{
  "best_metric": 0.34852299094200134,
  "best_model_checkpoint": "./convnext-base-3e-4/checkpoint-10990",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 10990,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "grad_norm": 17.357587814331055,
      "learning_rate": 0.00029993871755982685,
      "loss": 2.1854,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 28.103832244873047,
      "learning_rate": 0.0002997549203131404,
      "loss": 1.4323,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 15.243412017822266,
      "learning_rate": 0.0002994487584405243,
      "loss": 1.3262,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 13.571104049682617,
      "learning_rate": 0.00029902048210660057,
      "loss": 1.164,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.755253791809082,
      "learning_rate": 0.00029847044125561983,
      "loss": 1.1175,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.938959121704102,
      "learning_rate": 0.00029779908532552276,
      "loss": 1.0117,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.597779273986816,
      "learning_rate": 0.00029700696288070426,
      "loss": 1.0719,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.803329467773438,
      "learning_rate": 0.0002960947211637822,
      "loss": 1.0533,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.126448631286621,
      "learning_rate": 0.00029506310556673567,
      "loss": 0.9138,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.313648223876953,
      "learning_rate": 0.0002939129590218462,
      "loss": 0.8947,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7884691848906561,
      "eval_loss": 0.7363528609275818,
      "eval_runtime": 103.2916,
      "eval_samples_per_second": 24.349,
      "eval_steps_per_second": 1.53,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.587052345275879,
      "learning_rate": 0.00029264522131293815,
      "loss": 0.9902,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.60452938079834,
      "learning_rate": 0.00029126092830748215,
      "loss": 0.8517,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 14.598617553710938,
      "learning_rate": 0.00028976121111018877,
      "loss": 0.802,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.155284881591797,
      "learning_rate": 0.00028814729513878363,
      "loss": 0.7962,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 13.24478816986084,
      "learning_rate": 0.00028642049912271946,
      "loss": 0.7782,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.02603530883789,
      "learning_rate": 0.0002845822340256436,
      "loss": 0.7813,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.052382469177246,
      "learning_rate": 0.00028263400189250057,
      "loss": 0.8079,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.431123733520508,
      "learning_rate": 0.0002805773946222121,
      "loss": 0.8041,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.3062944412231445,
      "learning_rate": 0.00027841409266693835,
      "loss": 0.8019,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.146261215209961,
      "learning_rate": 0.0002761458636589813,
      "loss": 0.679,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 12.160822868347168,
      "learning_rate": 0.0002737745609664539,
      "loss": 0.7643,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8170974155069582,
      "eval_loss": 0.628582775592804,
      "eval_runtime": 103.2407,
      "eval_samples_per_second": 24.361,
      "eval_steps_per_second": 1.53,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.596354484558105,
      "learning_rate": 0.00027130212217889483,
      "loss": 0.7511,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 11.497203826904297,
      "learning_rate": 0.000268730567524065,
      "loss": 0.6527,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.976894855499268,
      "learning_rate": 0.00026606199821722166,
      "loss": 0.6289,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 9.718855857849121,
      "learning_rate": 0.0002632985947442167,
      "loss": 0.6755,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 6.231022357940674,
      "learning_rate": 0.00026044261507982355,
      "loss": 0.6377,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 15.673833847045898,
      "learning_rate": 0.0002574963928427478,
      "loss": 0.626,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 5.745795726776123,
      "learning_rate": 0.00025446233538882923,
      "loss": 0.6276,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 11.555608749389648,
      "learning_rate": 0.00025134292184399317,
      "loss": 0.695,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 7.4662089347839355,
      "learning_rate": 0.00024814070107855875,
      "loss": 0.6095,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 7.660247802734375,
      "learning_rate": 0.00024485828962455907,
      "loss": 0.631,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 9.731380462646484,
      "learning_rate": 0.00024149836953777485,
      "loss": 0.6036,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8481113320079523,
      "eval_loss": 0.5258452892303467,
      "eval_runtime": 104.1346,
      "eval_samples_per_second": 24.151,
      "eval_steps_per_second": 1.517,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.469219207763672,
      "learning_rate": 0.00023806368620622872,
      "loss": 0.5889,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 8.50100040435791,
      "learning_rate": 0.0002345570461069312,
      "loss": 0.5034,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 12.361306190490723,
      "learning_rate": 0.00023098131451271015,
      "loss": 0.5181,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 7.584137439727783,
      "learning_rate": 0.0002273394131509988,
      "loss": 0.5336,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 10.401506423950195,
      "learning_rate": 0.0002236343178164948,
      "loss": 0.5216,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 5.352778434753418,
      "learning_rate": 0.00021986905593964046,
      "loss": 0.4939,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 7.9993109703063965,
      "learning_rate": 0.0002160467041129117,
      "loss": 0.521,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 9.176218032836914,
      "learning_rate": 0.00021217038557693726,
      "loss": 0.5288,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 9.322657585144043,
      "learning_rate": 0.0002082432676685007,
      "loss": 0.5168,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 5.861387252807617,
      "learning_rate": 0.00020426855923251228,
      "loss": 0.5081,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 4.290045261383057,
      "learning_rate": 0.00020024950800006462,
      "loss": 0.5012,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.869582504970179,
      "eval_loss": 0.49109867215156555,
      "eval_runtime": 102.6443,
      "eval_samples_per_second": 24.502,
      "eval_steps_per_second": 1.539,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.063934803009033,
      "learning_rate": 0.0001961893979347137,
      "loss": 0.4754,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 5.163869380950928,
      "learning_rate": 0.00019209154654915522,
      "loss": 0.4471,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 3.733952522277832,
      "learning_rate": 0.0001879593021944875,
      "loss": 0.4004,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 7.615368366241455,
      "learning_rate": 0.00018379604132427648,
      "loss": 0.4076,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 9.600367546081543,
      "learning_rate": 0.0001796051657356582,
      "loss": 0.4035,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 9.41919231414795,
      "learning_rate": 0.0001753900997897331,
      "loss": 0.4281,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 13.647310256958008,
      "learning_rate": 0.00017115428761352327,
      "loss": 0.3674,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 2.1058132648468018,
      "learning_rate": 0.00016690119028577906,
      "loss": 0.3917,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 4.259520053863525,
      "learning_rate": 0.0001626342830089342,
      "loss": 0.3899,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 4.13034200668335,
      "learning_rate": 0.0001583570522695211,
      "loss": 0.4178,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 13.085577011108398,
      "learning_rate": 0.00015407299298936486,
      "loss": 0.3926,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8930417495029821,
      "eval_loss": 0.38039031624794006,
      "eval_runtime": 103.7463,
      "eval_samples_per_second": 24.242,
      "eval_steps_per_second": 1.523,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.522939443588257,
      "learning_rate": 0.000149785605669886,
      "loss": 0.3671,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 5.17425537109375,
      "learning_rate": 0.00014549839353184327,
      "loss": 0.3017,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 6.327219009399414,
      "learning_rate": 0.00014121485965285484,
      "loss": 0.2922,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.7092263102531433,
      "learning_rate": 0.00013693850410503614,
      "loss": 0.314,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.38373059034347534,
      "learning_rate": 0.0001326728210950942,
      "loss": 0.3141,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 5.553852558135986,
      "learning_rate": 0.00012842129610921376,
      "loss": 0.2821,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 3.678790330886841,
      "learning_rate": 0.00012418740306506922,
      "loss": 0.3359,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 4.428023338317871,
      "learning_rate": 0.00011997460147328983,
      "loss": 0.2825,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 3.3043198585510254,
      "learning_rate": 0.00011578633361069557,
      "loss": 0.3317,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 1.317456603050232,
      "learning_rate": 0.0001116260217076161,
      "loss": 0.2983,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 8.99087142944336,
      "learning_rate": 0.00010749706515158862,
      "loss": 0.3348,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8970178926441352,
      "eval_loss": 0.41324833035469055,
      "eval_runtime": 103.3001,
      "eval_samples_per_second": 24.347,
      "eval_steps_per_second": 1.53,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 4.097568988800049,
      "learning_rate": 0.00010340283770972167,
      "loss": 0.3045,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 10.98578929901123,
      "learning_rate": 9.93466847719919e-05,
      "loss": 0.2327,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 13.919866561889648,
      "learning_rate": 9.533192061772917e-05,
      "loss": 0.2696,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 6.072042942047119,
      "learning_rate": 9.136182570752152e-05,
      "loss": 0.2258,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.17451171576976776,
      "learning_rate": 8.743964400275302e-05,
      "loss": 0.2406,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.33122172951698303,
      "learning_rate": 8.356858031496595e-05,
      "loss": 0.2505,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 1.151172161102295,
      "learning_rate": 7.975179768721186e-05,
      "loss": 0.1903,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.7031873464584351,
      "learning_rate": 7.59924148095311e-05,
      "loss": 0.2085,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 6.131903171539307,
      "learning_rate": 7.229350347067424e-05,
      "loss": 0.2471,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 6.110349178314209,
      "learning_rate": 6.865808604814564e-05,
      "loss": 0.2085,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 4.413149833679199,
      "learning_rate": 6.508913303862143e-05,
      "loss": 0.2594,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9153081510934393,
      "eval_loss": 0.3626956641674042,
      "eval_runtime": 103.1814,
      "eval_samples_per_second": 24.375,
      "eval_steps_per_second": 1.531,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 3.3153018951416016,
      "learning_rate": 6.158956063075865e-05,
      "loss": 0.1743,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 2.5595600605010986,
      "learning_rate": 5.816222832238015e-05,
      "loss": 0.1699,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 3.9605636596679688,
      "learning_rate": 5.4809936583981286e-05,
      "loss": 0.2036,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.7597993612289429,
      "learning_rate": 5.1535424570467366e-05,
      "loss": 0.1829,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 5.694293022155762,
      "learning_rate": 4.834136788299248e-05,
      "loss": 0.2039,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.5163713097572327,
      "learning_rate": 4.523037638272821e-05,
      "loss": 0.1764,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 4.396867275238037,
      "learning_rate": 4.220499205834782e-05,
      "loss": 0.1862,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.054451316595077515,
      "learning_rate": 3.926768694896931e-05,
      "loss": 0.1773,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.23744052648544312,
      "learning_rate": 3.64208611242546e-05,
      "loss": 0.1648,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 3.540268659591675,
      "learning_rate": 3.366684072331414e-05,
      "loss": 0.1541,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.33744722604751587,
      "learning_rate": 3.100787605402072e-05,
      "loss": 0.1751,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9308151093439364,
      "eval_loss": 0.3506681025028229,
      "eval_runtime": 103.8384,
      "eval_samples_per_second": 24.22,
      "eval_steps_per_second": 1.522,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.13831892609596252,
      "learning_rate": 2.844613975428448e-05,
      "loss": 0.1472,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 2.3098576068878174,
      "learning_rate": 2.5983725016792572e-05,
      "loss": 0.1772,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.10428429394960403,
      "learning_rate": 2.3622643878662696e-05,
      "loss": 0.1524,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 9.646160125732422,
      "learning_rate": 2.1364825577409422e-05,
      "loss": 0.1023,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.3879956305027008,
      "learning_rate": 1.9212114974565664e-05,
      "loss": 0.1421,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.022449787706136703,
      "learning_rate": 1.7166271048247792e-05,
      "loss": 0.1101,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 4.682805061340332,
      "learning_rate": 1.5228965455896053e-05,
      "loss": 0.1355,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 4.212618350982666,
      "learning_rate": 1.3401781168364589e-05,
      "loss": 0.1465,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.17709462344646454,
      "learning_rate": 1.1686211176477206e-05,
      "loss": 0.1375,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 7.981707572937012,
      "learning_rate": 1.0083657271105799e-05,
      "loss": 0.1498,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 7.477297306060791,
      "learning_rate": 8.59542889776807e-06,
      "loss": 0.1613,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9300198807157057,
      "eval_loss": 0.34880414605140686,
      "eval_runtime": 103.3381,
      "eval_samples_per_second": 24.338,
      "eval_steps_per_second": 1.529,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 1.4414594173431396,
      "learning_rate": 7.222742086680755e-06,
      "loss": 0.1335,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 4.2091474533081055,
      "learning_rate": 5.966718459142195e-06,
      "loss": 0.1066,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 18.196033477783203,
      "learning_rate": 4.828384311056549e-06,
      "loss": 0.125,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 2.323212146759033,
      "learning_rate": 3.8086697743481664e-06,
      "loss": 0.1239,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.2004556953907013,
      "learning_rate": 2.9084080569515775e-06,
      "loss": 0.1076,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 12.655696868896484,
      "learning_rate": 2.128334761997924e-06,
      "loss": 0.1054,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.03721316158771515,
      "learning_rate": 1.469087286754289e-06,
      "loss": 0.125,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 3.7458605766296387,
      "learning_rate": 9.31204301806776e-07,
      "loss": 0.1161,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.63627690076828,
      "learning_rate": 5.151253109133391e-07,
      "loss": 0.1342,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 7.162803649902344,
      "learning_rate": 2.211902918855313e-07,
      "loss": 0.1365,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 1.5799212455749512,
      "learning_rate": 4.9639418792951634e-08,
      "loss": 0.1102,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9280318091451292,
      "eval_loss": 0.34852299094200134,
      "eval_runtime": 103.4235,
      "eval_samples_per_second": 24.317,
      "eval_steps_per_second": 1.528,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 4.09349935387607e+19,
      "train_loss": 0.45005689912540897,
      "train_runtime": 16363.7477,
      "train_samples_per_second": 10.744,
      "train_steps_per_second": 0.672
    }
  ],
  "logging_steps": 100,
  "max_steps": 10990,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 4.09349935387607e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}