End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 3.0,
-    "eval_accuracy": 0.6819923371647509,
-    "eval_loss": 0.8754438161849976,
-    "eval_runtime": 25.581,
-    "eval_samples_per_second": 112.232,
-    "eval_steps_per_second": 3.518,
-    "total_flos": 6.004415924974301e+18,
-    "train_loss": 1.159771179602091,
-    "train_runtime": 1779.2361,
-    "train_samples_per_second": 43.566,
-    "train_steps_per_second": 0.341
+    "epoch": 4.0,
+    "eval_accuracy": 0.6879136189481017,
+    "eval_loss": 0.8504465818405151,
+    "eval_runtime": 25.231,
+    "eval_samples_per_second": 113.789,
+    "eval_steps_per_second": 3.567,
+    "total_flos": 8.005887899965735e+18,
+    "train_loss": 1.122130044616095,
+    "train_runtime": 3022.4726,
+    "train_samples_per_second": 34.195,
+    "train_steps_per_second": 0.267
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "eval_accuracy": 0.6819923371647509,
-    "eval_loss": 0.8754438161849976,
-    "eval_runtime": 25.581,
-    "eval_samples_per_second": 112.232,
-    "eval_steps_per_second": 3.518
+    "epoch": 4.0,
+    "eval_accuracy": 0.6879136189481017,
+    "eval_loss": 0.8504465818405151,
+    "eval_runtime": 25.231,
+    "eval_samples_per_second": 113.789,
+    "eval_steps_per_second": 3.567
 }

runs/Jan08_10-49-11_9d7446287eef/events.out.tfevents.1673178226.9d7446287eef.23.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:934d59f7ea613c965daacaf93726a8fa602e6e33ebe986c055abe5f5aa686256
+size 363

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.0,
-    "total_flos": 6.004415924974301e+18,
-    "train_loss": 1.159771179602091,
-    "train_runtime": 1779.2361,
-    "train_samples_per_second": 43.566,
-    "train_steps_per_second": 0.341
+    "epoch": 4.0,
+    "total_flos": 8.005887899965735e+18,
+    "train_loss": 1.122130044616095,
+    "train_runtime": 3022.4726,
+    "train_samples_per_second": 34.195,
+    "train_steps_per_second": 0.267
 }

trainer_state.json

@@ -1,412 +1,541 @@
 {
-  "best_metric": 0.6819923371647509,
-  "best_model_checkpoint": "beit-base-patch16-224-pt22k-ft22k-finetuned-FER2013/checkpoint-606",
-  "epoch": 3.0,
-  "global_step": 606,
+  "best_metric": 0.6879136189481017,
+  "best_model_checkpoint": "beit-base-patch16-224-pt22k-ft22k-finetuned-FER2013/checkpoint-808",
+  "epoch": 4.0,
+  "global_step": 808,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.05,
-      "learning_rate": 8.196721311475409e-06,
-      "loss": 1.9392,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 2.0141,
       "step": 10
     },
     {
       "epoch": 0.1,
-      "learning_rate": 1.6393442622950818e-05,
-      "loss": 1.7671,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 1.7884,
       "step": 20
     },
     {
       "epoch": 0.15,
-      "learning_rate": 2.459016393442623e-05,
-      "loss": 1.6727,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 1.6854,
       "step": 30
     },
     {
       "epoch": 0.2,
-      "learning_rate": 3.2786885245901635e-05,
-      "loss": 1.4688,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 1.5246,
       "step": 40
     },
     {
       "epoch": 0.25,
-      "learning_rate": 4.098360655737705e-05,
-      "loss": 1.4389,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 1.4423,
       "step": 50
     },
     {
       "epoch": 0.3,
-      "learning_rate": 4.918032786885246e-05,
-      "loss": 1.3738,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 1.3828,
       "step": 60
     },
     {
       "epoch": 0.35,
-      "learning_rate": 4.917431192660551e-05,
-      "loss": 1.3653,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 1.3731,
       "step": 70
     },
     {
       "epoch": 0.4,
-      "learning_rate": 4.8256880733944956e-05,
-      "loss": 1.2999,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 1.313,
       "step": 80
     },
     {
       "epoch": 0.45,
-      "learning_rate": 4.733944954128441e-05,
-      "loss": 1.301,
+      "learning_rate": 4.9381017881705645e-05,
+      "loss": 1.3077,
       "step": 90
     },
     {
       "epoch": 0.5,
-      "learning_rate": 4.642201834862386e-05,
-      "loss": 1.3416,
+      "learning_rate": 4.8693259972489685e-05,
+      "loss": 1.3512,
       "step": 100
     },
     {
       "epoch": 0.54,
-      "learning_rate": 4.5504587155963305e-05,
-      "loss": 1.2908,
+      "learning_rate": 4.8005502063273726e-05,
+      "loss": 1.2712,
       "step": 110
     },
     {
       "epoch": 0.59,
-      "learning_rate": 4.458715596330276e-05,
-      "loss": 1.2205,
+      "learning_rate": 4.731774415405777e-05,
+      "loss": 1.2161,
       "step": 120
     },
     {
       "epoch": 0.64,
-      "learning_rate": 4.366972477064221e-05,
-      "loss": 1.2779,
+      "learning_rate": 4.662998624484182e-05,
+      "loss": 1.2698,
       "step": 130
     },
     {
       "epoch": 0.69,
-      "learning_rate": 4.2752293577981654e-05,
-      "loss": 1.2563,
+      "learning_rate": 4.594222833562587e-05,
+      "loss": 1.2618,
       "step": 140
     },
     {
       "epoch": 0.74,
-      "learning_rate": 4.1834862385321106e-05,
-      "loss": 1.2122,
+      "learning_rate": 4.525447042640991e-05,
+      "loss": 1.2162,
       "step": 150
     },
     {
       "epoch": 0.79,
-      "learning_rate": 4.091743119266056e-05,
-      "loss": 1.2289,
+      "learning_rate": 4.456671251719395e-05,
+      "loss": 1.2392,
       "step": 160
     },
     {
       "epoch": 0.84,
-      "learning_rate": 4e-05,
-      "loss": 1.1613,
+      "learning_rate": 4.3878954607977995e-05,
+      "loss": 1.1664,
       "step": 170
     },
     {
       "epoch": 0.89,
-      "learning_rate": 3.9082568807339455e-05,
-      "loss": 1.1646,
+      "learning_rate": 4.3191196698762035e-05,
+      "loss": 1.1736,
       "step": 180
     },
     {
       "epoch": 0.94,
-      "learning_rate": 3.81651376146789e-05,
-      "loss": 1.2029,
+      "learning_rate": 4.250343878954608e-05,
+      "loss": 1.2149,
       "step": 190
     },
     {
       "epoch": 0.99,
-      "learning_rate": 3.724770642201835e-05,
-      "loss": 1.1701,
+      "learning_rate": 4.181568088033013e-05,
+      "loss": 1.1617,
       "step": 200
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6179031696273075,
-      "eval_loss": 1.0163276195526123,
-      "eval_runtime": 25.3453,
-      "eval_samples_per_second": 113.276,
-      "eval_steps_per_second": 3.551,
+      "eval_accuracy": 0.6269592476489029,
+      "eval_loss": 1.0080664157867432,
+      "eval_runtime": 27.1012,
+      "eval_samples_per_second": 105.936,
+      "eval_steps_per_second": 3.321,
       "step": 202
     },
     {
       "epoch": 1.04,
-      "learning_rate": 3.6330275229357804e-05,
-      "loss": 1.1281,
+      "learning_rate": 4.112792297111417e-05,
+      "loss": 1.1415,
       "step": 210
     },
     {
       "epoch": 1.09,
-      "learning_rate": 3.541284403669725e-05,
-      "loss": 1.1595,
+      "learning_rate": 4.044016506189822e-05,
+      "loss": 1.1735,
       "step": 220
     },
     {
       "epoch": 1.14,
-      "learning_rate": 3.44954128440367e-05,
-      "loss": 1.1224,
+      "learning_rate": 3.975240715268226e-05,
+      "loss": 1.1356,
       "step": 230
     },
     {
       "epoch": 1.19,
-      "learning_rate": 3.3577981651376154e-05,
-      "loss": 1.0808,
+      "learning_rate": 3.90646492434663e-05,
+      "loss": 1.0938,
       "step": 240
     },
     {
       "epoch": 1.24,
-      "learning_rate": 3.26605504587156e-05,
-      "loss": 1.1312,
+      "learning_rate": 3.8376891334250345e-05,
+      "loss": 1.1276,
       "step": 250
     },
     {
       "epoch": 1.29,
-      "learning_rate": 3.174311926605505e-05,
-      "loss": 1.1423,
+      "learning_rate": 3.768913342503439e-05,
+      "loss": 1.1508,
       "step": 260
     },
     {
       "epoch": 1.34,
-      "learning_rate": 3.0825688073394496e-05,
-      "loss": 1.1882,
+      "learning_rate": 3.700137551581843e-05,
+      "loss": 1.2005,
       "step": 270
     },
     {
       "epoch": 1.39,
-      "learning_rate": 2.9908256880733948e-05,
-      "loss": 1.1086,
+      "learning_rate": 3.631361760660248e-05,
+      "loss": 1.1185,
       "step": 280
     },
     {
       "epoch": 1.44,
-      "learning_rate": 2.8990825688073397e-05,
-      "loss": 1.0772,
+      "learning_rate": 3.562585969738652e-05,
+      "loss": 1.0947,
       "step": 290
     },
     {
       "epoch": 1.49,
-      "learning_rate": 2.807339449541285e-05,
-      "loss": 1.1303,
+      "learning_rate": 3.493810178817056e-05,
+      "loss": 1.14,
       "step": 300
     },
     {
       "epoch": 1.53,
-      "learning_rate": 2.7155963302752297e-05,
-      "loss": 1.1025,
+      "learning_rate": 3.425034387895461e-05,
+      "loss": 1.1113,
       "step": 310
     },
     {
       "epoch": 1.58,
-      "learning_rate": 2.6238532110091746e-05,
-      "loss": 1.1117,
+      "learning_rate": 3.3562585969738655e-05,
+      "loss": 1.1291,
       "step": 320
     },
     {
       "epoch": 1.63,
-      "learning_rate": 2.5321100917431194e-05,
-      "loss": 1.0372,
+      "learning_rate": 3.28748280605227e-05,
+      "loss": 1.0612,
       "step": 330
     },
     {
       "epoch": 1.68,
-      "learning_rate": 2.4403669724770646e-05,
-      "loss": 1.0228,
+      "learning_rate": 3.218707015130674e-05,
+      "loss": 1.0498,
       "step": 340
     },
     {
       "epoch": 1.73,
-      "learning_rate": 2.3486238532110095e-05,
-      "loss": 1.0936,
+      "learning_rate": 3.149931224209078e-05,
+      "loss": 1.1312,
       "step": 350
     },
     {
       "epoch": 1.78,
-      "learning_rate": 2.2568807339449544e-05,
-      "loss": 1.0889,
+      "learning_rate": 3.081155433287483e-05,
+      "loss": 1.0964,
       "step": 360
     },
     {
       "epoch": 1.83,
-      "learning_rate": 2.1651376146788992e-05,
-      "loss": 1.0765,
+      "learning_rate": 3.0123796423658874e-05,
+      "loss": 1.1113,
       "step": 370
     },
     {
       "epoch": 1.88,
-      "learning_rate": 2.0733944954128444e-05,
-      "loss": 1.0954,
+      "learning_rate": 2.9436038514442914e-05,
+      "loss": 1.0912,
       "step": 380
     },
     {
       "epoch": 1.93,
-      "learning_rate": 1.9816513761467893e-05,
-      "loss": 1.1196,
+      "learning_rate": 2.874828060522696e-05,
+      "loss": 1.1223,
       "step": 390
     },
     {
       "epoch": 1.98,
-      "learning_rate": 1.889908256880734e-05,
-      "loss": 1.0447,
+      "learning_rate": 2.8060522696011005e-05,
+      "loss": 1.0604,
       "step": 400
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.6569139672587948,
-      "eval_loss": 0.9236605763435364,
-      "eval_runtime": 25.7906,
-      "eval_samples_per_second": 111.32,
-      "eval_steps_per_second": 3.49,
+      "eval_accuracy": 0.6523859282479972,
+      "eval_loss": 0.9515869617462158,
+      "eval_runtime": 24.9329,
+      "eval_samples_per_second": 115.149,
+      "eval_steps_per_second": 3.61,
       "step": 404
     },
     {
       "epoch": 2.03,
-      "learning_rate": 1.798165137614679e-05,
-      "loss": 1.0341,
+      "learning_rate": 2.7372764786795052e-05,
+      "loss": 1.0564,
       "step": 410
     },
     {
       "epoch": 2.08,
-      "learning_rate": 1.7064220183486242e-05,
-      "loss": 0.9925,
+      "learning_rate": 2.6685006877579092e-05,
+      "loss": 0.9928,
       "step": 420
     },
     {
       "epoch": 2.13,
-      "learning_rate": 1.614678899082569e-05,
-      "loss": 1.0571,
+      "learning_rate": 2.5997248968363136e-05,
+      "loss": 1.0681,
       "step": 430
     },
     {
       "epoch": 2.18,
-      "learning_rate": 1.5229357798165139e-05,
-      "loss": 1.0625,
+      "learning_rate": 2.5309491059147183e-05,
+      "loss": 1.0882,
       "step": 440
     },
     {
       "epoch": 2.23,
-      "learning_rate": 1.431192660550459e-05,
-      "loss": 1.0117,
+      "learning_rate": 2.4621733149931224e-05,
+      "loss": 1.0132,
       "step": 450
     },
     {
       "epoch": 2.28,
-      "learning_rate": 1.3394495412844038e-05,
-      "loss": 0.9992,
+      "learning_rate": 2.393397524071527e-05,
+      "loss": 1.0394,
       "step": 460
     },
     {
       "epoch": 2.33,
-      "learning_rate": 1.2477064220183488e-05,
-      "loss": 1.0723,
+      "learning_rate": 2.324621733149931e-05,
+      "loss": 1.0756,
       "step": 470
     },
     {
       "epoch": 2.38,
-      "learning_rate": 1.1559633027522937e-05,
-      "loss": 1.0284,
+      "learning_rate": 2.255845942228336e-05,
+      "loss": 1.0273,
       "step": 480
     },
     {
       "epoch": 2.43,
-      "learning_rate": 1.0642201834862387e-05,
-      "loss": 0.9993,
+      "learning_rate": 2.1870701513067402e-05,
+      "loss": 1.0051,
       "step": 490
     },
     {
       "epoch": 2.48,
-      "learning_rate": 9.724770642201836e-06,
-      "loss": 1.023,
+      "learning_rate": 2.1182943603851446e-05,
+      "loss": 1.0387,
       "step": 500
     },
     {
       "epoch": 2.52,
-      "learning_rate": 8.807339449541286e-06,
-      "loss": 1.0313,
+      "learning_rate": 2.049518569463549e-05,
+      "loss": 1.044,
       "step": 510
     },
     {
       "epoch": 2.57,
-      "learning_rate": 7.889908256880735e-06,
-      "loss": 1.0584,
+      "learning_rate": 1.9807427785419533e-05,
+      "loss": 1.0704,
       "step": 520
     },
     {
       "epoch": 2.62,
-      "learning_rate": 6.972477064220184e-06,
-      "loss": 1.003,
+      "learning_rate": 1.9119669876203577e-05,
+      "loss": 1.0369,
       "step": 530
     },
     {
       "epoch": 2.67,
-      "learning_rate": 6.0550458715596335e-06,
-      "loss": 1.0541,
+      "learning_rate": 1.843191196698762e-05,
+      "loss": 1.062,
       "step": 540
     },
     {
       "epoch": 2.72,
-      "learning_rate": 5.137614678899083e-06,
-      "loss": 1.0189,
+      "learning_rate": 1.7744154057771665e-05,
+      "loss": 1.0227,
       "step": 550
     },
     {
       "epoch": 2.77,
-      "learning_rate": 4.220183486238532e-06,
-      "loss": 0.9714,
+      "learning_rate": 1.705639614855571e-05,
+      "loss": 0.9935,
       "step": 560
     },
     {
       "epoch": 2.82,
-      "learning_rate": 3.302752293577982e-06,
-      "loss": 1.0046,
+      "learning_rate": 1.6368638239339752e-05,
+      "loss": 1.0015,
       "step": 570
     },
     {
       "epoch": 2.87,
-      "learning_rate": 2.3853211009174313e-06,
-      "loss": 1.0082,
+      "learning_rate": 1.56808803301238e-05,
+      "loss": 1.0193,
       "step": 580
     },
     {
       "epoch": 2.92,
-      "learning_rate": 1.467889908256881e-06,
-      "loss": 1.0499,
+      "learning_rate": 1.499312242090784e-05,
+      "loss": 1.0556,
       "step": 590
     },
     {
       "epoch": 2.97,
-      "learning_rate": 5.504587155963304e-07,
-      "loss": 0.9712,
+      "learning_rate": 1.4305364511691885e-05,
+      "loss": 0.998,
       "step": 600
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.6819923371647509,
-      "eval_loss": 0.8754438161849976,
-      "eval_runtime": 25.5354,
-      "eval_samples_per_second": 112.432,
-      "eval_steps_per_second": 3.525,
+      "eval_accuracy": 0.6809474050853361,
+      "eval_loss": 0.8857186436653137,
+      "eval_runtime": 27.0525,
+      "eval_samples_per_second": 106.127,
+      "eval_steps_per_second": 3.327,
       "step": 606
     },
     {
-      "epoch": 3.0,
-      "step": 606,
-      "total_flos": 6.004415924974301e+18,
-      "train_loss": 1.159771179602091,
-      "train_runtime": 1779.2361,
-      "train_samples_per_second": 43.566,
-      "train_steps_per_second": 0.341
+      "epoch": 3.02,
+      "learning_rate": 1.3617606602475929e-05,
+      "loss": 1.0409,
+      "step": 610
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2929848693259975e-05,
+      "loss": 1.006,
+      "step": 620
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.2242090784044018e-05,
+      "loss": 0.9774,
+      "step": 630
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.155433287482806e-05,
+      "loss": 0.9504,
+      "step": 640
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0866574965612106e-05,
+      "loss": 0.9869,
+      "step": 650
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.017881705639615e-05,
+      "loss": 1.0148,
+      "step": 660
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 9.491059147180192e-06,
+      "loss": 0.9741,
+      "step": 670
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.803301237964237e-06,
+      "loss": 0.9989,
+      "step": 680
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 8.115543328748281e-06,
+      "loss": 0.9169,
+      "step": 690
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 7.4277854195323255e-06,
+      "loss": 0.9754,
+      "step": 700
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 6.740027510316368e-06,
+      "loss": 0.9222,
+      "step": 710
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 6.052269601100413e-06,
+      "loss": 0.9857,
+      "step": 720
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.364511691884458e-06,
+      "loss": 0.9872,
+      "step": 730
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.676753782668501e-06,
+      "loss": 0.9698,
+      "step": 740
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.988995873452544e-06,
+      "loss": 0.9562,
+      "step": 750
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.3012379642365885e-06,
+      "loss": 0.9538,
+      "step": 760
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.613480055020633e-06,
+      "loss": 0.9707,
+      "step": 770
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.925722145804677e-06,
+      "loss": 0.9708,
+      "step": 780
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.2379642365887208e-06,
+      "loss": 0.9303,
+      "step": 790
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.502063273727648e-07,
+      "loss": 0.9971,
+      "step": 800
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6879136189481017,
+      "eval_loss": 0.8504465818405151,
+      "eval_runtime": 26.9878,
+      "eval_samples_per_second": 106.381,
+      "eval_steps_per_second": 3.335,
+      "step": 808
+    },
+    {
+      "epoch": 4.0,
+      "step": 808,
+      "total_flos": 8.005887899965735e+18,
+      "train_loss": 1.122130044616095,
+      "train_runtime": 3022.4726,
+      "train_samples_per_second": 34.195,
+      "train_steps_per_second": 0.267
     }
   ],
-  "max_steps": 606,
-  "num_train_epochs": 3,
-  "total_flos": 6.004415924974301e+18,
+  "max_steps": 808,
+  "num_train_epochs": 4,
+  "total_flos": 8.005887899965735e+18,
   "trial_name": null,
   "trial_params": null
 }