{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 2344,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_loss": 2.1641297340393066,
      "eval_runtime": 2.1264,
      "eval_samples_per_second": 9.405,
      "eval_steps_per_second": 0.941,
      "step": 0
    },
    {
      "epoch": 0.0008535154166222128,
      "grad_norm": 4.247310161590576,
      "learning_rate": 5.681818181818182e-08,
      "loss": 1.0032,
      "step": 1
    },
    {
      "epoch": 0.008535154166222128,
      "grad_norm": 4.285112380981445,
      "learning_rate": 5.681818181818182e-07,
      "loss": 0.9823,
      "step": 10
    },
    {
      "epoch": 0.017070308332444255,
      "grad_norm": 3.0184619426727295,
      "learning_rate": 1.1363636363636364e-06,
      "loss": 0.9382,
      "step": 20
    },
    {
      "epoch": 0.02560546249866638,
      "grad_norm": 2.0186476707458496,
      "learning_rate": 1.7045454545454546e-06,
      "loss": 0.8583,
      "step": 30
    },
    {
      "epoch": 0.03414061666488851,
      "grad_norm": 1.175138235092163,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.764,
      "step": 40
    },
    {
      "epoch": 0.04267577083111064,
      "grad_norm": 0.9045423269271851,
      "learning_rate": 2.8409090909090916e-06,
      "loss": 0.714,
      "step": 50
    },
    {
      "epoch": 0.05121092499733276,
      "grad_norm": 0.8305149078369141,
      "learning_rate": 3.409090909090909e-06,
      "loss": 0.6835,
      "step": 60
    },
    {
      "epoch": 0.05974607916355489,
      "grad_norm": 0.7908885478973389,
      "learning_rate": 3.9772727272727275e-06,
      "loss": 0.6549,
      "step": 70
    },
    {
      "epoch": 0.06828123332977702,
      "grad_norm": 0.8054904341697693,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.6368,
      "step": 80
    },
    {
      "epoch": 0.07681638749599914,
      "grad_norm": 0.7891983985900879,
      "learning_rate": 5.113636363636364e-06,
      "loss": 0.631,
      "step": 90
    },
    {
      "epoch": 0.08535154166222128,
      "grad_norm": 0.7775433659553528,
      "learning_rate": 5.681818181818183e-06,
      "loss": 0.6186,
      "step": 100
    },
    {
      "epoch": 0.0938866958284434,
      "grad_norm": 0.8340434432029724,
      "learning_rate": 6.25e-06,
      "loss": 0.6028,
      "step": 110
    },
    {
      "epoch": 0.10242184999466553,
      "grad_norm": 0.7749195098876953,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.5965,
      "step": 120
    },
    {
      "epoch": 0.11095700416088766,
      "grad_norm": 0.7949061393737793,
      "learning_rate": 7.386363636363637e-06,
      "loss": 0.5837,
      "step": 130
    },
    {
      "epoch": 0.11949215832710978,
      "grad_norm": 0.8714171051979065,
      "learning_rate": 7.954545454545455e-06,
      "loss": 0.5865,
      "step": 140
    },
    {
      "epoch": 0.1280273124933319,
      "grad_norm": 0.7890238761901855,
      "learning_rate": 8.522727272727273e-06,
      "loss": 0.5712,
      "step": 150
    },
    {
      "epoch": 0.13656246665955404,
      "grad_norm": 0.8195155262947083,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.5774,
      "step": 160
    },
    {
      "epoch": 0.14509762082577615,
      "grad_norm": 0.8219712972640991,
      "learning_rate": 9.65909090909091e-06,
      "loss": 0.5767,
      "step": 170
    },
    {
      "epoch": 0.1536327749919983,
      "grad_norm": 0.8124852776527405,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.5643,
      "step": 180
    },
    {
      "epoch": 0.16216792915822043,
      "grad_norm": 0.8383850455284119,
      "learning_rate": 1.0795454545454547e-05,
      "loss": 0.5734,
      "step": 190
    },
    {
      "epoch": 0.17070308332444256,
      "grad_norm": 0.7605228424072266,
      "learning_rate": 1.1363636363636366e-05,
      "loss": 0.5661,
      "step": 200
    },
    {
      "epoch": 0.17923823749066467,
      "grad_norm": 0.8585752844810486,
      "learning_rate": 1.1931818181818183e-05,
      "loss": 0.5534,
      "step": 210
    },
    {
      "epoch": 0.1877733916568868,
      "grad_norm": 0.8643070459365845,
      "learning_rate": 1.25e-05,
      "loss": 0.5549,
      "step": 220
    },
    {
      "epoch": 0.19630854582310894,
      "grad_norm": 0.9826616644859314,
      "learning_rate": 1.306818181818182e-05,
      "loss": 0.5531,
      "step": 230
    },
    {
      "epoch": 0.20484369998933105,
      "grad_norm": 0.8799106478691101,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.5371,
      "step": 240
    },
    {
      "epoch": 0.2133788541555532,
      "grad_norm": 0.757698118686676,
      "learning_rate": 1.4204545454545455e-05,
      "loss": 0.5459,
      "step": 250
    },
    {
      "epoch": 0.22191400832177532,
      "grad_norm": 0.8704412579536438,
      "learning_rate": 1.4772727272727274e-05,
      "loss": 0.5369,
      "step": 260
    },
    {
      "epoch": 0.23044916248799743,
      "grad_norm": 0.7941352725028992,
      "learning_rate": 1.5340909090909094e-05,
      "loss": 0.5359,
      "step": 270
    },
    {
      "epoch": 0.23898431665421957,
      "grad_norm": 0.7990615367889404,
      "learning_rate": 1.590909090909091e-05,
      "loss": 0.5285,
      "step": 280
    },
    {
      "epoch": 0.2475194708204417,
      "grad_norm": 0.7938647270202637,
      "learning_rate": 1.647727272727273e-05,
      "loss": 0.5391,
      "step": 290
    },
    {
      "epoch": 0.2560546249866638,
      "grad_norm": 0.7677845358848572,
      "learning_rate": 1.7045454545454546e-05,
      "loss": 0.5195,
      "step": 300
    },
    {
      "epoch": 0.26458977915288595,
      "grad_norm": 0.7977807521820068,
      "learning_rate": 1.7613636363636366e-05,
      "loss": 0.5246,
      "step": 310
    },
    {
      "epoch": 0.2731249333191081,
      "grad_norm": 0.819622814655304,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.5176,
      "step": 320
    },
    {
      "epoch": 0.2816600874853302,
      "grad_norm": 0.8428648114204407,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.513,
      "step": 330
    },
    {
      "epoch": 0.2901952416515523,
      "grad_norm": 0.7542017102241516,
      "learning_rate": 1.931818181818182e-05,
      "loss": 0.5196,
      "step": 340
    },
    {
      "epoch": 0.29873039581777444,
      "grad_norm": 0.8601102232933044,
      "learning_rate": 1.9886363636363638e-05,
      "loss": 0.5206,
      "step": 350
    },
    {
      "epoch": 0.3072655499839966,
      "grad_norm": 0.7532691955566406,
      "learning_rate": 1.9999683918961086e-05,
      "loss": 0.5162,
      "step": 360
    },
    {
      "epoch": 0.3158007041502187,
      "grad_norm": 0.7489638924598694,
      "learning_rate": 1.999839987398595e-05,
      "loss": 0.5122,
      "step": 370
    },
    {
      "epoch": 0.32433585831644085,
      "grad_norm": 0.7671491503715515,
      "learning_rate": 1.9996128236743682e-05,
      "loss": 0.5161,
      "step": 380
    },
    {
      "epoch": 0.332871012482663,
      "grad_norm": 0.7274801731109619,
      "learning_rate": 1.9992869231615323e-05,
      "loss": 0.5167,
      "step": 390
    },
    {
      "epoch": 0.3414061666488851,
      "grad_norm": 0.7289360165596008,
      "learning_rate": 1.9988623180509206e-05,
      "loss": 0.5127,
      "step": 400
    },
    {
      "epoch": 0.3499413208151072,
      "grad_norm": 0.7416843175888062,
      "learning_rate": 1.9983390502829168e-05,
      "loss": 0.5208,
      "step": 410
    },
    {
      "epoch": 0.35847647498132934,
      "grad_norm": 0.7530054450035095,
      "learning_rate": 1.997717171543311e-05,
      "loss": 0.5011,
      "step": 420
    },
    {
      "epoch": 0.3670116291475515,
      "grad_norm": 0.7372049689292908,
      "learning_rate": 1.9969967432581962e-05,
      "loss": 0.5091,
      "step": 430
    },
    {
      "epoch": 0.3755467833137736,
      "grad_norm": 0.8357003927230835,
      "learning_rate": 1.996177836587899e-05,
      "loss": 0.5114,
      "step": 440
    },
    {
      "epoch": 0.38408193747999575,
      "grad_norm": 0.8466181755065918,
      "learning_rate": 1.9952605324199516e-05,
      "loss": 0.5009,
      "step": 450
    },
    {
      "epoch": 0.3926170916462179,
      "grad_norm": 0.7260451316833496,
      "learning_rate": 1.9942449213611028e-05,
      "loss": 0.5087,
      "step": 460
    },
    {
      "epoch": 0.40115224581243997,
      "grad_norm": 0.6815033555030823,
      "learning_rate": 1.9931311037283673e-05,
      "loss": 0.5033,
      "step": 470
    },
    {
      "epoch": 0.4096873999786621,
      "grad_norm": 0.706380307674408,
      "learning_rate": 1.9919191895391176e-05,
      "loss": 0.4974,
      "step": 480
    },
    {
      "epoch": 0.41822255414488424,
      "grad_norm": 0.736223042011261,
      "learning_rate": 1.9906092985002163e-05,
      "loss": 0.4981,
      "step": 490
    },
    {
      "epoch": 0.4267577083111064,
      "grad_norm": 0.9467535614967346,
      "learning_rate": 1.9892015599961927e-05,
      "loss": 0.5017,
      "step": 500
    },
    {
      "epoch": 0.4352928624773285,
      "grad_norm": 0.7372705936431885,
      "learning_rate": 1.9876961130764624e-05,
      "loss": 0.5047,
      "step": 510
    },
    {
      "epoch": 0.44382801664355065,
      "grad_norm": 0.7471463680267334,
      "learning_rate": 1.9860931064415934e-05,
      "loss": 0.5009,
      "step": 520
    },
    {
      "epoch": 0.45236317080977273,
      "grad_norm": 0.6775240898132324,
      "learning_rate": 1.9843926984286165e-05,
      "loss": 0.5045,
      "step": 530
    },
    {
      "epoch": 0.46089832497599487,
      "grad_norm": 0.6793776750564575,
      "learning_rate": 1.9825950569953884e-05,
      "loss": 0.4978,
      "step": 540
    },
    {
      "epoch": 0.469433479142217,
      "grad_norm": 0.6851901412010193,
      "learning_rate": 1.980700359703999e-05,
      "loss": 0.494,
      "step": 550
    },
    {
      "epoch": 0.47796863330843914,
      "grad_norm": 0.6855452060699463,
      "learning_rate": 1.9787087937032333e-05,
      "loss": 0.4952,
      "step": 560
    },
    {
      "epoch": 0.4865037874746613,
      "grad_norm": 0.707453727722168,
      "learning_rate": 1.976620555710087e-05,
      "loss": 0.4949,
      "step": 570
    },
    {
      "epoch": 0.4950389416408834,
      "grad_norm": 0.6674401760101318,
      "learning_rate": 1.9744358519903343e-05,
      "loss": 0.4863,
      "step": 580
    },
    {
      "epoch": 0.5035740958071055,
      "grad_norm": 0.7680632472038269,
      "learning_rate": 1.9721548983381554e-05,
      "loss": 0.4882,
      "step": 590
    },
    {
      "epoch": 0.5121092499733276,
      "grad_norm": 0.6789171099662781,
      "learning_rate": 1.9697779200548202e-05,
      "loss": 0.4848,
      "step": 600
    },
    {
      "epoch": 0.5206444041395498,
      "grad_norm": 0.6551903486251831,
      "learning_rate": 1.9673051519264342e-05,
      "loss": 0.4951,
      "step": 610
    },
    {
      "epoch": 0.5291795583057719,
      "grad_norm": 0.7067411541938782,
      "learning_rate": 1.964736838200749e-05,
      "loss": 0.493,
      "step": 620
    },
    {
      "epoch": 0.537714712471994,
      "grad_norm": 0.6695910692214966,
      "learning_rate": 1.9620732325630342e-05,
      "loss": 0.4938,
      "step": 630
    },
    {
      "epoch": 0.5462498666382162,
      "grad_norm": 0.6625940203666687,
      "learning_rate": 1.9593145981110223e-05,
      "loss": 0.4873,
      "step": 640
    },
    {
      "epoch": 0.5547850208044383,
      "grad_norm": 0.7263137102127075,
      "learning_rate": 1.9564612073289192e-05,
      "loss": 0.4964,
      "step": 650
    },
    {
      "epoch": 0.5633201749706604,
      "grad_norm": 0.7034249901771545,
      "learning_rate": 1.9535133420604905e-05,
      "loss": 0.4952,
      "step": 660
    },
    {
      "epoch": 0.5718553291368825,
      "grad_norm": 0.633065402507782,
      "learning_rate": 1.9504712934812228e-05,
      "loss": 0.4982,
      "step": 670
    },
    {
      "epoch": 0.5803904833031046,
      "grad_norm": 0.7597299814224243,
      "learning_rate": 1.9473353620695614e-05,
      "loss": 0.4839,
      "step": 680
    },
    {
      "epoch": 0.5889256374693268,
      "grad_norm": 0.6461811065673828,
      "learning_rate": 1.9441058575772317e-05,
      "loss": 0.4853,
      "step": 690
    },
    {
      "epoch": 0.5974607916355489,
      "grad_norm": 0.6678339242935181,
      "learning_rate": 1.940783098998643e-05,
      "loss": 0.4814,
      "step": 700
    },
    {
      "epoch": 0.6059959458017711,
      "grad_norm": 0.6317865252494812,
      "learning_rate": 1.9373674145393804e-05,
      "loss": 0.4896,
      "step": 710
    },
    {
      "epoch": 0.6145310999679932,
      "grad_norm": 0.6536778211593628,
      "learning_rate": 1.9338591415837856e-05,
      "loss": 0.4795,
      "step": 720
    },
    {
      "epoch": 0.6230662541342153,
      "grad_norm": 0.653390645980835,
      "learning_rate": 1.9302586266616318e-05,
      "loss": 0.4862,
      "step": 730
    },
    {
      "epoch": 0.6316014083004374,
      "grad_norm": 0.6988620758056641,
      "learning_rate": 1.9265662254138958e-05,
      "loss": 0.4913,
      "step": 740
    },
    {
      "epoch": 0.6401365624666595,
      "grad_norm": 0.6603461503982544,
      "learning_rate": 1.922782302557628e-05,
      "loss": 0.4813,
      "step": 750
    },
    {
      "epoch": 0.6486717166328817,
      "grad_norm": 0.7291558980941772,
      "learning_rate": 1.918907231849931e-05,
      "loss": 0.4843,
      "step": 760
    },
    {
      "epoch": 0.6572068707991038,
      "grad_norm": 0.678535521030426,
      "learning_rate": 1.914941396051036e-05,
      "loss": 0.4819,
      "step": 770
    },
    {
      "epoch": 0.665742024965326,
      "grad_norm": 0.7419494390487671,
      "learning_rate": 1.910885186886502e-05,
      "loss": 0.4759,
      "step": 780
    },
    {
      "epoch": 0.674277179131548,
      "grad_norm": 0.7254301309585571,
      "learning_rate": 1.9067390050085183e-05,
      "loss": 0.4754,
      "step": 790
    },
    {
      "epoch": 0.6828123332977702,
      "grad_norm": 0.6892661452293396,
      "learning_rate": 1.902503259956333e-05,
      "loss": 0.4759,
      "step": 800
    },
    {
      "epoch": 0.6913474874639923,
      "grad_norm": 0.6353612542152405,
      "learning_rate": 1.8981783701157985e-05,
      "loss": 0.4787,
      "step": 810
    },
    {
      "epoch": 0.6998826416302144,
      "grad_norm": 0.578580915927887,
      "learning_rate": 1.8937647626780473e-05,
      "loss": 0.4748,
      "step": 820
    },
    {
      "epoch": 0.7084177957964366,
      "grad_norm": 0.6349884867668152,
      "learning_rate": 1.889262873597295e-05,
      "loss": 0.4817,
      "step": 830
    },
    {
      "epoch": 0.7169529499626587,
      "grad_norm": 0.6372865438461304,
      "learning_rate": 1.8846731475477796e-05,
      "loss": 0.4811,
      "step": 840
    },
    {
      "epoch": 0.7254881041288809,
      "grad_norm": 0.70332932472229,
      "learning_rate": 1.8799960378798382e-05,
      "loss": 0.4854,
      "step": 850
    },
    {
      "epoch": 0.734023258295103,
      "grad_norm": 0.6345937252044678,
      "learning_rate": 1.8752320065751276e-05,
      "loss": 0.4804,
      "step": 860
    },
    {
      "epoch": 0.742558412461325,
      "grad_norm": 0.7001857757568359,
      "learning_rate": 1.8703815242009927e-05,
      "loss": 0.4823,
      "step": 870
    },
    {
      "epoch": 0.7510935666275472,
      "grad_norm": 0.6997294425964355,
      "learning_rate": 1.8654450698639845e-05,
      "loss": 0.4772,
      "step": 880
    },
    {
      "epoch": 0.7596287207937693,
      "grad_norm": 0.6218723058700562,
      "learning_rate": 1.860423131162538e-05,
      "loss": 0.4741,
      "step": 890
    },
    {
      "epoch": 0.7681638749599915,
      "grad_norm": 0.6111359000205994,
      "learning_rate": 1.8553162041388096e-05,
      "loss": 0.4724,
      "step": 900
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 0.6032017469406128,
      "learning_rate": 1.8501247932296785e-05,
      "loss": 0.4769,
      "step": 910
    },
    {
      "epoch": 0.7852341832924358,
      "grad_norm": 0.6396908164024353,
      "learning_rate": 1.8448494112169234e-05,
      "loss": 0.4747,
      "step": 920
    },
    {
      "epoch": 0.7937693374586579,
      "grad_norm": 0.690779983997345,
      "learning_rate": 1.8394905791765714e-05,
      "loss": 0.4719,
      "step": 930
    },
    {
      "epoch": 0.8023044916248799,
      "grad_norm": 0.6640828251838684,
      "learning_rate": 1.8340488264274285e-05,
      "loss": 0.477,
      "step": 940
    },
    {
      "epoch": 0.8108396457911021,
      "grad_norm": 0.6492927074432373,
      "learning_rate": 1.8285246904787968e-05,
      "loss": 0.4626,
      "step": 950
    },
    {
      "epoch": 0.8193747999573242,
      "grad_norm": 0.6751061677932739,
      "learning_rate": 1.8229187169773805e-05,
      "loss": 0.4698,
      "step": 960
    },
    {
      "epoch": 0.8279099541235464,
      "grad_norm": 0.6475481390953064,
      "learning_rate": 1.8172314596533914e-05,
      "loss": 0.4698,
      "step": 970
    },
    {
      "epoch": 0.8364451082897685,
      "grad_norm": 0.6659871935844421,
      "learning_rate": 1.8114634802658542e-05,
      "loss": 0.4778,
      "step": 980
    },
    {
      "epoch": 0.8449802624559906,
      "grad_norm": 0.6479447484016418,
      "learning_rate": 1.8056153485471167e-05,
      "loss": 0.4703,
      "step": 990
    },
    {
      "epoch": 0.8535154166222128,
      "grad_norm": 0.7254726886749268,
      "learning_rate": 1.7996876421465764e-05,
      "loss": 0.4767,
      "step": 1000
    },
    {
      "epoch": 0.8620505707884348,
      "grad_norm": 0.6294846534729004,
      "learning_rate": 1.7936809465736223e-05,
      "loss": 0.4615,
      "step": 1010
    },
    {
      "epoch": 0.870585724954657,
      "grad_norm": 0.6530686020851135,
      "learning_rate": 1.7875958551398023e-05,
      "loss": 0.4642,
      "step": 1020
    },
    {
      "epoch": 0.8791208791208791,
      "grad_norm": 0.713152289390564,
      "learning_rate": 1.781432968900217e-05,
      "loss": 0.4692,
      "step": 1030
    },
    {
      "epoch": 0.8876560332871013,
      "grad_norm": 0.7651381492614746,
      "learning_rate": 1.775192896594151e-05,
      "loss": 0.4646,
      "step": 1040
    },
    {
      "epoch": 0.8961911874533234,
      "grad_norm": 0.5913043022155762,
      "learning_rate": 1.7688762545849466e-05,
      "loss": 0.4688,
      "step": 1050
    },
    {
      "epoch": 0.9047263416195455,
      "grad_norm": 0.6034150719642639,
      "learning_rate": 1.7624836667991195e-05,
      "loss": 0.4574,
      "step": 1060
    },
    {
      "epoch": 0.9132614957857677,
      "grad_norm": 0.6374826431274414,
      "learning_rate": 1.7560157646647335e-05,
      "loss": 0.4651,
      "step": 1070
    },
    {
      "epoch": 0.9217966499519897,
      "grad_norm": 0.6261687278747559,
      "learning_rate": 1.749473187049028e-05,
      "loss": 0.4608,
      "step": 1080
    },
    {
      "epoch": 0.9303318041182119,
      "grad_norm": 0.6067594885826111,
      "learning_rate": 1.742856580195316e-05,
      "loss": 0.4592,
      "step": 1090
    },
    {
      "epoch": 0.938866958284434,
      "grad_norm": 0.6240684986114502,
      "learning_rate": 1.7361665976591513e-05,
      "loss": 0.4663,
      "step": 1100
    },
    {
      "epoch": 0.9474021124506561,
      "grad_norm": 0.6192986965179443,
      "learning_rate": 1.7294039002437724e-05,
      "loss": 0.476,
      "step": 1110
    },
    {
      "epoch": 0.9559372666168783,
      "grad_norm": 0.6121706962585449,
      "learning_rate": 1.7225691559348333e-05,
      "loss": 0.4531,
      "step": 1120
    },
    {
      "epoch": 0.9644724207831004,
      "grad_norm": 0.5956526398658752,
      "learning_rate": 1.715663039834421e-05,
      "loss": 0.4643,
      "step": 1130
    },
    {
      "epoch": 0.9730075749493226,
      "grad_norm": 0.5699427127838135,
      "learning_rate": 1.7086862340943745e-05,
      "loss": 0.4601,
      "step": 1140
    },
    {
      "epoch": 0.9815427291155446,
      "grad_norm": 0.6840018630027771,
      "learning_rate": 1.701639427848903e-05,
      "loss": 0.4555,
      "step": 1150
    },
    {
      "epoch": 0.9900778832817668,
      "grad_norm": 0.6009624004364014,
      "learning_rate": 1.6945233171465193e-05,
      "loss": 0.4632,
      "step": 1160
    },
    {
      "epoch": 0.9986130374479889,
      "grad_norm": 0.5962963104248047,
      "learning_rate": 1.6873386048812854e-05,
      "loss": 0.4702,
      "step": 1170
    },
    {
      "epoch": 1.0068281233329777,
      "grad_norm": 0.679568350315094,
      "learning_rate": 1.680086000723385e-05,
      "loss": 0.4318,
      "step": 1180
    },
    {
      "epoch": 1.0153632774991999,
      "grad_norm": 0.6293564438819885,
      "learning_rate": 1.672766221049025e-05,
      "loss": 0.4231,
      "step": 1190
    },
    {
      "epoch": 1.023898431665422,
      "grad_norm": 0.6619155406951904,
      "learning_rate": 1.6653799888696777e-05,
      "loss": 0.4251,
      "step": 1200
    },
    {
      "epoch": 1.032433585831644,
      "grad_norm": 0.6258383393287659,
      "learning_rate": 1.6579280337606615e-05,
      "loss": 0.4237,
      "step": 1210
    },
    {
      "epoch": 1.0409687399978662,
      "grad_norm": 0.6029568314552307,
      "learning_rate": 1.650411091789082e-05,
      "loss": 0.42,
      "step": 1220
    },
    {
      "epoch": 1.0495038941640884,
      "grad_norm": 0.5933378338813782,
      "learning_rate": 1.6428299054411212e-05,
      "loss": 0.4325,
      "step": 1230
    },
    {
      "epoch": 1.0580390483303104,
      "grad_norm": 0.593717098236084,
      "learning_rate": 1.635185223548704e-05,
      "loss": 0.4289,
      "step": 1240
    },
    {
      "epoch": 1.0665742024965326,
      "grad_norm": 0.699511706829071,
      "learning_rate": 1.627477801215528e-05,
      "loss": 0.4218,
      "step": 1250
    },
    {
      "epoch": 1.0751093566627548,
      "grad_norm": 0.6585178971290588,
      "learning_rate": 1.619708399742481e-05,
      "loss": 0.4222,
      "step": 1260
    },
    {
      "epoch": 1.083644510828977,
      "grad_norm": 0.6042758822441101,
      "learning_rate": 1.6118777865524414e-05,
      "loss": 0.4231,
      "step": 1270
    },
    {
      "epoch": 1.092179664995199,
      "grad_norm": 0.5816319584846497,
      "learning_rate": 1.6039867351144778e-05,
      "loss": 0.4199,
      "step": 1280
    },
    {
      "epoch": 1.100714819161421,
      "grad_norm": 0.7652620673179626,
      "learning_rate": 1.5960360248674478e-05,
      "loss": 0.428,
      "step": 1290
    },
    {
      "epoch": 1.1092499733276433,
      "grad_norm": 0.6318002343177795,
      "learning_rate": 1.5880264411430106e-05,
      "loss": 0.4263,
      "step": 1300
    },
    {
      "epoch": 1.1177851274938653,
      "grad_norm": 0.5835772752761841,
      "learning_rate": 1.579958775088054e-05,
      "loss": 0.4277,
      "step": 1310
    },
    {
      "epoch": 1.1263202816600875,
      "grad_norm": 0.6293865442276001,
      "learning_rate": 1.5718338235865505e-05,
      "loss": 0.4226,
      "step": 1320
    },
    {
      "epoch": 1.1348554358263097,
      "grad_norm": 0.6008028984069824,
      "learning_rate": 1.5636523891808452e-05,
      "loss": 0.4247,
      "step": 1330
    },
    {
      "epoch": 1.1433905899925318,
      "grad_norm": 0.6461915373802185,
      "learning_rate": 1.5554152799923824e-05,
      "loss": 0.429,
      "step": 1340
    },
    {
      "epoch": 1.1519257441587538,
      "grad_norm": 0.606767475605011,
      "learning_rate": 1.547123309641885e-05,
      "loss": 0.4268,
      "step": 1350
    },
    {
      "epoch": 1.160460898324976,
      "grad_norm": 0.637474000453949,
      "learning_rate": 1.538777297168991e-05,
      "loss": 0.4378,
      "step": 1360
    },
    {
      "epoch": 1.1689960524911982,
      "grad_norm": 0.6810766458511353,
      "learning_rate": 1.5303780669513472e-05,
      "loss": 0.42,
      "step": 1370
    },
    {
      "epoch": 1.1775312066574202,
      "grad_norm": 0.6600694060325623,
      "learning_rate": 1.5219264486231882e-05,
      "loss": 0.4181,
      "step": 1380
    },
    {
      "epoch": 1.1860663608236424,
      "grad_norm": 0.7521950006484985,
      "learning_rate": 1.5134232769933836e-05,
      "loss": 0.4326,
      "step": 1390
    },
    {
      "epoch": 1.1946015149898646,
      "grad_norm": 0.6155983805656433,
      "learning_rate": 1.5048693919629837e-05,
      "loss": 0.4263,
      "step": 1400
    },
    {
      "epoch": 1.2031366691560867,
      "grad_norm": 0.6304906606674194,
      "learning_rate": 1.4962656384422555e-05,
      "loss": 0.4321,
      "step": 1410
    },
    {
      "epoch": 1.2116718233223087,
      "grad_norm": 0.5815775990486145,
      "learning_rate": 1.4876128662672277e-05,
      "loss": 0.4248,
      "step": 1420
    },
    {
      "epoch": 1.220206977488531,
      "grad_norm": 0.619872510433197,
      "learning_rate": 1.4789119301157491e-05,
      "loss": 0.4192,
      "step": 1430
    },
    {
      "epoch": 1.228742131654753,
      "grad_norm": 0.6162813305854797,
      "learning_rate": 1.4701636894230655e-05,
      "loss": 0.4199,
      "step": 1440
    },
    {
      "epoch": 1.237277285820975,
      "grad_norm": 0.6389777064323425,
      "learning_rate": 1.4613690082969311e-05,
      "loss": 0.4239,
      "step": 1450
    },
    {
      "epoch": 1.2458124399871973,
      "grad_norm": 0.6258131265640259,
      "learning_rate": 1.4525287554322558e-05,
      "loss": 0.4229,
      "step": 1460
    },
    {
      "epoch": 1.2543475941534195,
      "grad_norm": 0.678900957107544,
      "learning_rate": 1.4436438040252983e-05,
      "loss": 0.431,
      "step": 1470
    },
    {
      "epoch": 1.2628827483196416,
      "grad_norm": 0.564929723739624,
      "learning_rate": 1.4347150316874179e-05,
      "loss": 0.4178,
      "step": 1480
    },
    {
      "epoch": 1.2714179024858636,
      "grad_norm": 0.5994879603385925,
      "learning_rate": 1.4257433203583876e-05,
      "loss": 0.4271,
      "step": 1490
    },
    {
      "epoch": 1.2799530566520858,
      "grad_norm": 0.5823336839675903,
      "learning_rate": 1.4167295562192808e-05,
      "loss": 0.4246,
      "step": 1500
    },
    {
      "epoch": 1.2884882108183078,
      "grad_norm": 0.624218225479126,
      "learning_rate": 1.4076746296049387e-05,
      "loss": 0.4278,
      "step": 1510
    },
    {
      "epoch": 1.29702336498453,
      "grad_norm": 0.6522708535194397,
      "learning_rate": 1.3985794349160267e-05,
      "loss": 0.4225,
      "step": 1520
    },
    {
      "epoch": 1.3055585191507522,
      "grad_norm": 0.5826681852340698,
      "learning_rate": 1.3894448705306908e-05,
      "loss": 0.4195,
      "step": 1530
    },
    {
      "epoch": 1.3140936733169744,
      "grad_norm": 0.6364330053329468,
      "learning_rate": 1.3802718387158208e-05,
      "loss": 0.4167,
      "step": 1540
    },
    {
      "epoch": 1.3226288274831965,
      "grad_norm": 0.6075431704521179,
      "learning_rate": 1.3710612455379268e-05,
      "loss": 0.4216,
      "step": 1550
    },
    {
      "epoch": 1.3311639816494185,
      "grad_norm": 0.6225044131278992,
      "learning_rate": 1.3618140007736442e-05,
      "loss": 0.4181,
      "step": 1560
    },
    {
      "epoch": 1.3396991358156407,
      "grad_norm": 0.5846056938171387,
      "learning_rate": 1.3525310178198707e-05,
      "loss": 0.4264,
      "step": 1570
    },
    {
      "epoch": 1.3482342899818627,
      "grad_norm": 0.6489505171775818,
      "learning_rate": 1.3432132136035443e-05,
      "loss": 0.4267,
      "step": 1580
    },
    {
      "epoch": 1.3567694441480849,
      "grad_norm": 0.7888132929801941,
      "learning_rate": 1.3338615084910737e-05,
      "loss": 0.4173,
      "step": 1590
    },
    {
      "epoch": 1.365304598314307,
      "grad_norm": 0.6617109775543213,
      "learning_rate": 1.3244768261974307e-05,
      "loss": 0.422,
      "step": 1600
    },
    {
      "epoch": 1.3738397524805293,
      "grad_norm": 0.5967974066734314,
      "learning_rate": 1.3150600936949092e-05,
      "loss": 0.422,
      "step": 1610
    },
    {
      "epoch": 1.3823749066467514,
      "grad_norm": 0.6307809352874756,
      "learning_rate": 1.305612241121562e-05,
      "loss": 0.422,
      "step": 1620
    },
    {
      "epoch": 1.3909100608129734,
      "grad_norm": 0.5797590017318726,
      "learning_rate": 1.2961342016893302e-05,
      "loss": 0.4168,
      "step": 1630
    },
    {
      "epoch": 1.3994452149791956,
      "grad_norm": 0.5861985087394714,
      "learning_rate": 1.2866269115918606e-05,
      "loss": 0.4289,
      "step": 1640
    },
    {
      "epoch": 1.4079803691454176,
      "grad_norm": 0.6140124797821045,
      "learning_rate": 1.2770913099120374e-05,
      "loss": 0.4294,
      "step": 1650
    },
    {
      "epoch": 1.4165155233116398,
      "grad_norm": 0.5719994306564331,
      "learning_rate": 1.2675283385292212e-05,
      "loss": 0.4254,
      "step": 1660
    },
    {
      "epoch": 1.425050677477862,
      "grad_norm": 0.5942404866218567,
      "learning_rate": 1.2579389420262151e-05,
      "loss": 0.4212,
      "step": 1670
    },
    {
      "epoch": 1.4335858316440842,
      "grad_norm": 0.6488329172134399,
      "learning_rate": 1.248324067595966e-05,
      "loss": 0.4173,
      "step": 1680
    },
    {
      "epoch": 1.4421209858103061,
      "grad_norm": 0.5882269144058228,
      "learning_rate": 1.2386846649480036e-05,
      "loss": 0.4206,
      "step": 1690
    },
    {
      "epoch": 1.4506561399765283,
      "grad_norm": 0.6740261316299438,
      "learning_rate": 1.2290216862146309e-05,
      "loss": 0.4157,
      "step": 1700
    },
    {
      "epoch": 1.4591912941427505,
      "grad_norm": 0.6225579977035522,
      "learning_rate": 1.2193360858568824e-05,
      "loss": 0.4165,
      "step": 1710
    },
    {
      "epoch": 1.4677264483089725,
      "grad_norm": 0.6018004417419434,
      "learning_rate": 1.2096288205702431e-05,
      "loss": 0.4225,
      "step": 1720
    },
    {
      "epoch": 1.4762616024751947,
      "grad_norm": 0.6020993590354919,
      "learning_rate": 1.1999008491901511e-05,
      "loss": 0.4247,
      "step": 1730
    },
    {
      "epoch": 1.4847967566414169,
      "grad_norm": 0.6166216731071472,
      "learning_rate": 1.1901531325972911e-05,
      "loss": 0.4227,
      "step": 1740
    },
    {
      "epoch": 1.493331910807639,
      "grad_norm": 0.5822474956512451,
      "learning_rate": 1.180386633622681e-05,
      "loss": 0.4132,
      "step": 1750
    },
    {
      "epoch": 1.5018670649738612,
      "grad_norm": 0.6379313468933105,
      "learning_rate": 1.1706023169525691e-05,
      "loss": 0.4213,
      "step": 1760
    },
    {
      "epoch": 1.5104022191400832,
      "grad_norm": 0.5989879965782166,
      "learning_rate": 1.160801149033147e-05,
      "loss": 0.4222,
      "step": 1770
    },
    {
      "epoch": 1.5189373733063052,
      "grad_norm": 0.6491897702217102,
      "learning_rate": 1.1509840979750895e-05,
      "loss": 0.4186,
      "step": 1780
    },
    {
      "epoch": 1.5274725274725274,
      "grad_norm": 0.5892138481140137,
      "learning_rate": 1.1411521334579288e-05,
      "loss": 0.4215,
      "step": 1790
    },
    {
      "epoch": 1.5360076816387496,
      "grad_norm": 0.5960173606872559,
      "learning_rate": 1.131306226634274e-05,
      "loss": 0.418,
      "step": 1800
    },
    {
      "epoch": 1.5445428358049718,
      "grad_norm": 0.5612832307815552,
      "learning_rate": 1.1214473500338862e-05,
      "loss": 0.4203,
      "step": 1810
    },
    {
      "epoch": 1.553077989971194,
      "grad_norm": 0.5860463380813599,
      "learning_rate": 1.1115764774676172e-05,
      "loss": 0.422,
      "step": 1820
    },
    {
      "epoch": 1.5616131441374161,
      "grad_norm": 0.6095026135444641,
      "learning_rate": 1.1016945839312202e-05,
      "loss": 0.4153,
      "step": 1830
    },
    {
      "epoch": 1.5701482983036381,
      "grad_norm": 0.5922375321388245,
      "learning_rate": 1.0918026455090447e-05,
      "loss": 0.4195,
      "step": 1840
    },
    {
      "epoch": 1.57868345246986,
      "grad_norm": 0.5867443084716797,
      "learning_rate": 1.0819016392776245e-05,
      "loss": 0.4169,
      "step": 1850
    },
    {
      "epoch": 1.5872186066360823,
      "grad_norm": 0.6018590927124023,
      "learning_rate": 1.0719925432091671e-05,
      "loss": 0.4111,
      "step": 1860
    },
    {
      "epoch": 1.5957537608023045,
      "grad_norm": 0.5756944417953491,
      "learning_rate": 1.0620763360749534e-05,
      "loss": 0.4189,
      "step": 1870
    },
    {
      "epoch": 1.6042889149685267,
      "grad_norm": 0.6248449683189392,
      "learning_rate": 1.0521539973486592e-05,
      "loss": 0.4183,
      "step": 1880
    },
    {
      "epoch": 1.6128240691347489,
      "grad_norm": 0.6250705122947693,
      "learning_rate": 1.0422265071096101e-05,
      "loss": 0.4233,
      "step": 1890
    },
    {
      "epoch": 1.6213592233009708,
      "grad_norm": 0.5823164582252502,
      "learning_rate": 1.0322948459459716e-05,
      "loss": 0.4143,
      "step": 1900
    },
    {
      "epoch": 1.629894377467193,
      "grad_norm": 0.574417769908905,
      "learning_rate": 1.0223599948578923e-05,
      "loss": 0.423,
      "step": 1910
    },
    {
      "epoch": 1.638429531633415,
      "grad_norm": 0.547025203704834,
      "learning_rate": 1.0124229351606065e-05,
      "loss": 0.4164,
      "step": 1920
    },
    {
      "epoch": 1.6469646857996372,
      "grad_norm": 0.5772645473480225,
      "learning_rate": 1.002484648387503e-05,
      "loss": 0.4116,
      "step": 1930
    },
    {
      "epoch": 1.6554998399658594,
      "grad_norm": 0.5808480381965637,
      "learning_rate": 9.925461161931758e-06,
      "loss": 0.4147,
      "step": 1940
    },
    {
      "epoch": 1.6640349941320816,
      "grad_norm": 0.5725694894790649,
      "learning_rate": 9.826083202564596e-06,
      "loss": 0.4036,
      "step": 1950
    },
    {
      "epoch": 1.6725701482983037,
      "grad_norm": 0.563442587852478,
      "learning_rate": 9.726722421834664e-06,
      "loss": 0.421,
      "step": 1960
    },
    {
      "epoch": 1.6811053024645257,
      "grad_norm": 0.5920106172561646,
      "learning_rate": 9.627388634106245e-06,
      "loss": 0.4214,
      "step": 1970
    },
    {
      "epoch": 1.689640456630748,
      "grad_norm": 0.5918905138969421,
      "learning_rate": 9.528091651077404e-06,
      "loss": 0.417,
      "step": 1980
    },
    {
      "epoch": 1.6981756107969699,
      "grad_norm": 0.576627790927887,
      "learning_rate": 9.428841280810811e-06,
      "loss": 0.4151,
      "step": 1990
    },
    {
      "epoch": 1.706710764963192,
      "grad_norm": 0.5786689519882202,
      "learning_rate": 9.329647326764963e-06,
      "loss": 0.4076,
      "step": 2000
    },
    {
      "epoch": 1.7152459191294143,
      "grad_norm": 0.621134340763092,
      "learning_rate": 9.23051958682584e-06,
      "loss": 0.4165,
      "step": 2010
    },
    {
      "epoch": 1.7237810732956365,
      "grad_norm": 0.5719323754310608,
      "learning_rate": 9.131467852339123e-06,
      "loss": 0.4156,
      "step": 2020
    },
    {
      "epoch": 1.7323162274618586,
      "grad_norm": 0.5678684711456299,
      "learning_rate": 9.032501907143053e-06,
      "loss": 0.4191,
      "step": 2030
    },
    {
      "epoch": 1.7408513816280806,
      "grad_norm": 0.5779105424880981,
      "learning_rate": 8.933631526602028e-06,
      "loss": 0.4046,
      "step": 2040
    },
    {
      "epoch": 1.7493865357943028,
      "grad_norm": 0.5759092569351196,
      "learning_rate": 8.834866476641048e-06,
      "loss": 0.4143,
      "step": 2050
    },
    {
      "epoch": 1.7579216899605248,
      "grad_norm": 0.5681161284446716,
      "learning_rate": 8.73621651278108e-06,
      "loss": 0.424,
      "step": 2060
    },
    {
      "epoch": 1.766456844126747,
      "grad_norm": 0.614236056804657,
      "learning_rate": 8.637691379175453e-06,
      "loss": 0.4148,
      "step": 2070
    },
    {
      "epoch": 1.7749919982929692,
      "grad_norm": 0.5703302621841431,
      "learning_rate": 8.539300807647396e-06,
      "loss": 0.4179,
      "step": 2080
    },
    {
      "epoch": 1.7835271524591914,
      "grad_norm": 0.6135383248329163,
      "learning_rate": 8.441054516728747e-06,
      "loss": 0.4173,
      "step": 2090
    },
    {
      "epoch": 1.7920623066254135,
      "grad_norm": 0.5604820847511292,
      "learning_rate": 8.342962210700043e-06,
      "loss": 0.4167,
      "step": 2100
    },
    {
      "epoch": 1.8005974607916355,
      "grad_norm": 0.5805196166038513,
      "learning_rate": 8.245033578631939e-06,
      "loss": 0.4232,
      "step": 2110
    },
    {
      "epoch": 1.8091326149578577,
      "grad_norm": 0.5662888884544373,
      "learning_rate": 8.147278293428199e-06,
      "loss": 0.4066,
      "step": 2120
    },
    {
      "epoch": 1.8176677691240797,
      "grad_norm": 0.581598162651062,
      "learning_rate": 8.049706010870241e-06,
      "loss": 0.416,
      "step": 2130
    },
    {
      "epoch": 1.8262029232903019,
      "grad_norm": 0.5490183234214783,
      "learning_rate": 7.952326368663404e-06,
      "loss": 0.4088,
      "step": 2140
    },
    {
      "epoch": 1.834738077456524,
      "grad_norm": 0.6272237300872803,
      "learning_rate": 7.855148985484946e-06,
      "loss": 0.4099,
      "step": 2150
    },
    {
      "epoch": 1.8432732316227463,
      "grad_norm": 0.6257729530334473,
      "learning_rate": 7.758183460034005e-06,
      "loss": 0.4068,
      "step": 2160
    },
    {
      "epoch": 1.8518083857889684,
      "grad_norm": 0.6253445744514465,
      "learning_rate": 7.661439370083456e-06,
      "loss": 0.4137,
      "step": 2170
    },
    {
      "epoch": 1.8603435399551904,
      "grad_norm": 0.6083528995513916,
      "learning_rate": 7.564926271533876e-06,
      "loss": 0.4112,
      "step": 2180
    },
    {
      "epoch": 1.8688786941214126,
      "grad_norm": 0.598172128200531,
      "learning_rate": 7.468653697469655e-06,
      "loss": 0.4146,
      "step": 2190
    },
    {
      "epoch": 1.8774138482876346,
      "grad_norm": 0.5930103659629822,
      "learning_rate": 7.372631157217377e-06,
      "loss": 0.4093,
      "step": 2200
    },
    {
      "epoch": 1.8859490024538568,
      "grad_norm": 0.5641002655029297,
      "learning_rate": 7.276868135406523e-06,
      "loss": 0.4162,
      "step": 2210
    },
    {
      "epoch": 1.894484156620079,
      "grad_norm": 0.5636462569236755,
      "learning_rate": 7.181374091032635e-06,
      "loss": 0.408,
      "step": 2220
    },
    {
      "epoch": 1.9030193107863012,
      "grad_norm": 0.6011704802513123,
      "learning_rate": 7.0861584565229955e-06,
      "loss": 0.4119,
      "step": 2230
    },
    {
      "epoch": 1.9115544649525233,
      "grad_norm": 0.5807620882987976,
      "learning_rate": 6.991230636804953e-06,
      "loss": 0.4056,
      "step": 2240
    },
    {
      "epoch": 1.9200896191187453,
      "grad_norm": 0.6091267466545105,
      "learning_rate": 6.896600008376926e-06,
      "loss": 0.4095,
      "step": 2250
    },
    {
      "epoch": 1.9286247732849675,
      "grad_norm": 0.5541015267372131,
      "learning_rate": 6.802275918382266e-06,
      "loss": 0.4085,
      "step": 2260
    },
    {
      "epoch": 1.9371599274511895,
      "grad_norm": 0.5683854222297668,
      "learning_rate": 6.708267683685978e-06,
      "loss": 0.4098,
      "step": 2270
    },
    {
      "epoch": 1.9456950816174117,
      "grad_norm": 0.6314034461975098,
      "learning_rate": 6.614584589954447e-06,
      "loss": 0.4135,
      "step": 2280
    },
    {
      "epoch": 1.9542302357836339,
      "grad_norm": 0.5882005095481873,
      "learning_rate": 6.521235890738251e-06,
      "loss": 0.4102,
      "step": 2290
    },
    {
      "epoch": 1.962765389949856,
      "grad_norm": 0.5736626386642456,
      "learning_rate": 6.42823080655814e-06,
      "loss": 0.4037,
      "step": 2300
    },
    {
      "epoch": 1.9713005441160782,
      "grad_norm": 0.5581937432289124,
      "learning_rate": 6.33557852399428e-06,
      "loss": 0.4155,
      "step": 2310
    },
    {
      "epoch": 1.9798356982823002,
      "grad_norm": 0.5713927745819092,
      "learning_rate": 6.243288194778837e-06,
      "loss": 0.4058,
      "step": 2320
    },
    {
      "epoch": 1.9883708524485222,
      "grad_norm": 0.6096407771110535,
      "learning_rate": 6.151368934892028e-06,
      "loss": 0.4096,
      "step": 2330
    },
    {
      "epoch": 1.9969060066147444,
      "grad_norm": 0.5943323373794556,
      "learning_rate": 6.059829823661692e-06,
      "loss": 0.413,
      "step": 2340
    }
  ],
  "logging_steps": 10,
  "max_steps": 3513,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.459688578893742e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}