{
  "best_metric": 0.8172281558013501,
  "best_model_checkpoint": "outputs/t5-mini/weak_tiny_poe/qqp_87/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": 1.4921129941940308,
      "learning_rate": 4.956028493536189e-05,
      "loss": 0.731,
      "step": 500
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.265256643295288,
      "learning_rate": 4.9120569870723775e-05,
      "loss": 0.6589,
      "step": 1000
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7359956502914429,
      "learning_rate": 4.868085480608566e-05,
      "loss": 0.6362,
      "step": 1500
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5481916666030884,
      "learning_rate": 4.824113974144755e-05,
      "loss": 0.6255,
      "step": 2000
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.8537436723709106,
      "learning_rate": 4.780142467680943e-05,
      "loss": 0.6048,
      "step": 2500
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7439799308776855,
      "learning_rate": 4.736170961217132e-05,
      "loss": 0.5956,
      "step": 3000
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.284893274307251,
      "learning_rate": 4.6921994547533196e-05,
      "loss": 0.6007,
      "step": 3500
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.466986894607544,
      "learning_rate": 4.6482279482895086e-05,
      "loss": 0.6004,
      "step": 4000
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.0524117946624756,
      "learning_rate": 4.6042564418256975e-05,
      "loss": 0.5837,
      "step": 4500
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.0001771450042725,
      "learning_rate": 4.560284935361886e-05,
      "loss": 0.5894,
      "step": 5000
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.2638933658599854,
      "learning_rate": 4.516313428898074e-05,
      "loss": 0.5752,
      "step": 5500
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.2692410945892334,
      "learning_rate": 4.472341922434263e-05,
      "loss": 0.5761,
      "step": 6000
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.9219467639923096,
      "learning_rate": 4.428370415970451e-05,
      "loss": 0.5757,
      "step": 6500
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.599557876586914,
      "learning_rate": 4.38439890950664e-05,
      "loss": 0.5667,
      "step": 7000
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.1363203525543213,
      "learning_rate": 4.3404274030428286e-05,
      "loss": 0.5578,
      "step": 7500
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.675701141357422,
      "learning_rate": 4.296455896579017e-05,
      "loss": 0.5579,
      "step": 8000
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.613674283027649,
      "learning_rate": 4.252484390115206e-05,
      "loss": 0.5535,
      "step": 8500
    },
    {
      "epoch": 0.79,
      "grad_norm": 3.672745704650879,
      "learning_rate": 4.208512883651394e-05,
      "loss": 0.5465,
      "step": 9000
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.1016011238098145,
      "learning_rate": 4.1645413771875824e-05,
      "loss": 0.5457,
      "step": 9500
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.0293595790863037,
      "learning_rate": 4.1205698707237714e-05,
      "loss": 0.5425,
      "step": 10000
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.5119960308074951,
      "learning_rate": 4.0765983642599596e-05,
      "loss": 0.5389,
      "step": 10500
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.8837521076202393,
      "learning_rate": 4.0326268577961486e-05,
      "loss": 0.5411,
      "step": 11000
    },
    {
      "epoch": 1.0,
      "eval_combined_score": 0.7816986623933021,
      "eval_f1": 0.7816986623933021,
      "eval_loss": 0.37052249908447266,
      "eval_runtime": 38.1713,
      "eval_samples_per_second": 1059.173,
      "eval_steps_per_second": 4.139,
      "step": 11371
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.248117685317993,
      "learning_rate": 3.988655351332337e-05,
      "loss": 0.5388,
      "step": 11500
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.1271088123321533,
      "learning_rate": 3.944683844868525e-05,
      "loss": 0.5356,
      "step": 12000
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.2512476444244385,
      "learning_rate": 3.900712338404714e-05,
      "loss": 0.532,
      "step": 12500
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.627472996711731,
      "learning_rate": 3.8567408319409024e-05,
      "loss": 0.5307,
      "step": 13000
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.581695556640625,
      "learning_rate": 3.812769325477091e-05,
      "loss": 0.5291,
      "step": 13500
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6895989179611206,
      "learning_rate": 3.76879781901328e-05,
      "loss": 0.5168,
      "step": 14000
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.5524585247039795,
      "learning_rate": 3.724826312549468e-05,
      "loss": 0.5243,
      "step": 14500
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.877912163734436,
      "learning_rate": 3.680854806085657e-05,
      "loss": 0.5249,
      "step": 15000
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.8292911052703857,
      "learning_rate": 3.636883299621845e-05,
      "loss": 0.5125,
      "step": 15500
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.8245925903320312,
      "learning_rate": 3.5929117931580335e-05,
      "loss": 0.52,
      "step": 16000
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.095343112945557,
      "learning_rate": 3.5489402866942224e-05,
      "loss": 0.5165,
      "step": 16500
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.9062578678131104,
      "learning_rate": 3.504968780230411e-05,
      "loss": 0.5177,
      "step": 17000
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.735477924346924,
      "learning_rate": 3.4609972737666e-05,
      "loss": 0.5117,
      "step": 17500
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.8982046842575073,
      "learning_rate": 3.417025767302788e-05,
      "loss": 0.5061,
      "step": 18000
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.2440438270568848,
      "learning_rate": 3.373054260838976e-05,
      "loss": 0.5126,
      "step": 18500
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.681962728500366,
      "learning_rate": 3.329082754375165e-05,
      "loss": 0.5066,
      "step": 19000
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.9211692810058594,
      "learning_rate": 3.2851112479113535e-05,
      "loss": 0.5197,
      "step": 19500
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4573030471801758,
      "learning_rate": 3.241139741447542e-05,
      "loss": 0.507,
      "step": 20000
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.5176600217819214,
      "learning_rate": 3.197168234983731e-05,
      "loss": 0.5023,
      "step": 20500
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.4117181301116943,
      "learning_rate": 3.153196728519919e-05,
      "loss": 0.4966,
      "step": 21000
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.1353968381881714,
      "learning_rate": 3.109225222056108e-05,
      "loss": 0.5071,
      "step": 21500
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.829007863998413,
      "learning_rate": 3.065253715592296e-05,
      "loss": 0.5053,
      "step": 22000
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.5963010787963867,
      "learning_rate": 3.021282209128485e-05,
      "loss": 0.5084,
      "step": 22500
    },
    {
      "epoch": 2.0,
      "eval_combined_score": 0.7990887806112763,
      "eval_f1": 0.7990887806112763,
      "eval_loss": 0.35578909516334534,
      "eval_runtime": 38.0326,
      "eval_samples_per_second": 1063.035,
      "eval_steps_per_second": 4.154,
      "step": 22742
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.8584606647491455,
      "learning_rate": 2.9773107026646735e-05,
      "loss": 0.5022,
      "step": 23000
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.9419100284576416,
      "learning_rate": 2.933339196200862e-05,
      "loss": 0.4968,
      "step": 23500
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.221026659011841,
      "learning_rate": 2.8893676897370504e-05,
      "loss": 0.4884,
      "step": 24000
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.75677752494812,
      "learning_rate": 2.845396183273239e-05,
      "loss": 0.4963,
      "step": 24500
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.5008267164230347,
      "learning_rate": 2.8014246768094277e-05,
      "loss": 0.4963,
      "step": 25000
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.5264431238174438,
      "learning_rate": 2.7574531703456163e-05,
      "loss": 0.4981,
      "step": 25500
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.450463056564331,
      "learning_rate": 2.7134816638818046e-05,
      "loss": 0.4961,
      "step": 26000
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.3188283443450928,
      "learning_rate": 2.6695101574179932e-05,
      "loss": 0.4917,
      "step": 26500
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.173543691635132,
      "learning_rate": 2.625538650954182e-05,
      "loss": 0.4923,
      "step": 27000
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.7373517155647278,
      "learning_rate": 2.5815671444903705e-05,
      "loss": 0.489,
      "step": 27500
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.6519461870193481,
      "learning_rate": 2.5375956380265588e-05,
      "loss": 0.494,
      "step": 28000
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.8422359228134155,
      "learning_rate": 2.4936241315627474e-05,
      "loss": 0.5016,
      "step": 28500
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.1333723068237305,
      "learning_rate": 2.449652625098936e-05,
      "loss": 0.4868,
      "step": 29000
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.8119902610778809,
      "learning_rate": 2.4056811186351246e-05,
      "loss": 0.4885,
      "step": 29500
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.8639763593673706,
      "learning_rate": 2.3617096121713133e-05,
      "loss": 0.4911,
      "step": 30000
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.5926486253738403,
      "learning_rate": 2.317738105707502e-05,
      "loss": 0.4889,
      "step": 30500
    },
    {
      "epoch": 2.73,
      "grad_norm": 3.753464460372925,
      "learning_rate": 2.27376659924369e-05,
      "loss": 0.4905,
      "step": 31000
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.9747668504714966,
      "learning_rate": 2.2297950927798788e-05,
      "loss": 0.4871,
      "step": 31500
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.5802444219589233,
      "learning_rate": 2.1858235863160674e-05,
      "loss": 0.4846,
      "step": 32000
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.8520300388336182,
      "learning_rate": 2.141852079852256e-05,
      "loss": 0.4839,
      "step": 32500
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.0941038131713867,
      "learning_rate": 2.0978805733884443e-05,
      "loss": 0.4769,
      "step": 33000
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.694678783416748,
      "learning_rate": 2.053909066924633e-05,
      "loss": 0.4798,
      "step": 33500
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.0499818325042725,
      "learning_rate": 2.0099375604608216e-05,
      "loss": 0.4841,
      "step": 34000
    },
    {
      "epoch": 3.0,
      "eval_combined_score": 0.8122094824914782,
      "eval_f1": 0.8122094824914782,
      "eval_loss": 0.34033524990081787,
      "eval_runtime": 37.9138,
      "eval_samples_per_second": 1066.367,
      "eval_steps_per_second": 4.167,
      "step": 34113
    },
    {
      "epoch": 3.03,
      "grad_norm": 3.1498029232025146,
      "learning_rate": 1.9659660539970102e-05,
      "loss": 0.4738,
      "step": 34500
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.5172255039215088,
      "learning_rate": 1.9219945475331985e-05,
      "loss": 0.4803,
      "step": 35000
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.0573079586029053,
      "learning_rate": 1.878023041069387e-05,
      "loss": 0.4728,
      "step": 35500
    },
    {
      "epoch": 3.17,
      "grad_norm": 2.761320114135742,
      "learning_rate": 1.8340515346055757e-05,
      "loss": 0.4768,
      "step": 36000
    },
    {
      "epoch": 3.21,
      "grad_norm": 2.29789137840271,
      "learning_rate": 1.7900800281417643e-05,
      "loss": 0.4786,
      "step": 36500
    },
    {
      "epoch": 3.25,
      "grad_norm": 2.8867733478546143,
      "learning_rate": 1.7461085216779526e-05,
      "loss": 0.472,
      "step": 37000
    },
    {
      "epoch": 3.3,
      "grad_norm": 2.888866901397705,
      "learning_rate": 1.7021370152141413e-05,
      "loss": 0.4701,
      "step": 37500
    },
    {
      "epoch": 3.34,
      "grad_norm": 2.6121418476104736,
      "learning_rate": 1.65816550875033e-05,
      "loss": 0.4702,
      "step": 38000
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.7327868938446045,
      "learning_rate": 1.6141940022865185e-05,
      "loss": 0.4751,
      "step": 38500
    },
    {
      "epoch": 3.43,
      "grad_norm": 2.7664458751678467,
      "learning_rate": 1.5702224958227068e-05,
      "loss": 0.4772,
      "step": 39000
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.0849014520645142,
      "learning_rate": 1.5262509893588954e-05,
      "loss": 0.4738,
      "step": 39500
    },
    {
      "epoch": 3.52,
      "grad_norm": 3.0802319049835205,
      "learning_rate": 1.482279482895084e-05,
      "loss": 0.476,
      "step": 40000
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.9298323392868042,
      "learning_rate": 1.4383079764312727e-05,
      "loss": 0.4681,
      "step": 40500
    },
    {
      "epoch": 3.61,
      "grad_norm": 2.0291521549224854,
      "learning_rate": 1.3943364699674611e-05,
      "loss": 0.4769,
      "step": 41000
    },
    {
      "epoch": 3.65,
      "grad_norm": 2.866079330444336,
      "learning_rate": 1.3503649635036497e-05,
      "loss": 0.4771,
      "step": 41500
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.5067503452301025,
      "learning_rate": 1.3063934570398382e-05,
      "loss": 0.4754,
      "step": 42000
    },
    {
      "epoch": 3.74,
      "grad_norm": 2.293853521347046,
      "learning_rate": 1.2624219505760268e-05,
      "loss": 0.4791,
      "step": 42500
    },
    {
      "epoch": 3.78,
      "grad_norm": 2.38676381111145,
      "learning_rate": 1.2184504441122153e-05,
      "loss": 0.4674,
      "step": 43000
    },
    {
      "epoch": 3.83,
      "grad_norm": 2.829270601272583,
      "learning_rate": 1.1744789376484039e-05,
      "loss": 0.4715,
      "step": 43500
    },
    {
      "epoch": 3.87,
      "grad_norm": 2.3581595420837402,
      "learning_rate": 1.1305074311845925e-05,
      "loss": 0.4717,
      "step": 44000
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.803579568862915,
      "learning_rate": 1.086535924720781e-05,
      "loss": 0.4638,
      "step": 44500
    },
    {
      "epoch": 3.96,
      "grad_norm": 3.1875007152557373,
      "learning_rate": 1.0425644182569696e-05,
      "loss": 0.4762,
      "step": 45000
    },
    {
      "epoch": 4.0,
      "eval_combined_score": 0.8123888182973317,
      "eval_f1": 0.8123888182973317,
      "eval_loss": 0.33584973216056824,
      "eval_runtime": 37.8907,
      "eval_samples_per_second": 1067.017,
      "eval_steps_per_second": 4.17,
      "step": 45484
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.368762969970703,
      "learning_rate": 9.98592911793158e-06,
      "loss": 0.4707,
      "step": 45500
    },
    {
      "epoch": 4.05,
      "grad_norm": 2.4431540966033936,
      "learning_rate": 9.546214053293467e-06,
      "loss": 0.4654,
      "step": 46000
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.8363009691238403,
      "learning_rate": 9.106498988655351e-06,
      "loss": 0.4607,
      "step": 46500
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.9336574077606201,
      "learning_rate": 8.666783924017237e-06,
      "loss": 0.4661,
      "step": 47000
    },
    {
      "epoch": 4.18,
      "grad_norm": 1.5891854763031006,
      "learning_rate": 8.227068859379122e-06,
      "loss": 0.4637,
      "step": 47500
    },
    {
      "epoch": 4.22,
      "grad_norm": 1.3558140993118286,
      "learning_rate": 7.787353794741008e-06,
      "loss": 0.4599,
      "step": 48000
    },
    {
      "epoch": 4.27,
      "grad_norm": 1.899108648300171,
      "learning_rate": 7.347638730102893e-06,
      "loss": 0.4713,
      "step": 48500
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.9433221817016602,
      "learning_rate": 6.907923665464779e-06,
      "loss": 0.4655,
      "step": 49000
    },
    {
      "epoch": 4.35,
      "grad_norm": 3.1865389347076416,
      "learning_rate": 6.468208600826664e-06,
      "loss": 0.4598,
      "step": 49500
    },
    {
      "epoch": 4.4,
      "grad_norm": 2.2305116653442383,
      "learning_rate": 6.02849353618855e-06,
      "loss": 0.4659,
      "step": 50000
    },
    {
      "epoch": 4.44,
      "grad_norm": 2.6832566261291504,
      "learning_rate": 5.588778471550435e-06,
      "loss": 0.4597,
      "step": 50500
    },
    {
      "epoch": 4.49,
      "grad_norm": 2.4481277465820312,
      "learning_rate": 5.149063406912321e-06,
      "loss": 0.4721,
      "step": 51000
    },
    {
      "epoch": 4.53,
      "grad_norm": 2.5793635845184326,
      "learning_rate": 4.709348342274207e-06,
      "loss": 0.4661,
      "step": 51500
    },
    {
      "epoch": 4.57,
      "grad_norm": 2.133437395095825,
      "learning_rate": 4.269633277636092e-06,
      "loss": 0.4589,
      "step": 52000
    },
    {
      "epoch": 4.62,
      "grad_norm": 4.3351006507873535,
      "learning_rate": 3.8299182129979776e-06,
      "loss": 0.4697,
      "step": 52500
    },
    {
      "epoch": 4.66,
      "grad_norm": 3.0949718952178955,
      "learning_rate": 3.3902031483598625e-06,
      "loss": 0.4659,
      "step": 53000
    },
    {
      "epoch": 4.7,
      "grad_norm": 2.1113531589508057,
      "learning_rate": 2.9504880837217483e-06,
      "loss": 0.4675,
      "step": 53500
    },
    {
      "epoch": 4.75,
      "grad_norm": 2.830069065093994,
      "learning_rate": 2.5107730190836337e-06,
      "loss": 0.4736,
      "step": 54000
    },
    {
      "epoch": 4.79,
      "grad_norm": 1.8624334335327148,
      "learning_rate": 2.0710579544455195e-06,
      "loss": 0.4596,
      "step": 54500
    },
    {
      "epoch": 4.84,
      "grad_norm": 3.123922824859619,
      "learning_rate": 1.6313428898074047e-06,
      "loss": 0.4601,
      "step": 55000
    },
    {
      "epoch": 4.88,
      "grad_norm": 2.3546392917633057,
      "learning_rate": 1.1916278251692903e-06,
      "loss": 0.4605,
      "step": 55500
    },
    {
      "epoch": 4.92,
      "grad_norm": 2.721216917037964,
      "learning_rate": 7.519127605311758e-07,
      "loss": 0.4652,
      "step": 56000
    },
    {
      "epoch": 4.97,
      "grad_norm": 2.4257423877716064,
      "learning_rate": 3.121976958930613e-07,
      "loss": 0.4614,
      "step": 56500
    },
    {
      "epoch": 5.0,
      "eval_combined_score": 0.8172281558013501,
      "eval_f1": 0.8172281558013501,
      "eval_loss": 0.3320091664791107,
      "eval_runtime": 37.3626,
      "eval_samples_per_second": 1082.098,
      "eval_steps_per_second": 4.229,
      "step": 56855
    },
    {
      "epoch": 5.0,
      "step": 56855,
      "total_flos": 0.0,
      "train_loss": 0.5058023918567175,
      "train_runtime": 6368.9989,
      "train_samples_per_second": 285.638,
      "train_steps_per_second": 8.927
    }
  ],
  "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
}