{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 534,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01876172607879925,
      "grad_norm": 1.1220635175704956,
      "learning_rate": 1.791044776119403e-06,
      "loss": 1.3297,
      "step": 5
    },
    {
      "epoch": 0.0375234521575985,
      "grad_norm": 0.7967448234558105,
      "learning_rate": 4.029850746268657e-06,
      "loss": 1.3423,
      "step": 10
    },
    {
      "epoch": 0.05628517823639775,
      "grad_norm": 0.6561905145645142,
      "learning_rate": 6.268656716417911e-06,
      "loss": 1.3378,
      "step": 15
    },
    {
      "epoch": 0.075046904315197,
      "grad_norm": 0.5100926160812378,
      "learning_rate": 8.507462686567164e-06,
      "loss": 1.2788,
      "step": 20
    },
    {
      "epoch": 0.09380863039399624,
      "grad_norm": 0.5022882223129272,
      "learning_rate": 1.0746268656716418e-05,
      "loss": 1.2822,
      "step": 25
    },
    {
      "epoch": 0.1125703564727955,
      "grad_norm": 0.5697208642959595,
      "learning_rate": 1.2985074626865672e-05,
      "loss": 1.2556,
      "step": 30
    },
    {
      "epoch": 0.13133208255159476,
      "grad_norm": 0.5847183465957642,
      "learning_rate": 1.5223880597014927e-05,
      "loss": 1.2403,
      "step": 35
    },
    {
      "epoch": 0.150093808630394,
      "grad_norm": 0.47744235396385193,
      "learning_rate": 1.746268656716418e-05,
      "loss": 1.2096,
      "step": 40
    },
    {
      "epoch": 0.16885553470919323,
      "grad_norm": 0.47178810834884644,
      "learning_rate": 1.9701492537313435e-05,
      "loss": 1.2533,
      "step": 45
    },
    {
      "epoch": 0.18761726078799248,
      "grad_norm": 0.44852447509765625,
      "learning_rate": 2.194029850746269e-05,
      "loss": 1.2555,
      "step": 50
    },
    {
      "epoch": 0.20637898686679174,
      "grad_norm": 0.44277751445770264,
      "learning_rate": 2.417910447761194e-05,
      "loss": 1.1888,
      "step": 55
    },
    {
      "epoch": 0.225140712945591,
      "grad_norm": 0.4661575257778168,
      "learning_rate": 2.6417910447761193e-05,
      "loss": 1.2168,
      "step": 60
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.5210767984390259,
      "learning_rate": 2.8656716417910447e-05,
      "loss": 1.1508,
      "step": 65
    },
    {
      "epoch": 0.2626641651031895,
      "grad_norm": 0.5324119925498962,
      "learning_rate": 2.9999815845810124e-05,
      "loss": 1.1554,
      "step": 70
    },
    {
      "epoch": 0.28142589118198874,
      "grad_norm": 0.5865416526794434,
      "learning_rate": 2.9997744163102546e-05,
      "loss": 1.1346,
      "step": 75
    },
    {
      "epoch": 0.300187617260788,
      "grad_norm": 0.5545316338539124,
      "learning_rate": 2.9993370923930618e-05,
      "loss": 1.1047,
      "step": 80
    },
    {
      "epoch": 0.31894934333958724,
      "grad_norm": 0.6227930784225464,
      "learning_rate": 2.9986696799412394e-05,
      "loss": 1.1855,
      "step": 85
    },
    {
      "epoch": 0.33771106941838647,
      "grad_norm": 0.5252574682235718,
      "learning_rate": 2.9977722813760195e-05,
      "loss": 1.1261,
      "step": 90
    },
    {
      "epoch": 0.35647279549718575,
      "grad_norm": 0.6158416867256165,
      "learning_rate": 2.9966450344123444e-05,
      "loss": 1.1142,
      "step": 95
    },
    {
      "epoch": 0.37523452157598497,
      "grad_norm": 0.7427587509155273,
      "learning_rate": 2.9952881120377314e-05,
      "loss": 1.0015,
      "step": 100
    },
    {
      "epoch": 0.39399624765478425,
      "grad_norm": 0.6637995839118958,
      "learning_rate": 2.993701722485728e-05,
      "loss": 1.0785,
      "step": 105
    },
    {
      "epoch": 0.41275797373358347,
      "grad_norm": 0.7029932737350464,
      "learning_rate": 2.9918861092039548e-05,
      "loss": 1.057,
      "step": 110
    },
    {
      "epoch": 0.43151969981238275,
      "grad_norm": 0.6611917018890381,
      "learning_rate": 2.9898415508167465e-05,
      "loss": 1.0607,
      "step": 115
    },
    {
      "epoch": 0.450281425891182,
      "grad_norm": 0.539310872554779,
      "learning_rate": 2.987568361082392e-05,
      "loss": 1.0257,
      "step": 120
    },
    {
      "epoch": 0.46904315196998125,
      "grad_norm": 0.6124554872512817,
      "learning_rate": 2.9850668888449912e-05,
      "loss": 0.9757,
      "step": 125
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.6248557567596436,
      "learning_rate": 2.9823375179809135e-05,
      "loss": 1.0127,
      "step": 130
    },
    {
      "epoch": 0.5065666041275797,
      "grad_norm": 0.993503987789154,
      "learning_rate": 2.9793806673398934e-05,
      "loss": 1.0222,
      "step": 135
    },
    {
      "epoch": 0.525328330206379,
      "grad_norm": 0.7925176620483398,
      "learning_rate": 2.9761967906807514e-05,
      "loss": 0.9856,
      "step": 140
    },
    {
      "epoch": 0.5440900562851783,
      "grad_norm": 0.7589050531387329,
      "learning_rate": 2.9727863766017616e-05,
      "loss": 0.9949,
      "step": 145
    },
    {
      "epoch": 0.5628517823639775,
      "grad_norm": 0.7123038172721863,
      "learning_rate": 2.9691499484656697e-05,
      "loss": 0.939,
      "step": 150
    },
    {
      "epoch": 0.5816135084427767,
      "grad_norm": 0.8624748587608337,
      "learning_rate": 2.9652880643193788e-05,
      "loss": 0.9972,
      "step": 155
    },
    {
      "epoch": 0.600375234521576,
      "grad_norm": 0.732760488986969,
      "learning_rate": 2.9612013168083113e-05,
      "loss": 1.005,
      "step": 160
    },
    {
      "epoch": 0.6191369606003753,
      "grad_norm": 0.820073127746582,
      "learning_rate": 2.9568903330854612e-05,
      "loss": 0.9217,
      "step": 165
    },
    {
      "epoch": 0.6378986866791745,
      "grad_norm": 0.8361268043518066,
      "learning_rate": 2.9523557747151505e-05,
      "loss": 0.8491,
      "step": 170
    },
    {
      "epoch": 0.6566604127579737,
      "grad_norm": 0.8558568954467773,
      "learning_rate": 2.9475983375715068e-05,
      "loss": 0.8825,
      "step": 175
    },
    {
      "epoch": 0.6754221388367729,
      "grad_norm": 0.8939546942710876,
      "learning_rate": 2.9426187517316723e-05,
      "loss": 0.8438,
      "step": 180
    },
    {
      "epoch": 0.6941838649155723,
      "grad_norm": 0.8571194410324097,
      "learning_rate": 2.93741778136377e-05,
      "loss": 0.8708,
      "step": 185
    },
    {
      "epoch": 0.7129455909943715,
      "grad_norm": 0.906139075756073,
      "learning_rate": 2.931996224609628e-05,
      "loss": 0.8978,
      "step": 190
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 1.2007263898849487,
      "learning_rate": 2.926354913462303e-05,
      "loss": 0.8055,
      "step": 195
    },
    {
      "epoch": 0.7504690431519699,
      "grad_norm": 1.034984827041626,
      "learning_rate": 2.9204947136383984e-05,
      "loss": 0.8532,
      "step": 200
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.9235329031944275,
      "learning_rate": 2.9144165244452125e-05,
      "loss": 0.7931,
      "step": 205
    },
    {
      "epoch": 0.7879924953095685,
      "grad_norm": 0.8924856781959534,
      "learning_rate": 2.9081212786427314e-05,
      "loss": 0.8213,
      "step": 210
    },
    {
      "epoch": 0.8067542213883677,
      "grad_norm": 0.9607811570167542,
      "learning_rate": 2.9016099423004854e-05,
      "loss": 0.801,
      "step": 215
    },
    {
      "epoch": 0.8255159474671669,
      "grad_norm": 1.1123484373092651,
      "learning_rate": 2.8948835146492985e-05,
      "loss": 0.8307,
      "step": 220
    },
    {
      "epoch": 0.8442776735459663,
      "grad_norm": 0.9834749698638916,
      "learning_rate": 2.8879430279279443e-05,
      "loss": 0.7948,
      "step": 225
    },
    {
      "epoch": 0.8630393996247655,
      "grad_norm": 1.0126004219055176,
      "learning_rate": 2.8807895472247393e-05,
      "loss": 0.8065,
      "step": 230
    },
    {
      "epoch": 0.8818011257035647,
      "grad_norm": 1.1847103834152222,
      "learning_rate": 2.873424170314093e-05,
      "loss": 0.7728,
      "step": 235
    },
    {
      "epoch": 0.900562851782364,
      "grad_norm": 0.9110663533210754,
      "learning_rate": 2.8658480274880462e-05,
      "loss": 0.7374,
      "step": 240
    },
    {
      "epoch": 0.9193245778611632,
      "grad_norm": 1.0707428455352783,
      "learning_rate": 2.8580622813828123e-05,
      "loss": 0.7819,
      "step": 245
    },
    {
      "epoch": 0.9380863039399625,
      "grad_norm": 1.02041494846344,
      "learning_rate": 2.8500681268003624e-05,
      "loss": 0.8076,
      "step": 250
    },
    {
      "epoch": 0.9568480300187617,
      "grad_norm": 1.0658625364303589,
      "learning_rate": 2.8418667905250683e-05,
      "loss": 0.7103,
      "step": 255
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.9454028606414795,
      "learning_rate": 2.833459531135441e-05,
      "loss": 0.7534,
      "step": 260
    },
    {
      "epoch": 0.9943714821763602,
      "grad_norm": 1.212715744972229,
      "learning_rate": 2.8248476388109897e-05,
      "loss": 0.7299,
      "step": 265
    },
    {
      "epoch": 1.0112570356472796,
      "grad_norm": 1.012231469154358,
      "learning_rate": 2.8160324351342288e-05,
      "loss": 0.6632,
      "step": 270
    },
    {
      "epoch": 1.0300187617260788,
      "grad_norm": 1.054259181022644,
      "learning_rate": 2.80701527288787e-05,
      "loss": 0.6126,
      "step": 275
    },
    {
      "epoch": 1.048780487804878,
      "grad_norm": 1.0729856491088867,
      "learning_rate": 2.797797535847221e-05,
      "loss": 0.6336,
      "step": 280
    },
    {
      "epoch": 1.0675422138836772,
      "grad_norm": 1.1964389085769653,
      "learning_rate": 2.788380638567835e-05,
      "loss": 0.663,
      "step": 285
    },
    {
      "epoch": 1.0863039399624765,
      "grad_norm": 1.1786532402038574,
      "learning_rate": 2.778766026168428e-05,
      "loss": 0.5962,
      "step": 290
    },
    {
      "epoch": 1.1050656660412759,
      "grad_norm": 1.171257495880127,
      "learning_rate": 2.768955174109115e-05,
      "loss": 0.5722,
      "step": 295
    },
    {
      "epoch": 1.123827392120075,
      "grad_norm": 1.0211284160614014,
      "learning_rate": 2.758949587964983e-05,
      "loss": 0.5923,
      "step": 300
    },
    {
      "epoch": 1.1425891181988743,
      "grad_norm": 1.18953275680542,
      "learning_rate": 2.748750803195047e-05,
      "loss": 0.597,
      "step": 305
    },
    {
      "epoch": 1.1613508442776737,
      "grad_norm": 1.2630354166030884,
      "learning_rate": 2.7383603849066167e-05,
      "loss": 0.5956,
      "step": 310
    },
    {
      "epoch": 1.1801125703564728,
      "grad_norm": 1.1608085632324219,
      "learning_rate": 2.7277799276151137e-05,
      "loss": 0.6085,
      "step": 315
    },
    {
      "epoch": 1.198874296435272,
      "grad_norm": 1.1372995376586914,
      "learning_rate": 2.7170110549993817e-05,
      "loss": 0.5848,
      "step": 320
    },
    {
      "epoch": 1.2176360225140712,
      "grad_norm": 1.081286072731018,
      "learning_rate": 2.7060554196525105e-05,
      "loss": 0.5666,
      "step": 325
    },
    {
      "epoch": 1.2363977485928705,
      "grad_norm": 1.418300747871399,
      "learning_rate": 2.694914702828233e-05,
      "loss": 0.5474,
      "step": 330
    },
    {
      "epoch": 1.2551594746716699,
      "grad_norm": 1.1424428224563599,
      "learning_rate": 2.6835906141829173e-05,
      "loss": 0.5782,
      "step": 335
    },
    {
      "epoch": 1.273921200750469,
      "grad_norm": 1.208847165107727,
      "learning_rate": 2.6720848915132053e-05,
      "loss": 0.5132,
      "step": 340
    },
    {
      "epoch": 1.2926829268292683,
      "grad_norm": 1.0693706274032593,
      "learning_rate": 2.660399300489326e-05,
      "loss": 0.5607,
      "step": 345
    },
    {
      "epoch": 1.3114446529080674,
      "grad_norm": 1.2664192914962769,
      "learning_rate": 2.6485356343841395e-05,
      "loss": 0.5611,
      "step": 350
    },
    {
      "epoch": 1.3302063789868668,
      "grad_norm": 1.3361153602600098,
      "learning_rate": 2.636495713797938e-05,
      "loss": 0.5923,
      "step": 355
    },
    {
      "epoch": 1.3489681050656661,
      "grad_norm": 1.1549376249313354,
      "learning_rate": 2.624281386379059e-05,
      "loss": 0.5911,
      "step": 360
    },
    {
      "epoch": 1.3677298311444652,
      "grad_norm": 1.3522552251815796,
      "learning_rate": 2.6118945265403413e-05,
      "loss": 0.5859,
      "step": 365
    },
    {
      "epoch": 1.3864915572232646,
      "grad_norm": 1.36016047000885,
      "learning_rate": 2.599337035171481e-05,
      "loss": 0.5155,
      "step": 370
    },
    {
      "epoch": 1.4052532833020637,
      "grad_norm": 1.268315076828003,
      "learning_rate": 2.5866108393473182e-05,
      "loss": 0.4985,
      "step": 375
    },
    {
      "epoch": 1.424015009380863,
      "grad_norm": 1.075866937637329,
      "learning_rate": 2.573717892032109e-05,
      "loss": 0.4856,
      "step": 380
    },
    {
      "epoch": 1.4427767354596623,
      "grad_norm": 1.2592428922653198,
      "learning_rate": 2.5606601717798212e-05,
      "loss": 0.5281,
      "step": 385
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 1.2148767709732056,
      "learning_rate": 2.547439682430509e-05,
      "loss": 0.5133,
      "step": 390
    },
    {
      "epoch": 1.4803001876172608,
      "grad_norm": 1.2412337064743042,
      "learning_rate": 2.534058452802798e-05,
      "loss": 0.5065,
      "step": 395
    },
    {
      "epoch": 1.49906191369606,
      "grad_norm": 1.2149518728256226,
      "learning_rate": 2.5205185363825482e-05,
      "loss": 0.4689,
      "step": 400
    },
    {
      "epoch": 1.5178236397748592,
      "grad_norm": 1.579555869102478,
      "learning_rate": 2.506822011007721e-05,
      "loss": 0.5183,
      "step": 405
    },
    {
      "epoch": 1.5365853658536586,
      "grad_norm": 1.134460687637329,
      "learning_rate": 2.492970978549515e-05,
      "loss": 0.5026,
      "step": 410
    },
    {
      "epoch": 1.555347091932458,
      "grad_norm": 1.2218743562698364,
      "learning_rate": 2.478967564589816e-05,
      "loss": 0.5281,
      "step": 415
    },
    {
      "epoch": 1.574108818011257,
      "grad_norm": 1.143458366394043,
      "learning_rate": 2.464813918094999e-05,
      "loss": 0.4916,
      "step": 420
    },
    {
      "epoch": 1.5928705440900561,
      "grad_norm": 1.1755683422088623,
      "learning_rate": 2.450512211086151e-05,
      "loss": 0.5052,
      "step": 425
    },
    {
      "epoch": 1.6116322701688555,
      "grad_norm": 1.4553024768829346,
      "learning_rate": 2.4360646383057522e-05,
      "loss": 0.4152,
      "step": 430
    },
    {
      "epoch": 1.6303939962476548,
      "grad_norm": 1.1782385110855103,
      "learning_rate": 2.4214734168808693e-05,
      "loss": 0.4832,
      "step": 435
    },
    {
      "epoch": 1.6491557223264541,
      "grad_norm": 1.4304338693618774,
      "learning_rate": 2.4067407859829157e-05,
      "loss": 0.459,
      "step": 440
    },
    {
      "epoch": 1.6679174484052532,
      "grad_norm": 1.1110352277755737,
      "learning_rate": 2.3918690064840282e-05,
      "loss": 0.464,
      "step": 445
    },
    {
      "epoch": 1.6866791744840526,
      "grad_norm": 1.200399398803711,
      "learning_rate": 2.37686036061011e-05,
      "loss": 0.4133,
      "step": 450
    },
    {
      "epoch": 1.7054409005628517,
      "grad_norm": 1.1901819705963135,
      "learning_rate": 2.361717151590605e-05,
      "loss": 0.4412,
      "step": 455
    },
    {
      "epoch": 1.724202626641651,
      "grad_norm": 1.2039285898208618,
      "learning_rate": 2.3464417033050394e-05,
      "loss": 0.4763,
      "step": 460
    },
    {
      "epoch": 1.7429643527204504,
      "grad_norm": 1.3061219453811646,
      "learning_rate": 2.331036359926402e-05,
      "loss": 0.4514,
      "step": 465
    },
    {
      "epoch": 1.7617260787992497,
      "grad_norm": 1.2160491943359375,
      "learning_rate": 2.315503485561404e-05,
      "loss": 0.4169,
      "step": 470
    },
    {
      "epoch": 1.7804878048780488,
      "grad_norm": 1.1981985569000244,
      "learning_rate": 2.299845463887685e-05,
      "loss": 0.4207,
      "step": 475
    },
    {
      "epoch": 1.799249530956848,
      "grad_norm": 1.164445400238037,
      "learning_rate": 2.2840646977880132e-05,
      "loss": 0.4327,
      "step": 480
    },
    {
      "epoch": 1.8180112570356473,
      "grad_norm": 1.268277883529663,
      "learning_rate": 2.268163608981536e-05,
      "loss": 0.434,
      "step": 485
    },
    {
      "epoch": 1.8367729831144466,
      "grad_norm": 1.2116129398345947,
      "learning_rate": 2.2521446376521476e-05,
      "loss": 0.396,
      "step": 490
    },
    {
      "epoch": 1.855534709193246,
      "grad_norm": 1.208999752998352,
      "learning_rate": 2.236010242074013e-05,
      "loss": 0.472,
      "step": 495
    },
    {
      "epoch": 1.874296435272045,
      "grad_norm": 1.2668207883834839,
      "learning_rate": 2.2197628982343224e-05,
      "loss": 0.3733,
      "step": 500
    },
    {
      "epoch": 1.8930581613508441,
      "grad_norm": 1.347996711730957,
      "learning_rate": 2.2034050994533275e-05,
      "loss": 0.407,
      "step": 505
    },
    {
      "epoch": 1.9118198874296435,
      "grad_norm": 1.42654287815094,
      "learning_rate": 2.1869393560017143e-05,
      "loss": 0.3818,
      "step": 510
    },
    {
      "epoch": 1.9305816135084428,
      "grad_norm": 1.1722033023834229,
      "learning_rate": 2.1703681947153768e-05,
      "loss": 0.4143,
      "step": 515
    },
    {
      "epoch": 1.9493433395872422,
      "grad_norm": 1.4157124757766724,
      "learning_rate": 2.153694158607648e-05,
      "loss": 0.4325,
      "step": 520
    },
    {
      "epoch": 1.9681050656660413,
      "grad_norm": 1.2454453706741333,
      "learning_rate": 2.136919806479051e-05,
      "loss": 0.3982,
      "step": 525
    },
    {
      "epoch": 1.9868667917448404,
      "grad_norm": 1.444549798965454,
      "learning_rate": 2.1200477125246203e-05,
      "loss": 0.3563,
      "step": 530
    }
  ],
  "logging_steps": 5,
  "max_steps": 1335,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.513490532954276e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}