{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2052083842282699,
  "eval_steps": 500,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009771827820393806,
      "grad_norm": 0.5417118072509766,
      "learning_rate": 4.995602247740044e-05,
      "loss": 1.378,
      "step": 100
    },
    {
      "epoch": 0.001954365564078761,
      "grad_norm": 0.6493918895721436,
      "learning_rate": 4.990715856340093e-05,
      "loss": 1.3304,
      "step": 200
    },
    {
      "epoch": 0.0029315483461181415,
      "grad_norm": 0.9062462449073792,
      "learning_rate": 4.9858294649401425e-05,
      "loss": 1.3284,
      "step": 300
    },
    {
      "epoch": 0.003908731128157522,
      "grad_norm": 0.750052273273468,
      "learning_rate": 4.9809430735401906e-05,
      "loss": 1.3166,
      "step": 400
    },
    {
      "epoch": 0.004885913910196903,
      "grad_norm": 0.6602022051811218,
      "learning_rate": 4.97605668214024e-05,
      "loss": 1.3166,
      "step": 500
    },
    {
      "epoch": 0.005863096692236283,
      "grad_norm": 0.4193927049636841,
      "learning_rate": 4.971170290740288e-05,
      "loss": 1.3098,
      "step": 600
    },
    {
      "epoch": 0.006840279474275663,
      "grad_norm": 0.6095415949821472,
      "learning_rate": 4.966283899340338e-05,
      "loss": 1.3103,
      "step": 700
    },
    {
      "epoch": 0.007817462256315045,
      "grad_norm": 0.9943467378616333,
      "learning_rate": 4.9613975079403865e-05,
      "loss": 1.3096,
      "step": 800
    },
    {
      "epoch": 0.008794645038354424,
      "grad_norm": 1.2263585329055786,
      "learning_rate": 4.9565111165404346e-05,
      "loss": 1.3067,
      "step": 900
    },
    {
      "epoch": 0.009771827820393805,
      "grad_norm": 0.7198677659034729,
      "learning_rate": 4.951624725140484e-05,
      "loss": 1.3041,
      "step": 1000
    },
    {
      "epoch": 0.010749010602433185,
      "grad_norm": 0.7370775938034058,
      "learning_rate": 4.946738333740533e-05,
      "loss": 1.302,
      "step": 1100
    },
    {
      "epoch": 0.011726193384472566,
      "grad_norm": 0.5109437704086304,
      "learning_rate": 4.941851942340582e-05,
      "loss": 1.3089,
      "step": 1200
    },
    {
      "epoch": 0.012703376166511945,
      "grad_norm": 0.1879555583000183,
      "learning_rate": 4.9369655509406305e-05,
      "loss": 1.3043,
      "step": 1300
    },
    {
      "epoch": 0.013680558948551327,
      "grad_norm": 0.951046884059906,
      "learning_rate": 4.932079159540679e-05,
      "loss": 1.3098,
      "step": 1400
    },
    {
      "epoch": 0.014657741730590706,
      "grad_norm": 0.2478829026222229,
      "learning_rate": 4.927192768140728e-05,
      "loss": 1.3026,
      "step": 1500
    },
    {
      "epoch": 0.01563492451263009,
      "grad_norm": 0.5585843324661255,
      "learning_rate": 4.9223063767407776e-05,
      "loss": 1.3014,
      "step": 1600
    },
    {
      "epoch": 0.016612107294669467,
      "grad_norm": 0.48532453179359436,
      "learning_rate": 4.917419985340826e-05,
      "loss": 1.2981,
      "step": 1700
    },
    {
      "epoch": 0.017589290076708848,
      "grad_norm": 0.4233573079109192,
      "learning_rate": 4.912533593940875e-05,
      "loss": 1.2992,
      "step": 1800
    },
    {
      "epoch": 0.01856647285874823,
      "grad_norm": 0.3272475600242615,
      "learning_rate": 4.9076472025409234e-05,
      "loss": 1.292,
      "step": 1900
    },
    {
      "epoch": 0.01954365564078761,
      "grad_norm": 0.5299385786056519,
      "learning_rate": 4.902760811140973e-05,
      "loss": 1.2963,
      "step": 2000
    },
    {
      "epoch": 0.02052083842282699,
      "grad_norm": 0.1614024043083191,
      "learning_rate": 4.8978744197410216e-05,
      "loss": 1.2945,
      "step": 2100
    },
    {
      "epoch": 0.02149802120486637,
      "grad_norm": 0.6039963960647583,
      "learning_rate": 4.8929880283410705e-05,
      "loss": 1.2913,
      "step": 2200
    },
    {
      "epoch": 0.02247520398690575,
      "grad_norm": 0.5772804021835327,
      "learning_rate": 4.888101636941119e-05,
      "loss": 1.2895,
      "step": 2300
    },
    {
      "epoch": 0.023452386768945132,
      "grad_norm": 0.7489622235298157,
      "learning_rate": 4.883215245541168e-05,
      "loss": 1.2847,
      "step": 2400
    },
    {
      "epoch": 0.024429569550984513,
      "grad_norm": 0.30208253860473633,
      "learning_rate": 4.878328854141217e-05,
      "loss": 1.2924,
      "step": 2500
    },
    {
      "epoch": 0.02540675233302389,
      "grad_norm": 0.36944472789764404,
      "learning_rate": 4.873442462741266e-05,
      "loss": 1.2916,
      "step": 2600
    },
    {
      "epoch": 0.026383935115063272,
      "grad_norm": 0.3268676698207855,
      "learning_rate": 4.8685560713413145e-05,
      "loss": 1.2893,
      "step": 2700
    },
    {
      "epoch": 0.027361117897102653,
      "grad_norm": 0.2795974910259247,
      "learning_rate": 4.863669679941363e-05,
      "loss": 1.282,
      "step": 2800
    },
    {
      "epoch": 0.028338300679142035,
      "grad_norm": 0.36298853158950806,
      "learning_rate": 4.858783288541413e-05,
      "loss": 1.2832,
      "step": 2900
    },
    {
      "epoch": 0.029315483461181412,
      "grad_norm": 0.5242423415184021,
      "learning_rate": 4.853896897141461e-05,
      "loss": 1.2819,
      "step": 3000
    },
    {
      "epoch": 0.030292666243220794,
      "grad_norm": 0.25340864062309265,
      "learning_rate": 4.8490105057415104e-05,
      "loss": 1.2809,
      "step": 3100
    },
    {
      "epoch": 0.03126984902526018,
      "grad_norm": 0.7241976261138916,
      "learning_rate": 4.844124114341559e-05,
      "loss": 1.2802,
      "step": 3200
    },
    {
      "epoch": 0.032247031807299556,
      "grad_norm": 0.5154001712799072,
      "learning_rate": 4.839237722941608e-05,
      "loss": 1.2748,
      "step": 3300
    },
    {
      "epoch": 0.033224214589338934,
      "grad_norm": 0.5323473811149597,
      "learning_rate": 4.834351331541657e-05,
      "loss": 1.284,
      "step": 3400
    },
    {
      "epoch": 0.03420139737137832,
      "grad_norm": 0.3947168290615082,
      "learning_rate": 4.8294649401417056e-05,
      "loss": 1.276,
      "step": 3500
    },
    {
      "epoch": 0.035178580153417696,
      "grad_norm": 0.4776057302951813,
      "learning_rate": 4.8245785487417544e-05,
      "loss": 1.2783,
      "step": 3600
    },
    {
      "epoch": 0.036155762935457074,
      "grad_norm": 0.4884164035320282,
      "learning_rate": 4.819692157341804e-05,
      "loss": 1.2745,
      "step": 3700
    },
    {
      "epoch": 0.03713294571749646,
      "grad_norm": 0.5210428833961487,
      "learning_rate": 4.814805765941852e-05,
      "loss": 1.2707,
      "step": 3800
    },
    {
      "epoch": 0.038110128499535836,
      "grad_norm": 0.46214359998703003,
      "learning_rate": 4.809919374541901e-05,
      "loss": 1.2727,
      "step": 3900
    },
    {
      "epoch": 0.03908731128157522,
      "grad_norm": 0.2656782865524292,
      "learning_rate": 4.8050329831419496e-05,
      "loss": 1.2694,
      "step": 4000
    },
    {
      "epoch": 0.0400644940636146,
      "grad_norm": 0.4923059940338135,
      "learning_rate": 4.8001465917419985e-05,
      "loss": 1.2665,
      "step": 4100
    },
    {
      "epoch": 0.04104167684565398,
      "grad_norm": 0.92928147315979,
      "learning_rate": 4.795260200342048e-05,
      "loss": 1.2627,
      "step": 4200
    },
    {
      "epoch": 0.04201885962769336,
      "grad_norm": 1.0651229619979858,
      "learning_rate": 4.790373808942096e-05,
      "loss": 1.2623,
      "step": 4300
    },
    {
      "epoch": 0.04299604240973274,
      "grad_norm": 0.9612557888031006,
      "learning_rate": 4.7854874175421456e-05,
      "loss": 1.2482,
      "step": 4400
    },
    {
      "epoch": 0.043973225191772124,
      "grad_norm": 1.0120874643325806,
      "learning_rate": 4.7806010261421944e-05,
      "loss": 1.2589,
      "step": 4500
    },
    {
      "epoch": 0.0449504079738115,
      "grad_norm": 0.6250020861625671,
      "learning_rate": 4.775714634742243e-05,
      "loss": 1.2499,
      "step": 4600
    },
    {
      "epoch": 0.04592759075585088,
      "grad_norm": 0.2850038707256317,
      "learning_rate": 4.770828243342292e-05,
      "loss": 1.2446,
      "step": 4700
    },
    {
      "epoch": 0.046904773537890264,
      "grad_norm": 1.2032625675201416,
      "learning_rate": 4.765941851942341e-05,
      "loss": 1.2238,
      "step": 4800
    },
    {
      "epoch": 0.04788195631992964,
      "grad_norm": 0.42024949193000793,
      "learning_rate": 4.7610554605423896e-05,
      "loss": 1.2255,
      "step": 4900
    },
    {
      "epoch": 0.048859139101969026,
      "grad_norm": 0.7451406121253967,
      "learning_rate": 4.756169069142439e-05,
      "loss": 1.2071,
      "step": 5000
    },
    {
      "epoch": 0.049836321884008404,
      "grad_norm": 0.8735096454620361,
      "learning_rate": 4.751282677742487e-05,
      "loss": 1.2126,
      "step": 5100
    },
    {
      "epoch": 0.05081350466604778,
      "grad_norm": 0.73675137758255,
      "learning_rate": 4.746396286342537e-05,
      "loss": 1.2036,
      "step": 5200
    },
    {
      "epoch": 0.051790687448087167,
      "grad_norm": 0.6540606617927551,
      "learning_rate": 4.741509894942585e-05,
      "loss": 1.1825,
      "step": 5300
    },
    {
      "epoch": 0.052767870230126544,
      "grad_norm": 0.825066864490509,
      "learning_rate": 4.7366235035426336e-05,
      "loss": 1.1655,
      "step": 5400
    },
    {
      "epoch": 0.05374505301216593,
      "grad_norm": 1.6421219110488892,
      "learning_rate": 4.731737112142683e-05,
      "loss": 1.1716,
      "step": 5500
    },
    {
      "epoch": 0.05472223579420531,
      "grad_norm": 1.0644057989120483,
      "learning_rate": 4.726850720742731e-05,
      "loss": 1.1384,
      "step": 5600
    },
    {
      "epoch": 0.055699418576244684,
      "grad_norm": 1.1611616611480713,
      "learning_rate": 4.721964329342781e-05,
      "loss": 1.1499,
      "step": 5700
    },
    {
      "epoch": 0.05667660135828407,
      "grad_norm": 2.0900723934173584,
      "learning_rate": 4.7170779379428295e-05,
      "loss": 1.1323,
      "step": 5800
    },
    {
      "epoch": 0.05765378414032345,
      "grad_norm": 1.0580404996871948,
      "learning_rate": 4.712191546542878e-05,
      "loss": 1.112,
      "step": 5900
    },
    {
      "epoch": 0.058630966922362825,
      "grad_norm": 0.6299407482147217,
      "learning_rate": 4.707305155142927e-05,
      "loss": 1.104,
      "step": 6000
    },
    {
      "epoch": 0.05960814970440221,
      "grad_norm": 0.6816271543502808,
      "learning_rate": 4.702418763742976e-05,
      "loss": 1.1128,
      "step": 6100
    },
    {
      "epoch": 0.06058533248644159,
      "grad_norm": 0.654796302318573,
      "learning_rate": 4.697532372343025e-05,
      "loss": 1.0942,
      "step": 6200
    },
    {
      "epoch": 0.06156251526848097,
      "grad_norm": 1.0433884859085083,
      "learning_rate": 4.692645980943074e-05,
      "loss": 1.0862,
      "step": 6300
    },
    {
      "epoch": 0.06253969805052036,
      "grad_norm": 0.6256537437438965,
      "learning_rate": 4.6877595895431224e-05,
      "loss": 1.081,
      "step": 6400
    },
    {
      "epoch": 0.06351688083255973,
      "grad_norm": 0.8173975348472595,
      "learning_rate": 4.682873198143172e-05,
      "loss": 1.0767,
      "step": 6500
    },
    {
      "epoch": 0.06449406361459911,
      "grad_norm": 0.7856473922729492,
      "learning_rate": 4.6779868067432206e-05,
      "loss": 1.0767,
      "step": 6600
    },
    {
      "epoch": 0.0654712463966385,
      "grad_norm": 0.6337741017341614,
      "learning_rate": 4.6731004153432695e-05,
      "loss": 1.0829,
      "step": 6700
    },
    {
      "epoch": 0.06644842917867787,
      "grad_norm": 0.5813809037208557,
      "learning_rate": 4.668214023943318e-05,
      "loss": 1.0571,
      "step": 6800
    },
    {
      "epoch": 0.06742561196071725,
      "grad_norm": 0.4155445992946625,
      "learning_rate": 4.6633276325433664e-05,
      "loss": 1.0707,
      "step": 6900
    },
    {
      "epoch": 0.06840279474275664,
      "grad_norm": 0.6730567812919617,
      "learning_rate": 4.658441241143416e-05,
      "loss": 1.0477,
      "step": 7000
    },
    {
      "epoch": 0.06937997752479601,
      "grad_norm": 0.8348300457000732,
      "learning_rate": 4.653554849743465e-05,
      "loss": 1.0644,
      "step": 7100
    },
    {
      "epoch": 0.07035716030683539,
      "grad_norm": 2.2414326667785645,
      "learning_rate": 4.6486684583435135e-05,
      "loss": 1.0577,
      "step": 7200
    },
    {
      "epoch": 0.07133434308887478,
      "grad_norm": 1.6573911905288696,
      "learning_rate": 4.643782066943562e-05,
      "loss": 1.0836,
      "step": 7300
    },
    {
      "epoch": 0.07231152587091415,
      "grad_norm": 0.5690039396286011,
      "learning_rate": 4.638895675543611e-05,
      "loss": 1.0541,
      "step": 7400
    },
    {
      "epoch": 0.07328870865295353,
      "grad_norm": 0.527215301990509,
      "learning_rate": 4.63400928414366e-05,
      "loss": 1.0164,
      "step": 7500
    },
    {
      "epoch": 0.07426589143499292,
      "grad_norm": 0.7997362613677979,
      "learning_rate": 4.6291228927437094e-05,
      "loss": 1.0447,
      "step": 7600
    },
    {
      "epoch": 0.0752430742170323,
      "grad_norm": 2.257143259048462,
      "learning_rate": 4.6242365013437575e-05,
      "loss": 1.0365,
      "step": 7700
    },
    {
      "epoch": 0.07622025699907167,
      "grad_norm": 0.9132490158081055,
      "learning_rate": 4.619350109943807e-05,
      "loss": 1.0498,
      "step": 7800
    },
    {
      "epoch": 0.07719743978111106,
      "grad_norm": 0.5229859948158264,
      "learning_rate": 4.614463718543856e-05,
      "loss": 1.0342,
      "step": 7900
    },
    {
      "epoch": 0.07817462256315044,
      "grad_norm": 0.6948792338371277,
      "learning_rate": 4.6095773271439046e-05,
      "loss": 1.0325,
      "step": 8000
    },
    {
      "epoch": 0.07915180534518981,
      "grad_norm": 0.8526360988616943,
      "learning_rate": 4.6046909357439534e-05,
      "loss": 1.0183,
      "step": 8100
    },
    {
      "epoch": 0.0801289881272292,
      "grad_norm": 1.1457374095916748,
      "learning_rate": 4.599804544344002e-05,
      "loss": 1.0243,
      "step": 8200
    },
    {
      "epoch": 0.08110617090926858,
      "grad_norm": 0.9335997700691223,
      "learning_rate": 4.594918152944051e-05,
      "loss": 1.046,
      "step": 8300
    },
    {
      "epoch": 0.08208335369130795,
      "grad_norm": 0.8367229700088501,
      "learning_rate": 4.5900317615441e-05,
      "loss": 1.0176,
      "step": 8400
    },
    {
      "epoch": 0.08306053647334734,
      "grad_norm": 3.7648801803588867,
      "learning_rate": 4.5851453701441486e-05,
      "loss": 1.0047,
      "step": 8500
    },
    {
      "epoch": 0.08403771925538672,
      "grad_norm": 0.5877612829208374,
      "learning_rate": 4.5802589787441975e-05,
      "loss": 1.0346,
      "step": 8600
    },
    {
      "epoch": 0.08501490203742611,
      "grad_norm": 0.5145990252494812,
      "learning_rate": 4.575372587344246e-05,
      "loss": 1.0268,
      "step": 8700
    },
    {
      "epoch": 0.08599208481946548,
      "grad_norm": 0.9310688376426697,
      "learning_rate": 4.570486195944295e-05,
      "loss": 1.0109,
      "step": 8800
    },
    {
      "epoch": 0.08696926760150486,
      "grad_norm": 0.5182886719703674,
      "learning_rate": 4.5655998045443445e-05,
      "loss": 1.0117,
      "step": 8900
    },
    {
      "epoch": 0.08794645038354425,
      "grad_norm": 0.4319695234298706,
      "learning_rate": 4.560713413144393e-05,
      "loss": 1.0053,
      "step": 9000
    },
    {
      "epoch": 0.08892363316558362,
      "grad_norm": 4.307732582092285,
      "learning_rate": 4.555827021744442e-05,
      "loss": 1.0151,
      "step": 9100
    },
    {
      "epoch": 0.089900815947623,
      "grad_norm": 0.46516236662864685,
      "learning_rate": 4.550940630344491e-05,
      "loss": 0.9945,
      "step": 9200
    },
    {
      "epoch": 0.09087799872966239,
      "grad_norm": 1.2372952699661255,
      "learning_rate": 4.54605423894454e-05,
      "loss": 0.9865,
      "step": 9300
    },
    {
      "epoch": 0.09185518151170176,
      "grad_norm": 0.7494595646858215,
      "learning_rate": 4.5411678475445886e-05,
      "loss": 0.9824,
      "step": 9400
    },
    {
      "epoch": 0.09283236429374114,
      "grad_norm": 0.5540333390235901,
      "learning_rate": 4.5362814561446374e-05,
      "loss": 1.0132,
      "step": 9500
    },
    {
      "epoch": 0.09380954707578053,
      "grad_norm": 0.48533427715301514,
      "learning_rate": 4.531395064744686e-05,
      "loss": 1.0173,
      "step": 9600
    },
    {
      "epoch": 0.0947867298578199,
      "grad_norm": 0.4972572922706604,
      "learning_rate": 4.526508673344736e-05,
      "loss": 1.0078,
      "step": 9700
    },
    {
      "epoch": 0.09576391263985928,
      "grad_norm": 0.6748878955841064,
      "learning_rate": 4.521622281944784e-05,
      "loss": 1.0172,
      "step": 9800
    },
    {
      "epoch": 0.09674109542189867,
      "grad_norm": 0.5261876583099365,
      "learning_rate": 4.5167358905448326e-05,
      "loss": 1.0189,
      "step": 9900
    },
    {
      "epoch": 0.09771827820393805,
      "grad_norm": 0.4164600670337677,
      "learning_rate": 4.5118494991448814e-05,
      "loss": 0.9978,
      "step": 10000
    },
    {
      "epoch": 0.09869546098597742,
      "grad_norm": 0.40417763590812683,
      "learning_rate": 4.50696310774493e-05,
      "loss": 1.0103,
      "step": 10100
    },
    {
      "epoch": 0.09967264376801681,
      "grad_norm": 0.8591890931129456,
      "learning_rate": 4.50207671634498e-05,
      "loss": 1.0065,
      "step": 10200
    },
    {
      "epoch": 0.10064982655005619,
      "grad_norm": 0.5676371455192566,
      "learning_rate": 4.497190324945028e-05,
      "loss": 1.0089,
      "step": 10300
    },
    {
      "epoch": 0.10162700933209556,
      "grad_norm": 0.616646945476532,
      "learning_rate": 4.492303933545077e-05,
      "loss": 0.9897,
      "step": 10400
    },
    {
      "epoch": 0.10260419211413495,
      "grad_norm": 0.37536484003067017,
      "learning_rate": 4.487417542145126e-05,
      "loss": 0.9989,
      "step": 10500
    },
    {
      "epoch": 0.10358137489617433,
      "grad_norm": 0.6801789402961731,
      "learning_rate": 4.482531150745175e-05,
      "loss": 0.9923,
      "step": 10600
    },
    {
      "epoch": 0.1045585576782137,
      "grad_norm": 0.5848776698112488,
      "learning_rate": 4.477644759345224e-05,
      "loss": 0.9919,
      "step": 10700
    },
    {
      "epoch": 0.10553574046025309,
      "grad_norm": 0.7715157866477966,
      "learning_rate": 4.4727583679452725e-05,
      "loss": 0.9814,
      "step": 10800
    },
    {
      "epoch": 0.10651292324229247,
      "grad_norm": 0.8080986142158508,
      "learning_rate": 4.4678719765453214e-05,
      "loss": 0.9935,
      "step": 10900
    },
    {
      "epoch": 0.10749010602433186,
      "grad_norm": 0.4375016391277313,
      "learning_rate": 4.462985585145371e-05,
      "loss": 0.988,
      "step": 11000
    },
    {
      "epoch": 0.10846728880637123,
      "grad_norm": 0.8055805563926697,
      "learning_rate": 4.458099193745419e-05,
      "loss": 0.9861,
      "step": 11100
    },
    {
      "epoch": 0.10944447158841061,
      "grad_norm": 1.1914618015289307,
      "learning_rate": 4.4532128023454685e-05,
      "loss": 0.9622,
      "step": 11200
    },
    {
      "epoch": 0.11042165437045,
      "grad_norm": 0.4247540533542633,
      "learning_rate": 4.448326410945517e-05,
      "loss": 0.9602,
      "step": 11300
    },
    {
      "epoch": 0.11139883715248937,
      "grad_norm": 0.5454650521278381,
      "learning_rate": 4.4434400195455654e-05,
      "loss": 0.9696,
      "step": 11400
    },
    {
      "epoch": 0.11237601993452875,
      "grad_norm": 0.5259748697280884,
      "learning_rate": 4.438553628145615e-05,
      "loss": 1.0021,
      "step": 11500
    },
    {
      "epoch": 0.11335320271656814,
      "grad_norm": 0.5165246725082397,
      "learning_rate": 4.433667236745663e-05,
      "loss": 0.982,
      "step": 11600
    },
    {
      "epoch": 0.11433038549860751,
      "grad_norm": 0.6768147945404053,
      "learning_rate": 4.4287808453457125e-05,
      "loss": 0.9398,
      "step": 11700
    },
    {
      "epoch": 0.1153075682806469,
      "grad_norm": 1.0245041847229004,
      "learning_rate": 4.423894453945761e-05,
      "loss": 0.9934,
      "step": 11800
    },
    {
      "epoch": 0.11628475106268628,
      "grad_norm": 0.6241583228111267,
      "learning_rate": 4.41900806254581e-05,
      "loss": 0.9697,
      "step": 11900
    },
    {
      "epoch": 0.11726193384472565,
      "grad_norm": 0.4234873652458191,
      "learning_rate": 4.414121671145859e-05,
      "loss": 0.9723,
      "step": 12000
    },
    {
      "epoch": 0.11823911662676503,
      "grad_norm": 0.3932545781135559,
      "learning_rate": 4.409235279745908e-05,
      "loss": 0.9826,
      "step": 12100
    },
    {
      "epoch": 0.11921629940880442,
      "grad_norm": 1.5067880153656006,
      "learning_rate": 4.4043488883459565e-05,
      "loss": 0.9581,
      "step": 12200
    },
    {
      "epoch": 0.1201934821908438,
      "grad_norm": 0.41707366704940796,
      "learning_rate": 4.399462496946006e-05,
      "loss": 0.9666,
      "step": 12300
    },
    {
      "epoch": 0.12117066497288317,
      "grad_norm": 1.1278653144836426,
      "learning_rate": 4.394576105546054e-05,
      "loss": 0.9553,
      "step": 12400
    },
    {
      "epoch": 0.12214784775492256,
      "grad_norm": 0.350543737411499,
      "learning_rate": 4.3896897141461036e-05,
      "loss": 0.9422,
      "step": 12500
    },
    {
      "epoch": 0.12312503053696194,
      "grad_norm": 0.3775838315486908,
      "learning_rate": 4.3848033227461524e-05,
      "loss": 0.9626,
      "step": 12600
    },
    {
      "epoch": 0.12410221331900131,
      "grad_norm": 0.8341017365455627,
      "learning_rate": 4.379916931346201e-05,
      "loss": 0.9289,
      "step": 12700
    },
    {
      "epoch": 0.1250793961010407,
      "grad_norm": 0.805614173412323,
      "learning_rate": 4.37503053994625e-05,
      "loss": 0.9474,
      "step": 12800
    },
    {
      "epoch": 0.12605657888308008,
      "grad_norm": 0.8439397215843201,
      "learning_rate": 4.370144148546299e-05,
      "loss": 0.9661,
      "step": 12900
    },
    {
      "epoch": 0.12703376166511945,
      "grad_norm": 1.1272892951965332,
      "learning_rate": 4.3652577571463476e-05,
      "loss": 0.9514,
      "step": 13000
    },
    {
      "epoch": 0.12801094444715885,
      "grad_norm": 0.6426375508308411,
      "learning_rate": 4.3603713657463965e-05,
      "loss": 0.9448,
      "step": 13100
    },
    {
      "epoch": 0.12898812722919822,
      "grad_norm": 1.3205431699752808,
      "learning_rate": 4.355484974346445e-05,
      "loss": 0.9511,
      "step": 13200
    },
    {
      "epoch": 0.1299653100112376,
      "grad_norm": 0.3671954870223999,
      "learning_rate": 4.350598582946494e-05,
      "loss": 0.9506,
      "step": 13300
    },
    {
      "epoch": 0.130942492793277,
      "grad_norm": 0.7566332817077637,
      "learning_rate": 4.345712191546543e-05,
      "loss": 0.9363,
      "step": 13400
    },
    {
      "epoch": 0.13191967557531636,
      "grad_norm": 0.8800159692764282,
      "learning_rate": 4.340825800146592e-05,
      "loss": 0.9388,
      "step": 13500
    },
    {
      "epoch": 0.13289685835735573,
      "grad_norm": 0.7134628891944885,
      "learning_rate": 4.335939408746641e-05,
      "loss": 0.9162,
      "step": 13600
    },
    {
      "epoch": 0.13387404113939513,
      "grad_norm": 0.5555543899536133,
      "learning_rate": 4.331053017346689e-05,
      "loss": 0.9366,
      "step": 13700
    },
    {
      "epoch": 0.1348512239214345,
      "grad_norm": 0.4485512375831604,
      "learning_rate": 4.326166625946739e-05,
      "loss": 0.9286,
      "step": 13800
    },
    {
      "epoch": 0.13582840670347388,
      "grad_norm": 0.8888948559761047,
      "learning_rate": 4.3212802345467876e-05,
      "loss": 0.943,
      "step": 13900
    },
    {
      "epoch": 0.13680558948551327,
      "grad_norm": 0.6719749569892883,
      "learning_rate": 4.3163938431468364e-05,
      "loss": 0.9217,
      "step": 14000
    },
    {
      "epoch": 0.13778277226755264,
      "grad_norm": 0.695377767086029,
      "learning_rate": 4.311507451746885e-05,
      "loss": 0.9093,
      "step": 14100
    },
    {
      "epoch": 0.13875995504959202,
      "grad_norm": 0.5966312885284424,
      "learning_rate": 4.306621060346934e-05,
      "loss": 0.9195,
      "step": 14200
    },
    {
      "epoch": 0.13973713783163141,
      "grad_norm": 0.8073310256004333,
      "learning_rate": 4.301734668946983e-05,
      "loss": 0.9309,
      "step": 14300
    },
    {
      "epoch": 0.14071432061367078,
      "grad_norm": 0.6303800940513611,
      "learning_rate": 4.2968482775470316e-05,
      "loss": 0.9458,
      "step": 14400
    },
    {
      "epoch": 0.14169150339571016,
      "grad_norm": 0.7043970823287964,
      "learning_rate": 4.2919618861470804e-05,
      "loss": 0.9132,
      "step": 14500
    },
    {
      "epoch": 0.14266868617774955,
      "grad_norm": 0.9100736379623413,
      "learning_rate": 4.287075494747129e-05,
      "loss": 0.9296,
      "step": 14600
    },
    {
      "epoch": 0.14364586895978892,
      "grad_norm": 0.787862241268158,
      "learning_rate": 4.282189103347179e-05,
      "loss": 0.9643,
      "step": 14700
    },
    {
      "epoch": 0.1446230517418283,
      "grad_norm": 0.8169028162956238,
      "learning_rate": 4.277302711947227e-05,
      "loss": 0.9244,
      "step": 14800
    },
    {
      "epoch": 0.1456002345238677,
      "grad_norm": 0.9544184803962708,
      "learning_rate": 4.272416320547276e-05,
      "loss": 0.918,
      "step": 14900
    },
    {
      "epoch": 0.14657741730590707,
      "grad_norm": 0.5325574278831482,
      "learning_rate": 4.2675299291473245e-05,
      "loss": 0.9273,
      "step": 15000
    },
    {
      "epoch": 0.14755460008794646,
      "grad_norm": 1.1403323411941528,
      "learning_rate": 4.262643537747374e-05,
      "loss": 0.9095,
      "step": 15100
    },
    {
      "epoch": 0.14853178286998583,
      "grad_norm": 1.0411937236785889,
      "learning_rate": 4.257757146347423e-05,
      "loss": 0.8967,
      "step": 15200
    },
    {
      "epoch": 0.1495089656520252,
      "grad_norm": 0.630393922328949,
      "learning_rate": 4.2528707549474715e-05,
      "loss": 0.8883,
      "step": 15300
    },
    {
      "epoch": 0.1504861484340646,
      "grad_norm": 0.9445775747299194,
      "learning_rate": 4.2479843635475204e-05,
      "loss": 0.9253,
      "step": 15400
    },
    {
      "epoch": 0.15146333121610397,
      "grad_norm": 0.5689444541931152,
      "learning_rate": 4.243097972147569e-05,
      "loss": 0.8983,
      "step": 15500
    },
    {
      "epoch": 0.15244051399814335,
      "grad_norm": 0.7726677656173706,
      "learning_rate": 4.238211580747618e-05,
      "loss": 0.9228,
      "step": 15600
    },
    {
      "epoch": 0.15341769678018274,
      "grad_norm": 0.8260165452957153,
      "learning_rate": 4.2333251893476675e-05,
      "loss": 0.9202,
      "step": 15700
    },
    {
      "epoch": 0.15439487956222211,
      "grad_norm": 0.4869302809238434,
      "learning_rate": 4.2284387979477156e-05,
      "loss": 0.9283,
      "step": 15800
    },
    {
      "epoch": 0.15537206234426149,
      "grad_norm": 0.5768991708755493,
      "learning_rate": 4.2235524065477644e-05,
      "loss": 0.9233,
      "step": 15900
    },
    {
      "epoch": 0.15634924512630088,
      "grad_norm": 0.8856435418128967,
      "learning_rate": 4.218666015147814e-05,
      "loss": 0.8825,
      "step": 16000
    },
    {
      "epoch": 0.15732642790834026,
      "grad_norm": 0.5258185267448425,
      "learning_rate": 4.213779623747862e-05,
      "loss": 0.8834,
      "step": 16100
    },
    {
      "epoch": 0.15830361069037963,
      "grad_norm": 0.8340526223182678,
      "learning_rate": 4.2088932323479115e-05,
      "loss": 0.8856,
      "step": 16200
    },
    {
      "epoch": 0.15928079347241902,
      "grad_norm": 0.4123723804950714,
      "learning_rate": 4.2040068409479596e-05,
      "loss": 0.8957,
      "step": 16300
    },
    {
      "epoch": 0.1602579762544584,
      "grad_norm": 0.8336274027824402,
      "learning_rate": 4.199120449548009e-05,
      "loss": 0.9053,
      "step": 16400
    },
    {
      "epoch": 0.16123515903649777,
      "grad_norm": 0.7977516055107117,
      "learning_rate": 4.194234058148058e-05,
      "loss": 0.8698,
      "step": 16500
    },
    {
      "epoch": 0.16221234181853716,
      "grad_norm": 0.5064985156059265,
      "learning_rate": 4.189347666748107e-05,
      "loss": 0.8945,
      "step": 16600
    },
    {
      "epoch": 0.16318952460057654,
      "grad_norm": 0.8241267204284668,
      "learning_rate": 4.1844612753481555e-05,
      "loss": 0.8875,
      "step": 16700
    },
    {
      "epoch": 0.1641667073826159,
      "grad_norm": 0.7517113089561462,
      "learning_rate": 4.179574883948204e-05,
      "loss": 0.8845,
      "step": 16800
    },
    {
      "epoch": 0.1651438901646553,
      "grad_norm": 0.6297169923782349,
      "learning_rate": 4.174688492548253e-05,
      "loss": 0.9303,
      "step": 16900
    },
    {
      "epoch": 0.16612107294669468,
      "grad_norm": 0.5828490257263184,
      "learning_rate": 4.1698021011483026e-05,
      "loss": 0.8654,
      "step": 17000
    },
    {
      "epoch": 0.16709825572873405,
      "grad_norm": 0.3038561940193176,
      "learning_rate": 4.164915709748351e-05,
      "loss": 0.8933,
      "step": 17100
    },
    {
      "epoch": 0.16807543851077344,
      "grad_norm": 0.8928827047348022,
      "learning_rate": 4.1600293183484e-05,
      "loss": 0.8509,
      "step": 17200
    },
    {
      "epoch": 0.16905262129281282,
      "grad_norm": 0.7055086493492126,
      "learning_rate": 4.155142926948449e-05,
      "loss": 0.8814,
      "step": 17300
    },
    {
      "epoch": 0.17002980407485221,
      "grad_norm": 0.5377823710441589,
      "learning_rate": 4.150256535548497e-05,
      "loss": 0.888,
      "step": 17400
    },
    {
      "epoch": 0.17100698685689159,
      "grad_norm": 0.6319778561592102,
      "learning_rate": 4.1453701441485466e-05,
      "loss": 0.8575,
      "step": 17500
    },
    {
      "epoch": 0.17198416963893096,
      "grad_norm": 0.8756042122840881,
      "learning_rate": 4.1404837527485954e-05,
      "loss": 0.8805,
      "step": 17600
    },
    {
      "epoch": 0.17296135242097035,
      "grad_norm": 0.5293178558349609,
      "learning_rate": 4.135597361348644e-05,
      "loss": 0.8471,
      "step": 17700
    },
    {
      "epoch": 0.17393853520300973,
      "grad_norm": 0.9118284583091736,
      "learning_rate": 4.130710969948693e-05,
      "loss": 0.8426,
      "step": 17800
    },
    {
      "epoch": 0.1749157179850491,
      "grad_norm": 1.0211195945739746,
      "learning_rate": 4.125824578548742e-05,
      "loss": 0.8877,
      "step": 17900
    },
    {
      "epoch": 0.1758929007670885,
      "grad_norm": 1.4174985885620117,
      "learning_rate": 4.120938187148791e-05,
      "loss": 0.8731,
      "step": 18000
    },
    {
      "epoch": 0.17687008354912787,
      "grad_norm": 0.8243415951728821,
      "learning_rate": 4.1160517957488395e-05,
      "loss": 0.8852,
      "step": 18100
    },
    {
      "epoch": 0.17784726633116724,
      "grad_norm": 0.8385602235794067,
      "learning_rate": 4.111165404348888e-05,
      "loss": 0.8361,
      "step": 18200
    },
    {
      "epoch": 0.17882444911320663,
      "grad_norm": 1.003968358039856,
      "learning_rate": 4.106279012948938e-05,
      "loss": 0.8738,
      "step": 18300
    },
    {
      "epoch": 0.179801631895246,
      "grad_norm": 0.7428449988365173,
      "learning_rate": 4.101392621548986e-05,
      "loss": 0.8563,
      "step": 18400
    },
    {
      "epoch": 0.18077881467728538,
      "grad_norm": 1.8963735103607178,
      "learning_rate": 4.0965062301490354e-05,
      "loss": 0.8428,
      "step": 18500
    },
    {
      "epoch": 0.18175599745932478,
      "grad_norm": 0.6868895888328552,
      "learning_rate": 4.091619838749084e-05,
      "loss": 0.8727,
      "step": 18600
    },
    {
      "epoch": 0.18273318024136415,
      "grad_norm": 1.8936256170272827,
      "learning_rate": 4.086733447349133e-05,
      "loss": 0.9211,
      "step": 18700
    },
    {
      "epoch": 0.18371036302340352,
      "grad_norm": 1.004941463470459,
      "learning_rate": 4.081847055949182e-05,
      "loss": 0.8404,
      "step": 18800
    },
    {
      "epoch": 0.18468754580544292,
      "grad_norm": 1.4084818363189697,
      "learning_rate": 4.0769606645492306e-05,
      "loss": 0.868,
      "step": 18900
    },
    {
      "epoch": 0.1856647285874823,
      "grad_norm": 0.6459541320800781,
      "learning_rate": 4.0720742731492794e-05,
      "loss": 0.8583,
      "step": 19000
    },
    {
      "epoch": 0.18664191136952166,
      "grad_norm": 0.7335548996925354,
      "learning_rate": 4.067187881749328e-05,
      "loss": 0.8622,
      "step": 19100
    },
    {
      "epoch": 0.18761909415156106,
      "grad_norm": 0.6783348321914673,
      "learning_rate": 4.062301490349377e-05,
      "loss": 0.8572,
      "step": 19200
    },
    {
      "epoch": 0.18859627693360043,
      "grad_norm": 0.6323419809341431,
      "learning_rate": 4.057415098949426e-05,
      "loss": 0.8763,
      "step": 19300
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 0.963927686214447,
      "learning_rate": 4.052528707549475e-05,
      "loss": 0.8543,
      "step": 19400
    },
    {
      "epoch": 0.1905506424976792,
      "grad_norm": 0.4785550832748413,
      "learning_rate": 4.0476423161495234e-05,
      "loss": 0.863,
      "step": 19500
    },
    {
      "epoch": 0.19152782527971857,
      "grad_norm": 0.6358627080917358,
      "learning_rate": 4.042755924749573e-05,
      "loss": 0.8842,
      "step": 19600
    },
    {
      "epoch": 0.19250500806175797,
      "grad_norm": 0.7857956886291504,
      "learning_rate": 4.037869533349621e-05,
      "loss": 0.8698,
      "step": 19700
    },
    {
      "epoch": 0.19348219084379734,
      "grad_norm": 0.5225537419319153,
      "learning_rate": 4.0329831419496705e-05,
      "loss": 0.8842,
      "step": 19800
    },
    {
      "epoch": 0.1944593736258367,
      "grad_norm": 0.582313597202301,
      "learning_rate": 4.0280967505497194e-05,
      "loss": 0.8506,
      "step": 19900
    },
    {
      "epoch": 0.1954365564078761,
      "grad_norm": 0.7206740379333496,
      "learning_rate": 4.023210359149768e-05,
      "loss": 0.8529,
      "step": 20000
    },
    {
      "epoch": 0.19641373918991548,
      "grad_norm": 0.45054760575294495,
      "learning_rate": 4.018323967749817e-05,
      "loss": 0.8564,
      "step": 20100
    },
    {
      "epoch": 0.19739092197195485,
      "grad_norm": 0.9214595556259155,
      "learning_rate": 4.013437576349866e-05,
      "loss": 0.8443,
      "step": 20200
    },
    {
      "epoch": 0.19836810475399425,
      "grad_norm": 0.9843263626098633,
      "learning_rate": 4.0085511849499146e-05,
      "loss": 0.856,
      "step": 20300
    },
    {
      "epoch": 0.19934528753603362,
      "grad_norm": 0.6508098840713501,
      "learning_rate": 4.0036647935499634e-05,
      "loss": 0.8532,
      "step": 20400
    },
    {
      "epoch": 0.200322470318073,
      "grad_norm": 0.8091655969619751,
      "learning_rate": 3.998778402150012e-05,
      "loss": 0.8691,
      "step": 20500
    },
    {
      "epoch": 0.20129965310011239,
      "grad_norm": 0.8139657378196716,
      "learning_rate": 3.993892010750061e-05,
      "loss": 0.8608,
      "step": 20600
    },
    {
      "epoch": 0.20227683588215176,
      "grad_norm": 0.628423273563385,
      "learning_rate": 3.9890056193501105e-05,
      "loss": 0.8369,
      "step": 20700
    },
    {
      "epoch": 0.20325401866419113,
      "grad_norm": 1.737331748008728,
      "learning_rate": 3.9841192279501586e-05,
      "loss": 0.8363,
      "step": 20800
    },
    {
      "epoch": 0.20423120144623053,
      "grad_norm": 1.036280870437622,
      "learning_rate": 3.979232836550208e-05,
      "loss": 0.8387,
      "step": 20900
    },
    {
      "epoch": 0.2052083842282699,
      "grad_norm": 0.35834863781929016,
      "learning_rate": 3.974346445150256e-05,
      "loss": 0.8565,
      "step": 21000
    }
  ],
  "logging_steps": 100,
  "max_steps": 102335,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.6340138657251328e+17,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}