{
  "best_metric": 0.2452014982700348,
  "best_model_checkpoint": "./convnext-tiny-new-5e-5/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": 31.272228240966797,
      "learning_rate": 4.998978625997115e-05,
      "loss": 2.7647,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 25.303544998168945,
      "learning_rate": 4.995915338552341e-05,
      "loss": 1.6979,
      "step": 200
    },
    {
      "epoch": 0.27,
      "grad_norm": 23.634435653686523,
      "learning_rate": 4.9908126406754066e-05,
      "loss": 1.3318,
      "step": 300
    },
    {
      "epoch": 0.36,
      "grad_norm": 43.90665817260742,
      "learning_rate": 4.9836747017766765e-05,
      "loss": 1.0622,
      "step": 400
    },
    {
      "epoch": 0.45,
      "grad_norm": 27.907812118530273,
      "learning_rate": 4.9745073542603314e-05,
      "loss": 0.9937,
      "step": 500
    },
    {
      "epoch": 0.55,
      "grad_norm": 24.89875030517578,
      "learning_rate": 4.963318088758713e-05,
      "loss": 0.9094,
      "step": 600
    },
    {
      "epoch": 0.64,
      "grad_norm": 30.73819923400879,
      "learning_rate": 4.950116048011739e-05,
      "loss": 0.8766,
      "step": 700
    },
    {
      "epoch": 0.73,
      "grad_norm": 34.316131591796875,
      "learning_rate": 4.934912019396371e-05,
      "loss": 0.8155,
      "step": 800
    },
    {
      "epoch": 0.82,
      "grad_norm": 25.591245651245117,
      "learning_rate": 4.917718426112262e-05,
      "loss": 0.7545,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 24.966123580932617,
      "learning_rate": 4.898549317030771e-05,
      "loss": 0.7921,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8679920477137177,
      "eval_loss": 0.4486232101917267,
      "eval_runtime": 70.8234,
      "eval_samples_per_second": 35.511,
      "eval_steps_per_second": 2.231,
      "step": 1099
    },
    {
      "epoch": 1.0,
      "grad_norm": 39.45865249633789,
      "learning_rate": 4.877420355215637e-05,
      "loss": 0.8097,
      "step": 1100
    },
    {
      "epoch": 1.09,
      "grad_norm": 20.27395248413086,
      "learning_rate": 4.854348805124703e-05,
      "loss": 0.689,
      "step": 1200
    },
    {
      "epoch": 1.18,
      "grad_norm": 29.029136657714844,
      "learning_rate": 4.829353518503147e-05,
      "loss": 0.6716,
      "step": 1300
    },
    {
      "epoch": 1.27,
      "grad_norm": 21.9611759185791,
      "learning_rate": 4.8024549189797276e-05,
      "loss": 0.6204,
      "step": 1400
    },
    {
      "epoch": 1.36,
      "grad_norm": 34.769710540771484,
      "learning_rate": 4.7736749853786585e-05,
      "loss": 0.6118,
      "step": 1500
    },
    {
      "epoch": 1.46,
      "grad_norm": 24.207292556762695,
      "learning_rate": 4.743037233760728e-05,
      "loss": 0.6005,
      "step": 1600
    },
    {
      "epoch": 1.55,
      "grad_norm": 19.889554977416992,
      "learning_rate": 4.710566698208343e-05,
      "loss": 0.6599,
      "step": 1700
    },
    {
      "epoch": 1.64,
      "grad_norm": 21.616836547851562,
      "learning_rate": 4.676289910370202e-05,
      "loss": 0.6449,
      "step": 1800
    },
    {
      "epoch": 1.73,
      "grad_norm": 24.98408317565918,
      "learning_rate": 4.6402348777823063e-05,
      "loss": 0.5651,
      "step": 1900
    },
    {
      "epoch": 1.82,
      "grad_norm": 16.174781799316406,
      "learning_rate": 4.602431060983022e-05,
      "loss": 0.5636,
      "step": 2000
    },
    {
      "epoch": 1.91,
      "grad_norm": 34.6273193359375,
      "learning_rate": 4.562909349440899e-05,
      "loss": 0.6062,
      "step": 2100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8926441351888668,
      "eval_loss": 0.37534618377685547,
      "eval_runtime": 70.0431,
      "eval_samples_per_second": 35.906,
      "eval_steps_per_second": 2.256,
      "step": 2198
    },
    {
      "epoch": 2.0,
      "grad_norm": 32.93877410888672,
      "learning_rate": 4.5217020363149146e-05,
      "loss": 0.5771,
      "step": 2200
    },
    {
      "epoch": 2.09,
      "grad_norm": 21.045169830322266,
      "learning_rate": 4.478842792067751e-05,
      "loss": 0.4969,
      "step": 2300
    },
    {
      "epoch": 2.18,
      "grad_norm": 21.596593856811523,
      "learning_rate": 4.434366636953695e-05,
      "loss": 0.536,
      "step": 2400
    },
    {
      "epoch": 2.27,
      "grad_norm": 22.82591438293457,
      "learning_rate": 4.388309912403612e-05,
      "loss": 0.496,
      "step": 2500
    },
    {
      "epoch": 2.37,
      "grad_norm": 17.93474006652832,
      "learning_rate": 4.340710251330393e-05,
      "loss": 0.5515,
      "step": 2600
    },
    {
      "epoch": 2.46,
      "grad_norm": 26.939273834228516,
      "learning_rate": 4.2916065473791306e-05,
      "loss": 0.4736,
      "step": 2700
    },
    {
      "epoch": 2.55,
      "grad_norm": 19.711278915405273,
      "learning_rate": 4.241038923147154e-05,
      "loss": 0.5195,
      "step": 2800
    },
    {
      "epoch": 2.64,
      "grad_norm": 27.58098793029785,
      "learning_rate": 4.1890486973998866e-05,
      "loss": 0.5065,
      "step": 2900
    },
    {
      "epoch": 2.73,
      "grad_norm": 10.552973747253418,
      "learning_rate": 4.135678351309313e-05,
      "loss": 0.5157,
      "step": 3000
    },
    {
      "epoch": 2.82,
      "grad_norm": 16.484024047851562,
      "learning_rate": 4.0809714937426516e-05,
      "loss": 0.5058,
      "step": 3100
    },
    {
      "epoch": 2.91,
      "grad_norm": 36.20509338378906,
      "learning_rate": 4.024972825629582e-05,
      "loss": 0.4537,
      "step": 3200
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9113320079522863,
      "eval_loss": 0.3195631802082062,
      "eval_runtime": 69.7597,
      "eval_samples_per_second": 36.052,
      "eval_steps_per_second": 2.265,
      "step": 3297
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.331707000732422,
      "learning_rate": 3.967728103437146e-05,
      "loss": 0.5216,
      "step": 3300
    },
    {
      "epoch": 3.09,
      "grad_norm": 21.994306564331055,
      "learning_rate": 3.909284101782187e-05,
      "loss": 0.413,
      "step": 3400
    },
    {
      "epoch": 3.18,
      "grad_norm": 21.89015007019043,
      "learning_rate": 3.849688575211836e-05,
      "loss": 0.4547,
      "step": 3500
    },
    {
      "epoch": 3.28,
      "grad_norm": 21.92621421813965,
      "learning_rate": 3.788990219183314e-05,
      "loss": 0.445,
      "step": 3600
    },
    {
      "epoch": 3.37,
      "grad_norm": 14.9326810836792,
      "learning_rate": 3.727238630274914e-05,
      "loss": 0.3784,
      "step": 3700
    },
    {
      "epoch": 3.46,
      "grad_norm": 14.167369842529297,
      "learning_rate": 3.664484265660675e-05,
      "loss": 0.4614,
      "step": 3800
    },
    {
      "epoch": 3.55,
      "grad_norm": 15.961216926574707,
      "learning_rate": 3.6007784018818626e-05,
      "loss": 0.4641,
      "step": 3900
    },
    {
      "epoch": 3.64,
      "grad_norm": 21.290969848632812,
      "learning_rate": 3.536173092948955e-05,
      "loss": 0.4814,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "grad_norm": 21.05672264099121,
      "learning_rate": 3.470721127808345e-05,
      "loss": 0.4465,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "grad_norm": 2.6313436031341553,
      "learning_rate": 3.404475987208539e-05,
      "loss": 0.4207,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "grad_norm": 38.63340377807617,
      "learning_rate": 3.3374918000010775e-05,
      "loss": 0.4313,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9240556660039762,
      "eval_loss": 0.27500826120376587,
      "eval_runtime": 68.9229,
      "eval_samples_per_second": 36.49,
      "eval_steps_per_second": 2.292,
      "step": 4396
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.710648059844971,
      "learning_rate": 3.2698232989118956e-05,
      "loss": 0.3547,
      "step": 4400
    },
    {
      "epoch": 4.09,
      "grad_norm": 16.290864944458008,
      "learning_rate": 3.201525775819254e-05,
      "loss": 0.3683,
      "step": 4500
    },
    {
      "epoch": 4.19,
      "grad_norm": 32.59161376953125,
      "learning_rate": 3.132655036574792e-05,
      "loss": 0.3951,
      "step": 4600
    },
    {
      "epoch": 4.28,
      "grad_norm": 33.41773986816406,
      "learning_rate": 3.063267355404608e-05,
      "loss": 0.3669,
      "step": 4700
    },
    {
      "epoch": 4.37,
      "grad_norm": 14.899664878845215,
      "learning_rate": 2.9934194289276367e-05,
      "loss": 0.3503,
      "step": 4800
    },
    {
      "epoch": 4.46,
      "grad_norm": 34.56674575805664,
      "learning_rate": 2.9231683298288853e-05,
      "loss": 0.3445,
      "step": 4900
    },
    {
      "epoch": 4.55,
      "grad_norm": 19.835254669189453,
      "learning_rate": 2.8525714602253885e-05,
      "loss": 0.3648,
      "step": 5000
    },
    {
      "epoch": 4.64,
      "grad_norm": 31.69833755493164,
      "learning_rate": 2.7816865047629847e-05,
      "loss": 0.3827,
      "step": 5100
    },
    {
      "epoch": 4.73,
      "grad_norm": 19.364742279052734,
      "learning_rate": 2.710571383482237e-05,
      "loss": 0.4087,
      "step": 5200
    },
    {
      "epoch": 4.82,
      "grad_norm": 17.92793083190918,
      "learning_rate": 2.6392842044920184e-05,
      "loss": 0.3993,
      "step": 5300
    },
    {
      "epoch": 4.91,
      "grad_norm": 7.249355792999268,
      "learning_rate": 2.5678832164894147e-05,
      "loss": 0.3747,
      "step": 5400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.920079522862823,
      "eval_loss": 0.292158842086792,
      "eval_runtime": 68.7276,
      "eval_samples_per_second": 36.594,
      "eval_steps_per_second": 2.299,
      "step": 5495
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.56585693359375,
      "learning_rate": 2.4964267611647672e-05,
      "loss": 0.3785,
      "step": 5500
    },
    {
      "epoch": 5.1,
      "grad_norm": 23.507110595703125,
      "learning_rate": 2.4249732255307216e-05,
      "loss": 0.3241,
      "step": 5600
    },
    {
      "epoch": 5.19,
      "grad_norm": 25.035778045654297,
      "learning_rate": 2.3535809942142476e-05,
      "loss": 0.324,
      "step": 5700
    },
    {
      "epoch": 5.28,
      "grad_norm": 16.61673355102539,
      "learning_rate": 2.2823084017506024e-05,
      "loss": 0.3357,
      "step": 5800
    },
    {
      "epoch": 5.37,
      "grad_norm": 18.352272033691406,
      "learning_rate": 2.2112136849182368e-05,
      "loss": 0.3176,
      "step": 5900
    },
    {
      "epoch": 5.46,
      "grad_norm": 11.678404808044434,
      "learning_rate": 2.1403549351535627e-05,
      "loss": 0.3258,
      "step": 6000
    },
    {
      "epoch": 5.55,
      "grad_norm": 32.66041564941406,
      "learning_rate": 2.0697900510844873e-05,
      "loss": 0.345,
      "step": 6100
    },
    {
      "epoch": 5.64,
      "grad_norm": 17.961326599121094,
      "learning_rate": 1.9995766912214975e-05,
      "loss": 0.3102,
      "step": 6200
    },
    {
      "epoch": 5.73,
      "grad_norm": 22.43958854675293,
      "learning_rate": 1.9297722268449264e-05,
      "loss": 0.3459,
      "step": 6300
    },
    {
      "epoch": 5.82,
      "grad_norm": 16.841283798217773,
      "learning_rate": 1.8604336951269353e-05,
      "loss": 0.3226,
      "step": 6400
    },
    {
      "epoch": 5.91,
      "grad_norm": 28.8320255279541,
      "learning_rate": 1.7916177525264775e-05,
      "loss": 0.3382,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.927634194831014,
      "eval_loss": 0.2779538929462433,
      "eval_runtime": 68.4746,
      "eval_samples_per_second": 36.729,
      "eval_steps_per_second": 2.307,
      "step": 6594
    },
    {
      "epoch": 6.01,
      "grad_norm": 13.064939498901367,
      "learning_rate": 1.7233806284953614e-05,
      "loss": 0.3289,
      "step": 6600
    },
    {
      "epoch": 6.1,
      "grad_norm": 19.51109504699707,
      "learning_rate": 1.6557780795331985e-05,
      "loss": 0.2805,
      "step": 6700
    },
    {
      "epoch": 6.19,
      "grad_norm": 15.29958724975586,
      "learning_rate": 1.5888653436288196e-05,
      "loss": 0.3253,
      "step": 6800
    },
    {
      "epoch": 6.28,
      "grad_norm": 6.749192714691162,
      "learning_rate": 1.5226970951253588e-05,
      "loss": 0.3179,
      "step": 6900
    },
    {
      "epoch": 6.37,
      "grad_norm": 18.46894073486328,
      "learning_rate": 1.457327400045884e-05,
      "loss": 0.26,
      "step": 7000
    },
    {
      "epoch": 6.46,
      "grad_norm": 2.205888032913208,
      "learning_rate": 1.3928096719160993e-05,
      "loss": 0.3339,
      "step": 7100
    },
    {
      "epoch": 6.55,
      "grad_norm": 6.4344916343688965,
      "learning_rate": 1.3291966281201978e-05,
      "loss": 0.2705,
      "step": 7200
    },
    {
      "epoch": 6.64,
      "grad_norm": 14.4483003616333,
      "learning_rate": 1.2665402468255186e-05,
      "loss": 0.2662,
      "step": 7300
    },
    {
      "epoch": 6.73,
      "grad_norm": 26.225765228271484,
      "learning_rate": 1.2048917245112376e-05,
      "loss": 0.2827,
      "step": 7400
    },
    {
      "epoch": 6.82,
      "grad_norm": 7.608391761779785,
      "learning_rate": 1.1443014341357609e-05,
      "loss": 0.279,
      "step": 7500
    },
    {
      "epoch": 6.92,
      "grad_norm": 1.276964545249939,
      "learning_rate": 1.084818883977024e-05,
      "loss": 0.3065,
      "step": 7600
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9292246520874752,
      "eval_loss": 0.256943941116333,
      "eval_runtime": 69.2749,
      "eval_samples_per_second": 36.305,
      "eval_steps_per_second": 2.281,
      "step": 7693
    },
    {
      "epoch": 7.01,
      "grad_norm": 22.203845977783203,
      "learning_rate": 1.026492677179311e-05,
      "loss": 0.228,
      "step": 7700
    },
    {
      "epoch": 7.1,
      "grad_norm": 29.86297035217285,
      "learning_rate": 9.693704720396693e-06,
      "loss": 0.2649,
      "step": 7800
    },
    {
      "epoch": 7.19,
      "grad_norm": 12.928833961486816,
      "learning_rate": 9.13498943066355e-06,
      "loss": 0.2724,
      "step": 7900
    },
    {
      "epoch": 7.28,
      "grad_norm": 16.028324127197266,
      "learning_rate": 8.589237428411228e-06,
      "loss": 0.2809,
      "step": 8000
    },
    {
      "epoch": 7.37,
      "grad_norm": 28.42373275756836,
      "learning_rate": 8.056894647165414e-06,
      "loss": 0.2733,
      "step": 8100
    },
    {
      "epoch": 7.46,
      "grad_norm": 25.67112159729004,
      "learning_rate": 7.538396063788037e-06,
      "loss": 0.2672,
      "step": 8200
    },
    {
      "epoch": 7.55,
      "grad_norm": 42.678218841552734,
      "learning_rate": 7.0341653430579715e-06,
      "loss": 0.2422,
      "step": 8300
    },
    {
      "epoch": 7.64,
      "grad_norm": 14.172524452209473,
      "learning_rate": 6.544614491494885e-06,
      "loss": 0.2836,
      "step": 8400
    },
    {
      "epoch": 7.73,
      "grad_norm": 6.9926910400390625,
      "learning_rate": 6.070143520709101e-06,
      "loss": 0.2416,
      "step": 8500
    },
    {
      "epoch": 7.83,
      "grad_norm": 25.349214553833008,
      "learning_rate": 5.611140120552358e-06,
      "loss": 0.2877,
      "step": 8600
    },
    {
      "epoch": 7.92,
      "grad_norm": 14.601572036743164,
      "learning_rate": 5.167979342336787e-06,
      "loss": 0.2331,
      "step": 8700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9343936381709742,
      "eval_loss": 0.2538711428642273,
      "eval_runtime": 68.9903,
      "eval_samples_per_second": 36.454,
      "eval_steps_per_second": 2.29,
      "step": 8792
    },
    {
      "epoch": 8.01,
      "grad_norm": 12.48675537109375,
      "learning_rate": 4.7410232923807475e-06,
      "loss": 0.2297,
      "step": 8800
    },
    {
      "epoch": 8.1,
      "grad_norm": 14.78099250793457,
      "learning_rate": 4.330620836132096e-06,
      "loss": 0.2545,
      "step": 8900
    },
    {
      "epoch": 8.19,
      "grad_norm": 12.66365909576416,
      "learning_rate": 3.937107313110449e-06,
      "loss": 0.2315,
      "step": 9000
    },
    {
      "epoch": 8.28,
      "grad_norm": 9.719255447387695,
      "learning_rate": 3.5608042629015707e-06,
      "loss": 0.2011,
      "step": 9100
    },
    {
      "epoch": 8.37,
      "grad_norm": 1.5284979343414307,
      "learning_rate": 3.202019162427611e-06,
      "loss": 0.2473,
      "step": 9200
    },
    {
      "epoch": 8.46,
      "grad_norm": 13.79902458190918,
      "learning_rate": 2.8610451747079658e-06,
      "loss": 0.2237,
      "step": 9300
    },
    {
      "epoch": 8.55,
      "grad_norm": 9.03069019317627,
      "learning_rate": 2.538160909316009e-06,
      "loss": 0.2538,
      "step": 9400
    },
    {
      "epoch": 8.64,
      "grad_norm": 16.241405487060547,
      "learning_rate": 2.233630194727432e-06,
      "loss": 0.2521,
      "step": 9500
    },
    {
      "epoch": 8.74,
      "grad_norm": 18.589529037475586,
      "learning_rate": 1.947701862746201e-06,
      "loss": 0.2537,
      "step": 9600
    },
    {
      "epoch": 8.83,
      "grad_norm": 15.931888580322266,
      "learning_rate": 1.6806095451843002e-06,
      "loss": 0.2759,
      "step": 9700
    },
    {
      "epoch": 8.92,
      "grad_norm": 39.139305114746094,
      "learning_rate": 1.432571482961345e-06,
      "loss": 0.2127,
      "step": 9800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9351888667992048,
      "eval_loss": 0.2480221688747406,
      "eval_runtime": 68.9739,
      "eval_samples_per_second": 36.463,
      "eval_steps_per_second": 2.291,
      "step": 9891
    },
    {
      "epoch": 9.01,
      "grad_norm": 14.2534761428833,
      "learning_rate": 1.203790347780126e-06,
      "loss": 0.2111,
      "step": 9900
    },
    {
      "epoch": 9.1,
      "grad_norm": 9.488052368164062,
      "learning_rate": 9.944530765236993e-07,
      "loss": 0.2245,
      "step": 10000
    },
    {
      "epoch": 9.19,
      "grad_norm": 16.650590896606445,
      "learning_rate": 8.047307185094249e-07,
      "loss": 0.2208,
      "step": 10100
    },
    {
      "epoch": 9.28,
      "grad_norm": 20.264984130859375,
      "learning_rate": 6.347782957246945e-07,
      "loss": 0.2058,
      "step": 10200
    },
    {
      "epoch": 9.37,
      "grad_norm": 16.82746124267578,
      "learning_rate": 4.847346761585963e-07,
      "loss": 0.2441,
      "step": 10300
    },
    {
      "epoch": 9.46,
      "grad_norm": 13.732030868530273,
      "learning_rate": 3.547224603329874e-07,
      "loss": 0.2307,
      "step": 10400
    },
    {
      "epoch": 9.55,
      "grad_norm": 16.514938354492188,
      "learning_rate": 2.448478811257149e-07,
      "loss": 0.262,
      "step": 10500
    },
    {
      "epoch": 9.65,
      "grad_norm": 19.0447940826416,
      "learning_rate": 1.5520071696779604e-07,
      "loss": 0.2637,
      "step": 10600
    },
    {
      "epoch": 9.74,
      "grad_norm": 20.074918746948242,
      "learning_rate": 8.585421848555652e-08,
      "loss": 0.2517,
      "step": 10700
    },
    {
      "epoch": 9.83,
      "grad_norm": 17.162107467651367,
      "learning_rate": 3.686504864758855e-08,
      "loss": 0.2432,
      "step": 10800
    },
    {
      "epoch": 9.92,
      "grad_norm": 22.354610443115234,
      "learning_rate": 8.27323646549194e-09,
      "loss": 0.2409,
      "step": 10900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9343936381709742,
      "eval_loss": 0.2452014982700348,
      "eval_runtime": 69.0419,
      "eval_samples_per_second": 36.427,
      "eval_steps_per_second": 2.288,
      "step": 10990
    },
    {
      "epoch": 10.0,
      "step": 10990,
      "total_flos": 1.301428412334932e+19,
      "train_loss": 0.4456085243259808,
      "train_runtime": 10502.9024,
      "train_samples_per_second": 16.739,
      "train_steps_per_second": 1.046
    }
  ],
  "logging_steps": 100,
  "max_steps": 10990,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.301428412334932e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}