End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9803703703703703,
-    "eval_loss": 0.059159792959690094,
-    "eval_runtime": 4.8853,
-    "eval_samples_per_second": 552.68,
-    "eval_steps_per_second": 17.399,
+    "eval_accuracy": 0.9740740740740741,
+    "eval_loss": 0.07266821712255478,
+    "eval_runtime": 5.0363,
+    "eval_samples_per_second": 536.108,
+    "eval_steps_per_second": 16.877,
     "total_flos": 1.8124066505760768e+18,
-    "train_loss": 0.32973566536317794,
-    "train_runtime": 250.3967,
-    "train_samples_per_second": 291.138,
-    "train_steps_per_second": 2.276
+    "train_loss": 0.34982479978026004,
+    "train_runtime": 261.9976,
+    "train_samples_per_second": 278.247,
+    "train_steps_per_second": 2.176
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9803703703703703,
-    "eval_loss": 0.059159792959690094,
-    "eval_runtime": 4.8853,
-    "eval_samples_per_second": 552.68,
-    "eval_steps_per_second": 17.399
+    "eval_accuracy": 0.9740740740740741,
+    "eval_loss": 0.07266821712255478,
+    "eval_runtime": 5.0363,
+    "eval_samples_per_second": 536.108,
+    "eval_steps_per_second": 16.877
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
     "total_flos": 1.8124066505760768e+18,
-    "train_loss": 0.32973566536317794,
-    "train_runtime": 250.3967,
-    "train_samples_per_second": 291.138,
-    "train_steps_per_second": 2.276
+    "train_loss": 0.34982479978026004,
+    "train_runtime": 261.9976,
+    "train_samples_per_second": 278.247,
+    "train_steps_per_second": 2.176
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.9803703703703703,
+  "best_metric": 0.9740740740740741,
   "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-570",
   "epoch": 3.0,
   "eval_steps": 500,
@@ -11,380 +11,380 @@
     {
       "epoch": 0.05,
       "learning_rate": 8.771929824561403e-06,
-      "loss": 2.2734,
+      "loss": 2.3809,
       "step": 10
     },
     {
       "epoch": 0.11,
       "learning_rate": 1.7543859649122806e-05,
-      "loss": 2.0923,
+      "loss": 2.1931,
       "step": 20
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.6315789473684212e-05,
-      "loss": 1.6401,
+      "loss": 1.7658,
       "step": 30
     },
     {
       "epoch": 0.21,
       "learning_rate": 3.508771929824561e-05,
-      "loss": 0.9479,
+      "loss": 1.1384,
       "step": 40
     },
     {
       "epoch": 0.26,
       "learning_rate": 4.3859649122807014e-05,
-      "loss": 0.5664,
+      "loss": 0.7236,
       "step": 50
     },
     {
       "epoch": 0.32,
       "learning_rate": 4.970760233918128e-05,
-      "loss": 0.476,
+      "loss": 0.5221,
       "step": 60
     },
     {
       "epoch": 0.37,
       "learning_rate": 4.8732943469785574e-05,
-      "loss": 0.4555,
+      "loss": 0.503,
       "step": 70
     },
     {
       "epoch": 0.42,
       "learning_rate": 4.7758284600389865e-05,
-      "loss": 0.3879,
+      "loss": 0.4116,
       "step": 80
     },
     {
       "epoch": 0.47,
       "learning_rate": 4.678362573099415e-05,
-      "loss": 0.3706,
+      "loss": 0.399,
       "step": 90
     },
     {
       "epoch": 0.53,
       "learning_rate": 4.580896686159844e-05,
-      "loss": 0.394,
+      "loss": 0.3777,
       "step": 100
     },
     {
       "epoch": 0.58,
       "learning_rate": 4.483430799220273e-05,
-      "loss": 0.344,
+      "loss": 0.3544,
       "step": 110
     },
     {
       "epoch": 0.63,
       "learning_rate": 4.3859649122807014e-05,
-      "loss": 0.2958,
+      "loss": 0.3187,
       "step": 120
     },
     {
       "epoch": 0.68,
       "learning_rate": 4.2884990253411305e-05,
-      "loss": 0.2769,
+      "loss": 0.2529,
       "step": 130
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.1910331384015596e-05,
-      "loss": 0.3316,
+      "loss": 0.2952,
       "step": 140
     },
     {
       "epoch": 0.79,
       "learning_rate": 4.093567251461988e-05,
-      "loss": 0.2917,
+      "loss": 0.3222,
       "step": 150
     },
     {
       "epoch": 0.84,
       "learning_rate": 3.996101364522417e-05,
-      "loss": 0.2905,
+      "loss": 0.2986,
       "step": 160
     },
     {
       "epoch": 0.89,
       "learning_rate": 3.898635477582846e-05,
-      "loss": 0.2667,
+      "loss": 0.2572,
       "step": 170
     },
     {
       "epoch": 0.95,
       "learning_rate": 3.8011695906432746e-05,
-      "loss": 0.2338,
+      "loss": 0.2493,
       "step": 180
     },
     {
       "epoch": 1.0,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.2484,
+      "loss": 0.252,
       "step": 190
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9685185185185186,
-      "eval_loss": 0.1036255806684494,
-      "eval_runtime": 5.1297,
-      "eval_samples_per_second": 526.346,
-      "eval_steps_per_second": 16.57,
+      "eval_accuracy": 0.95,
+      "eval_loss": 0.14173315465450287,
+      "eval_runtime": 5.2278,
+      "eval_samples_per_second": 516.472,
+      "eval_steps_per_second": 16.259,
       "step": 190
     },
     {
       "epoch": 1.05,
       "learning_rate": 3.606237816764133e-05,
-      "loss": 0.2245,
+      "loss": 0.2277,
       "step": 200
     },
     {
       "epoch": 1.11,
       "learning_rate": 3.508771929824561e-05,
-      "loss": 0.2082,
+      "loss": 0.2509,
       "step": 210
     },
     {
       "epoch": 1.16,
       "learning_rate": 3.41130604288499e-05,
-      "loss": 0.2787,
+      "loss": 0.2161,
       "step": 220
     },
     {
       "epoch": 1.21,
       "learning_rate": 3.313840155945419e-05,
-      "loss": 0.2265,
+      "loss": 0.2069,
       "step": 230
     },
     {
       "epoch": 1.26,
       "learning_rate": 3.216374269005848e-05,
-      "loss": 0.2218,
+      "loss": 0.2356,
       "step": 240
     },
     {
       "epoch": 1.32,
       "learning_rate": 3.118908382066277e-05,
-      "loss": 0.2426,
+      "loss": 0.2241,
       "step": 250
     },
     {
       "epoch": 1.37,
       "learning_rate": 3.0214424951267055e-05,
-      "loss": 0.2044,
+      "loss": 0.199,
       "step": 260
     },
     {
       "epoch": 1.42,
       "learning_rate": 2.9239766081871346e-05,
-      "loss": 0.2025,
+      "loss": 0.2151,
       "step": 270
     },
     {
       "epoch": 1.47,
       "learning_rate": 2.8265107212475634e-05,
-      "loss": 0.2021,
+      "loss": 0.2074,
       "step": 280
     },
     {
       "epoch": 1.53,
       "learning_rate": 2.729044834307992e-05,
-      "loss": 0.1925,
+      "loss": 0.1828,
       "step": 290
     },
     {
       "epoch": 1.58,
       "learning_rate": 2.6315789473684212e-05,
-      "loss": 0.2151,
+      "loss": 0.2399,
       "step": 300
     },
     {
       "epoch": 1.63,
       "learning_rate": 2.53411306042885e-05,
-      "loss": 0.1906,
+      "loss": 0.2006,
       "step": 310
     },
     {
       "epoch": 1.68,
       "learning_rate": 2.4366471734892787e-05,
-      "loss": 0.1787,
+      "loss": 0.2121,
       "step": 320
     },
     {
       "epoch": 1.74,
       "learning_rate": 2.3391812865497074e-05,
-      "loss": 0.1393,
+      "loss": 0.188,
       "step": 330
     },
     {
       "epoch": 1.79,
       "learning_rate": 2.2417153996101365e-05,
-      "loss": 0.1516,
+      "loss": 0.1615,
       "step": 340
     },
     {
       "epoch": 1.84,
       "learning_rate": 2.1442495126705653e-05,
-      "loss": 0.164,
+      "loss": 0.184,
       "step": 350
     },
     {
       "epoch": 1.89,
       "learning_rate": 2.046783625730994e-05,
-      "loss": 0.1747,
+      "loss": 0.1969,
       "step": 360
     },
     {
       "epoch": 1.95,
       "learning_rate": 1.949317738791423e-05,
-      "loss": 0.1964,
+      "loss": 0.2069,
       "step": 370
     },
     {
       "epoch": 2.0,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.219,
+      "loss": 0.1846,
       "step": 380
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9729629629629629,
-      "eval_loss": 0.08252211660146713,
-      "eval_runtime": 4.4943,
-      "eval_samples_per_second": 600.762,
-      "eval_steps_per_second": 18.913,
+      "eval_accuracy": 0.9714814814814815,
+      "eval_loss": 0.07922197133302689,
+      "eval_runtime": 4.6807,
+      "eval_samples_per_second": 576.841,
+      "eval_steps_per_second": 18.16,
       "step": 380
     },
     {
       "epoch": 2.05,
       "learning_rate": 1.7543859649122806e-05,
-      "loss": 0.2044,
+      "loss": 0.169,
       "step": 390
     },
     {
       "epoch": 2.11,
       "learning_rate": 1.6569200779727097e-05,
-      "loss": 0.1499,
+      "loss": 0.1869,
       "step": 400
     },
     {
       "epoch": 2.16,
       "learning_rate": 1.5594541910331384e-05,
-      "loss": 0.1401,
+      "loss": 0.145,
       "step": 410
     },
     {
       "epoch": 2.21,
       "learning_rate": 1.4619883040935673e-05,
-      "loss": 0.1596,
+      "loss": 0.1967,
       "step": 420
     },
     {
       "epoch": 2.26,
       "learning_rate": 1.364522417153996e-05,
-      "loss": 0.163,
+      "loss": 0.188,
       "step": 430
     },
     {
       "epoch": 2.32,
       "learning_rate": 1.267056530214425e-05,
-      "loss": 0.1403,
+      "loss": 0.1593,
       "step": 440
     },
     {
       "epoch": 2.37,
       "learning_rate": 1.1695906432748537e-05,
-      "loss": 0.1767,
+      "loss": 0.1371,
       "step": 450
     },
     {
       "epoch": 2.42,
       "learning_rate": 1.0721247563352826e-05,
-      "loss": 0.134,
+      "loss": 0.1841,
       "step": 460
     },
     {
       "epoch": 2.47,
       "learning_rate": 9.746588693957115e-06,
-      "loss": 0.1522,
+      "loss": 0.1659,
       "step": 470
     },
     {
       "epoch": 2.53,
       "learning_rate": 8.771929824561403e-06,
-      "loss": 0.14,
+      "loss": 0.1354,
       "step": 480
     },
     {
       "epoch": 2.58,
       "learning_rate": 7.797270955165692e-06,
-      "loss": 0.129,
+      "loss": 0.1381,
       "step": 490
     },
     {
       "epoch": 2.63,
       "learning_rate": 6.82261208576998e-06,
-      "loss": 0.1437,
+      "loss": 0.1331,
       "step": 500
     },
     {
       "epoch": 2.68,
       "learning_rate": 5.8479532163742686e-06,
-      "loss": 0.1594,
+      "loss": 0.1494,
       "step": 510
     },
     {
       "epoch": 2.74,
       "learning_rate": 4.873294346978558e-06,
-      "loss": 0.1436,
+      "loss": 0.12,
       "step": 520
     },
     {
       "epoch": 2.79,
       "learning_rate": 3.898635477582846e-06,
-      "loss": 0.1386,
+      "loss": 0.1648,
       "step": 530
     },
     {
       "epoch": 2.84,
       "learning_rate": 2.9239766081871343e-06,
-      "loss": 0.1363,
+      "loss": 0.1532,
       "step": 540
     },
     {
       "epoch": 2.89,
       "learning_rate": 1.949317738791423e-06,
-      "loss": 0.105,
+      "loss": 0.1392,
       "step": 550
     },
     {
       "epoch": 2.95,
       "learning_rate": 9.746588693957115e-07,
-      "loss": 0.1413,
+      "loss": 0.1302,
       "step": 560
     },
     {
       "epoch": 3.0,
       "learning_rate": 0.0,
-      "loss": 0.121,
+      "loss": 0.1887,
       "step": 570
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9803703703703703,
-      "eval_loss": 0.059159792959690094,
-      "eval_runtime": 4.6082,
-      "eval_samples_per_second": 585.917,
-      "eval_steps_per_second": 18.446,
+      "eval_accuracy": 0.9740740740740741,
+      "eval_loss": 0.07266821712255478,
+      "eval_runtime": 5.2195,
+      "eval_samples_per_second": 517.289,
+      "eval_steps_per_second": 16.285,
       "step": 570
     },
     {
       "epoch": 3.0,
       "step": 570,
       "total_flos": 1.8124066505760768e+18,
-      "train_loss": 0.32973566536317794,
-      "train_runtime": 250.3967,
-      "train_samples_per_second": 291.138,
-      "train_steps_per_second": 2.276
+      "train_loss": 0.34982479978026004,
+      "train_runtime": 261.9976,
+      "train_samples_per_second": 278.247,
+      "train_steps_per_second": 2.176
     }
   ],
   "logging_steps": 10,