{
  "best_metric": 0.8002366534419155,
  "best_model_checkpoint": "outputs/t5-mini/weak_tiny_poe/qqp_21/checkpoint-56855",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 56855,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "grad_norm": 1.4761992692947388,
      "learning_rate": 4.956028493536189e-05,
      "loss": 0.7583,
      "step": 500
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.450453519821167,
      "learning_rate": 4.9120569870723775e-05,
      "loss": 0.6838,
      "step": 1000
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.613905668258667,
      "learning_rate": 4.868085480608566e-05,
      "loss": 0.6626,
      "step": 1500
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.006730079650879,
      "learning_rate": 4.824113974144755e-05,
      "loss": 0.6505,
      "step": 2000
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.0432426929473877,
      "learning_rate": 4.780142467680943e-05,
      "loss": 0.6357,
      "step": 2500
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.369479179382324,
      "learning_rate": 4.736170961217132e-05,
      "loss": 0.6192,
      "step": 3000
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.6632206439971924,
      "learning_rate": 4.6921994547533196e-05,
      "loss": 0.6158,
      "step": 3500
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.6710329055786133,
      "learning_rate": 4.6482279482895086e-05,
      "loss": 0.6163,
      "step": 4000
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6218762397766113,
      "learning_rate": 4.6042564418256975e-05,
      "loss": 0.6077,
      "step": 4500
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.753265142440796,
      "learning_rate": 4.560284935361886e-05,
      "loss": 0.6084,
      "step": 5000
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.213019371032715,
      "learning_rate": 4.516313428898074e-05,
      "loss": 0.5951,
      "step": 5500
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.2585835456848145,
      "learning_rate": 4.472341922434263e-05,
      "loss": 0.5929,
      "step": 6000
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.8762115240097046,
      "learning_rate": 4.428370415970451e-05,
      "loss": 0.5815,
      "step": 6500
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8141961097717285,
      "learning_rate": 4.38439890950664e-05,
      "loss": 0.5799,
      "step": 7000
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5737587213516235,
      "learning_rate": 4.3404274030428286e-05,
      "loss": 0.582,
      "step": 7500
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7630951404571533,
      "learning_rate": 4.296455896579017e-05,
      "loss": 0.5833,
      "step": 8000
    },
    {
      "epoch": 0.75,
      "grad_norm": 3.324857234954834,
      "learning_rate": 4.252484390115206e-05,
      "loss": 0.569,
      "step": 8500
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.0849748849868774,
      "learning_rate": 4.208512883651394e-05,
      "loss": 0.5664,
      "step": 9000
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.513084650039673,
      "learning_rate": 4.1645413771875824e-05,
      "loss": 0.5655,
      "step": 9500
    },
    {
      "epoch": 0.88,
      "grad_norm": 3.041181802749634,
      "learning_rate": 4.1205698707237714e-05,
      "loss": 0.5666,
      "step": 10000
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.9409087896347046,
      "learning_rate": 4.0765983642599596e-05,
      "loss": 0.5612,
      "step": 10500
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.8327817916870117,
      "learning_rate": 4.0326268577961486e-05,
      "loss": 0.5616,
      "step": 11000
    },
    {
      "epoch": 1.0,
      "eval_combined_score": 0.75431762556448,
      "eval_f1": 0.75431762556448,
      "eval_loss": 0.39584288001060486,
      "eval_runtime": 35.7699,
      "eval_samples_per_second": 1130.28,
      "eval_steps_per_second": 4.417,
      "step": 11371
    },
    {
      "epoch": 1.01,
      "grad_norm": 3.406697988510132,
      "learning_rate": 3.988655351332337e-05,
      "loss": 0.5595,
      "step": 11500
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.2096521854400635,
      "learning_rate": 3.944683844868525e-05,
      "loss": 0.5484,
      "step": 12000
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.2976303100585938,
      "learning_rate": 3.900712338404714e-05,
      "loss": 0.5485,
      "step": 12500
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.539963960647583,
      "learning_rate": 3.8567408319409024e-05,
      "loss": 0.5468,
      "step": 13000
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.3027236461639404,
      "learning_rate": 3.812769325477091e-05,
      "loss": 0.5472,
      "step": 13500
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.070688247680664,
      "learning_rate": 3.76879781901328e-05,
      "loss": 0.5425,
      "step": 14000
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.171999931335449,
      "learning_rate": 3.724826312549468e-05,
      "loss": 0.5409,
      "step": 14500
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.4110608100891113,
      "learning_rate": 3.680854806085657e-05,
      "loss": 0.5482,
      "step": 15000
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.1505320072174072,
      "learning_rate": 3.636883299621845e-05,
      "loss": 0.5381,
      "step": 15500
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.6976615190505981,
      "learning_rate": 3.5929117931580335e-05,
      "loss": 0.5403,
      "step": 16000
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.6042710542678833,
      "learning_rate": 3.5489402866942224e-05,
      "loss": 0.5408,
      "step": 16500
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.5285661220550537,
      "learning_rate": 3.504968780230411e-05,
      "loss": 0.5372,
      "step": 17000
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.611952543258667,
      "learning_rate": 3.4609972737666e-05,
      "loss": 0.5316,
      "step": 17500
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.74088978767395,
      "learning_rate": 3.417025767302788e-05,
      "loss": 0.5417,
      "step": 18000
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.620376706123352,
      "learning_rate": 3.373054260838976e-05,
      "loss": 0.5291,
      "step": 18500
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.3796972036361694,
      "learning_rate": 3.329082754375165e-05,
      "loss": 0.536,
      "step": 19000
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.2904984951019287,
      "learning_rate": 3.2851112479113535e-05,
      "loss": 0.5188,
      "step": 19500
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.1823525428771973,
      "learning_rate": 3.241139741447542e-05,
      "loss": 0.5252,
      "step": 20000
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.8540767431259155,
      "learning_rate": 3.197168234983731e-05,
      "loss": 0.532,
      "step": 20500
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.533015251159668,
      "learning_rate": 3.153196728519919e-05,
      "loss": 0.5316,
      "step": 21000
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.5881567001342773,
      "learning_rate": 3.109225222056108e-05,
      "loss": 0.5338,
      "step": 21500
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.6812840700149536,
      "learning_rate": 3.065253715592296e-05,
      "loss": 0.5298,
      "step": 22000
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.3194057941436768,
      "learning_rate": 3.021282209128485e-05,
      "loss": 0.5277,
      "step": 22500
    },
    {
      "epoch": 2.0,
      "eval_combined_score": 0.7885976005742216,
      "eval_f1": 0.7885976005742216,
      "eval_loss": 0.3457885980606079,
      "eval_runtime": 35.6671,
      "eval_samples_per_second": 1133.538,
      "eval_steps_per_second": 4.43,
      "step": 22742
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.574708938598633,
      "learning_rate": 2.9773107026646735e-05,
      "loss": 0.5285,
      "step": 23000
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.895447254180908,
      "learning_rate": 2.933339196200862e-05,
      "loss": 0.513,
      "step": 23500
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.720695972442627,
      "learning_rate": 2.8893676897370504e-05,
      "loss": 0.5252,
      "step": 24000
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.181859016418457,
      "learning_rate": 2.845396183273239e-05,
      "loss": 0.5148,
      "step": 24500
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.2445052862167358,
      "learning_rate": 2.8014246768094277e-05,
      "loss": 0.5174,
      "step": 25000
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.124485731124878,
      "learning_rate": 2.7574531703456163e-05,
      "loss": 0.5052,
      "step": 25500
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.8577401638031006,
      "learning_rate": 2.7134816638818046e-05,
      "loss": 0.5134,
      "step": 26000
    },
    {
      "epoch": 2.33,
      "grad_norm": 3.3754642009735107,
      "learning_rate": 2.6695101574179932e-05,
      "loss": 0.5106,
      "step": 26500
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.662766695022583,
      "learning_rate": 2.625538650954182e-05,
      "loss": 0.5131,
      "step": 27000
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.0520458221435547,
      "learning_rate": 2.5815671444903705e-05,
      "loss": 0.5048,
      "step": 27500
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.1979711055755615,
      "learning_rate": 2.5375956380265588e-05,
      "loss": 0.5161,
      "step": 28000
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.9927352666854858,
      "learning_rate": 2.4936241315627474e-05,
      "loss": 0.5083,
      "step": 28500
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.1663730144500732,
      "learning_rate": 2.449652625098936e-05,
      "loss": 0.5131,
      "step": 29000
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.8424278497695923,
      "learning_rate": 2.4056811186351246e-05,
      "loss": 0.5076,
      "step": 29500
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.6461939811706543,
      "learning_rate": 2.3617096121713133e-05,
      "loss": 0.5064,
      "step": 30000
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.1023998260498047,
      "learning_rate": 2.317738105707502e-05,
      "loss": 0.5085,
      "step": 30500
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.8294382095336914,
      "learning_rate": 2.27376659924369e-05,
      "loss": 0.503,
      "step": 31000
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.6910927295684814,
      "learning_rate": 2.2297950927798788e-05,
      "loss": 0.5075,
      "step": 31500
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.7387686967849731,
      "learning_rate": 2.1858235863160674e-05,
      "loss": 0.5057,
      "step": 32000
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.62984037399292,
      "learning_rate": 2.141852079852256e-05,
      "loss": 0.5058,
      "step": 32500
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.4338366985321045,
      "learning_rate": 2.0978805733884443e-05,
      "loss": 0.5105,
      "step": 33000
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.7565324306488037,
      "learning_rate": 2.053909066924633e-05,
      "loss": 0.5046,
      "step": 33500
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.6135342121124268,
      "learning_rate": 2.0099375604608216e-05,
      "loss": 0.5071,
      "step": 34000
    },
    {
      "epoch": 3.0,
      "eval_combined_score": 0.7936085219707057,
      "eval_f1": 0.7936085219707057,
      "eval_loss": 0.3366371691226959,
      "eval_runtime": 35.2824,
      "eval_samples_per_second": 1145.898,
      "eval_steps_per_second": 4.478,
      "step": 34113
    },
    {
      "epoch": 3.03,
      "grad_norm": 3.032174587249756,
      "learning_rate": 1.9659660539970102e-05,
      "loss": 0.4951,
      "step": 34500
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.865243673324585,
      "learning_rate": 1.9219945475331985e-05,
      "loss": 0.4966,
      "step": 35000
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.8979029655456543,
      "learning_rate": 1.878023041069387e-05,
      "loss": 0.4971,
      "step": 35500
    },
    {
      "epoch": 3.17,
      "grad_norm": 2.6854591369628906,
      "learning_rate": 1.8340515346055757e-05,
      "loss": 0.4928,
      "step": 36000
    },
    {
      "epoch": 3.21,
      "grad_norm": 2.190274953842163,
      "learning_rate": 1.7900800281417643e-05,
      "loss": 0.499,
      "step": 36500
    },
    {
      "epoch": 3.25,
      "grad_norm": 2.532625913619995,
      "learning_rate": 1.7461085216779526e-05,
      "loss": 0.4995,
      "step": 37000
    },
    {
      "epoch": 3.3,
      "grad_norm": 2.1517629623413086,
      "learning_rate": 1.7021370152141413e-05,
      "loss": 0.493,
      "step": 37500
    },
    {
      "epoch": 3.34,
      "grad_norm": 2.9633235931396484,
      "learning_rate": 1.65816550875033e-05,
      "loss": 0.5014,
      "step": 38000
    },
    {
      "epoch": 3.39,
      "grad_norm": 2.1532862186431885,
      "learning_rate": 1.6141940022865185e-05,
      "loss": 0.499,
      "step": 38500
    },
    {
      "epoch": 3.43,
      "grad_norm": 2.8993899822235107,
      "learning_rate": 1.5702224958227068e-05,
      "loss": 0.4934,
      "step": 39000
    },
    {
      "epoch": 3.47,
      "grad_norm": 4.290128231048584,
      "learning_rate": 1.5262509893588954e-05,
      "loss": 0.4923,
      "step": 39500
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.6720905303955078,
      "learning_rate": 1.482279482895084e-05,
      "loss": 0.4873,
      "step": 40000
    },
    {
      "epoch": 3.56,
      "grad_norm": 2.33063006401062,
      "learning_rate": 1.4383079764312727e-05,
      "loss": 0.4979,
      "step": 40500
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.8272016048431396,
      "learning_rate": 1.3943364699674611e-05,
      "loss": 0.4912,
      "step": 41000
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.9617942571640015,
      "learning_rate": 1.3503649635036497e-05,
      "loss": 0.5045,
      "step": 41500
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.7576007843017578,
      "learning_rate": 1.3063934570398382e-05,
      "loss": 0.4834,
      "step": 42000
    },
    {
      "epoch": 3.74,
      "grad_norm": 3.1541335582733154,
      "learning_rate": 1.2624219505760268e-05,
      "loss": 0.4893,
      "step": 42500
    },
    {
      "epoch": 3.78,
      "grad_norm": 2.4728527069091797,
      "learning_rate": 1.2184504441122153e-05,
      "loss": 0.4908,
      "step": 43000
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.833243727684021,
      "learning_rate": 1.1744789376484039e-05,
      "loss": 0.4905,
      "step": 43500
    },
    {
      "epoch": 3.87,
      "grad_norm": 2.1114351749420166,
      "learning_rate": 1.1305074311845925e-05,
      "loss": 0.4971,
      "step": 44000
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.988950490951538,
      "learning_rate": 1.086535924720781e-05,
      "loss": 0.4954,
      "step": 44500
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.7815330028533936,
      "learning_rate": 1.0425644182569696e-05,
      "loss": 0.4861,
      "step": 45000
    },
    {
      "epoch": 4.0,
      "eval_combined_score": 0.7983325743510702,
      "eval_f1": 0.7983325743510702,
      "eval_loss": 0.329210489988327,
      "eval_runtime": 36.1124,
      "eval_samples_per_second": 1119.561,
      "eval_steps_per_second": 4.375,
      "step": 45484
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9500830173492432,
      "learning_rate": 9.98592911793158e-06,
      "loss": 0.4975,
      "step": 45500
    },
    {
      "epoch": 4.05,
      "grad_norm": 2.2637670040130615,
      "learning_rate": 9.546214053293467e-06,
      "loss": 0.4919,
      "step": 46000
    },
    {
      "epoch": 4.09,
      "grad_norm": 2.1222739219665527,
      "learning_rate": 9.106498988655351e-06,
      "loss": 0.4891,
      "step": 46500
    },
    {
      "epoch": 4.13,
      "grad_norm": 2.9774413108825684,
      "learning_rate": 8.666783924017237e-06,
      "loss": 0.4823,
      "step": 47000
    },
    {
      "epoch": 4.18,
      "grad_norm": 1.4093037843704224,
      "learning_rate": 8.227068859379122e-06,
      "loss": 0.4945,
      "step": 47500
    },
    {
      "epoch": 4.22,
      "grad_norm": 2.2752432823181152,
      "learning_rate": 7.787353794741008e-06,
      "loss": 0.4803,
      "step": 48000
    },
    {
      "epoch": 4.27,
      "grad_norm": 3.996915340423584,
      "learning_rate": 7.347638730102893e-06,
      "loss": 0.4786,
      "step": 48500
    },
    {
      "epoch": 4.31,
      "grad_norm": 2.5842177867889404,
      "learning_rate": 6.907923665464779e-06,
      "loss": 0.4902,
      "step": 49000
    },
    {
      "epoch": 4.35,
      "grad_norm": 1.688774824142456,
      "learning_rate": 6.468208600826664e-06,
      "loss": 0.4828,
      "step": 49500
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.9284757375717163,
      "learning_rate": 6.02849353618855e-06,
      "loss": 0.4737,
      "step": 50000
    },
    {
      "epoch": 4.44,
      "grad_norm": 2.362175226211548,
      "learning_rate": 5.588778471550435e-06,
      "loss": 0.4893,
      "step": 50500
    },
    {
      "epoch": 4.49,
      "grad_norm": 2.7047996520996094,
      "learning_rate": 5.149063406912321e-06,
      "loss": 0.4824,
      "step": 51000
    },
    {
      "epoch": 4.53,
      "grad_norm": 2.2833824157714844,
      "learning_rate": 4.709348342274207e-06,
      "loss": 0.4862,
      "step": 51500
    },
    {
      "epoch": 4.57,
      "grad_norm": 2.7893359661102295,
      "learning_rate": 4.269633277636092e-06,
      "loss": 0.4706,
      "step": 52000
    },
    {
      "epoch": 4.62,
      "grad_norm": 1.4342833757400513,
      "learning_rate": 3.8299182129979776e-06,
      "loss": 0.4892,
      "step": 52500
    },
    {
      "epoch": 4.66,
      "grad_norm": 1.9848226308822632,
      "learning_rate": 3.3902031483598625e-06,
      "loss": 0.4876,
      "step": 53000
    },
    {
      "epoch": 4.7,
      "grad_norm": 2.6507408618927,
      "learning_rate": 2.9504880837217483e-06,
      "loss": 0.4764,
      "step": 53500
    },
    {
      "epoch": 4.75,
      "grad_norm": 2.80248761177063,
      "learning_rate": 2.5107730190836337e-06,
      "loss": 0.4847,
      "step": 54000
    },
    {
      "epoch": 4.79,
      "grad_norm": 2.777669668197632,
      "learning_rate": 2.0710579544455195e-06,
      "loss": 0.4924,
      "step": 54500
    },
    {
      "epoch": 4.84,
      "grad_norm": 2.0418293476104736,
      "learning_rate": 1.6313428898074047e-06,
      "loss": 0.485,
      "step": 55000
    },
    {
      "epoch": 4.88,
      "grad_norm": 2.3356010913848877,
      "learning_rate": 1.1916278251692903e-06,
      "loss": 0.4844,
      "step": 55500
    },
    {
      "epoch": 4.92,
      "grad_norm": 3.416121244430542,
      "learning_rate": 7.519127605311758e-07,
      "loss": 0.4877,
      "step": 56000
    },
    {
      "epoch": 4.97,
      "grad_norm": 1.7180390357971191,
      "learning_rate": 3.121976958930613e-07,
      "loss": 0.4822,
      "step": 56500
    },
    {
      "epoch": 5.0,
      "eval_combined_score": 0.8002366534419155,
      "eval_f1": 0.8002366534419155,
      "eval_loss": 0.3343241214752197,
      "eval_runtime": 35.6088,
      "eval_samples_per_second": 1135.394,
      "eval_steps_per_second": 4.437,
      "step": 56855
    },
    {
      "epoch": 5.0,
      "step": 56855,
      "total_flos": 0.0,
      "train_loss": 0.5264342698119119,
      "train_runtime": 6200.063,
      "train_samples_per_second": 293.421,
      "train_steps_per_second": 9.17
    }
  ],
  "logging_steps": 500,
  "max_steps": 56855,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}