Upload trainer_state.json

trainer_state.json

@@ -2,261 +2,261 @@
   "best_metric": null,
   "best_model_checkpoint": null,
   "epoch": 5.0,
-  "global_step": 16110,
+  "global_step": 16080,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.16,
-      "learning_rate": 5.17063081695967e-06,
-      "loss": 0.9438,
+      "learning_rate": 5.1813471502590676e-06,
+      "loss": 0.9606,
       "step": 500
     },
     {
       "epoch": 0.31,
-      "learning_rate": 9.97820775275705e-06,
-      "loss": 0.6981,
+      "learning_rate": 9.976844194508766e-06,
+      "loss": 0.7024,
       "step": 1000
     },
     {
       "epoch": 0.47,
-      "learning_rate": 9.648022188469921e-06,
-      "loss": 0.6641,
+      "learning_rate": 9.646046973205426e-06,
+      "loss": 0.6643,
       "step": 1500
     },
     {
       "epoch": 0.62,
-      "learning_rate": 9.317836624182791e-06,
-      "loss": 0.6351,
+      "learning_rate": 9.315249751902085e-06,
+      "loss": 0.6306,
       "step": 2000
     },
     {
       "epoch": 0.78,
-      "learning_rate": 8.987651059895663e-06,
-      "loss": 0.6069,
+      "learning_rate": 8.984452530598744e-06,
+      "loss": 0.6088,
       "step": 2500
     },
     {
       "epoch": 0.93,
-      "learning_rate": 8.657465495608533e-06,
-      "loss": 0.6,
+      "learning_rate": 8.653655309295402e-06,
+      "loss": 0.6011,
       "step": 3000
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7530310750007629,
-      "eval_loss": 0.5826240181922913,
-      "eval_runtime": 74.1302,
-      "eval_samples_per_second": 132.402,
-      "eval_steps_per_second": 4.141,
-      "step": 3222
+      "eval_accuracy": 0.7731024026870728,
+      "eval_loss": 0.5318970084190369,
+      "eval_runtime": 73.9957,
+      "eval_samples_per_second": 132.643,
+      "eval_steps_per_second": 4.149,
+      "step": 3216
     },
     {
       "epoch": 1.09,
-      "learning_rate": 8.327279931321404e-06,
-      "loss": 0.5475,
+      "learning_rate": 8.322858087992062e-06,
+      "loss": 0.5493,
       "step": 3500
     },
     {
       "epoch": 1.24,
-      "learning_rate": 7.997094367034274e-06,
-      "loss": 0.5118,
+      "learning_rate": 7.99206086668872e-06,
+      "loss": 0.5099,
       "step": 4000
     },
     {
       "epoch": 1.4,
-      "learning_rate": 7.666908802747146e-06,
-      "loss": 0.5113,
+      "learning_rate": 7.66126364538538e-06,
+      "loss": 0.5124,
       "step": 4500
     },
     {
       "epoch": 1.55,
-      "learning_rate": 7.336723238460015e-06,
-      "loss": 0.5134,
+      "learning_rate": 7.330466424082037e-06,
+      "loss": 0.5131,
       "step": 5000
     },
     {
       "epoch": 1.71,
-      "learning_rate": 7.006537674172885e-06,
-      "loss": 0.5022,
+      "learning_rate": 6.999669202778697e-06,
+      "loss": 0.5117,
       "step": 5500
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 6.676352109885757e-06,
-      "loss": 0.5082,
+      "epoch": 1.87,
+      "learning_rate": 6.668871981475356e-06,
+      "loss": 0.5138,
       "step": 6000
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7587366104125977,
-      "eval_loss": 0.5997494459152222,
-      "eval_runtime": 74.1755,
-      "eval_samples_per_second": 132.321,
-      "eval_steps_per_second": 4.139,
-      "step": 6444
+      "eval_accuracy": 0.764034628868103,
+      "eval_loss": 0.6075196862220764,
+      "eval_runtime": 74.2,
+      "eval_samples_per_second": 132.278,
+      "eval_steps_per_second": 4.137,
+      "step": 6432
     },
     {
       "epoch": 2.02,
-      "learning_rate": 6.3461665455986265e-06,
-      "loss": 0.4847,
+      "learning_rate": 6.338074760172015e-06,
+      "loss": 0.4949,
       "step": 6500
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 6.015980981311498e-06,
-      "loss": 0.3947,
+      "epoch": 2.18,
+      "learning_rate": 6.007277538868674e-06,
+      "loss": 0.3903,
       "step": 7000
     },
     {
       "epoch": 2.33,
-      "learning_rate": 5.685795417024368e-06,
-      "loss": 0.3968,
+      "learning_rate": 5.676480317565333e-06,
+      "loss": 0.3965,
       "step": 7500
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 5.3556098527372394e-06,
-      "loss": 0.3938,
+      "epoch": 2.49,
+      "learning_rate": 5.345683096261992e-06,
+      "loss": 0.3876,
       "step": 8000
     },
     {
       "epoch": 2.64,
-      "learning_rate": 5.025424288450109e-06,
-      "loss": 0.3877,
+      "learning_rate": 5.01488587495865e-06,
+      "loss": 0.3981,
       "step": 8500
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 4.69523872416298e-06,
-      "loss": 0.3839,
+      "epoch": 2.8,
+      "learning_rate": 4.6840886536553096e-06,
+      "loss": 0.3883,
       "step": 9000
     },
     {
       "epoch": 2.95,
-      "learning_rate": 4.365053159875851e-06,
-      "loss": 0.3949,
+      "learning_rate": 4.353291432351969e-06,
+      "loss": 0.3833,
       "step": 9500
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8090677261352539,
-      "eval_loss": 0.5072706937789917,
-      "eval_runtime": 74.3802,
-      "eval_samples_per_second": 131.957,
-      "eval_steps_per_second": 4.127,
-      "step": 9666
+      "eval_accuracy": 0.8112073540687561,
+      "eval_loss": 0.4962076246738434,
+      "eval_runtime": 74.1434,
+      "eval_samples_per_second": 132.379,
+      "eval_steps_per_second": 4.141,
+      "step": 9648
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 4.034867595588721e-06,
-      "loss": 0.321,
+      "epoch": 3.11,
+      "learning_rate": 4.022494211048628e-06,
+      "loss": 0.3235,
       "step": 10000
     },
     {
       "epoch": 3.26,
-      "learning_rate": 3.704682031301592e-06,
-      "loss": 0.2894,
+      "learning_rate": 3.6916969897452866e-06,
+      "loss": 0.2933,
       "step": 10500
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 3.3744964670144627e-06,
-      "loss": 0.2923,
+      "epoch": 3.42,
+      "learning_rate": 3.3608997684419452e-06,
+      "loss": 0.2844,
       "step": 11000
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 3.0443109027273334e-06,
-      "loss": 0.2906,
+      "epoch": 3.58,
+      "learning_rate": 3.0301025471386043e-06,
+      "loss": 0.2992,
       "step": 11500
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 2.714125338440204e-06,
-      "loss": 0.2913,
+      "epoch": 3.73,
+      "learning_rate": 2.699305325835263e-06,
+      "loss": 0.2941,
       "step": 12000
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 2.3839397741530743e-06,
-      "loss": 0.2969,
+      "epoch": 3.89,
+      "learning_rate": 2.3685081045319223e-06,
+      "loss": 0.2856,
       "step": 12500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7917473316192627,
-      "eval_loss": 0.6656549572944641,
-      "eval_runtime": 74.3184,
-      "eval_samples_per_second": 132.067,
-      "eval_steps_per_second": 4.131,
-      "step": 12888
+      "eval_accuracy": 0.7993887066841125,
+      "eval_loss": 0.6498498320579529,
+      "eval_runtime": 74.0745,
+      "eval_samples_per_second": 132.502,
+      "eval_steps_per_second": 4.144,
+      "step": 12864
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 2.053754209865945e-06,
-      "loss": 0.2657,
+      "epoch": 4.04,
+      "learning_rate": 2.037710883228581e-06,
+      "loss": 0.2637,
       "step": 13000
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 1.7235686455788153e-06,
-      "loss": 0.2215,
+      "epoch": 4.2,
+      "learning_rate": 1.70691366192524e-06,
+      "loss": 0.225,
       "step": 13500
     },
     {
       "epoch": 4.35,
-      "learning_rate": 1.393383081291686e-06,
-      "loss": 0.2228,
+      "learning_rate": 1.3761164406218988e-06,
+      "loss": 0.2185,
       "step": 14000
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.0631975170045567e-06,
-      "loss": 0.215,
+      "epoch": 4.51,
+      "learning_rate": 1.0453192193185578e-06,
+      "loss": 0.2189,
       "step": 14500
     },
     {
       "epoch": 4.66,
-      "learning_rate": 7.330119527174273e-07,
-      "loss": 0.2306,
+      "learning_rate": 7.145219980152167e-07,
+      "loss": 0.2196,
       "step": 15000
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 4.0282638843029786e-07,
-      "loss": 0.2168,
+      "epoch": 4.82,
+      "learning_rate": 3.837247767118756e-07,
+      "loss": 0.2201,
       "step": 15500
     },
     {
-      "epoch": 4.97,
-      "learning_rate": 7.264082414316847e-08,
-      "loss": 0.2267,
+      "epoch": 4.98,
+      "learning_rate": 5.2927555408534575e-08,
+      "loss": 0.2194,
       "step": 16000
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.801732063293457,
-      "eval_loss": 0.7047895789146423,
-      "eval_runtime": 74.3187,
-      "eval_samples_per_second": 132.066,
-      "eval_steps_per_second": 4.131,
-      "step": 16110
+      "eval_accuracy": 0.8041772842407227,
+      "eval_loss": 0.6977987885475159,
+      "eval_runtime": 74.188,
+      "eval_samples_per_second": 132.299,
+      "eval_steps_per_second": 4.138,
+      "step": 16080
     },
     {
       "epoch": 5.0,
-      "step": 16110,
-      "total_flos": 1.2007859696955264e+17,
-      "train_loss": 0.41926650207727906,
-      "train_runtime": 12549.4046,
-      "train_samples_per_second": 41.069,
-      "train_steps_per_second": 1.284
+      "step": 16080,
+      "total_flos": 1.198526028503616e+17,
+      "train_loss": 0.4203688863498061,
+      "train_runtime": 12593.4384,
+      "train_samples_per_second": 40.849,
+      "train_steps_per_second": 1.277
     }
   ],
-  "max_steps": 16110,
+  "max_steps": 16080,
   "num_train_epochs": 5,
-  "total_flos": 1.2007859696955264e+17,
+  "total_flos": 1.198526028503616e+17,
   "trial_name": null,
   "trial_params": null
 }