate": 1.057488972864179e-06,
      "loss": 0.6346,
      "num_input_tokens_seen": 22808112,
      "step": 34970
    },
    {
      "epoch": 18.33071278825996,
      "grad_norm": 0.45044904947280884,
      "learning_rate": 1.0542010669823855e-06,
      "loss": 0.4689,
      "num_input_tokens_seen": 22814544,
      "step": 34975
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.5880870819091797,
      "learning_rate": 1.0509181702885395e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 22817200,
      "step": 34980
    },
    {
      "epoch": 18.33595387840671,
      "grad_norm": 0.654508113861084,
      "learning_rate": 1.047640283469392e-06,
      "loss": 0.3604,
      "num_input_tokens_seen": 22820048,
      "step": 34985
    },
    {
      "epoch": 18.338574423480082,
      "grad_norm": 0.750743567943573,
      "learning_rate": 1.0443674072106308e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 22823280,
      "step": 34990
    },
    {
      "epoch": 18.34119496855346,
      "grad_norm": 0.6557040214538574,
      "learning_rate": 1.041099542196905e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 22826224,
      "step": 34995
    },
    {
      "epoch": 18.343815513626833,
      "grad_norm": 0.4875090718269348,
      "learning_rate": 1.0378366891118234e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 22829360,
      "step": 35000
    },
    {
      "epoch": 18.34643605870021,
      "grad_norm": 0.6402535438537598,
      "learning_rate": 1.0345788486379265e-06,
      "loss": 0.4357,
      "num_input_tokens_seen": 22832688,
      "step": 35005
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 0.5178892016410828,
      "learning_rate": 1.0313260214567272e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 22836176,
      "step": 35010
    },
    {
      "epoch": 18.35167714884696,
      "grad_norm": 0.36442631483078003,
      "learning_rate": 1.0280782082486729e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 22839088,
      "step": 35015
    },
    {
      "epoch": 18.354297693920337,
      "grad_norm": 0.6699310541152954,
      "learning_rate": 1.02483540969317e-06,
      "loss": 0.4081,
      "num_input_tokens_seen": 22842384,
      "step": 35020
    },
    {
      "epoch": 18.35691823899371,
      "grad_norm": 0.6189755797386169,
      "learning_rate": 1.0215976264685763e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 22845232,
      "step": 35025
    },
    {
      "epoch": 18.359538784067087,
      "grad_norm": 0.6443455219268799,
      "learning_rate": 1.0183648592522021e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 22848016,
      "step": 35030
    },
    {
      "epoch": 18.36215932914046,
      "grad_norm": 0.770226240158081,
      "learning_rate": 1.0151371087203015e-06,
      "loss": 0.5274,
      "num_input_tokens_seen": 22850672,
      "step": 35035
    },
    {
      "epoch": 18.364779874213838,
      "grad_norm": 0.4397909343242645,
      "learning_rate": 1.0119143755480815e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 22853584,
      "step": 35040
    },
    {
      "epoch": 18.36740041928721,
      "grad_norm": 0.23715253174304962,
      "learning_rate": 1.0086966604097026e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 22857424,
      "step": 35045
    },
    {
      "epoch": 18.370020964360588,
      "grad_norm": 0.38406801223754883,
      "learning_rate": 1.0054839639782766e-06,
      "loss": 0.4039,
      "num_input_tokens_seen": 22861456,
      "step": 35050
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 0.4839761555194855,
      "learning_rate": 1.00227628692586e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 22864752,
      "step": 35055
    },
    {
      "epoch": 18.37526205450734,
      "grad_norm": 0.342966765165329,
      "learning_rate": 9.990736299234692e-07,
      "loss": 0.3793,
      "num_input_tokens_seen": 22868624,
      "step": 35060
    },
    {
      "epoch": 18.377882599580712,
      "grad_norm": 0.5083631277084351,
      "learning_rate": 9.958759936410573e-07,
      "loss": 0.5099,
      "num_input_tokens_seen": 22871760,
      "step": 35065
    },
    {
      "epoch": 18.38050314465409,
      "grad_norm": 0.4202345609664917,
      "learning_rate": 9.926833787475304e-07,
      "loss": 0.3738,
      "num_input_tokens_seen": 22875248,
      "step": 35070
    },
    {
      "epoch": 18.383123689727462,
      "grad_norm": 0.4308411777019501,
      "learning_rate": 9.89495785910749e-07,
      "loss": 0.4988,
      "num_input_tokens_seen": 22878640,
      "step": 35075
    },
    {
      "epoch": 18.38574423480084,
      "grad_norm": 0.3650533854961395,
      "learning_rate": 9.86313215797524e-07,
      "loss": 0.401,
      "num_input_tokens_seen": 22881584,
      "step": 35080
    },
    {
      "epoch": 18.388364779874212,
      "grad_norm": 0.4912787973880768,
      "learning_rate": 9.831356690736149e-07,
      "loss": 0.4037,
      "num_input_tokens_seen": 22884368,
      "step": 35085
    },
    {
      "epoch": 18.39098532494759,
      "grad_norm": 0.8945860862731934,
      "learning_rate": 9.799631464037256e-07,
      "loss": 0.555,
      "num_input_tokens_seen": 22886896,
      "step": 35090
    },
    {
      "epoch": 18.393605870020963,
      "grad_norm": 0.3092974126338959,
      "learning_rate": 9.76795648451509e-07,
      "loss": 0.4858,
      "num_input_tokens_seen": 22890448,
      "step": 35095
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 0.32925254106521606,
      "learning_rate": 9.73633175879568e-07,
      "loss": 0.3951,
      "num_input_tokens_seen": 22894224,
      "step": 35100
    },
    {
      "epoch": 18.398846960167713,
      "grad_norm": 0.4870154857635498,
      "learning_rate": 9.704757293494627e-07,
      "loss": 0.5085,
      "num_input_tokens_seen": 22897680,
      "step": 35105
    },
    {
      "epoch": 18.40146750524109,
      "grad_norm": 0.5467864871025085,
      "learning_rate": 9.673233095216922e-07,
      "loss": 0.4202,
      "num_input_tokens_seen": 22900880,
      "step": 35110
    },
    {
      "epoch": 18.404088050314467,
      "grad_norm": 0.5233452916145325,
      "learning_rate": 9.641759170557068e-07,
      "loss": 0.4392,
      "num_input_tokens_seen": 22903984,
      "step": 35115
    },
    {
      "epoch": 18.40670859538784,
      "grad_norm": 0.48630136251449585,
      "learning_rate": 9.610335526099024e-07,
      "loss": 0.5219,
      "num_input_tokens_seen": 22907216,
      "step": 35120
    },
    {
      "epoch": 18.409329140461217,
      "grad_norm": 0.46830254793167114,
      "learning_rate": 9.578962168416278e-07,
      "loss": 0.5869,
      "num_input_tokens_seen": 22910352,
      "step": 35125
    },
    {
      "epoch": 18.41194968553459,
      "grad_norm": 0.34792929887771606,
      "learning_rate": 9.547639104071775e-07,
      "loss": 0.6359,
      "num_input_tokens_seen": 22913776,
      "step": 35130
    },
    {
      "epoch": 18.414570230607968,
      "grad_norm": 0.48056793212890625,
      "learning_rate": 9.516366339617944e-07,
      "loss": 0.5756,
      "num_input_tokens_seen": 22917584,
      "step": 35135
    },
    {
      "epoch": 18.41719077568134,
      "grad_norm": 0.9091428518295288,
      "learning_rate": 9.485143881596714e-07,
      "loss": 0.4244,
      "num_input_tokens_seen": 22920592,
      "step": 35140
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 0.5575458407402039,
      "learning_rate": 9.453971736539447e-07,
      "loss": 0.4123,
      "num_input_tokens_seen": 22923248,
      "step": 35145
    },
    {
      "epoch": 18.42243186582809,
      "grad_norm": 0.6248784065246582,
      "learning_rate": 9.42284991096698e-07,
      "loss": 0.3749,
      "num_input_tokens_seen": 22926832,
      "step": 35150
    },
    {
      "epoch": 18.42505241090147,
      "grad_norm": 1.1093761920928955,
      "learning_rate": 9.391778411389634e-07,
      "loss": 0.4207,
      "num_input_tokens_seen": 22930256,
      "step": 35155
    },
    {
      "epoch": 18.427672955974842,
      "grad_norm": 0.6188889145851135,
      "learning_rate": 9.360757244307267e-07,
      "loss": 0.5778,
      "num_input_tokens_seen": 22933552,
      "step": 35160
    },
    {
      "epoch": 18.43029350104822,
      "grad_norm": 0.9262675046920776,
      "learning_rate": 9.329786416209157e-07,
      "loss": 0.4174,
      "num_input_tokens_seen": 22936112,
      "step": 35165
    },
    {
      "epoch": 18.432914046121592,
      "grad_norm": 1.2622770071029663,
      "learning_rate": 9.298865933574014e-07,
      "loss": 0.5999,
      "num_input_tokens_seen": 22939376,
      "step": 35170
    },
    {
      "epoch": 18.43553459119497,
      "grad_norm": 0.40664276480674744,
      "learning_rate": 9.267995802870022e-07,
      "loss": 0.3953,
      "num_input_tokens_seen": 22942416,
      "step": 35175
    },
    {
      "epoch": 18.438155136268342,
      "grad_norm": 0.5406326055526733,
      "learning_rate": 9.237176030554878e-07,
      "loss": 0.4916,
      "num_input_tokens_seen": 22945520,
      "step": 35180
    },
    {
      "epoch": 18.44077568134172,
      "grad_norm": 0.539644718170166,
      "learning_rate": 9.206406623075758e-07,
      "loss": 0.4979,
      "num_input_tokens_seen": 22948240,
      "step": 35185
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 0.37247174978256226,
      "learning_rate": 9.175687586869292e-07,
      "loss": 0.4945,
      "num_input_tokens_seen": 22951536,
      "step": 35190
    },
    {
      "epoch": 18.44601677148847,
      "grad_norm": 0.4372307062149048,
      "learning_rate": 9.145018928361476e-07,
      "loss": 0.4225,
      "num_input_tokens_seen": 22955280,
      "step": 35195
    },
    {
      "epoch": 18.448637316561843,
      "grad_norm": 0.33054685592651367,
      "learning_rate": 9.114400653967903e-07,
      "loss": 0.4385,
      "num_input_tokens_seen": 22958576,
      "step": 35200
    },
    {
      "epoch": 18.45125786163522,
      "grad_norm": 0.6090857982635498,
      "learning_rate": 9.083832770093531e-07,
      "loss": 0.462,
      "num_input_tokens_seen": 22961328,
      "step": 35205
    },
    {
      "epoch": 18.453878406708597,
      "grad_norm": 0.31664639711380005,
      "learning_rate": 9.053315283132802e-07,
      "loss": 0.3234,
      "num_input_tokens_seen": 22964880,
      "step": 35210
    },
    {
      "epoch": 18.45649895178197,
      "grad_norm": 0.5998038053512573,
      "learning_rate": 9.022848199469663e-07,
      "loss": 0.4095,
      "num_input_tokens_seen": 22968848,
      "step": 35215
    },
    {
      "epoch": 18.459119496855347,
      "grad_norm": 0.45094045996665955,
      "learning_rate": 8.992431525477518e-07,
      "loss": 0.5262,
      "num_input_tokens_seen": 22972944,
      "step": 35220
    },
    {
      "epoch": 18.46174004192872,
      "grad_norm": 0.5060280561447144,
      "learning_rate": 8.962065267519082e-07,
      "loss": 0.4894,
      "num_input_tokens_seen": 22976208,
      "step": 35225
    },
    {
      "epoch": 18.464360587002098,
      "grad_norm": 0.756167471408844,
      "learning_rate": 8.931749431946718e-07,
      "loss": 0.4324,
      "num_input_tokens_seen": 22978832,
      "step": 35230
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 0.5216332674026489,
      "learning_rate": 8.901484025102103e-07,
      "loss": 0.5165,
      "num_input_tokens_seen": 22982096,
      "step": 35235
    },
    {
      "epoch": 18.469601677148848,
      "grad_norm": 0.5987080335617065,
      "learning_rate": 8.871269053316422e-07,
      "loss": 0.6036,
      "num_input_tokens_seen": 22986032,
      "step": 35240
    },
    {
      "epoch": 18.47222222222222,
      "grad_norm": 0.4677388072013855,
      "learning_rate": 8.841104522910343e-07,
      "loss": 0.3751,
      "num_input_tokens_seen": 22989168,
      "step": 35245
    },
    {
      "epoch": 18.4748427672956,
      "grad_norm": 0.4010280966758728,
      "learning_rate": 8.8109904401939e-07,
      "loss": 0.4898,
      "num_input_tokens_seen": 22992784,
      "step": 35250
    },
    {
      "epoch": 18.47746331236897,
      "grad_norm": 0.3382872939109802,
      "learning_rate": 8.780926811466639e-07,
      "loss": 0.3587,
      "num_input_tokens_seen": 22995824,
      "step": 35255
    },
    {
      "epoch": 18.48008385744235,
      "grad_norm": 0.6519536972045898,
      "learning_rate": 8.7509136430175e-07,
      "loss": 0.5116,
      "num_input_tokens_seen": 22998928,
      "step": 35260
    },
    {
      "epoch": 18.482704402515722,
      "grad_norm": 0.40120360255241394,
      "learning_rate": 8.720950941124877e-07,
      "loss": 0.3597,
      "num_input_tokens_seen": 23001424,
      "step": 35265
    },
    {
      "epoch": 18.4853249475891,
      "grad_norm": 0.41633346676826477,
      "learning_rate": 8.691038712056732e-07,
      "loss": 0.4178,
      "num_input_tokens_seen": 23004784,
      "step": 35270
    },
    {
      "epoch": 18.487945492662472,
      "grad_norm": 0.48115500807762146,
      "learning_rate": 8.66117696207025e-07,
      "loss": 0.3876,
      "num_input_tokens_seen": 23008048,
      "step": 35275
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 0.7342041730880737,
      "learning_rate": 8.631365697412186e-07,
      "loss": 0.3259,
      "num_input_tokens_seen": 23010416,
      "step": 35280
    },
    {
      "epoch": 18.493186582809223,
      "grad_norm": 0.49433186650276184,
      "learning_rate": 8.601604924318802e-07,
      "loss": 0.4428,
      "num_input_tokens_seen": 23013616,
      "step": 35285
    },
    {
      "epoch": 18.4958071278826,
      "grad_norm": 0.4332135021686554,
      "learning_rate": 8.571894649015588e-07,
      "loss": 0.4212,
      "num_input_tokens_seen": 23017808,
      "step": 35290
    },
    {
      "epoch": 18.498427672955973,
      "grad_norm": 0.4699711799621582,
      "learning_rate": 8.542234877717686e-07,
      "loss": 0.3451,
      "num_input_tokens_seen": 23020720,
      "step": 35295
    },
    {
      "epoch": 18.50104821802935,
      "grad_norm": 0.7291731238365173,
      "learning_rate": 8.512625616629549e-07,
      "loss": 0.3122,
      "num_input_tokens_seen": 23024560,
      "step": 35300
    },
    {
      "epoch": 18.503668763102727,
      "grad_norm": 0.4075316786766052,
      "learning_rate": 8.483066871945083e-07,
      "loss": 0.4316,
      "num_input_tokens_seen": 23027856,
      "step": 35305
    },
    {
      "epoch": 18.5062893081761,
      "grad_norm": 0.6707350611686707,
      "learning_rate": 8.453558649847676e-07,
      "loss": 0.3892,
      "num_input_tokens_seen": 23031024,
      "step": 35310
    },
    {
      "epoch": 18.508909853249477,
      "grad_norm": 0.5978732705116272,
      "learning_rate": 8.424100956510112e-07,
      "loss": 0.3769,
      "num_input_tokens_seen": 23034192,
      "step": 35315
    },
    {
      "epoch": 18.51153039832285,
      "grad_norm": 0.6928167939186096,
      "learning_rate": 8.394693798094544e-07,
      "loss": 0.465,
      "num_input_tokens_seen": 23037680,
      "step": 35320
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 0.46922552585601807,
      "learning_rate": 8.365337180752691e-07,
      "loss": 0.4248,
      "num_input_tokens_seen": 23040784,
      "step": 35325
    },
    {
      "epoch": 18.5167714884696,
      "grad_norm": 0.34728819131851196,
      "learning_rate": 8.336031110625558e-07,
      "loss": 0.4048,
      "num_input_tokens_seen": 23043888,
      "step": 35330
    },
    {
      "epoch": 18.519392033542978,
      "grad_norm": 0.7420543432235718,
      "learning_rate": 8.306775593843657e-07,
      "loss": 0.4814,
      "num_input_tokens_seen": 23046928,
      "step": 35335
    },
    {
      "epoch": 18.52201257861635,
      "grad_norm": 0.30302461981773376,
      "learning_rate": 8.277570636526955e-07,
      "loss": 0.3325,
      "num_input_tokens_seen": 23052016,
      "step": 35340
    },
    {
      "epoch": 18.52463312368973,
      "grad_norm": 0.6719740629196167,
      "learning_rate": 8.24841624478473e-07,
      "loss": 0.4239,
      "num_input_tokens_seen": 23055632,
      "step": 35345
    },
    {
      "epoch": 18.5272536687631,
      "grad_norm": 0.40008920431137085,
      "learning_rate": 8.219312424715825e-07,
      "loss": 0.45,
      "num_input_tokens_seen": 23058640,
      "step": 35350
    },
    {
      "epoch": 18.52987421383648,
      "grad_norm": 0.7574991583824158,
      "learning_rate": 8.190259182408345e-07,
      "loss": 0.48,
      "num_input_tokens_seen": 23061488,
      "step": 35355
    },
    {
      "epoch": 18.532494758909852,
      "grad_norm": 0.5840998291969299,
      "learning_rate": 8.161256523939925e-07,
      "loss": 0.3457,
      "num_input_tokens_seen": 23064016,
      "step": 35360
    },
    {
      "epoch": 18.53511530398323,
      "grad_norm": 0.41131192445755005,
      "learning_rate": 8.132304455377576e-07,
      "loss": 0.3913,
      "num_input_tokens_seen": 23066576,
      "step": 35365
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 0.41095274686813354,
      "learning_rate": 8.103402982777813e-07,
      "loss": 0.454,
      "num_input_tokens_seen": 23069936,
      "step": 35370
    },
    {
      "epoch": 18.54035639412998,
      "grad_norm": 0.558268129825592,
      "learning_rate": 8.07455211218644e-07,
      "loss": 0.7261,
      "num_input_tokens_seen": 23073392,
      "step": 35375
    },
    {
      "epoch": 18.542976939203353,
      "grad_norm": 0.4178115427494049,
      "learning_rate": 8.045751849638738e-07,
      "loss": 0.4018,
      "num_input_tokens_seen": 23079568,
      "step": 35380
    },
    {
      "epoch": 18.54559748427673,
      "grad_norm": 0.9413891434669495,
      "learning_rate": 8.017002201159363e-07,
      "loss": 0.3738,
      "num_input_tokens_seen": 23082480,
      "step": 35385
    },
    {
      "epoch": 18.548218029350103,
      "grad_norm": 0.5428705215454102,
      "learning_rate": 7.988303172762418e-07,
      "loss": 0.4058,
      "num_input_tokens_seen": 23085328,
      "step": 35390
    },
    {
      "epoch": 18.55083857442348,
      "grad_norm": 0.4266148507595062,
      "learning_rate": 7.959654770451463e-07,
      "loss": 0.4597,
      "num_input_tokens_seen": 23088208,
      "step": 35395
    },
    {
      "epoch": 18.553459119496857,
      "grad_norm": 0.4586085081100464,
      "learning_rate": 7.9310570002194e-07,
      "loss": 0.3612,
      "num_input_tokens_seen": 23091376,
      "step": 35400
    },
    {
      "epoch": 18.55607966457023,
      "grad_norm": 0.8235530853271484,
      "learning_rate": 7.902509868048552e-07,
      "loss": 0.5263,
      "num_input_tokens_seen": 23094832,
      "step": 35405
    },
    {
      "epoch": 18.558700209643607,
      "grad_norm": 0.42981642484664917,
      "learning_rate": 7.874013379910616e-07,
      "loss": 0.3852,
      "num_input_tokens_seen": 23097808,
      "step": 35410
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 1.2706140279769897,
      "learning_rate": 7.845567541766769e-07,
      "loss": 0.5473,
      "num_input_tokens_seen": 23100464,
      "step": 35415
    },
    {
      "epoch": 18.563941299790358,
      "grad_norm": 0.29698115587234497,
      "learning_rate": 7.817172359567526e-07,
      "loss": 0.3933,
      "num_input_tokens_seen": 23103760,
      "step": 35420
    },
    {
      "epoch": 18.56656184486373,
      "grad_norm": 0.5939328074455261,
      "learning_rate": 7.788827839252888e-07,
      "loss": 0.4352,
      "num_input_tokens_seen": 23106832,
      "step": 35425
    },
    {
      "epoch": 18.569182389937108,
      "grad_norm": 0.3620974123477936,
      "learning_rate": 7.760533986752222e-07,
      "loss": 0.4779,
      "num_input_tokens_seen": 23110832,
      "step": 35430
    },
    {
      "epoch": 18.57180293501048,
      "grad_norm": 0.7338303327560425,
      "learning_rate": 7.732290807984155e-07,
      "loss": 0.3294,
      "num_input_tokens_seen": 23113968,
      "step": 35435
    },
    {
      "epoch": 18.57442348008386,
      "grad_norm": 1.168734073638916,
      "learning_rate": 7.704098308856933e-07,
      "loss": 0.4481,
      "num_input_tokens_seen": 23116944,
      "step": 35440
    },
    {
      "epoch": 18.57704402515723,
      "grad_norm": 0.6340978741645813,
      "learning_rate": 7.675956495268088e-07,
      "loss": 0.5104,
      "num_input_tokens_seen": 23120112,
      "step": 35445
    },
    {
      "epoch": 18.57966457023061,
      "grad_norm": 0.4188145101070404,
      "learning_rate": 7.647865373104551e-07,
      "loss": 0.4886,
      "num_input_tokens_seen": 23124176,
      "step": 35450
    },
    {
      "epoch": 18.582285115303982,
      "grad_norm": 0.567279040813446,
      "learning_rate": 7.619824948242732e-07,
      "loss": 0.4027,
      "num_input_tokens_seen": 23126832,
      "step": 35455
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 0.40183225274086,
      "learning_rate": 7.5918352265483e-07,
      "loss": 0.4374,
      "num_input_tokens_seen": 23129904,
      "step": 35460
    },
    {
      "epoch": 18.587526205450732,
      "grad_norm": 0.40011072158813477,
      "learning_rate": 7.563896213876348e-07,
      "loss": 0.4364,
      "num_input_tokens_seen": 23133200,
      "step": 35465
    },
    {
      "epoch": 18.59014675052411,
      "grad_norm": 0.6658798456192017,
      "learning_rate": 7.53600791607148e-07,
      "loss": 0.3733,
      "num_input_tokens_seen": 23136496,
      "step": 35470
    },
    {
      "epoch": 18.592767295597483,
      "grad_norm": 0.4902965724468231,
      "learning_rate": 7.508170338967585e-07,
      "loss": 0.372,
      "num_input_tokens_seen": 23139472,
      "step": 35475
    },
    {
      "epoch": 18.59538784067086,
      "grad_norm": 0.8102616667747498,
      "learning_rate": 7.480383488387976e-07,
      "loss": 0.423,
      "num_input_tokens_seen": 23142384,
      "step": 35480
    },
    {
      "epoch": 18.598008385744233,
      "grad_norm": 0.38254591822624207,
      "learning_rate": 7.452647370145338e-07,
      "loss": 0.4726,
      "num_input_tokens_seen": 23145328,
      "step": 35485
    },
    {
      "epoch": 18.60062893081761,
      "grad_norm": 0.667789876461029,
      "learning_rate": 7.424961990041696e-07,
      "loss": 0.4392,
      "num_input_tokens_seen": 23148912,
      "step": 35490
    },
    {
      "epoch": 18.603249475890987,
      "grad_norm": 0.6741603016853333,
      "learning_rate": 7.397327353868583e-07,
      "loss": 0.4447,
      "num_input_tokens_seen": 23151632,
      "step": 35495
    },
    {
      "epoch": 18.60587002096436,
      "grad_norm": 0.6126498579978943,
      "learning_rate": 7.369743467406848e-07,
      "loss": 0.4343,
      "num_input_tokens_seen": 23154032,
      "step": 35500
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 0.5368441939353943,
      "learning_rate": 7.34221033642668e-07,
      "loss": 0.4546,
      "num_input_tokens_seen": 23156976,
      "step": 35505
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.44683560729026794,
      "learning_rate": 7.314727966687751e-07,
      "loss": 0.438,
      "num_input_tokens_seen": 23159472,
      "step": 35510
    },
    {
      "epoch": 18.613731656184488,
      "grad_norm": 0.47850528359413147,
      "learning_rate": 7.287296363939045e-07,
      "loss": 0.5321,
      "num_input_tokens_seen": 23162672,
      "step": 35515
    },
    {
      "epoch": 18.61635220125786,
      "grad_norm": 0.6442386507987976,
      "learning_rate": 7.259915533918888e-07,
      "loss": 0.5169,
      "num_input_tokens_seen": 23165456,
      "step": 35520
    },
    {
      "epoch": 18.618972746331238,
      "grad_norm": 0.6219649910926819,
      "learning_rate": 7.232585482355086e-07,
      "loss": 0.4166,
      "num_input_tokens_seen": 23168656,
      "step": 35525
    },
    {
      "epoch": 18.62159329140461,
      "grad_norm": 0.4257546067237854,
      "learning_rate": 7.205306214964736e-07,
      "loss": 0.4483,
      "num_input_tokens_seen": 23171856,
      "step": 35530
    },
    {
      "epoch": 18.62421383647799,
      "grad_norm": 0.5304633378982544,
      "learning_rate": 7.178077737454436e-07,
      "loss": 0.3833,
      "num_input_tokens_seen": 23175856,
      "step": 35535
    },
    {
      "epoch": 18.62683438155136,
      "grad_norm": 0.591601550579071,
      "learning_rate": 7.150900055519993e-07,
      "loss": 0.5055,
      "num_input_tokens_seen": 23178928,
      "step": 35540
    },
    {
      "epoch": 18.62945492662474,
      "grad_norm": 0.40280550718307495,
      "learning_rate": 7.123773174846693e-07,
      "loss": 0.4864,
      "num_input_tokens_seen": 23182384,
      "step": 35545
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 0.523026168346405,
      "learning_rate": 7.096697101109134e-07,
      "loss": 0.3202,
      "num_input_tokens_seen": 23186192,
      "step": 35550
    },
    {
      "epoch": 18.63469601677149,
      "grad_norm": 0.44681739807128906,
      "learning_rate": 7.069671839971371e-07,
      "loss": 0.4486,
      "num_input_tokens_seen": 23191056,
      "step": 35555
    },
    {
      "epoch": 18.637316561844862,
      "grad_norm": 0.6335291266441345,
      "learning_rate": 7.042697397086795e-07,
      "loss": 0.4333,
      "num_input_tokens_seen": 23194160,
      "step": 35560
    },
    {
      "epoch": 18.63993710691824,
      "grad_norm": 0.564699649810791,
      "learning_rate": 7.015773778098089e-07,
      "loss": 0.4149,
      "num_input_tokens_seen": 23197392,
      "step": 35565
    },
    {
      "epoch": 18.642557651991613,
      "grad_norm": 0.34705591201782227,
      "learning_rate": 6.988900988637415e-07,
      "loss": 0.3718,
      "num_input_tokens_seen": 23201008,
      "step": 35570
    },
    {
      "epoch": 18.64517819706499,
      "grad_norm": 0.2803017199039459,
      "learning_rate": 6.962079034326218e-07,
      "loss": 0.3572,
      "num_input_tokens_seen": 23208176,
      "step": 35575
    },
    {
      "epoch": 18.647798742138363,
      "grad_norm": 0.6083044409751892,
      "learning_rate": 6.935307920775346e-07,
      "loss": 0.4559,
      "num_input_tokens_seen": 23210768,
      "step": 35580
    },
    {
      "epoch": 18.65041928721174,
      "grad_norm": 0.5107142329216003,
      "learning_rate": 6.90858765358507e-07,
      "loss": 0.6661,
      "num_input_tokens_seen": 23214480,
      "step": 35585
    },
    {
      "epoch": 18.653039832285117,
      "grad_norm": 1.0476958751678467,
      "learning_rate": 6.881918238344915e-07,
      "loss": 0.4089,
      "num_input_tokens_seen": 23217296,
      "step": 35590
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 0.5535264611244202,
      "learning_rate": 6.855299680633808e-07,
      "loss": 0.3589,
      "num_input_tokens_seen": 23220272,
      "step": 35595
    },
    {
      "epoch": 18.658280922431867,
      "grad_norm": 0.3134688436985016,
      "learning_rate": 6.828731986020104e-07,
      "loss": 0.3458,
      "num_input_tokens_seen": 23224912,
      "step": 35600
    },
    {
      "epoch": 18.66090146750524,
      "grad_norm": 0.37197190523147583,
      "learning_rate": 6.802215160061381e-07,
      "loss": 0.364,
      "num_input_tokens_seen": 23227600,
      "step": 35605
    },
    {
      "epoch": 18.663522012578618,
      "grad_norm": 0.7997124791145325,
      "learning_rate": 6.775749208304704e-07,
      "loss": 0.5836,
      "num_input_tokens_seen": 23230576,
      "step": 35610
    },
    {
      "epoch": 18.66614255765199,
      "grad_norm": 0.7806811928749084,
      "learning_rate": 6.749334136286478e-07,
      "loss": 0.5351,
      "num_input_tokens_seen": 23233360,
      "step": 35615
    },
    {
      "epoch": 18.668763102725368,
      "grad_norm": 0.4336949586868286,
      "learning_rate": 6.722969949532365e-07,
      "loss": 0.6187,
      "num_input_tokens_seen": 23237328,
      "step": 35620
    },
    {
      "epoch": 18.67138364779874,
      "grad_norm": 0.41632479429244995,
      "learning_rate": 6.69665665355751e-07,
      "loss": 0.3687,
      "num_input_tokens_seen": 23240592,
      "step": 35625
    },
    {
      "epoch": 18.67400419287212,
      "grad_norm": 0.35316693782806396,
      "learning_rate": 6.670394253866314e-07,
      "loss": 0.4412,
      "num_input_tokens_seen": 23243472,
      "step": 35630
    },
    {
      "epoch": 18.67662473794549,
      "grad_norm": 0.39734822511672974,
      "learning_rate": 6.644182755952605e-07,
      "loss": 0.4758,
      "num_input_tokens_seen": 23246032,
      "step": 35635
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 0.3878830671310425,
      "learning_rate": 6.618022165299526e-07,
      "loss": 0.4152,
      "num_input_tokens_seen": 23248304,
      "step": 35640
    },
    {
      "epoch": 18.681865828092242,
      "grad_norm": 0.44628962874412537,
      "learning_rate": 6.591912487379531e-07,
      "loss": 0.4441,
      "num_input_tokens_seen": 23251632,
      "step": 35645
    },
    {
      "epoch": 18.68448637316562,
      "grad_norm": 0.7111796736717224,
      "learning_rate": 6.565853727654503e-07,
      "loss": 0.394,
      "num_input_tokens_seen": 23255216,
      "step": 35650
    },
    {
      "epoch": 18.687106918238992,
      "grad_norm": 0.6694605946540833,
      "learning_rate": 6.539845891575636e-07,
      "loss": 0.5269,
      "num_input_tokens_seen": 23258736,
      "step": 35655
    },
    {
      "epoch": 18.68972746331237,
      "grad_norm": 0.5081431865692139,
      "learning_rate": 6.513888984583466e-07,
      "loss": 0.3379,
      "num_input_tokens_seen": 23261872,
      "step": 35660
    },
    {
      "epoch": 18.692348008385743,
      "grad_norm": 0.27531981468200684,
      "learning_rate": 6.487983012107873e-07,
      "loss": 0.5605,
      "num_input_tokens_seen": 23265008,
      "step": 35665
    },
    {
      "epoch": 18.69496855345912,
      "grad_norm": 0.5295745730400085,
      "learning_rate": 6.462127979568133e-07,
      "loss": 0.4015,
      "num_input_tokens_seen": 23268912,
      "step": 35670
    },
    {
      "epoch": 18.697589098532493,
      "grad_norm": 0.621428370475769,
      "learning_rate": 6.436323892372781e-07,
      "loss": 0.2731,
      "num_input_tokens_seen": 23271344,
      "step": 35675
    },
    {
      "epoch": 18.70020964360587,
      "grad_norm": 0.30677446722984314,
      "learning_rate": 6.410570755919721e-07,
      "loss": 0.425,
      "num_input_tokens_seen": 23274960,
      "step": 35680
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 0.6074567437171936,
      "learning_rate": 6.384868575596281e-07,
      "loss": 0.4663,
      "num_input_tokens_seen": 23278160,
      "step": 35685
    },
    {
      "epoch": 18.70545073375262,
      "grad_norm": 0.400661826133728,
      "learning_rate": 6.359217356779024e-07,
      "loss": 0.4706,
      "num_input_tokens_seen": 23282192,
      "step": 35690
    },
    {
      "epoch": 18.708071278825997,
      "grad_norm": 0.5431098937988281,
      "learning_rate": 6.333617104833878e-07,
      "loss": 0.4086,
      "num_input_tokens_seen": 23285648,
      "step": 35695
    },
    {
      "epoch": 18.71069182389937,
      "grad_norm": 0.8068427443504333,
      "learning_rate": 6.308067825116142e-07,
      "loss": 0.6497,
      "num_input_tokens_seen": 23288304,
      "step": 35700
    },
    {
      "epoch": 18.713312368972748,
      "grad_norm": 0.5823047161102295,
      "learning_rate": 6.282569522970433e-07,
      "loss": 0.3967,
      "num_input_tokens_seen": 23291408,
      "step": 35705
    },
    {
      "epoch": 18.71593291404612,
      "grad_norm": 0.49986761808395386,
      "learning_rate": 6.257122203730731e-07,
      "loss": 0.4513,
      "num_input_tokens_seen": 23294448,
      "step": 35710
    },
    {
      "epoch": 18.718553459119498,
      "grad_norm": 0.5048725605010986,
      "learning_rate": 6.231725872720251e-07,
      "loss": 0.4865,
      "num_input_tokens_seen": 23297232,
      "step": 35715
    },
    {
      "epoch": 18.72117400419287,
      "grad_norm": 0.4899386167526245,
      "learning_rate": 6.206380535251716e-07,
      "loss": 0.4811,
      "num_input_tokens_seen": 23301264,
      "step": 35720
    },
    {
      "epoch": 18.72379454926625,
      "grad_norm": 0.3958679437637329,
      "learning_rate": 6.181086196626995e-07,
      "loss": 0.4712,
      "num_input_tokens_seen": 23304720,
      "step": 35725
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 0.20242249965667725,
      "learning_rate": 6.155842862137384e-07,
      "loss": 0.4484,
      "num_input_tokens_seen": 23308592,
      "step": 35730
    },
    {
      "epoch": 18.729035639413,
      "grad_norm": 0.49590423703193665,
      "learning_rate": 6.130650537063548e-07,
      "loss": 0.8644,
      "num_input_tokens_seen": 23312080,
      "step": 35735
    },
    {
      "epoch": 18.731656184486372,
      "grad_norm": 0.6247934699058533,
      "learning_rate": 6.105509226675438e-07,
      "loss": 0.6742,
      "num_input_tokens_seen": 23314800,
      "step": 35740
    },
    {
      "epoch": 18.73427672955975,
      "grad_norm": 0.35762330889701843,
      "learning_rate": 6.08041893623229e-07,
      "loss": 0.5491,
      "num_input_tokens_seen": 23317840,
      "step": 35745
    },
    {
      "epoch": 18.736897274633122,
      "grad_norm": 1.2518013715744019,
      "learning_rate": 6.055379670982713e-07,
      "loss": 0.4608,
      "num_input_tokens_seen": 23321200,
      "step": 35750
    },
    {
      "epoch": 18.7395178197065,
      "grad_norm": 0.9669104218482971,
      "learning_rate": 6.030391436164628e-07,
      "loss": 0.5586,
      "num_input_tokens_seen": 23323824,
      "step": 35755
    },
    {
      "epoch": 18.742138364779873,
      "grad_norm": 0.7589288353919983,
      "learning_rate": 6.005454237005298e-07,
      "loss": 0.4653,
      "num_input_tokens_seen": 23326928,
      "step": 35760
    },
    {
      "epoch": 18.74475890985325,
      "grad_norm": 0.5459305644035339,
      "learning_rate": 5.980568078721327e-07,
      "loss": 0.5428,
      "num_input_tokens_seen": 23330576,
      "step": 35765
    },
    {
      "epoch": 18.747379454926623,
      "grad_norm": 0.3356189727783203,
      "learning_rate": 5.955732966518606e-07,
      "loss": 0.4973,
      "num_input_tokens_seen": 23334832,
      "step": 35770
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.34872373938560486,
      "learning_rate": 5.930948905592343e-07,
      "loss": 0.3642,
      "num_input_tokens_seen": 23337488,
      "step": 35775
    },
    {
      "epoch": 18.752620545073377,
      "grad_norm": 0.609638512134552,
      "learning_rate": 5.906215901127054e-07,
      "loss": 0.3485,
      "num_input_tokens_seen": 23341296,
      "step": 35780
    },
    {
      "epoch": 18.75524109014675,
      "grad_norm": 0.4339747428894043,
      "learning_rate": 5.881533958296631e-07,
      "loss": 0.4967,
      "num_input_tokens_seen": 23348144,
      "step": 35785
    },
    {
      "epoch": 18.757861635220127,
      "grad_norm": 0.3450342118740082,
      "learning_rate": 5.856903082264248e-07,
      "loss": 0.4249,
      "num_input_tokens_seen": 23352080,
      "step": 35790
    },
    {
      "epoch": 18.7604821802935,
      "grad_norm": 0.40688830614089966,
      "learning_rate": 5.832323278182422e-07,
      "loss": 0.2718,
      "num_input_tokens_seen": 23355536,
      "step": 35795
    },
    {
      "epoch": 18.763102725366878,
      "grad_norm": 0.5194177031517029,
      "learning_rate": 5.807794551192986e-07,
      "loss": 0.366,
      "num_input_tokens_seen": 23358160,
      "step": 35800
    },
    {
      "epoch": 18.76572327044025,
      "grad_norm": 0.45821115374565125,
      "learning_rate": 5.783316906427e-07,
      "loss": 0.5362,
      "num_input_tokens_seen": 23361296,
      "step": 35805
    },
    {
      "epoch": 18.768343815513628,
      "grad_norm": 0.5453272461891174,
      "learning_rate": 5.758890349004925e-07,
      "loss": 0.4356,
      "num_input_tokens_seen": 23365264,
      "step": 35810
    },
    {
      "epoch": 18.770964360587,
      "grad_norm": 0.5241059064865112,
      "learning_rate": 5.734514884036563e-07,
      "loss": 0.3183,
      "num_input_tokens_seen": 23368208,
      "step": 35815
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 0.925557017326355,
      "learning_rate": 5.710190516620945e-07,
      "loss": 0.4821,
      "num_input_tokens_seen": 23371248,
      "step": 35820
    },
    {
      "epoch": 18.77620545073375,
      "grad_norm": 0.5548497438430786,
      "learning_rate": 5.685917251846501e-07,
      "loss": 0.361,
      "num_input_tokens_seen": 23374960,
      "step": 35825
    },
    {
      "epoch": 18.77882599580713,
      "grad_norm": 0.48317402601242065,
      "learning_rate": 5.661695094790892e-07,
      "loss": 0.3709,
      "num_input_tokens_seen": 23378832,
      "step": 35830
    },
    {
      "epoch": 18.781446540880502,
      "grad_norm": 0.4660939872264862,
      "learning_rate": 5.637524050521064e-07,
      "loss": 0.5405,
      "num_input_tokens_seen": 23382032,
      "step": 35835
    },
    {
      "epoch": 18.78406708595388,
      "grad_norm": 0.4000738263130188,
      "learning_rate": 5.613404124093391e-07,
      "loss": 0.4074,
      "num_input_tokens_seen": 23385456,
      "step": 35840
    },
    {
      "epoch": 18.786687631027252,
      "grad_norm": 0.4180501401424408,
      "learning_rate": 5.589335320553446e-07,
      "loss": 0.5052,
      "num_input_tokens_seen": 23388592,
      "step": 35845
    },
    {
      "epoch": 18.78930817610063,
      "grad_norm": 0.5102282762527466,
      "learning_rate": 5.565317644936202e-07,
      "loss": 0.4154,
      "num_input_tokens_seen": 23391824,
      "step": 35850
    },
    {
      "epoch": 18.791928721174003,
      "grad_norm": 0.4417542815208435,
      "learning_rate": 5.541351102265863e-07,
      "loss": 0.4294,
      "num_input_tokens_seen": 23394896,
      "step": 35855
    },
    {
      "epoch": 18.79454926624738,
      "grad_norm": 0.6129765510559082,
      "learning_rate": 5.51743569755589e-07,
      "loss": 0.3716,
      "num_input_tokens_seen": 23399888,
      "step": 35860
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 0.7357019186019897,
      "learning_rate": 5.493571435809197e-07,
      "loss": 0.412,
      "num_input_tokens_seen": 23402832,
      "step": 35865
    },
    {
      "epoch": 18.79979035639413,
      "grad_norm": 0.3899708390235901,
      "learning_rate": 5.469758322017849e-07,
      "loss": 0.5458,
      "num_input_tokens_seen": 23405968,
      "step": 35870
    },
    {
      "epoch": 18.802410901467507,
      "grad_norm": 0.5004414319992065,
      "learning_rate": 5.445996361163358e-07,
      "loss": 0.5773,
      "num_input_tokens_seen": 23408656,
      "step": 35875
    },
    {
      "epoch": 18.80503144654088,
      "grad_norm": 0.37673941254615784,
      "learning_rate": 5.422285558216389e-07,
      "loss": 0.6112,
      "num_input_tokens_seen": 23411952,
      "step": 35880
    },
    {
      "epoch": 18.807651991614257,
      "grad_norm": 0.6888284683227539,
      "learning_rate": 5.398625918137001e-07,
      "loss": 0.4761,
      "num_input_tokens_seen": 23414896,
      "step": 35885
    },
    {
      "epoch": 18.81027253668763,
      "grad_norm": 0.6560801863670349,
      "learning_rate": 5.375017445874486e-07,
      "loss": 0.511,
      "num_input_tokens_seen": 23418224,
      "step": 35890
    },
    {
      "epoch": 18.812893081761008,
      "grad_norm": 0.6642569303512573,
      "learning_rate": 5.351460146367504e-07,
      "loss": 0.3941,
      "num_input_tokens_seen": 23421008,
      "step": 35895
    },
    {
      "epoch": 18.81551362683438,
      "grad_norm": 0.3209807574748993,
      "learning_rate": 5.327954024543946e-07,
      "loss": 0.402,
      "num_input_tokens_seen": 23423792,
      "step": 35900
    },
    {
      "epoch": 18.818134171907758,
      "grad_norm": 0.4580826759338379,
      "learning_rate": 5.304499085321074e-07,
      "loss": 0.5038,
      "num_input_tokens_seen": 23426320,
      "step": 35905
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 0.933629035949707,
      "learning_rate": 5.281095333605352e-07,
      "loss": 0.5311,
      "num_input_tokens_seen": 23429840,
      "step": 35910
    },
    {
      "epoch": 18.82337526205451,
      "grad_norm": 0.8739490509033203,
      "learning_rate": 5.257742774292556e-07,
      "loss": 0.3514,
      "num_input_tokens_seen": 23432720,
      "step": 35915
    },
    {
      "epoch": 18.82599580712788,
      "grad_norm": 0.3122164011001587,
      "learning_rate": 5.234441412267837e-07,
      "loss": 0.3517,
      "num_input_tokens_seen": 23436592,
      "step": 35920
    },
    {
      "epoch": 18.82861635220126,
      "grad_norm": 1.0898067951202393,
      "learning_rate": 5.211191252405517e-07,
      "loss": 0.3858,
      "num_input_tokens_seen": 23439856,
      "step": 35925
    },
    {
      "epoch": 18.831236897274632,
      "grad_norm": 0.4372369349002838,
      "learning_rate": 5.187992299569316e-07,
      "loss": 0.3528,
      "num_input_tokens_seen": 23442832,
      "step": 35930
    },
    {
      "epoch": 18.83385744234801,
      "grad_norm": 0.3423387110233307,
      "learning_rate": 5.164844558612131e-07,
      "loss": 0.3833,
      "num_input_tokens_seen": 23446416,
      "step": 35935
    },
    {
      "epoch": 18.836477987421382,
      "grad_norm": 0.4821758568286896,
      "learning_rate": 5.141748034376254e-07,
      "loss": 0.4193,
      "num_input_tokens_seen": 23449104,
      "step": 35940
    },
    {
      "epoch": 18.83909853249476,
      "grad_norm": 0.49348554015159607,
      "learning_rate": 5.118702731693209e-07,
      "loss": 0.4913,
      "num_input_tokens_seen": 23452272,
      "step": 35945
    },
    {
      "epoch": 18.841719077568133,
      "grad_norm": 0.7652196288108826,
      "learning_rate": 5.095708655383751e-07,
      "loss": 0.4608,
      "num_input_tokens_seen": 23455088,
      "step": 35950
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 0.4152604937553406,
      "learning_rate": 5.07276581025809e-07,
      "loss": 0.4821,
      "num_input_tokens_seen": 23458704,
      "step": 35955
    },
    {
      "epoch": 18.846960167714883,
      "grad_norm": 0.38663437962532043,
      "learning_rate": 5.049874201115495e-07,
      "loss": 0.4435,
      "num_input_tokens_seen": 23462704,
      "step": 35960
    },
    {
      "epoch": 18.84958071278826,
      "grad_norm": 0.42247381806373596,
      "learning_rate": 5.027033832744693e-07,
      "loss": 0.3511,
      "num_input_tokens_seen": 23465328,
      "step": 35965
    },
    {
      "epoch": 18.852201257861637,
      "grad_norm": 0.4538993835449219,
      "learning_rate": 5.004244709923639e-07,
      "loss": 0.4906,
      "num_input_tokens_seen": 23468912,
      "step": 35970
    },
    {
      "epoch": 18.85482180293501,
      "grad_norm": 0.42391616106033325,
      "learning_rate": 4.981506837419519e-07,
      "loss": 0.3816,
      "num_input_tokens_seen": 23471984,
      "step": 35975
    },
    {
      "epoch": 18.857442348008387,
      "grad_norm": 0.42437517642974854,
      "learning_rate": 4.958820219988836e-07,
      "loss": 0.3534,
      "num_input_tokens_seen": 23475888,
      "step": 35980
    },
    {
      "epoch": 18.86006289308176,
      "grad_norm": 0.6178160309791565,
      "learning_rate": 4.936184862377402e-07,
      "loss": 0.3934,
      "num_input_tokens_seen": 23478704,
      "step": 35985
    },
    {
      "epoch": 18.862683438155138,
      "grad_norm": 0.6643914580345154,
      "learning_rate": 4.913600769320237e-07,
      "loss": 0.3882,
      "num_input_tokens_seen": 23481136,
      "step": 35990
    },
    {
      "epoch": 18.86530398322851,
      "grad_norm": 0.4203452467918396,
      "learning_rate": 4.891067945541728e-07,
      "loss": 0.3468,
      "num_input_tokens_seen": 23483408,
      "step": 35995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.43709537386894226,
      "learning_rate": 4.868586395755437e-07,
      "loss": 0.4778,
      "num_input_tokens_seen": 23487568,
      "step": 36000
    },
    {
      "epoch": 18.87054507337526,
      "grad_norm": 0.6724894642829895,
      "learning_rate": 4.846156124664242e-07,
      "loss": 0.516,
      "num_input_tokens_seen": 23490512,
      "step": 36005
    },
    {
      "epoch": 18.87316561844864,
      "grad_norm": 0.5993010401725769,
      "learning_rate": 4.823777136960362e-07,
      "loss": 0.5134,
      "num_input_tokens_seen": 23493680,
      "step": 36010
    },
    {
      "epoch": 18.87578616352201,
      "grad_norm": 0.5429602861404419,
      "learning_rate": 4.801449437325162e-07,
      "loss": 0.4557,
      "num_input_tokens_seen": 23497680,
      "step": 36015
    },
    {
      "epoch": 18.87840670859539,
      "grad_norm": 0.6637938022613525,
      "learning_rate": 4.779173030429379e-07,
      "loss": 0.4877,
      "num_input_tokens_seen": 23500464,
      "step": 36020
    },
    {
      "epoch": 18.881027253668762,
      "grad_norm": 0.6028876304626465,
      "learning_rate": 4.7569479209329804e-07,
      "loss": 0.3849,
      "num_input_tokens_seen": 23503184,
      "step": 36025
    },
    {
      "epoch": 18.88364779874214,
      "grad_norm": 0.7169002294540405,
      "learning_rate": 4.734774113485163e-07,
      "loss": 0.4648,
      "num_input_tokens_seen": 23506064,
      "step": 36030
    },
    {
      "epoch": 18.886268343815512,
      "grad_norm": 0.46276798844337463,
      "learning_rate": 4.7126516127244945e-07,
      "loss": 0.4986,
      "num_input_tokens_seen": 23508848,
      "step": 36035
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.7717196941375732,
      "learning_rate": 4.690580423278718e-07,
      "loss": 0.512,
      "num_input_tokens_seen": 23512432,
      "step": 36040
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 1.9861186742782593,
      "learning_rate": 4.668560549764861e-07,
      "loss": 0.5373,
      "num_input_tokens_seen": 23515312,
      "step": 36045
    },
    {
      "epoch": 18.89412997903564,
      "grad_norm": 0.4623790979385376,
      "learning_rate": 4.646591996789268e-07,
      "loss": 0.4845,
      "num_input_tokens_seen": 23519024,
      "step": 36050
    },
    {
      "epoch": 18.896750524109013,
      "grad_norm": 0.5423641800880432,
      "learning_rate": 4.6246747689474847e-07,
      "loss": 0.3326,
      "num_input_tokens_seen": 23522032,
      "step": 36055
    },
    {
      "epoch": 18.89937106918239,
      "grad_norm": 0.843543529510498,
      "learning_rate": 4.6028088708243434e-07,
      "loss": 0.555,
      "num_input_tokens_seen": 23524912,
      "step": 36060
    },
    {
      "epoch": 18.901991614255767,
      "grad_norm": 0.4623037874698639,
      "learning_rate": 4.580994306993991e-07,
      "loss": 0.3056,
      "num_input_tokens_seen": 23527984,
      "step": 36065
    },
    {
      "epoch": 18.90461215932914,
      "grad_norm": 0.375565767288208,
      "learning_rate": 4.5592310820196936e-07,
      "loss": 0.4972,
      "num_input_tokens_seen": 23531792,
      "step": 36070
    },
    {
      "epoch": 18.907232704402517,
      "grad_norm": 0.3612499237060547,
      "learning_rate": 4.537519200454116e-07,
      "loss": 0.4061,
      "num_input_tokens_seen": 23535504,
      "step": 36075
    },
    {
      "epoch": 18.90985324947589,
      "grad_norm": 0.4042753577232361,
      "learning_rate": 4.515858666839151e-07,
      "loss": 0.4108,
      "num_input_tokens_seen": 23538864,
      "step": 36080
    },
    {
      "epoch": 18.912473794549268,
      "grad_norm": 0.546899676322937,
      "learning_rate": 4.494249485705926e-07,
      "loss": 0.5116,
      "num_input_tokens_seen": 23542736,
      "step": 36085
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 0.43214064836502075,
      "learning_rate": 4.472691661574824e-07,
      "loss": 0.4812,
      "num_input_tokens_seen": 23547504,
      "step": 36090
    },
    {
      "epoch": 18.917714884696018,
      "grad_norm": 0.7765440940856934,
      "learning_rate": 4.4511851989555156e-07,
      "loss": 0.4165,
      "num_input_tokens_seen": 23550224,
      "step": 36095
    },
    {
      "epoch": 18.92033542976939,
      "grad_norm": 0.39922574162483215,
      "learning_rate": 4.429730102346874e-07,
      "loss": 0.5173,
      "num_input_tokens_seen": 23553488,
      "step": 36100
    },
    {
      "epoch": 18.92295597484277,
      "grad_norm": 0.5254065990447998,
      "learning_rate": 4.4083263762370873e-07,
      "loss": 0.5374,
      "num_input_tokens_seen": 23556816,
      "step": 36105
    },
    {
      "epoch": 18.92557651991614,
      "grad_norm": 0.904911458492279,
      "learning_rate": 4.386974025103574e-07,
      "loss": 0.3441,
      "num_input_tokens_seen": 23559280,
      "step": 36110
    },
    {
      "epoch": 18.92819706498952,
      "grad_norm": 0.3380356431007385,
      "learning_rate": 4.365673053413011e-07,
      "loss": 0.5179,
      "num_input_tokens_seen": 23562416,
      "step": 36115
    },
    {
      "epoch": 18.930817610062892,
      "grad_norm": 0.33435457944869995,
      "learning_rate": 4.34442346562125e-07,
      "loss": 0.3399,
      "num_input_tokens_seen": 23566032,
      "step": 36120
    },
    {
      "epoch": 18.93343815513627,
      "grad_norm": 0.476179838180542,
      "learning_rate": 4.323225266173542e-07,
      "loss": 0.4872,
      "num_input_tokens_seen": 23568880,
      "step": 36125
    },
    {
      "epoch": 18.936058700209642,
      "grad_norm": 0.4642607867717743,
      "learning_rate": 4.302078459504283e-07,
      "loss": 0.3963,
      "num_input_tokens_seen": 23571216,
      "step": 36130
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 0.37870174646377563,
      "learning_rate": 4.2809830500371294e-07,
      "loss": 0.5555,
      "num_input_tokens_seen": 23575216,
      "step": 36135
    },
    {
      "epoch": 18.941299790356393,
      "grad_norm": 0.6334014534950256,
      "learning_rate": 4.2599390421850236e-07,
      "loss": 0.4149,
      "num_input_tokens_seen": 23578576,
      "step": 36140
    },
    {
      "epoch": 18.94392033542977,
      "grad_norm": 0.40960973501205444,
      "learning_rate": 4.2389464403501113e-07,
      "loss": 0.3885,
      "num_input_tokens_seen": 23580848,
      "step": 36145
    },
    {
      "epoch": 18.946540880503143,
      "grad_norm": 0.4352792203426361,
      "learning_rate": 4.218005248923823e-07,
      "loss": 0.5191,
      "num_input_tokens_seen": 23583440,
      "step": 36150
    },
    {
      "epoch": 18.94916142557652,
      "grad_norm": 0.4190107583999634,
      "learning_rate": 4.1971154722867946e-07,
      "loss": 0.3535,
      "num_input_tokens_seen": 23586640,
      "step": 36155
    },
    {
      "epoch": 18.951781970649897,
      "grad_norm": 0.517744779586792,
      "learning_rate": 4.176277114808946e-07,
      "loss": 0.4109,
      "num_input_tokens_seen": 23589968,
      "step": 36160
    },
    {
      "epoch": 18.95440251572327,
      "grad_norm": 0.6423959136009216,
      "learning_rate": 4.155490180849431e-07,
      "loss": 0.3669,
      "num_input_tokens_seen": 23592368,
      "step": 36165
    },
    {
      "epoch": 18.957023060796647,
      "grad_norm": 0.7242658138275146,
      "learning_rate": 4.1347546747566304e-07,
      "loss": 0.3694,
      "num_input_tokens_seen": 23595248,
      "step": 36170
    },
    {
      "epoch": 18.95964360587002,
      "grad_norm": 0.48373833298683167,
      "learning_rate": 4.114070600868131e-07,
      "loss": 0.4514,
      "num_input_tokens_seen": 23598384,
      "step": 36175
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 0.4495646357536316,
      "learning_rate": 4.093437963510888e-07,
      "loss": 0.482,
      "num_input_tokens_seen": 23602128,
      "step": 36180
    },
    {
      "epoch": 18.96488469601677,
      "grad_norm": 0.2781803011894226,
      "learning_rate": 4.072856767000949e-07,
      "loss": 0.3376,
      "num_input_tokens_seen": 23605232,
      "step": 36185
    },
    {
      "epoch": 18.967505241090148,
      "grad_norm": 0.5944094061851501,
      "learning_rate": 4.052327015643703e-07,
      "loss": 0.3937,
      "num_input_tokens_seen": 23608272,
      "step": 36190
    },
    {
      "epoch": 18.97012578616352,
      "grad_norm": 0.5783764123916626,
      "learning_rate": 4.031848713733744e-07,
      "loss": 0.3977,
      "num_input_tokens_seen": 23610576,
      "step": 36195
    },
    {
      "epoch": 18.9727463312369,
      "grad_norm": 0.5894608497619629,
      "learning_rate": 4.011421865554837e-07,
      "loss": 0.4867,
      "num_input_tokens_seen": 23614000,
      "step": 36200
    },
    {
      "epoch": 18.97536687631027,
      "grad_norm": 0.37869444489479065,
      "learning_rate": 3.9910464753801224e-07,
      "loss": 0.3802,
      "num_input_tokens_seen": 23617264,
      "step": 36205
    },
    {
      "epoch": 18.97798742138365,
      "grad_norm": 0.5171501040458679,
      "learning_rate": 3.9707225474718557e-07,
      "loss": 0.488,
      "num_input_tokens_seen": 23620240,
      "step": 36210
    },
    {
      "epoch": 18.980607966457022,
      "grad_norm": 0.520415186882019,
      "learning_rate": 3.950450086081553e-07,
      "loss": 0.4083,
      "num_input_tokens_seen": 23623216,
      "step": 36215
    },
    {
      "epoch": 18.9832285115304,
      "grad_norm": 1.1117563247680664,
      "learning_rate": 3.930229095450072e-07,
      "loss": 0.3987,
      "num_input_tokens_seen": 23626064,
      "step": 36220
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 0.543402373790741,
      "learning_rate": 3.9100595798073345e-07,
      "loss": 0.4616,
      "num_input_tokens_seen": 23629360,
      "step": 36225
    },
    {
      "epoch": 18.98846960167715,
      "grad_norm": 0.723065972328186,
      "learning_rate": 3.8899415433725493e-07,
      "loss": 0.4599,
      "num_input_tokens_seen": 23631792,
      "step": 36230
    },
    {
      "epoch": 18.991090146750523,
      "grad_norm": 0.36893272399902344,
      "learning_rate": 3.8698749903542666e-07,
      "loss": 0.377,
      "num_input_tokens_seen": 23636400,
      "step": 36235
    },
    {
      "epoch": 18.9937106918239,
      "grad_norm": 0.7629532814025879,
      "learning_rate": 3.8498599249501e-07,
      "loss": 0.4144,
      "num_input_tokens_seen": 23639056,
      "step": 36240
    },
    {
      "epoch": 18.996331236897273,
      "grad_norm": 0.660149097442627,
      "learning_rate": 3.8298963513470345e-07,
      "loss": 0.3694,
      "num_input_tokens_seen": 23642672,
      "step": 36245
    },
    {
      "epoch": 18.99895178197065,
      "grad_norm": 0.7075915932655334,
      "learning_rate": 3.8099842737211734e-07,
      "loss": 0.4204,
      "num_input_tokens_seen": 23644848,
      "step": 36250
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.45250704884529114,
      "eval_runtime": 9.9019,
      "eval_samples_per_second": 85.64,
      "eval_steps_per_second": 21.41,
      "num_input_tokens_seen": 23645440,
      "step": 36252
    },
    {
      "epoch": 19.001572327044027,
      "grad_norm": 0.584495484828949,
      "learning_rate": 3.7901236962379627e-07,
      "loss": 0.3204,
      "num_input_tokens_seen": 23647040,
      "step": 36255
    },
    {
      "epoch": 19.0041928721174,
      "grad_norm": 0.4337828457355499,
      "learning_rate": 3.7703146230519126e-07,
      "loss": 0.4551,
      "num_input_tokens_seen": 23649760,
      "step": 36260
    },
    {
      "epoch": 19.006813417190777,
      "grad_norm": 0.4754452705383301,
      "learning_rate": 3.7505570583069306e-07,
      "loss": 0.351,
      "num_input_tokens_seen": 23652672,
      "step": 36265
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 0.43879434466362,
      "learning_rate": 3.7308510061360435e-07,
      "loss": 0.4181,
      "num_input_tokens_seen": 23655584,
      "step": 36270
    },
    {
      "epoch": 19.012054507337528,
      "grad_norm": 0.38877859711647034,
      "learning_rate": 3.7111964706615374e-07,
      "loss": 0.4966,
      "num_input_tokens_seen": 23658496,
      "step": 36275
    },
    {
      "epoch": 19.0146750524109,
      "grad_norm": 0.6386126279830933,
      "learning_rate": 3.691593455994929e-07,
      "loss": 0.3885,
      "num_input_tokens_seen": 23660928,
      "step": 36280
    },
    {
      "epoch": 19.017295597484278,
      "grad_norm": 0.66725093126297,
      "learning_rate": 3.672041966236911e-07,
      "loss": 0.6202,
      "num_input_tokens_seen": 23664448,
      "step": 36285
    },
    {
      "epoch": 19.01991614255765,
      "grad_norm": 0.36439579725265503,
      "learning_rate": 3.6525420054774884e-07,
      "loss": 0.5492,
      "num_input_tokens_seen": 23668128,
      "step": 36290
    },
    {
      "epoch": 19.02253668763103,
      "grad_norm": 0.6128097772598267,
      "learning_rate": 3.6330935777957597e-07,
      "loss": 0.4497,
      "num_input_tokens_seen": 23672128,
      "step": 36295
    },
    {
      "epoch": 19.0251572327044,
      "grad_norm": 0.34392714500427246,
      "learning_rate": 3.613696687260193e-07,
      "loss": 0.443,
      "num_input_tokens_seen": 23675392,
      "step": 36300
    },
    {
      "epoch": 19.02777777777778,
      "grad_norm": 0.4766956865787506,
      "learning_rate": 3.59435133792832e-07,
      "loss": 0.4943,
      "num_input_tokens_seen": 23679136,
      "step": 36305
    },
    {
      "epoch": 19.030398322851152,
      "grad_norm": 0.8385258913040161,
      "learning_rate": 3.575057533847043e-07,
      "loss": 0.4636,
      "num_input_tokens_seen": 23681920,
      "step": 36310
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 0.32874205708503723,
      "learning_rate": 3.555815279052327e-07,
      "loss": 0.3889,
      "num_input_tokens_seen": 23685824,
      "step": 36315
    },
    {
      "epoch": 19.035639412997902,
      "grad_norm": 0.5694165229797363,
      "learning_rate": 3.5366245775694805e-07,
      "loss": 0.5027,
      "num_input_tokens_seen": 23689344,
      "step": 36320
    },
    {
      "epoch": 19.03825995807128,
      "grad_norm": 0.6625328063964844,
      "learning_rate": 3.517485433412987e-07,
      "loss": 0.3912,
      "num_input_tokens_seen": 23692384,
      "step": 36325
    },
    {
      "epoch": 19.040880503144653,
      "grad_norm": 0.5516113042831421,
      "learning_rate": 3.4983978505865044e-07,
      "loss": 0.4091,
      "num_input_tokens_seen": 23695936,
      "step": 36330
    },
    {
      "epoch": 19.04350104821803,
      "grad_norm": 0.4813024401664734,
      "learning_rate": 3.47936183308295e-07,
      "loss": 0.482,
      "num_input_tokens_seen": 23698848,
      "step": 36335
    },
    {
      "epoch": 19.046121593291403,
      "grad_norm": 0.4069702625274658,
      "learning_rate": 3.460377384884472e-07,
      "loss": 0.3994,
      "num_input_tokens_seen": 23702464,
      "step": 36340
    },
    {
      "epoch": 19.04874213836478,
      "grad_norm": 0.926628828048706,
      "learning_rate": 3.4414445099623384e-07,
      "loss": 0.3682,
      "num_input_tokens_seen": 23705344,
      "step": 36345
    },
    {
      "epoch": 19.051362683438157,
      "grad_norm": 0.40354254841804504,
      "learning_rate": 3.42256321227713e-07,
      "loss": 0.4654,
      "num_input_tokens_seen": 23708992,
      "step": 36350
    },
    {
      "epoch": 19.05398322851153,
      "grad_norm": 0.5458836555480957,
      "learning_rate": 3.403733495778605e-07,
      "loss": 0.3632,
      "num_input_tokens_seen": 23711808,
      "step": 36355
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 0.3304532766342163,
      "learning_rate": 3.3849553644056955e-07,
      "loss": 0.5599,
      "num_input_tokens_seen": 23714720,
      "step": 36360
    },
    {
      "epoch": 19.05922431865828,
      "grad_norm": 0.49292194843292236,
      "learning_rate": 3.36622882208662e-07,
      "loss": 0.5543,
      "num_input_tokens_seen": 23717984,
      "step": 36365
    },
    {
      "epoch": 19.061844863731658,
      "grad_norm": 0.4189697504043579,
      "learning_rate": 3.347553872738718e-07,
      "loss": 0.473,
      "num_input_tokens_seen": 23721184,
      "step": 36370
    },
    {
      "epoch": 19.06446540880503,
      "grad_norm": 0.27750998735427856,
      "learning_rate": 3.3289305202685583e-07,
      "loss": 0.4508,
      "num_input_tokens_seen": 23724384,
      "step": 36375
    },
    {
      "epoch": 19.067085953878408,
      "grad_norm": 0.4667239189147949,
      "learning_rate": 3.310358768571997e-07,
      "loss": 0.4812,
      "num_input_tokens_seen": 23727680,
      "step": 36380
    },
    {
      "epoch": 19.06970649895178,
      "grad_norm": 0.7057337164878845,
      "learning_rate": 3.2918386215339827e-07,
      "loss": 0.5444,
      "num_input_tokens_seen": 23730464,
      "step": 36385
    },
    {
      "epoch": 19.072327044025158,
      "grad_norm": 0.5226178169250488,
      "learning_rate": 3.2733700830287205e-07,
      "loss": 0.3538,
      "num_input_tokens_seen": 23733888,
      "step": 36390
    },
    {
      "epoch": 19.07494758909853,
      "grad_norm": 0.5423871278762817,
      "learning_rate": 3.254953156919649e-07,
      "loss": 0.4645,
      "num_input_tokens_seen": 23736736,
      "step": 36395
    },
    {
      "epoch": 19.07756813417191,
      "grad_norm": 0.37383192777633667,
      "learning_rate": 3.2365878470593527e-07,
      "loss": 0.4598,
      "num_input_tokens_seen": 23739872,
      "step": 36400
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 0.46082985401153564,
      "learning_rate": 3.2182741572896747e-07,
      "loss": 0.3743,
      "num_input_tokens_seen": 23742592,
      "step": 36405
    },
    {
      "epoch": 19.08280922431866,
      "grad_norm": 0.33007779717445374,
      "learning_rate": 3.2000120914415786e-07,
      "loss": 0.5486,
      "num_input_tokens_seen": 23747296,
      "step": 36410
    },
    {
      "epoch": 19.085429769392032,
      "grad_norm": 0.46219491958618164,
      "learning_rate": 3.181801653335315e-07,
      "loss": 0.4476,
      "num_input_tokens_seen": 23751040,
      "step": 36415
    },
    {
      "epoch": 19.08805031446541,
      "grad_norm": 0.7078852653503418,
      "learning_rate": 3.163642846780307e-07,
      "loss": 0.4831,
      "num_input_tokens_seen": 23754976,
      "step": 36420
    },
    {
      "epoch": 19.090670859538783,
      "grad_norm": 0.8429471254348755,
      "learning_rate": 3.1455356755751565e-07,
      "loss": 0.5216,
      "num_input_tokens_seen": 23758464,
      "step": 36425
    },
    {
      "epoch": 19.09329140461216,
      "grad_norm": 0.7562459111213684,
      "learning_rate": 3.127480143507666e-07,
      "loss": 0.4865,
      "num_input_tokens_seen": 23762880,
      "step": 36430
    },
    {
      "epoch": 19.095911949685533,
      "grad_norm": 0.9667474031448364,
      "learning_rate": 3.1094762543548704e-07,
      "loss": 0.5949,
      "num_input_tokens_seen": 23765664,
      "step": 36435
    },
    {
      "epoch": 19.09853249475891,
      "grad_norm": 0.6230937242507935,
      "learning_rate": 3.091524011882951e-07,
      "loss": 0.5936,
      "num_input_tokens_seen": 23770016,
      "step": 36440
    },
    {
      "epoch": 19.101153039832287,
      "grad_norm": 0.33354637026786804,
      "learning_rate": 3.073623419847349e-07,
      "loss": 0.4691,
      "num_input_tokens_seen": 23773408,
      "step": 36445
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 0.34391915798187256,
      "learning_rate": 3.0557744819926235e-07,
      "loss": 0.3873,
      "num_input_tokens_seen": 23776672,
      "step": 36450
    },
    {
      "epoch": 19.106394129979037,
      "grad_norm": 0.4747996926307678,
      "learning_rate": 3.037977202052594e-07,
      "loss": 0.3582,
      "num_input_tokens_seen": 23779264,
      "step": 36455
    },
    {
      "epoch": 19.10901467505241,
      "grad_norm": 0.29389822483062744,
      "learning_rate": 3.020231583750255e-07,
      "loss": 0.4667,
      "num_input_tokens_seen": 23783104,
      "step": 36460
    },
    {
      "epoch": 19.111635220125788,
      "grad_norm": 0.5649950504302979,
      "learning_rate": 3.002537630797747e-07,
      "loss": 0.3727,
      "num_input_tokens_seen": 23785856,
      "step": 36465
    },
    {
      "epoch": 19.11425576519916,
      "grad_norm": 0.7002025246620178,
      "learning_rate": 2.984895346896499e-07,
      "loss": 0.3765,
      "num_input_tokens_seen": 23789344,
      "step": 36470
    },
    {
      "epoch": 19.116876310272538,
      "grad_norm": 0.35928022861480713,
      "learning_rate": 2.967304735737031e-07,
      "loss": 0.4456,
      "num_input_tokens_seen": 23793056,
      "step": 36475
    },
    {
      "epoch": 19.11949685534591,
      "grad_norm": 1.0884746313095093,
      "learning_rate": 2.9497658009991214e-07,
      "loss": 0.5191,
      "num_input_tokens_seen": 23795552,
      "step": 36480
    },
    {
      "epoch": 19.122117400419288,
      "grad_norm": 0.5037884712219238,
      "learning_rate": 2.932278546351724e-07,
      "loss": 0.4049,
      "num_input_tokens_seen": 23797920,
      "step": 36485
    },
    {
      "epoch": 19.12473794549266,
      "grad_norm": 0.42497551441192627,
      "learning_rate": 2.9148429754529137e-07,
      "loss": 0.4991,
      "num_input_tokens_seen": 23800864,
      "step": 36490
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 0.4502081573009491,
      "learning_rate": 2.897459091950078e-07,
      "loss": 0.4242,
      "num_input_tokens_seen": 23804480,
      "step": 36495
    },
    {
      "epoch": 19.129979035639412,
      "grad_norm": 0.6794552206993103,
      "learning_rate": 2.880126899479724e-07,
      "loss": 0.3549,
      "num_input_tokens_seen": 23807296,
      "step": 36500
    },
    {
      "epoch": 19.13259958071279,
      "grad_norm": 0.4674725830554962,
      "learning_rate": 2.862846401667507e-07,
      "loss": 0.34,
      "num_input_tokens_seen": 23810432,
      "step": 36505
    },
    {
      "epoch": 19.135220125786162,
      "grad_norm": 0.7091056108474731,
      "learning_rate": 2.8456176021283697e-07,
      "loss": 0.3589,
      "num_input_tokens_seen": 23813088,
      "step": 36510
    },
    {
      "epoch": 19.13784067085954,
      "grad_norm": 0.6649689078330994,
      "learning_rate": 2.8284405044662897e-07,
      "loss": 0.6559,
      "num_input_tokens_seen": 23815968,
      "step": 36515
    },
    {
      "epoch": 19.140461215932913,
      "grad_norm": 0.6074590086936951,
      "learning_rate": 2.8113151122745874e-07,
      "loss": 0.4468,
      "num_input_tokens_seen": 23819392,
      "step": 36520
    },
    {
      "epoch": 19.14308176100629,
      "grad_norm": 0.6010943055152893,
      "learning_rate": 2.7942414291357023e-07,
      "loss": 0.5651,
      "num_input_tokens_seen": 23822144,
      "step": 36525
    },
    {
      "epoch": 19.145702306079663,
      "grad_norm": 0.35022667050361633,
      "learning_rate": 2.777219458621222e-07,
      "loss": 0.4692,
      "num_input_tokens_seen": 23826144,
      "step": 36530
    },
    {
      "epoch": 19.14832285115304,
      "grad_norm": 1.7953015565872192,
      "learning_rate": 2.7602492042919923e-07,
      "loss": 0.4729,
      "num_input_tokens_seen": 23829536,
      "step": 36535
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 0.3798462450504303,
      "learning_rate": 2.7433306696979523e-07,
      "loss": 0.3178,
      "num_input_tokens_seen": 23833408,
      "step": 36540
    },
    {
      "epoch": 19.15356394129979,
      "grad_norm": 0.6343478560447693,
      "learning_rate": 2.7264638583782973e-07,
      "loss": 0.3575,
      "num_input_tokens_seen": 23835808,
      "step": 36545
    },
    {
      "epoch": 19.156184486373167,
      "grad_norm": 0.437093585729599,
      "learning_rate": 2.7096487738613177e-07,
      "loss": 0.55,
      "num_input_tokens_seen": 23838624,
      "step": 36550
    },
    {
      "epoch": 19.15880503144654,
      "grad_norm": 0.5662854909896851,
      "learning_rate": 2.6928854196646147e-07,
      "loss": 0.3225,
      "num_input_tokens_seen": 23841312,
      "step": 36555
    },
    {
      "epoch": 19.161425576519918,
      "grad_norm": 0.6112024784088135,
      "learning_rate": 2.6761737992948565e-07,
      "loss": 0.4112,
      "num_input_tokens_seen": 23844416,
      "step": 36560
    },
    {
      "epoch": 19.16404612159329,
      "grad_norm": 0.31254103779792786,
      "learning_rate": 2.6595139162479134e-07,
      "loss": 0.3755,
      "num_input_tokens_seen": 23847744,
      "step": 36565
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.8744021058082581,
      "learning_rate": 2.6429057740088303e-07,
      "loss": 0.5487,
      "num_input_tokens_seen": 23850144,
      "step": 36570
    },
    {
      "epoch": 19.16928721174004,
      "grad_norm": 0.38635706901550293,
      "learning_rate": 2.6263493760518564e-07,
      "loss": 0.5928,
      "num_input_tokens_seen": 23853952,
      "step": 36575
    },
    {
      "epoch": 19.171907756813418,
      "grad_norm": 0.5206465125083923,
      "learning_rate": 2.6098447258404157e-07,
      "loss": 0.4532,
      "num_input_tokens_seen": 23857376,
      "step": 36580
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 0.6204423904418945,
      "learning_rate": 2.593391826827107e-07,
      "loss": 0.4595,
      "num_input_tokens_seen": 23861056,
      "step": 36585
    },
    {
      "epoch": 19.17714884696017,
      "grad_norm": 0.45164281129837036,
      "learning_rate": 2.5769906824536225e-07,
      "loss": 0.3402,
      "num_input_tokens_seen": 23863968,
      "step": 36590
    },
    {
      "epoch": 19.179769392033542,
      "grad_norm": 0.3013990819454193,
      "learning_rate": 2.560641296150967e-07,
      "loss": 0.4079,
      "num_input_tokens_seen": 23866912,
      "step": 36595
    },
    {
      "epoch": 19.18238993710692,
      "grad_norm": 0.5061550736427307,
      "learning_rate": 2.544343671339211e-07,
      "loss": 0.3594,
      "num_input_tokens_seen": 23869856,
      "step": 36600
    },
    {
      "epoch": 19.185010482180292,
      "grad_norm": 0.3633125126361847,
      "learning_rate": 2.5280978114276264e-07,
      "loss": 0.4413,
      "num_input_tokens_seen": 23873088,
      "step": 36605
    },
    {
      "epoch": 19.18763102725367,
      "grad_norm": 0.38449448347091675,
      "learning_rate": 2.511903719814662e-07,
      "loss": 0.3807,
      "num_input_tokens_seen": 23876160,
      "step": 36610
    },
    {
      "epoch": 19.190251572327043,
      "grad_norm": 0.591256856918335,
      "learning_rate": 2.4957613998879695e-07,
      "loss": 0.5029,
      "num_input_tokens_seen": 23878880,
      "step": 36615
    },
    {
      "epoch": 19.19287211740042,
      "grad_norm": 0.6061390042304993,
      "learning_rate": 2.479670855024291e-07,
      "loss": 0.3567,
      "num_input_tokens_seen": 23882912,
      "step": 36620
    },
    {
      "epoch": 19.195492662473793,
      "grad_norm": 0.4701060354709625,
      "learning_rate": 2.463632088589629e-07,
      "loss": 0.4124,
      "num_input_tokens_seen": 23886560,
      "step": 36625
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 0.5873157382011414,
      "learning_rate": 2.447645103939078e-07,
      "loss": 0.5505,
      "num_input_tokens_seen": 23889728,
      "step": 36630
    },
    {
      "epoch": 19.200733752620547,
      "grad_norm": 0.40385544300079346,
      "learning_rate": 2.431709904416962e-07,
      "loss": 0.4317,
      "num_input_tokens_seen": 23893408,
      "step": 36635
    },
    {
      "epoch": 19.20335429769392,
      "grad_norm": 0.4765896499156952,
      "learning_rate": 2.4158264933567264e-07,
      "loss": 0.3561,
      "num_input_tokens_seen": 23896960,
      "step": 36640
    },
    {
      "epoch": 19.205974842767297,
      "grad_norm": 0.519435703754425,
      "learning_rate": 2.399994874080963e-07,
      "loss": 0.3826,
      "num_input_tokens_seen": 23899936,
      "step": 36645
    },
    {
      "epoch": 19.20859538784067,
      "grad_norm": 0.6495832800865173,
      "learning_rate": 2.3842150499015504e-07,
      "loss": 0.3809,
      "num_input_tokens_seen": 23903360,
      "step": 36650
    },
    {
      "epoch": 19.211215932914047,
      "grad_norm": 0.9860399961471558,
      "learning_rate": 2.368487024119348e-07,
      "loss": 0.3886,
      "num_input_tokens_seen": 23906784,
      "step": 36655
    },
    {
      "epoch": 19.21383647798742,
      "grad_norm": 0.3596516251564026,
      "learning_rate": 2.3528108000245573e-07,
      "loss": 0.3117,
      "num_input_tokens_seen": 23910240,
      "step": 36660
    },
    {
      "epoch": 19.216457023060798,
      "grad_norm": 0.31619852781295776,
      "learning_rate": 2.3371863808964445e-07,
      "loss": 0.438,
      "num_input_tokens_seen": 23913888,
      "step": 36665
    },
    {
      "epoch": 19.21907756813417,
      "grad_norm": 0.2997482717037201,
      "learning_rate": 2.32161377000345e-07,
      "loss": 0.3718,
      "num_input_tokens_seen": 23917312,
      "step": 36670
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 0.7703347206115723,
      "learning_rate": 2.306092970603163e-07,
      "loss": 0.4489,
      "num_input_tokens_seen": 23920032,
      "step": 36675
    },
    {
      "epoch": 19.22431865828092,
      "grad_norm": 0.287280410528183,
      "learning_rate": 2.2906239859424306e-07,
      "loss": 0.5069,
      "num_input_tokens_seen": 23924000,
      "step": 36680
    },
    {
      "epoch": 19.2269392033543,
      "grad_norm": 0.4699797034263611,
      "learning_rate": 2.2752068192571084e-07,
      "loss": 0.3237,
      "num_input_tokens_seen": 23928128,
      "step": 36685
    },
    {
      "epoch": 19.229559748427672,
      "grad_norm": 0.4338015019893646,
      "learning_rate": 2.2598414737723396e-07,
      "loss": 0.4221,
      "num_input_tokens_seen": 23931712,
      "step": 36690
    },
    {
      "epoch": 19.23218029350105,
      "grad_norm": 0.41238561272621155,
      "learning_rate": 2.2445279527023578e-07,
      "loss": 0.4877,
      "num_input_tokens_seen": 23934912,
      "step": 36695
    },
    {
      "epoch": 19.234800838574422,
      "grad_norm": 0.3383120000362396,
      "learning_rate": 2.2292662592506007e-07,
      "loss": 0.4932,
      "num_input_tokens_seen": 23938560,
      "step": 36700
    },
    {
      "epoch": 19.2374213836478,
      "grad_norm": 0.41214996576309204,
      "learning_rate": 2.2140563966095985e-07,
      "loss": 0.4577,
      "num_input_tokens_seen": 23941440,
      "step": 36705
    },
    {
      "epoch": 19.240041928721173,
      "grad_norm": 0.4668619930744171,
      "learning_rate": 2.1988983679611386e-07,
      "loss": 0.4324,
      "num_input_tokens_seen": 23944736,
      "step": 36710
    },
    {
      "epoch": 19.24266247379455,
      "grad_norm": 0.3645295798778534,
      "learning_rate": 2.1837921764760462e-07,
      "loss": 0.6455,
      "num_input_tokens_seen": 23949376,
      "step": 36715
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 0.31406307220458984,
      "learning_rate": 2.1687378253144043e-07,
      "loss": 0.4898,
      "num_input_tokens_seen": 23952640,
      "step": 36720
    },
    {
      "epoch": 19.2479035639413,
      "grad_norm": 0.5892524719238281,
      "learning_rate": 2.153735317625416e-07,
      "loss": 0.4118,
      "num_input_tokens_seen": 23956992,
      "step": 36725
    },
    {
      "epoch": 19.250524109014677,
      "grad_norm": 0.45248138904571533,
      "learning_rate": 2.1387846565474045e-07,
      "loss": 0.3622,
      "num_input_tokens_seen": 23960192,
      "step": 36730
    },
    {
      "epoch": 19.25314465408805,
      "grad_norm": 0.3524947166442871,
      "learning_rate": 2.123885845207896e-07,
      "loss": 0.3153,
      "num_input_tokens_seen": 23963040,
      "step": 36735
    },
    {
      "epoch": 19.255765199161427,
      "grad_norm": 0.401949405670166,
      "learning_rate": 2.1090388867235356e-07,
      "loss": 0.4536,
      "num_input_tokens_seen": 23965664,
      "step": 36740
    },
    {
      "epoch": 19.2583857442348,
      "grad_norm": 0.5795449614524841,
      "learning_rate": 2.0942437842001728e-07,
      "loss": 0.4232,
      "num_input_tokens_seen": 23968992,
      "step": 36745
    },
    {
      "epoch": 19.261006289308177,
      "grad_norm": 0.3974466025829315,
      "learning_rate": 2.0795005407327483e-07,
      "loss": 0.801,
      "num_input_tokens_seen": 23972032,
      "step": 36750
    },
    {
      "epoch": 19.26362683438155,
      "grad_norm": 0.4835573732852936,
      "learning_rate": 2.0648091594053508e-07,
      "loss": 0.5674,
      "num_input_tokens_seen": 23975040,
      "step": 36755
    },
    {
      "epoch": 19.266247379454928,
      "grad_norm": 0.570563018321991,
      "learning_rate": 2.0501696432913275e-07,
      "loss": 0.449,
      "num_input_tokens_seen": 23977408,
      "step": 36760
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 0.5674493908882141,
      "learning_rate": 2.035581995453034e-07,
      "loss": 0.4585,
      "num_input_tokens_seen": 23980032,
      "step": 36765
    },
    {
      "epoch": 19.271488469601678,
      "grad_norm": 0.4715767204761505,
      "learning_rate": 2.0210462189420577e-07,
      "loss": 0.4089,
      "num_input_tokens_seen": 23982720,
      "step": 36770
    },
    {
      "epoch": 19.27410901467505,
      "grad_norm": 0.7600159049034119,
      "learning_rate": 2.006562316799132e-07,
      "loss": 0.4764,
      "num_input_tokens_seen": 23985984,
      "step": 36775
    },
    {
      "epoch": 19.27672955974843,
      "grad_norm": 0.45235827565193176,
      "learning_rate": 1.992130292054084e-07,
      "loss": 0.5341,
      "num_input_tokens_seen": 23989504,
      "step": 36780
    },
    {
      "epoch": 19.279350104821802,
      "grad_norm": 0.5541817545890808,
      "learning_rate": 1.9777501477259708e-07,
      "loss": 0.4474,
      "num_input_tokens_seen": 23992160,
      "step": 36785
    },
    {
      "epoch": 19.28197064989518,
      "grad_norm": 0.4691372513771057,
      "learning_rate": 1.9634218868229414e-07,
      "loss": 0.5125,
      "num_input_tokens_seen": 23998816,
      "step": 36790
    },
    {
      "epoch": 19.284591194968552,
      "grad_norm": 0.26914364099502563,
      "learning_rate": 1.9491455123423207e-07,
      "loss": 0.428,
      "num_input_tokens_seen": 24003040,
      "step": 36795
    },
    {
      "epoch": 19.28721174004193,
      "grad_norm": 0.3518806993961334,
      "learning_rate": 1.9349210272705253e-07,
      "loss": 0.4748,
      "num_input_tokens_seen": 24006272,
      "step": 36800
    },
    {
      "epoch": 19.289832285115303,
      "grad_norm": 0.48455360531806946,
      "learning_rate": 1.920748434583175e-07,
      "loss": 0.4052,
      "num_input_tokens_seen": 24009792,
      "step": 36805
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 0.5003572106361389,
      "learning_rate": 1.906627737245009e-07,
      "loss": 0.341,
      "num_input_tokens_seen": 24012960,
      "step": 36810
    },
    {
      "epoch": 19.295073375262053,
      "grad_norm": 0.49488934874534607,
      "learning_rate": 1.8925589382099152e-07,
      "loss": 0.5638,
      "num_input_tokens_seen": 24016416,
      "step": 36815
    },
    {
      "epoch": 19.29769392033543,
      "grad_norm": 0.6960572600364685,
      "learning_rate": 1.8785420404209553e-07,
      "loss": 0.3442,
      "num_input_tokens_seen": 24019520,
      "step": 36820
    },
    {
      "epoch": 19.300314465408803,
      "grad_norm": 0.7706698775291443,
      "learning_rate": 1.864577046810284e-07,
      "loss": 0.4207,
      "num_input_tokens_seen": 24023168,
      "step": 36825
    },
    {
      "epoch": 19.30293501048218,
      "grad_norm": 0.45122045278549194,
      "learning_rate": 1.8506639602992038e-07,
      "loss": 0.4488,
      "num_input_tokens_seen": 24026368,
      "step": 36830
    },
    {
      "epoch": 19.305555555555557,
      "grad_norm": 1.1784892082214355,
      "learning_rate": 1.836802783798164e-07,
      "loss": 0.3799,
      "num_input_tokens_seen": 24028960,
      "step": 36835
    },
    {
      "epoch": 19.30817610062893,
      "grad_norm": 0.4272902011871338,
      "learning_rate": 1.822993520206817e-07,
      "loss": 0.514,
      "num_input_tokens_seen": 24031392,
      "step": 36840
    },
    {
      "epoch": 19.310796645702307,
      "grad_norm": 0.5464162230491638,
      "learning_rate": 1.8092361724138528e-07,
      "loss": 0.4397,
      "num_input_tokens_seen": 24034368,
      "step": 36845
    },
    {
      "epoch": 19.31341719077568,
      "grad_norm": 0.2982584536075592,
      "learning_rate": 1.7955307432971914e-07,
      "loss": 0.4662,
      "num_input_tokens_seen": 24037472,
      "step": 36850
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 0.5449378490447998,
      "learning_rate": 1.7818772357238445e-07,
      "loss": 0.4756,
      "num_input_tokens_seen": 24040672,
      "step": 36855
    },
    {
      "epoch": 19.31865828092243,
      "grad_norm": 0.43939271569252014,
      "learning_rate": 1.7682756525499166e-07,
      "loss": 0.4272,
      "num_input_tokens_seen": 24044192,
      "step": 36860
    },
    {
      "epoch": 19.321278825995808,
      "grad_norm": 0.45500075817108154,
      "learning_rate": 1.7547259966207708e-07,
      "loss": 0.4433,
      "num_input_tokens_seen": 24047520,
      "step": 36865
    },
    {
      "epoch": 19.32389937106918,
      "grad_norm": 0.5632569789886475,
      "learning_rate": 1.7412282707708062e-07,
      "loss": 0.5614,
      "num_input_tokens_seen": 24050944,
      "step": 36870
    },
    {
      "epoch": 19.32651991614256,
      "grad_norm": 0.6163976192474365,
      "learning_rate": 1.7277824778235974e-07,
      "loss": 0.4197,
      "num_input_tokens_seen": 24053440,
      "step": 36875
    },
    {
      "epoch": 19.329140461215932,
      "grad_norm": 1.0222883224487305,
      "learning_rate": 1.714388620591867e-07,
      "loss": 0.4496,
      "num_input_tokens_seen": 24056480,
      "step": 36880
    },
    {
      "epoch": 19.33176100628931,
      "grad_norm": 0.5755986571311951,
      "learning_rate": 1.701046701877429e-07,
      "loss": 0.3657,
      "num_input_tokens_seen": 24060768,
      "step": 36885
    },
    {
      "epoch": 19.334381551362682,
      "grad_norm": 0.7212134599685669,
      "learning_rate": 1.6877567244713011e-07,
      "loss": 0.5097,
      "num_input_tokens_seen": 24063200,
      "step": 36890
    },
    {
      "epoch": 19.33700209643606,
      "grad_norm": 0.9813724160194397,
      "learning_rate": 1.6745186911535372e-07,
      "loss": 0.6268,
      "num_input_tokens_seen": 24066272,
      "step": 36895
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 0.44676703214645386,
      "learning_rate": 1.6613326046934497e-07,
      "loss": 0.4213,
      "num_input_tokens_seen": 24068960,
      "step": 36900
    },
    {
      "epoch": 19.34224318658281,
      "grad_norm": 0.46048590540885925,
      "learning_rate": 1.64819846784936e-07,
      "loss": 0.5764,
      "num_input_tokens_seen": 24071936,
      "step": 36905
    },
    {
      "epoch": 19.344863731656183,
      "grad_norm": 1.0051629543304443,
      "learning_rate": 1.6351162833688204e-07,
      "loss": 0.4457,
      "num_input_tokens_seen": 24074752,
      "step": 36910
    },
    {
      "epoch": 19.34748427672956,
      "grad_norm": 0.3734484910964966,
      "learning_rate": 1.6220860539884475e-07,
      "loss": 0.5153,
      "num_input_tokens_seen": 24078240,
      "step": 36915
    },
    {
      "epoch": 19.350104821802937,
      "grad_norm": 0.5701689124107361,
      "learning_rate": 1.6091077824340327e-07,
      "loss": 0.4102,
      "num_input_tokens_seen": 24081920,
      "step": 36920
    },
    {
      "epoch": 19.35272536687631,
      "grad_norm": 0.5333217978477478,
      "learning_rate": 1.5961814714204605e-07,
      "loss": 0.4896,
      "num_input_tokens_seen": 24085952,
      "step": 36925
    },
    {
      "epoch": 19.355345911949687,
      "grad_norm": 0.6508482098579407,
      "learning_rate": 1.5833071236518172e-07,
      "loss": 0.4226,
      "num_input_tokens_seen": 24088672,
      "step": 36930
    },
    {
      "epoch": 19.35796645702306,
      "grad_norm": 0.2947154939174652,
      "learning_rate": 1.5704847418212266e-07,
      "loss": 0.4716,
      "num_input_tokens_seen": 24092064,
      "step": 36935
    },
    {
      "epoch": 19.360587002096437,
      "grad_norm": 0.40894758701324463,
      "learning_rate": 1.5577143286109875e-07,
      "loss": 0.4961,
      "num_input_tokens_seen": 24095328,
      "step": 36940
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 0.5103383660316467,
      "learning_rate": 1.5449958866925185e-07,
      "loss": 0.4929,
      "num_input_tokens_seen": 24098560,
      "step": 36945
    },
    {
      "epoch": 19.365828092243188,
      "grad_norm": 0.4060070216655731,
      "learning_rate": 1.532329418726386e-07,
      "loss": 0.4451,
      "num_input_tokens_seen": 24101696,
      "step": 36950
    },
    {
      "epoch": 19.36844863731656,
      "grad_norm": 0.8762305378913879,
      "learning_rate": 1.5197149273623036e-07,
      "loss": 0.537,
      "num_input_tokens_seen": 24104448,
      "step": 36955
    },
    {
      "epoch": 19.371069182389938,
      "grad_norm": 0.47899118065834045,
      "learning_rate": 1.5071524152389936e-07,
      "loss": 0.44,
      "num_input_tokens_seen": 24107520,
      "step": 36960
    },
    {
      "epoch": 19.37368972746331,
      "grad_norm": 0.8400583863258362,
      "learning_rate": 1.4946418849844934e-07,
      "loss": 0.3841,
      "num_input_tokens_seen": 24110304,
      "step": 36965
    },
    {
      "epoch": 19.37631027253669,
      "grad_norm": 0.49245336651802063,
      "learning_rate": 1.4821833392157923e-07,
      "loss": 0.5214,
      "num_input_tokens_seen": 24113792,
      "step": 36970
    },
    {
      "epoch": 19.378930817610062,
      "grad_norm": 1.265389323234558,
      "learning_rate": 1.4697767805390838e-07,
      "loss": 0.4737,
      "num_input_tokens_seen": 24116224,
      "step": 36975
    },
    {
      "epoch": 19.38155136268344,
      "grad_norm": 0.3315073549747467,
      "learning_rate": 1.4574222115497084e-07,
      "loss": 0.4157,
      "num_input_tokens_seen": 24120480,
      "step": 36980
    },
    {
      "epoch": 19.384171907756812,
      "grad_norm": 1.1613396406173706,
      "learning_rate": 1.4451196348320706e-07,
      "loss": 0.548,
      "num_input_tokens_seen": 24123424,
      "step": 36985
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 0.6709330081939697,
      "learning_rate": 1.4328690529597232e-07,
      "loss": 0.3848,
      "num_input_tokens_seen": 24126528,
      "step": 36990
    },
    {
      "epoch": 19.389412997903563,
      "grad_norm": 0.5574541091918945,
      "learning_rate": 1.4206704684953943e-07,
      "loss": 0.3567,
      "num_input_tokens_seen": 24131360,
      "step": 36995
    },
    {
      "epoch": 19.39203354297694,
      "grad_norm": 0.6487683057785034,
      "learning_rate": 1.4085238839908476e-07,
      "loss": 0.4481,
      "num_input_tokens_seen": 24133920,
      "step": 37000
    },
    {
      "epoch": 19.394654088050313,
      "grad_norm": 0.4627656936645508,
      "learning_rate": 1.3964293019870235e-07,
      "loss": 0.3679,
      "num_input_tokens_seen": 24137536,
      "step": 37005
    },
    {
      "epoch": 19.39727463312369,
      "grad_norm": 0.686931312084198,
      "learning_rate": 1.3843867250139807e-07,
      "loss": 0.4199,
      "num_input_tokens_seen": 24140448,
      "step": 37010
    },
    {
      "epoch": 19.399895178197063,
      "grad_norm": 0.516965389251709,
      "learning_rate": 1.3723961555908715e-07,
      "loss": 0.5105,
      "num_input_tokens_seen": 24143488,
      "step": 37015
    },
    {
      "epoch": 19.40251572327044,
      "grad_norm": 1.5911636352539062,
      "learning_rate": 1.3604575962259947e-07,
      "loss": 0.4219,
      "num_input_tokens_seen": 24146944,
      "step": 37020
    },
    {
      "epoch": 19.405136268343817,
      "grad_norm": 0.30555975437164307,
      "learning_rate": 1.3485710494167692e-07,
      "loss": 0.4293,
      "num_input_tokens_seen": 24150944,
      "step": 37025
    },
    {
      "epoch": 19.40775681341719,
      "grad_norm": 0.9555433392524719,
      "learning_rate": 1.3367365176496783e-07,
      "loss": 0.4568,
      "num_input_tokens_seen": 24153504,
      "step": 37030
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 0.3297436535358429,
      "learning_rate": 1.324954003400436e-07,
      "loss": 0.4209,
      "num_input_tokens_seen": 24156480,
      "step": 37035
    },
    {
      "epoch": 19.41299790356394,
      "grad_norm": 0.5656729340553284,
      "learning_rate": 1.313223509133793e-07,
      "loss": 0.5879,
      "num_input_tokens_seen": 24160192,
      "step": 37040
    },
    {
      "epoch": 19.415618448637318,
      "grad_norm": 0.41203153133392334,
      "learning_rate": 1.301545037303592e-07,
      "loss": 0.5,
      "num_input_tokens_seen": 24162560,
      "step": 37045
    },
    {
      "epoch": 19.41823899371069,
      "grad_norm": 0.36101093888282776,
      "learning_rate": 1.2899185903529066e-07,
      "loss": 0.4222,
      "num_input_tokens_seen": 24165568,
      "step": 37050
    },
    {
      "epoch": 19.420859538784068,
      "grad_norm": 0.37154191732406616,
      "learning_rate": 1.278344170713791e-07,
      "loss": 0.3658,
      "num_input_tokens_seen": 24168608,
      "step": 37055
    },
    {
      "epoch": 19.42348008385744,
      "grad_norm": 0.3561137616634369,
      "learning_rate": 1.2668217808075033e-07,
      "loss": 0.5302,
      "num_input_tokens_seen": 24171552,
      "step": 37060
    },
    {
      "epoch": 19.42610062893082,
      "grad_norm": 0.6682414412498474,
      "learning_rate": 1.255351423044393e-07,
      "loss": 0.4294,
      "num_input_tokens_seen": 24174336,
      "step": 37065
    },
    {
      "epoch": 19.428721174004192,
      "grad_norm": 0.3153277039527893,
      "learning_rate": 1.2439330998239572e-07,
      "loss": 0.3106,
      "num_input_tokens_seen": 24178400,
      "step": 37070
    },
    {
      "epoch": 19.43134171907757,
      "grad_norm": 0.5040766596794128,
      "learning_rate": 1.2325668135347578e-07,
      "loss": 0.3901,
      "num_input_tokens_seen": 24181312,
      "step": 37075
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 0.5399276614189148,
      "learning_rate": 1.221252566554476e-07,
      "loss": 0.4529,
      "num_input_tokens_seen": 24184544,
      "step": 37080
    },
    {
      "epoch": 19.43658280922432,
      "grad_norm": 0.6003967523574829,
      "learning_rate": 1.2099903612499685e-07,
      "loss": 0.3033,
      "num_input_tokens_seen": 24186720,
      "step": 37085
    },
    {
      "epoch": 19.439203354297693,
      "grad_norm": 0.6091405153274536,
      "learning_rate": 1.198780199977101e-07,
      "loss": 0.4925,
      "num_input_tokens_seen": 24189984,
      "step": 37090
    },
    {
      "epoch": 19.44182389937107,
      "grad_norm": 0.5389623641967773,
      "learning_rate": 1.1876220850809416e-07,
      "loss": 0.3939,
      "num_input_tokens_seen": 24192448,
      "step": 37095
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.34851735830307007,
      "learning_rate": 1.1765160188956514e-07,
      "loss": 0.3947,
      "num_input_tokens_seen": 24196128,
      "step": 37100
    },
    {
      "epoch": 19.44706498951782,
      "grad_norm": 0.4595688581466675,
      "learning_rate": 1.1654620037445107e-07,
      "loss": 0.3766,
      "num_input_tokens_seen": 24199104,
      "step": 37105
    },
    {
      "epoch": 19.449685534591197,
      "grad_norm": 0.555652379989624,
      "learning_rate": 1.1544600419398088e-07,
      "loss": 0.5637,
      "num_input_tokens_seen": 24202176,
      "step": 37110
    },
    {
      "epoch": 19.45230607966457,
      "grad_norm": 0.5175102949142456,
      "learning_rate": 1.1435101357831213e-07,
      "loss": 0.5029,
      "num_input_tokens_seen": 24205824,
      "step": 37115
    },
    {
      "epoch": 19.454926624737947,
      "grad_norm": 0.37692785263061523,
      "learning_rate": 1.1326122875650047e-07,
      "loss": 0.54,
      "num_input_tokens_seen": 24208640,
      "step": 37120
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 0.07961906492710114,
      "learning_rate": 1.1217664995651633e-07,
      "loss": 0.2731,
      "num_input_tokens_seen": 24215456,
      "step": 37125
    },
    {
      "epoch": 19.460167714884697,
      "grad_norm": 0.562080442905426,
      "learning_rate": 1.1109727740523935e-07,
      "loss": 0.4409,
      "num_input_tokens_seen": 24218752,
      "step": 37130
    },
    {
      "epoch": 19.46278825995807,
      "grad_norm": 0.4870162308216095,
      "learning_rate": 1.1002311132846944e-07,
      "loss": 0.4153,
      "num_input_tokens_seen": 24221632,
      "step": 37135
    },
    {
      "epoch": 19.465408805031448,
      "grad_norm": 0.6059625148773193,
      "learning_rate": 1.0895415195090186e-07,
      "loss": 0.5238,
      "num_input_tokens_seen": 24225824,
      "step": 37140
    },
    {
      "epoch": 19.46802935010482,
      "grad_norm": 0.633686900138855,
      "learning_rate": 1.0789039949615221e-07,
      "loss": 0.4784,
      "num_input_tokens_seen": 24228960,
      "step": 37145
    },
    {
      "epoch": 19.470649895178198,
      "grad_norm": 0.4113811254501343,
      "learning_rate": 1.0683185418674801e-07,
      "loss": 0.3906,
      "num_input_tokens_seen": 24232352,
      "step": 37150
    },
    {
      "epoch": 19.47327044025157,
      "grad_norm": 0.3622499108314514,
      "learning_rate": 1.0577851624412604e-07,
      "loss": 0.3734,
      "num_input_tokens_seen": 24235456,
      "step": 37155
    },
    {
      "epoch": 19.47589098532495,
      "grad_norm": 0.6446032524108887,
      "learning_rate": 1.0473038588862672e-07,
      "loss": 0.4678,
      "num_input_tokens_seen": 24238016,
      "step": 37160
    },
    {
      "epoch": 19.478511530398322,
      "grad_norm": 0.5835837125778198,
      "learning_rate": 1.0368746333951351e-07,
      "loss": 0.399,
      "num_input_tokens_seen": 24240672,
      "step": 37165
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 0.6273515820503235,
      "learning_rate": 1.0264974881494805e-07,
      "loss": 0.3449,
      "num_input_tokens_seen": 24243520,
      "step": 37170
    },
    {
      "epoch": 19.483752620545072,
      "grad_norm": 0.7860379815101624,
      "learning_rate": 1.0161724253201222e-07,
      "loss": 0.464,
      "num_input_tokens_seen": 24246784,
      "step": 37175
    },
    {
      "epoch": 19.48637316561845,
      "grad_norm": 0.37935322523117065,
      "learning_rate": 1.0058994470669436e-07,
      "loss": 0.6442,
      "num_input_tokens_seen": 24250848,
      "step": 37180
    },
    {
      "epoch": 19.488993710691823,
      "grad_norm": 0.48471730947494507,
      "learning_rate": 9.956785555388926e-08,
      "loss": 0.4591,
      "num_input_tokens_seen": 24254208,
      "step": 37185
    },
    {
      "epoch": 19.4916142557652,
      "grad_norm": 0.6068388819694519,
      "learning_rate": 9.855097528741475e-08,
      "loss": 0.4382,
      "num_input_tokens_seen": 24260896,
      "step": 37190
    },
    {
      "epoch": 19.494234800838573,
      "grad_norm": 3.477729082107544,
      "learning_rate": 9.753930411998402e-08,
      "loss": 0.5293,
      "num_input_tokens_seen": 24263232,
      "step": 37195
    },
    {
      "epoch": 19.49685534591195,
      "grad_norm": 0.4060947895050049,
      "learning_rate": 9.653284226322501e-08,
      "loss": 0.5072,
      "num_input_tokens_seen": 24266784,
      "step": 37200
    },
    {
      "epoch": 19.499475890985323,
      "grad_norm": 0.5075659155845642,
      "learning_rate": 9.553158992768318e-08,
      "loss": 0.3947,
      "num_input_tokens_seen": 24270112,
      "step": 37205
    },
    {
      "epoch": 19.5020964360587,
      "grad_norm": 0.48843902349472046,
      "learning_rate": 9.453554732281044e-08,
      "loss": 0.4979,
      "num_input_tokens_seen": 24273184,
      "step": 37210
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 0.4016445279121399,
      "learning_rate": 9.354471465696236e-08,
      "loss": 0.4719,
      "num_input_tokens_seen": 24276224,
      "step": 37215
    },
    {
      "epoch": 19.50733752620545,
      "grad_norm": 0.3674727976322174,
      "learning_rate": 9.255909213741199e-08,
      "loss": 0.3808,
      "num_input_tokens_seen": 24281024,
      "step": 37220
    },
    {
      "epoch": 19.509958071278827,
      "grad_norm": 0.38430705666542053,
      "learning_rate": 9.157867997034164e-08,
      "loss": 0.5475,
      "num_input_tokens_seen": 24283840,
      "step": 37225
    },
    {
      "epoch": 19.5125786163522,
      "grad_norm": 0.5741512775421143,
      "learning_rate": 9.060347836084004e-08,
      "loss": 0.2606,
      "num_input_tokens_seen": 24286240,
      "step": 37230
    },
    {
      "epoch": 19.515199161425578,
      "grad_norm": 0.5704386830329895,
      "learning_rate": 8.96334875129079e-08,
      "loss": 0.3964,
      "num_input_tokens_seen": 24288992,
      "step": 37235
    },
    {
      "epoch": 19.51781970649895,
      "grad_norm": 0.5330469608306885,
      "learning_rate": 8.866870762946067e-08,
      "loss": 0.4195,
      "num_input_tokens_seen": 24292288,
      "step": 37240
    },
    {
      "epoch": 19.520440251572328,
      "grad_norm": 0.3623182475566864,
      "learning_rate": 8.77091389123147e-08,
      "loss": 0.4941,
      "num_input_tokens_seen": 24295584,
      "step": 37245
    },
    {
      "epoch": 19.5230607966457,
      "grad_norm": 0.3176884651184082,
      "learning_rate": 8.675478156220385e-08,
      "loss": 0.4076,
      "num_input_tokens_seen": 24298400,
      "step": 37250
    },
    {
      "epoch": 19.52568134171908,
      "grad_norm": 0.8256052732467651,
      "learning_rate": 8.580563577876566e-08,
      "loss": 0.3672,
      "num_input_tokens_seen": 24301280,
      "step": 37255
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 0.6460780501365662,
      "learning_rate": 8.486170176055242e-08,
      "loss": 0.4716,
      "num_input_tokens_seen": 24303968,
      "step": 37260
    },
    {
      "epoch": 19.53092243186583,
      "grad_norm": 0.4698036015033722,
      "learning_rate": 8.392297970502282e-08,
      "loss": 0.4324,
      "num_input_tokens_seen": 24307456,
      "step": 37265
    },
    {
      "epoch": 19.533542976939202,
      "grad_norm": 0.682587206363678,
      "learning_rate": 8.298946980855315e-08,
      "loss": 0.4087,
      "num_input_tokens_seen": 24310912,
      "step": 37270
    },
    {
      "epoch": 19.53616352201258,
      "grad_norm": 0.5042396783828735,
      "learning_rate": 8.206117226641497e-08,
      "loss": 0.4092,
      "num_input_tokens_seen": 24313728,
      "step": 37275
    },
    {
      "epoch": 19.538784067085953,
      "grad_norm": 0.5477098226547241,
      "learning_rate": 8.113808727280015e-08,
      "loss": 0.4125,
      "num_input_tokens_seen": 24316928,
      "step": 37280
    },
    {
      "epoch": 19.54140461215933,
      "grad_norm": 1.2243428230285645,
      "learning_rate": 8.022021502080979e-08,
      "loss": 0.4082,
      "num_input_tokens_seen": 24319552,
      "step": 37285
    },
    {
      "epoch": 19.544025157232703,
      "grad_norm": 0.5210872888565063,
      "learning_rate": 7.930755570244863e-08,
      "loss": 0.4697,
      "num_input_tokens_seen": 24321984,
      "step": 37290
    },
    {
      "epoch": 19.54664570230608,
      "grad_norm": 0.5365453958511353,
      "learning_rate": 7.840010950863619e-08,
      "loss": 0.4778,
      "num_input_tokens_seen": 24324928,
      "step": 37295
    },
    {
      "epoch": 19.549266247379457,
      "grad_norm": 0.43530771136283875,
      "learning_rate": 7.749787662920116e-08,
      "loss": 0.3304,
      "num_input_tokens_seen": 24328064,
      "step": 37300
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 1.0537605285644531,
      "learning_rate": 7.660085725288146e-08,
      "loss": 0.5502,
      "num_input_tokens_seen": 24331136,
      "step": 37305
    },
    {
      "epoch": 19.554507337526207,
      "grad_norm": 0.6279534101486206,
      "learning_rate": 7.570905156731867e-08,
      "loss": 0.4505,
      "num_input_tokens_seen": 24334112,
      "step": 37310
    },
    {
      "epoch": 19.55712788259958,
      "grad_norm": 0.23194724321365356,
      "learning_rate": 7.482245975907188e-08,
      "loss": 0.3983,
      "num_input_tokens_seen": 24338048,
      "step": 37315
    },
    {
      "epoch": 19.559748427672957,
      "grad_norm": 0.34704580903053284,
      "learning_rate": 7.394108201360383e-08,
      "loss": 0.543,
      "num_input_tokens_seen": 24341632,
      "step": 37320
    },
    {
      "epoch": 19.56236897274633,
      "grad_norm": 0.41040822863578796,
      "learning_rate": 7.306491851529206e-08,
      "loss": 0.4247,
      "num_input_tokens_seen": 24345088,
      "step": 37325
    },
    {
      "epoch": 19.564989517819708,
      "grad_norm": 0.304351806640625,
      "learning_rate": 7.219396944741774e-08,
      "loss": 0.4504,
      "num_input_tokens_seen": 24348736,
      "step": 37330
    },
    {
      "epoch": 19.56761006289308,
      "grad_norm": 0.3758210241794586,
      "learning_rate": 7.132823499217123e-08,
      "loss": 0.345,
      "num_input_tokens_seen": 24351776,
      "step": 37335
    },
    {
      "epoch": 19.570230607966458,
      "grad_norm": 0.3342866599559784,
      "learning_rate": 7.046771533065766e-08,
      "loss": 0.3697,
      "num_input_tokens_seen": 24355296,
      "step": 37340
    },
    {
      "epoch": 19.57285115303983,
      "grad_norm": 0.36630555987358093,
      "learning_rate": 6.961241064288859e-08,
      "loss": 0.4321,
      "num_input_tokens_seen": 24359168,
      "step": 37345
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 0.5036506056785583,
      "learning_rate": 6.876232110778202e-08,
      "loss": 0.5048,
      "num_input_tokens_seen": 24362720,
      "step": 37350
    },
    {
      "epoch": 19.578092243186582,
      "grad_norm": 0.382807195186615,
      "learning_rate": 6.791744690316515e-08,
      "loss": 0.3854,
      "num_input_tokens_seen": 24365824,
      "step": 37355
    },
    {
      "epoch": 19.58071278825996,
      "grad_norm": 0.43702685832977295,
      "learning_rate": 6.707778820577715e-08,
      "loss": 0.3637,
      "num_input_tokens_seen": 24368928,
      "step": 37360
    },
    {
      "epoch": 19.583333333333332,
      "grad_norm": 0.6153278350830078,
      "learning_rate": 6.624334519126919e-08,
      "loss": 0.4267,
      "num_input_tokens_seen": 24372192,
      "step": 37365
    },
    {
      "epoch": 19.58595387840671,
      "grad_norm": 0.6754017472267151,
      "learning_rate": 6.541411803419329e-08,
      "loss": 0.4676,
      "num_input_tokens_seen": 24374912,
      "step": 37370
    },
    {
      "epoch": 19.588574423480082,
      "grad_norm": 0.4342288374900818,
      "learning_rate": 6.459010690801626e-08,
      "loss": 0.4187,
      "num_input_tokens_seen": 24378176,
      "step": 37375
    },
    {
      "epoch": 19.59119496855346,
      "grad_norm": 0.49679529666900635,
      "learning_rate": 6.377131198510855e-08,
      "loss": 0.3858,
      "num_input_tokens_seen": 24381760,
      "step": 37380
    },
    {
      "epoch": 19.593815513626833,
      "grad_norm": 0.3582403361797333,
      "learning_rate": 6.295773343675537e-08,
      "loss": 0.4084,
      "num_input_tokens_seen": 24386144,
      "step": 37385
    },
    {
      "epoch": 19.59643605870021,
      "grad_norm": 0.2706049382686615,
      "learning_rate": 6.214937143314837e-08,
      "loss": 0.4268,
      "num_input_tokens_seen": 24389600,
      "step": 37390
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 0.8358598351478577,
      "learning_rate": 6.134622614338559e-08,
      "loss": 0.4261,
      "num_input_tokens_seen": 24392160,
      "step": 37395
    },
    {
      "epoch": 19.60167714884696,
      "grad_norm": 0.4456375241279602,
      "learning_rate": 6.05482977354771e-08,
      "loss": 0.4275,
      "num_input_tokens_seen": 24395584,
      "step": 37400
    },
    {
      "epoch": 19.604297693920337,
      "grad_norm": 1.1125199794769287,
      "learning_rate": 5.975558637634215e-08,
      "loss": 0.4038,
      "num_input_tokens_seen": 24397696,
      "step": 37405
    },
    {
      "epoch": 19.60691823899371,
      "grad_norm": 0.3585033416748047,
      "learning_rate": 5.896809223180366e-08,
      "loss": 0.4226,
      "num_input_tokens_seen": 24401696,
      "step": 37410
    },
    {
      "epoch": 19.609538784067087,
      "grad_norm": 0.4592333436012268,
      "learning_rate": 5.8185815466596516e-08,
      "loss": 0.4305,
      "num_input_tokens_seen": 24404416,
      "step": 37415
    },
    {
      "epoch": 19.61215932914046,
      "grad_norm": 0.5817984938621521,
      "learning_rate": 5.7408756244370365e-08,
      "loss": 0.3965,
      "num_input_tokens_seen": 24407008,
      "step": 37420
    },
    {
      "epoch": 19.614779874213838,
      "grad_norm": 0.4024326205253601,
      "learning_rate": 5.663691472766741e-08,
      "loss": 0.3253,
      "num_input_tokens_seen": 24410208,
      "step": 37425
    },
    {
      "epoch": 19.61740041928721,
      "grad_norm": 0.23328787088394165,
      "learning_rate": 5.587029107795572e-08,
      "loss": 0.3069,
      "num_input_tokens_seen": 24415168,
      "step": 37430
    },
    {
      "epoch": 19.620020964360588,
      "grad_norm": 0.39117565751075745,
      "learning_rate": 5.510888545559867e-08,
      "loss": 0.3781,
      "num_input_tokens_seen": 24417984,
      "step": 37435
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 0.6809456944465637,
      "learning_rate": 5.435269801987997e-08,
      "loss": 0.48,
      "num_input_tokens_seen": 24421376,
      "step": 37440
    },
    {
      "epoch": 19.62526205450734,
      "grad_norm": 0.3493563234806061,
      "learning_rate": 5.360172892897863e-08,
      "loss": 0.2882,
      "num_input_tokens_seen": 24424544,
      "step": 37445
    },
    {
      "epoch": 19.627882599580712,
      "grad_norm": 0.425764799118042,
      "learning_rate": 5.285597833999401e-08,
      "loss": 0.3822,
      "num_input_tokens_seen": 24427488,
      "step": 37450
    },
    {
      "epoch": 19.63050314465409,
      "grad_norm": 0.5874269008636475,
      "learning_rate": 5.2115446408926295e-08,
      "loss": 0.4544,
      "num_input_tokens_seen": 24430336,
      "step": 37455
    },
    {
      "epoch": 19.633123689727462,
      "grad_norm": 0.48121845722198486,
      "learning_rate": 5.138013329068769e-08,
      "loss": 0.4268,
      "num_input_tokens_seen": 24435136,
      "step": 37460
    },
    {
      "epoch": 19.63574423480084,
      "grad_norm": 0.9088745713233948,
      "learning_rate": 5.065003913909405e-08,
      "loss": 0.5297,
      "num_input_tokens_seen": 24438848,
      "step": 37465
    },
    {
      "epoch": 19.638364779874212,
      "grad_norm": 0.4572279155254364,
      "learning_rate": 4.9925164106875975e-08,
      "loss": 0.5143,
      "num_input_tokens_seen": 24441248,
      "step": 37470
    },
    {
      "epoch": 19.64098532494759,
      "grad_norm": 0.6239016652107239,
      "learning_rate": 4.9205508345670506e-08,
      "loss": 0.419,
      "num_input_tokens_seen": 24443840,
      "step": 37475
    },
    {
      "epoch": 19.643605870020963,
      "grad_norm": 0.4519958794116974,
      "learning_rate": 4.8491072006015546e-08,
      "loss": 0.4814,
      "num_input_tokens_seen": 24447200,
      "step": 37480
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 1.964296579360962,
      "learning_rate": 4.778185523736933e-08,
      "loss": 0.5949,
      "num_input_tokens_seen": 24450368,
      "step": 37485
    },
    {
      "epoch": 19.648846960167717,
      "grad_norm": 0.4200560450553894,
      "learning_rate": 4.7077858188085414e-08,
      "loss": 0.3281,
      "num_input_tokens_seen": 24452992,
      "step": 37490
    },
    {
      "epoch": 19.65146750524109,
      "grad_norm": 0.6578401923179626,
      "learning_rate": 4.6379081005437644e-08,
      "loss": 0.4985,
      "num_input_tokens_seen": 24455264,
      "step": 37495
    },
    {
      "epoch": 19.654088050314467,
      "grad_norm": 0.5600469708442688,
      "learning_rate": 4.568552383560076e-08,
      "loss": 0.5384,
      "num_input_tokens_seen": 24458304,
      "step": 37500
    },
    {
      "epoch": 19.65670859538784,
      "grad_norm": 0.5319085121154785,
      "learning_rate": 4.4997186823655945e-08,
      "loss": 0.5216,
      "num_input_tokens_seen": 24461824,
      "step": 37505
    },
    {
      "epoch": 19.659329140461217,
      "grad_norm": 0.9782479405403137,
      "learning_rate": 4.431407011360189e-08,
      "loss": 0.4024,
      "num_input_tokens_seen": 24464736,
      "step": 37510
    },
    {
      "epoch": 19.66194968553459,
      "grad_norm": 0.6548820734024048,
      "learning_rate": 4.3636173848332627e-08,
      "loss": 0.3957,
      "num_input_tokens_seen": 24467264,
      "step": 37515
    },
    {
      "epoch": 19.664570230607968,
      "grad_norm": 0.405209481716156,
      "learning_rate": 4.2963498169656946e-08,
      "loss": 0.4884,
      "num_input_tokens_seen": 24470912,
      "step": 37520
    },
    {
      "epoch": 19.66719077568134,
      "grad_norm": 0.3113040626049042,
      "learning_rate": 4.229604321829561e-08,
      "loss": 0.4813,
      "num_input_tokens_seen": 24474304,
      "step": 37525
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 1.0481318235397339,
      "learning_rate": 4.16338091338675e-08,
      "loss": 0.3881,
      "num_input_tokens_seen": 24477280,
      "step": 37530
    },
    {
      "epoch": 19.67243186582809,
      "grad_norm": 0.6360366940498352,
      "learning_rate": 4.097679605490623e-08,
      "loss": 0.417,
      "num_input_tokens_seen": 24479456,
      "step": 37535
    },
    {
      "epoch": 19.67505241090147,
      "grad_norm": 0.7593510150909424,
      "learning_rate": 4.032500411885465e-08,
      "loss": 0.4986,
      "num_input_tokens_seen": 24482624,
      "step": 37540
    },
    {
      "epoch": 19.677672955974842,
      "grad_norm": 0.6972701549530029,
      "learning_rate": 3.967843346205369e-08,
      "loss": 0.3954,
      "num_input_tokens_seen": 24485792,
      "step": 37545
    },
    {
      "epoch": 19.68029350104822,
      "grad_norm": 0.6065839529037476,
      "learning_rate": 3.903708421976182e-08,
      "loss": 0.4143,
      "num_input_tokens_seen": 24488576,
      "step": 37550
    },
    {
      "epoch": 19.682914046121592,
      "grad_norm": 0.7620758414268494,
      "learning_rate": 3.840095652614395e-08,
      "loss": 0.6129,
      "num_input_tokens_seen": 24491136,
      "step": 37555
    },
    {
      "epoch": 19.68553459119497,
      "grad_norm": 0.4322444498538971,
      "learning_rate": 3.777005051426863e-08,
      "loss": 0.6154,
      "num_input_tokens_seen": 24494528,
      "step": 37560
    },
    {
      "epoch": 19.688155136268342,
      "grad_norm": 0.30878034234046936,
      "learning_rate": 3.71443663161164e-08,
      "loss": 0.5289,
      "num_input_tokens_seen": 24498816,
      "step": 37565
    },
    {
      "epoch": 19.69077568134172,
      "grad_norm": 0.5015395283699036,
      "learning_rate": 3.652390406256867e-08,
      "loss": 0.4004,
      "num_input_tokens_seen": 24502368,
      "step": 37570
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 0.5228737592697144,
      "learning_rate": 3.5908663883421624e-08,
      "loss": 0.4316,
      "num_input_tokens_seen": 24505440,
      "step": 37575
    },
    {
      "epoch": 19.69601677148847,
      "grad_norm": 0.3489134609699249,
      "learning_rate": 3.529864590737786e-08,
      "loss": 0.5083,
      "num_input_tokens_seen": 24508512,
      "step": 37580
    },
    {
      "epoch": 19.698637316561843,
      "grad_norm": 0.3379765748977661,
      "learning_rate": 3.4693850262046415e-08,
      "loss": 0.4073,
      "num_input_tokens_seen": 24511360,
      "step": 37585
    },
    {
      "epoch": 19.70125786163522,
      "grad_norm": 0.39136993885040283,
      "learning_rate": 3.409427707394275e-08,
      "loss": 0.4807,
      "num_input_tokens_seen": 24514048,
      "step": 37590
    },
    {
      "epoch": 19.703878406708597,
      "grad_norm": 0.9739487171173096,
      "learning_rate": 3.349992646848599e-08,
      "loss": 0.413,
      "num_input_tokens_seen": 24517088,
      "step": 37595
    },
    {
      "epoch": 19.70649895178197,
      "grad_norm": 1.560955286026001,
      "learning_rate": 3.291079857001556e-08,
      "loss": 0.5403,
      "num_input_tokens_seen": 24519808,
      "step": 37600
    },
    {
      "epoch": 19.709119496855347,
      "grad_norm": 0.7468889355659485,
      "learning_rate": 3.2326893501763436e-08,
      "loss": 0.5125,
      "num_input_tokens_seen": 24524320,
      "step": 37605
    },
    {
      "epoch": 19.71174004192872,
      "grad_norm": 0.44558393955230713,
      "learning_rate": 3.174821138588191e-08,
      "loss": 0.445,
      "num_input_tokens_seen": 24528192,
      "step": 37610
    },
    {
      "epoch": 19.714360587002098,
      "grad_norm": 0.543097734451294,
      "learning_rate": 3.11747523434186e-08,
      "loss": 0.5124,
      "num_input_tokens_seen": 24530880,
      "step": 37615
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 0.38290897011756897,
      "learning_rate": 3.060651649434143e-08,
      "loss": 0.4907,
      "num_input_tokens_seen": 24533952,
      "step": 37620
    },
    {
      "epoch": 19.719601677148848,
      "grad_norm": 0.454085111618042,
      "learning_rate": 3.004350395751088e-08,
      "loss": 0.4248,
      "num_input_tokens_seen": 24537152,
      "step": 37625
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.910185694694519,
      "learning_rate": 2.94857148507105e-08,
      "loss": 0.3216,
      "num_input_tokens_seen": 24539776,
      "step": 37630
    },
    {
      "epoch": 19.7248427672956,
      "grad_norm": 0.31429532170295715,
      "learning_rate": 2.893314929061641e-08,
      "loss": 0.4427,
      "num_input_tokens_seen": 24542944,
      "step": 37635
    },
    {
      "epoch": 19.72746331236897,
      "grad_norm": 0.6323398351669312,
      "learning_rate": 2.8385807392825037e-08,
      "loss": 0.408,
      "num_input_tokens_seen": 24545504,
      "step": 37640
    },
    {
      "epoch": 19.73008385744235,
      "grad_norm": 0.6976080536842346,
      "learning_rate": 2.7843689271830897e-08,
      "loss": 0.4729,
      "num_input_tokens_seen": 24548736,
      "step": 37645
    },
    {
      "epoch": 19.732704402515722,
      "grad_norm": 0.7588242292404175,
      "learning_rate": 2.7306795041040502e-08,
      "loss": 0.6357,
      "num_input_tokens_seen": 24552000,
      "step": 37650
    },
    {
      "epoch": 19.7353249475891,
      "grad_norm": 0.28610509634017944,
      "learning_rate": 2.677512481276401e-08,
      "loss": 0.3916,
      "num_input_tokens_seen": 24555968,
      "step": 37655
    },
    {
      "epoch": 19.737945492662472,
      "grad_norm": 0.3881831765174866,
      "learning_rate": 2.6248678698220785e-08,
      "loss": 0.4809,
      "num_input_tokens_seen": 24559104,
      "step": 37660
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 0.9659132957458496,
      "learning_rate": 2.5727456807542183e-08,
      "loss": 0.4834,
      "num_input_tokens_seen": 24561824,
      "step": 37665
    },
    {
      "epoch": 19.743186582809223,
      "grad_norm": 2.1297049522399902,
      "learning_rate": 2.5211459249754875e-08,
      "loss": 0.4293,
      "num_input_tokens_seen": 24564864,
      "step": 37670
    },
    {
      "epoch": 19.7458071278826,
      "grad_norm": 0.22070178389549255,
      "learning_rate": 2.4700686132803076e-08,
      "loss": 0.3825,
      "num_input_tokens_seen": 24568224,
      "step": 37675
    },
    {
      "epoch": 19.748427672955973,
      "grad_norm": 0.3420773446559906,
      "learning_rate": 2.4195137563534643e-08,
      "loss": 0.4174,
      "num_input_tokens_seen": 24573312,
      "step": 37680
    },
    {
      "epoch": 19.75104821802935,
      "grad_norm": 0.3925844132900238,
      "learning_rate": 2.369481364770387e-08,
      "loss": 0.4099,
      "num_input_tokens_seen": 24577248,
      "step": 37685
    },
    {
      "epoch": 19.753668763102727,
      "grad_norm": 0.3880830407142639,
      "learning_rate": 2.3199714489974266e-08,
      "loss": 0.3998,
      "num_input_tokens_seen": 24580544,
      "step": 37690
    },
    {
      "epoch": 19.7562893081761,
      "grad_norm": 0.44492560625076294,
      "learning_rate": 2.2709840193912978e-08,
      "loss": 0.4648,
      "num_input_tokens_seen": 24582944,
      "step": 37695
    },
    {
      "epoch": 19.758909853249477,
      "grad_norm": 0.39322635531425476,
      "learning_rate": 2.2225190861999146e-08,
      "loss": 0.6128,
      "num_input_tokens_seen": 24585728,
      "step": 37700
    },
    {
      "epoch": 19.76153039832285,
      "grad_norm": 0.4684711992740631,
      "learning_rate": 2.1745766595612783e-08,
      "loss": 0.5245,
      "num_input_tokens_seen": 24588640,
      "step": 37705
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 0.34401431679725647,
      "learning_rate": 2.1271567495045886e-08,
      "loss": 0.4571,
      "num_input_tokens_seen": 24592064,
      "step": 37710
    },
    {
      "epoch": 19.7667714884696,
      "grad_norm": 0.39471909403800964,
      "learning_rate": 2.0802593659491333e-08,
      "loss": 0.3903,
      "num_input_tokens_seen": 24598944,
      "step": 37715
    },
    {
      "epoch": 19.769392033542978,
      "grad_norm": 0.5432056188583374,
      "learning_rate": 2.0338845187059528e-08,
      "loss": 0.4142,
      "num_input_tokens_seen": 24601152,
      "step": 37720
    },
    {
      "epoch": 19.77201257861635,
      "grad_norm": 0.6496204733848572,
      "learning_rate": 1.988032217475899e-08,
      "loss": 0.4357,
      "num_input_tokens_seen": 24603904,
      "step": 37725
    },
    {
      "epoch": 19.77463312368973,
      "grad_norm": 0.593947172164917,
      "learning_rate": 1.942702471850466e-08,
      "loss": 0.4051,
      "num_input_tokens_seen": 24606240,
      "step": 37730
    },
    {
      "epoch": 19.7772536687631,
      "grad_norm": 0.6401872038841248,
      "learning_rate": 1.897895291312346e-08,
      "loss": 0.4532,
      "num_input_tokens_seen": 24609088,
      "step": 37735
    },
    {
      "epoch": 19.77987421383648,
      "grad_norm": 0.43535315990448,
      "learning_rate": 1.8536106852348746e-08,
      "loss": 0.4044,
      "num_input_tokens_seen": 24612320,
      "step": 37740
    },
    {
      "epoch": 19.782494758909852,
      "grad_norm": 0.3365311324596405,
      "learning_rate": 1.8098486628814748e-08,
      "loss": 0.411,
      "num_input_tokens_seen": 24617664,
      "step": 37745
    },
    {
      "epoch": 19.78511530398323,
      "grad_norm": 0.4946593642234802,
      "learning_rate": 1.7666092334067684e-08,
      "loss": 0.4169,
      "num_input_tokens_seen": 24622304,
      "step": 37750
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 0.41339460015296936,
      "learning_rate": 1.7238924058562977e-08,
      "loss": 0.4327,
      "num_input_tokens_seen": 24625088,
      "step": 37755
    },
    {
      "epoch": 19.79035639412998,
      "grad_norm": 0.4236402213573456,
      "learning_rate": 1.6816981891656924e-08,
      "loss": 0.45,
      "num_input_tokens_seen": 24628544,
      "step": 37760
    },
    {
      "epoch": 19.792976939203353,
      "grad_norm": 1.2953873872756958,
      "learning_rate": 1.6400265921615032e-08,
      "loss": 0.6125,
      "num_input_tokens_seen": 24631776,
      "step": 37765
    },
    {
      "epoch": 19.79559748427673,
      "grad_norm": 0.68556809425354,
      "learning_rate": 1.598877623560924e-08,
      "loss": 0.4453,
      "num_input_tokens_seen": 24634976,
      "step": 37770
    },
    {
      "epoch": 19.798218029350103,
      "grad_norm": 0.5565481781959534,
      "learning_rate": 1.5582512919717908e-08,
      "loss": 0.5011,
      "num_input_tokens_seen": 24637696,
      "step": 37775
    },
    {
      "epoch": 19.80083857442348,
      "grad_norm": 0.5783411264419556,
      "learning_rate": 1.5181476058928613e-08,
      "loss": 0.3713,
      "num_input_tokens_seen": 24640864,
      "step": 37780
    },
    {
      "epoch": 19.803459119496857,
      "grad_norm": 0.3301094174385071,
      "learning_rate": 1.4785665737129805e-08,
      "loss": 0.2711,
      "num_input_tokens_seen": 24646784,
      "step": 37785
    },
    {
      "epoch": 19.80607966457023,
      "grad_norm": 0.4152378737926483,
      "learning_rate": 1.4395082037127471e-08,
      "loss": 0.4055,
      "num_input_tokens_seen": 24650048,
      "step": 37790
    },
    {
      "epoch": 19.808700209643607,
      "grad_norm": 0.45641836524009705,
      "learning_rate": 1.4009725040620148e-08,
      "loss": 0.407,
      "num_input_tokens_seen": 24653728,
      "step": 37795
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 0.6344526410102844,
      "learning_rate": 1.3629594828223902e-08,
      "loss": 0.377,
      "num_input_tokens_seen": 24656288,
      "step": 37800
    },
    {
      "epoch": 19.813941299790358,
      "grad_norm": 0.3788401186466217,
      "learning_rate": 1.3254691479458458e-08,
      "loss": 0.3986,
      "num_input_tokens_seen": 24659744,
      "step": 37805
    },
    {
      "epoch": 19.81656184486373,
      "grad_norm": 0.31988537311553955,
      "learning_rate": 1.2885015072747197e-08,
      "loss": 0.5295,
      "num_input_tokens_seen": 24663488,
      "step": 37810
    },
    {
      "epoch": 19.819182389937108,
      "grad_norm": 0.43164998292922974,
      "learning_rate": 1.2520565685419927e-08,
      "loss": 0.3847,
      "num_input_tokens_seen": 24666272,
      "step": 37815
    },
    {
      "epoch": 19.82180293501048,
      "grad_norm": 0.3578822612762451,
      "learning_rate": 1.2161343393721214e-08,
      "loss": 0.2614,
      "num_input_tokens_seen": 24670400,
      "step": 37820
    },
    {
      "epoch": 19.82442348008386,
      "grad_norm": 0.6345581412315369,
      "learning_rate": 1.1807348272790952e-08,
      "loss": 0.426,
      "num_input_tokens_seen": 24673280,
      "step": 37825
    },
    {
      "epoch": 19.82704402515723,
      "grad_norm": 0.49027541279792786,
      "learning_rate": 1.1458580396683794e-08,
      "loss": 0.4929,
      "num_input_tokens_seen": 24676320,
      "step": 37830
    },
    {
      "epoch": 19.82966457023061,
      "grad_norm": 0.7813634276390076,
      "learning_rate": 1.1115039838358044e-08,
      "loss": 0.3194,
      "num_input_tokens_seen": 24680960,
      "step": 37835
    },
    {
      "epoch": 19.832285115303982,
      "grad_norm": 0.45716536045074463,
      "learning_rate": 1.0776726669678439e-08,
      "loss": 0.4665,
      "num_input_tokens_seen": 24685792,
      "step": 37840
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 0.44992053508758545,
      "learning_rate": 1.044364096141337e-08,
      "loss": 0.3895,
      "num_input_tokens_seen": 24688448,
      "step": 37845
    },
    {
      "epoch": 19.837526205450732,
      "grad_norm": 0.3984825015068054,
      "learning_rate": 1.0115782783243211e-08,
      "loss": 0.3479,
      "num_input_tokens_seen": 24691552,
      "step": 37850
    },
    {
      "epoch": 19.84014675052411,
      "grad_norm": 0.703706681728363,
      "learning_rate": 9.793152203751988e-09,
      "loss": 0.3977,
      "num_input_tokens_seen": 24694368,
      "step": 37855
    },
    {
      "epoch": 19.842767295597483,
      "grad_norm": 0.6008557081222534,
      "learning_rate": 9.475749290430158e-09,
      "loss": 0.4847,
      "num_input_tokens_seen": 24696608,
      "step": 37860
    },
    {
      "epoch": 19.84538784067086,
      "grad_norm": 0.5194452404975891,
      "learning_rate": 9.16357410967461e-09,
      "loss": 0.4221,
      "num_input_tokens_seen": 24699872,
      "step": 37865
    },
    {
      "epoch": 19.848008385744233,
      "grad_norm": 0.5456796288490295,
      "learning_rate": 8.856626726788664e-09,
      "loss": 0.3766,
      "num_input_tokens_seen": 24704608,
      "step": 37870
    },
    {
      "epoch": 19.85062893081761,
      "grad_norm": 0.43412265181541443,
      "learning_rate": 8.554907205982066e-09,
      "loss": 0.4594,
      "num_input_tokens_seen": 24708544,
      "step": 37875
    },
    {
      "epoch": 19.853249475890987,
      "grad_norm": 1.0319030284881592,
      "learning_rate": 8.258415610373772e-09,
      "loss": 0.4688,
      "num_input_tokens_seen": 24710880,
      "step": 37880
    },
    {
      "epoch": 19.85587002096436,
      "grad_norm": 0.41108351945877075,
      "learning_rate": 7.967152001980838e-09,
      "loss": 0.378,
      "num_input_tokens_seen": 24715008,
      "step": 37885
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 0.798280656337738,
      "learning_rate": 7.681116441737856e-09,
      "loss": 0.4545,
      "num_input_tokens_seen": 24718240,
      "step": 37890
    },
    {
      "epoch": 19.86111111111111,
      "grad_norm": 0.7841141819953918,
      "learning_rate": 7.4003089894747464e-09,
      "loss": 0.3528,
      "num_input_tokens_seen": 24720800,
      "step": 37895
    },
    {
      "epoch": 19.863731656184488,
      "grad_norm": 0.3889091908931732,
      "learning_rate": 7.12472970393896e-09,
      "loss": 0.4531,
      "num_input_tokens_seen": 24723392,
      "step": 37900
    },
    {
      "epoch": 19.86635220125786,
      "grad_norm": 0.6980143189430237,
      "learning_rate": 6.854378642776049e-09,
      "loss": 0.4664,
      "num_input_tokens_seen": 24725792,
      "step": 37905
    },
    {
      "epoch": 19.868972746331238,
      "grad_norm": 0.7410687804222107,
      "learning_rate": 6.589255862538002e-09,
      "loss": 0.4037,
      "num_input_tokens_seen": 24728672,
      "step": 37910
    },
    {
      "epoch": 19.87159329140461,
      "grad_norm": 0.41266652941703796,
      "learning_rate": 6.32936141869156e-09,
      "loss": 0.4757,
      "num_input_tokens_seen": 24732384,
      "step": 37915
    },
    {
      "epoch": 19.87421383647799,
      "grad_norm": 1.7541497945785522,
      "learning_rate": 6.074695365596017e-09,
      "loss": 0.4111,
      "num_input_tokens_seen": 24735744,
      "step": 37920
    },
    {
      "epoch": 19.87683438155136,
      "grad_norm": 0.7235763072967529,
      "learning_rate": 5.8252577565309775e-09,
      "loss": 0.4357,
      "num_input_tokens_seen": 24738496,
      "step": 37925
    },
    {
      "epoch": 19.87945492662474,
      "grad_norm": 0.3753299117088318,
      "learning_rate": 5.581048643671371e-09,
      "loss": 0.3982,
      "num_input_tokens_seen": 24742080,
      "step": 37930
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 0.5499361753463745,
      "learning_rate": 5.342068078106888e-09,
      "loss": 0.4019,
      "num_input_tokens_seen": 24744512,
      "step": 37935
    },
    {
      "epoch": 19.88469601677149,
      "grad_norm": 0.3549118638038635,
      "learning_rate": 5.1083161098253175e-09,
      "loss": 0.3878,
      "num_input_tokens_seen": 24748224,
      "step": 37940
    },
    {
      "epoch": 19.887316561844862,
      "grad_norm": 0.6132162809371948,
      "learning_rate": 4.87979278772921e-09,
      "loss": 0.5536,
      "num_input_tokens_seen": 24751232,
      "step": 37945
    },
    {
      "epoch": 19.88993710691824,
      "grad_norm": 0.38414159417152405,
      "learning_rate": 4.656498159619216e-09,
      "loss": 0.4185,
      "num_input_tokens_seen": 24754208,
      "step": 37950
    },
    {
      "epoch": 19.892557651991613,
      "grad_norm": 0.5412895679473877,
      "learning_rate": 4.438432272210746e-09,
      "loss": 0.4207,
      "num_input_tokens_seen": 24757440,
      "step": 37955
    },
    {
      "epoch": 19.89517819706499,
      "grad_norm": 0.8652777075767517,
      "learning_rate": 4.225595171114538e-09,
      "loss": 0.4674,
      "num_input_tokens_seen": 24759808,
      "step": 37960
    },
    {
      "epoch": 19.897798742138363,
      "grad_norm": 0.6515485048294067,
      "learning_rate": 4.017986900858861e-09,
      "loss": 0.4335,
      "num_input_tokens_seen": 24761952,
      "step": 37965
    },
    {
      "epoch": 19.90041928721174,
      "grad_norm": 0.2407110333442688,
      "learning_rate": 3.815607504870089e-09,
      "loss": 0.5217,
      "num_input_tokens_seen": 24765280,
      "step": 37970
    },
    {
      "epoch": 19.903039832285117,
      "grad_norm": 0.4605909287929535,
      "learning_rate": 3.618457025486577e-09,
      "loss": 0.4166,
      "num_input_tokens_seen": 24768320,
      "step": 37975
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 1.6404768228530884,
      "learning_rate": 3.426535503947559e-09,
      "loss": 0.4478,
      "num_input_tokens_seen": 24770912,
      "step": 37980
    },
    {
      "epoch": 19.908280922431867,
      "grad_norm": 0.4693704843521118,
      "learning_rate": 3.2398429803986997e-09,
      "loss": 0.5554,
      "num_input_tokens_seen": 24774176,
      "step": 37985
    },
    {
      "epoch": 19.91090146750524,
      "grad_norm": 0.4477767050266266,
      "learning_rate": 3.058379493900421e-09,
      "loss": 0.4922,
      "num_input_tokens_seen": 24777120,
      "step": 37990
    },
    {
      "epoch": 19.913522012578618,
      "grad_norm": 0.6628676652908325,
      "learning_rate": 2.882145082405696e-09,
      "loss": 0.4129,
      "num_input_tokens_seen": 24779712,
      "step": 37995
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 0.5249236822128296,
      "learning_rate": 2.711139782785033e-09,
      "loss": 0.4838,
      "num_input_tokens_seen": 24783200,
      "step": 38000
    },
    {
      "epoch": 19.918763102725368,
      "grad_norm": 0.45470860600471497,
      "learning_rate": 2.5453636308098163e-09,
      "loss": 0.5346,
      "num_input_tokens_seen": 24786592,
      "step": 38005
    },
    {
      "epoch": 19.92138364779874,
      "grad_norm": 0.5089201331138611,
      "learning_rate": 2.3848166611606383e-09,
      "loss": 0.3333,
      "num_input_tokens_seen": 24789664,
      "step": 38010
    },
    {
      "epoch": 19.92400419287212,
      "grad_norm": 0.5279310941696167,
      "learning_rate": 2.2294989074161943e-09,
      "loss": 0.4644,
      "num_input_tokens_seen": 24792288,
      "step": 38015
    },
    {
      "epoch": 19.92662473794549,
      "grad_norm": 0.418610155582428,
      "learning_rate": 2.0794104020727125e-09,
      "loss": 0.3671,
      "num_input_tokens_seen": 24796000,
      "step": 38020
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 0.40173062682151794,
      "learning_rate": 1.9345511765272993e-09,
      "loss": 0.4026,
      "num_input_tokens_seen": 24798976,
      "step": 38025
    },
    {
      "epoch": 19.931865828092242,
      "grad_norm": 0.721526026725769,
      "learning_rate": 1.794921261077942e-09,
      "loss": 0.3942,
      "num_input_tokens_seen": 24801760,
      "step": 38030
    },
    {
      "epoch": 19.93448637316562,
      "grad_norm": 0.4896986782550812,
      "learning_rate": 1.6605206849373833e-09,
      "loss": 0.4506,
      "num_input_tokens_seen": 24805376,
      "step": 38035
    },
    {
      "epoch": 19.937106918238992,
      "grad_norm": 0.5634881854057312,
      "learning_rate": 1.531349476219246e-09,
      "loss": 0.493,
      "num_input_tokens_seen": 24809440,
      "step": 38040
    },
    {
      "epoch": 19.93972746331237,
      "grad_norm": 0.5604363083839417,
      "learning_rate": 1.4074076619463584e-09,
      "loss": 0.4503,
      "num_input_tokens_seen": 24813472,
      "step": 38045
    },
    {
      "epoch": 19.942348008385743,
      "grad_norm": 0.5153073072433472,
      "learning_rate": 1.2886952680452036e-09,
      "loss": 0.48,
      "num_input_tokens_seen": 24816608,
      "step": 38050
    },
    {
      "epoch": 19.94496855345912,
      "grad_norm": 0.6625958681106567,
      "learning_rate": 1.1752123193459197e-09,
      "loss": 0.4982,
      "num_input_tokens_seen": 24819776,
      "step": 38055
    },
    {
      "epoch": 19.947589098532493,
      "grad_norm": 0.6305820345878601,
      "learning_rate": 1.0669588395934017e-09,
      "loss": 0.397,
      "num_input_tokens_seen": 24822560,
      "step": 38060
    },
    {
      "epoch": 19.95020964360587,
      "grad_norm": 0.47705158591270447,
      "learning_rate": 9.639348514278722e-10,
      "loss": 0.3019,
      "num_input_tokens_seen": 24825600,
      "step": 38065
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 0.5407901406288147,
      "learning_rate": 8.661403764043119e-10,
      "loss": 0.4735,
      "num_input_tokens_seen": 24828224,
      "step": 38070
    },
    {
      "epoch": 19.95545073375262,
      "grad_norm": 0.4643896818161011,
      "learning_rate": 7.735754349785795e-10,
      "loss": 0.3975,
      "num_input_tokens_seen": 24830848,
      "step": 38075
    },
    {
      "epoch": 19.958071278825997,
      "grad_norm": 0.5957532525062561,
      "learning_rate": 6.862400465157403e-10,
      "loss": 0.4249,
      "num_input_tokens_seen": 24834304,
      "step": 38080
    },
    {
      "epoch": 19.96069182389937,
      "grad_norm": 0.4173380434513092,
      "learning_rate": 6.041342292817387e-10,
      "loss": 0.3288,
      "num_input_tokens_seen": 24838112,
      "step": 38085
    },
    {
      "epoch": 19.963312368972748,
      "grad_norm": 0.5695293545722961,
      "learning_rate": 5.272580004572758e-10,
      "loss": 0.4167,
      "num_input_tokens_seen": 24842240,
      "step": 38090
    },
    {
      "epoch": 19.96593291404612,
      "grad_norm": 0.39933130145072937,
      "learning_rate": 4.556113761211567e-10,
      "loss": 0.3549,
      "num_input_tokens_seen": 24845824,
      "step": 38095
    },
    {
      "epoch": 19.968553459119498,
      "grad_norm": 0.5239704847335815,
      "learning_rate": 3.8919437125861657e-10,
      "loss": 0.515,
      "num_input_tokens_seen": 24848704,
      "step": 38100
    },
    {
      "epoch": 19.97117400419287,
      "grad_norm": 0.49125316739082336,
      "learning_rate": 3.280069997696478e-10,
      "loss": 0.2932,
      "num_input_tokens_seen": 24853120,
      "step": 38105
    },
    {
      "epoch": 19.97379454926625,
      "grad_norm": 0.3826046586036682,
      "learning_rate": 2.720492744467951e-10,
      "loss": 0.4956,
      "num_input_tokens_seen": 24856800,
      "step": 38110
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 0.5678246021270752,
      "learning_rate": 2.2132120700291136e-10,
      "loss": 0.4984,
      "num_input_tokens_seen": 24859392,
      "step": 38115
    },
    {
      "epoch": 19.979035639413,
      "grad_norm": 0.43351617455482483,
      "learning_rate": 1.7582280804340213e-10,
      "loss": 0.4418,
      "num_input_tokens_seen": 24863328,
      "step": 38120
    },
    {
      "epoch": 19.981656184486372,
      "grad_norm": 0.34245535731315613,
      "learning_rate": 1.3555408708842977e-10,
      "loss": 0.3971,
      "num_input_tokens_seen": 24866496,
      "step": 38125
    },
    {
      "epoch": 19.98427672955975,
      "grad_norm": 0.39432668685913086,
      "learning_rate": 1.0051505256458704e-10,
      "loss": 0.3763,
      "num_input_tokens_seen": 24871328,
      "step": 38130
    },
    {
      "epoch": 19.986897274633122,
      "grad_norm": 0.5288941860198975,
      "learning_rate": 7.070571179657037e-11,
      "loss": 0.3732,
      "num_input_tokens_seen": 24874336,
      "step": 38135
    },
    {
      "epoch": 19.9895178197065,
      "grad_norm": 0.5768799185752869,
      "learning_rate": 4.612607102383315e-11,
      "loss": 0.4185,
      "num_input_tokens_seen": 24877024,
      "step": 38140
    },
    {
      "epoch": 19.992138364779873,
      "grad_norm": 1.2447319030761719,
      "learning_rate": 2.677613538670798e-11,
      "loss": 0.4707,
      "num_input_tokens_seen": 24880096,
      "step": 38145
    },
    {
      "epoch": 19.99475890985325,
      "grad_norm": 0.5109580159187317,
      "learning_rate": 1.2655908931957782e-11,
      "loss": 0.4536,
      "num_input_tokens_seen": 24883200,
      "step": 38150
    },
    {
      "epoch": 19.997379454926623,
      "grad_norm": 0.8021366596221924,
      "learning_rate": 3.765394615551366e-12,
      "loss": 0.4497,
      "num_input_tokens_seen": 24885664,
      "step": 38155
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0754106193780899,
      "learning_rate": 1.0459429711229262e-13,
      "loss": 0.3394,
      "num_input_tokens_seen": 24887720,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.45270562171936035,
      "eval_runtime": 9.7874,
      "eval_samples_per_second": 86.642,
      "eval_steps_per_second": 21.661,
      "num_input_tokens_seen": 24887720,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 24887720,
      "step": 38160,
      "total_flos": 1.1207225397864038e+18,
      "train_loss": 0.4843763030143904,
      "train_runtime": 6609.4499,
      "train_samples_per_second": 23.085,
      "train_steps_per_second": 5.774
    }
  ],
  "logging_steps": 5,
  "max_steps": 38160,
  "num_input_tokens_seen": 24887720,
  "num_train_epochs": 20,
  "save_steps": 1908,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1207225397864038e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}