End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9751851851851852,
-    "eval_loss": 0.07488196343183517,
-    "eval_runtime": 5.0968,
-    "eval_samples_per_second": 529.746,
-    "eval_steps_per_second": 16.677,
+    "eval_accuracy": 0.9748148148148148,
+    "eval_loss": 0.0716383084654808,
+    "eval_runtime": 5.2267,
+    "eval_samples_per_second": 516.583,
+    "eval_steps_per_second": 16.263,
     "total_flos": 1.8124066505760768e+18,
-    "train_loss": 0.34760786274023225,
-    "train_runtime": 263.5996,
-    "train_samples_per_second": 276.556,
-    "train_steps_per_second": 2.162
+    "train_loss": 0.3320666231607136,
+    "train_runtime": 260.3624,
+    "train_samples_per_second": 279.994,
+    "train_steps_per_second": 2.189
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.9751851851851852,
-    "eval_loss": 0.07488196343183517,
-    "eval_runtime": 5.0968,
-    "eval_samples_per_second": 529.746,
-    "eval_steps_per_second": 16.677
+    "eval_accuracy": 0.9748148148148148,
+    "eval_loss": 0.0716383084654808,
+    "eval_runtime": 5.2267,
+    "eval_samples_per_second": 516.583,
+    "eval_steps_per_second": 16.263
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
     "total_flos": 1.8124066505760768e+18,
-    "train_loss": 0.34760786274023225,
-    "train_runtime": 263.5996,
-    "train_samples_per_second": 276.556,
-    "train_steps_per_second": 2.162
+    "train_loss": 0.3320666231607136,
+    "train_runtime": 260.3624,
+    "train_samples_per_second": 279.994,
+    "train_steps_per_second": 2.189
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.9751851851851852,
+  "best_metric": 0.9748148148148148,
   "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": 2.3445,
+      "loss": 2.2881,
       "step": 10
     },
     {
       "epoch": 0.11,
       "learning_rate": 1.7543859649122806e-05,
-      "loss": 2.1563,
+      "loss": 2.0518,
       "step": 20
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.6315789473684212e-05,
-      "loss": 1.7305,
+      "loss": 1.5928,
       "step": 30
     },
     {
       "epoch": 0.21,
       "learning_rate": 3.508771929824561e-05,
-      "loss": 1.0732,
+      "loss": 0.9705,
       "step": 40
     },
     {
       "epoch": 0.26,
       "learning_rate": 4.3859649122807014e-05,
-      "loss": 0.7244,
+      "loss": 0.6146,
       "step": 50
     },
     {
       "epoch": 0.32,
       "learning_rate": 4.970760233918128e-05,
-      "loss": 0.5671,
+      "loss": 0.4841,
       "step": 60
     },
     {
       "epoch": 0.37,
       "learning_rate": 4.8732943469785574e-05,
-      "loss": 0.4963,
+      "loss": 0.462,
       "step": 70
     },
     {
       "epoch": 0.42,
       "learning_rate": 4.7758284600389865e-05,
-      "loss": 0.5004,
+      "loss": 0.3651,
       "step": 80
     },
     {
       "epoch": 0.47,
       "learning_rate": 4.678362573099415e-05,
-      "loss": 0.4085,
+      "loss": 0.3779,
       "step": 90
     },
     {
       "epoch": 0.53,
       "learning_rate": 4.580896686159844e-05,
-      "loss": 0.4172,
+      "loss": 0.3984,
       "step": 100
     },
     {
       "epoch": 0.58,
       "learning_rate": 4.483430799220273e-05,
-      "loss": 0.3287,
+      "loss": 0.3431,
       "step": 110
     },
     {
       "epoch": 0.63,
       "learning_rate": 4.3859649122807014e-05,
-      "loss": 0.3083,
+      "loss": 0.3088,
       "step": 120
     },
     {
       "epoch": 0.68,
       "learning_rate": 4.2884990253411305e-05,
-      "loss": 0.3073,
+      "loss": 0.2594,
       "step": 130
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.1910331384015596e-05,
-      "loss": 0.2898,
+      "loss": 0.3027,
       "step": 140
     },
     {
       "epoch": 0.79,
       "learning_rate": 4.093567251461988e-05,
-      "loss": 0.2867,
+      "loss": 0.2738,
       "step": 150
     },
     {
       "epoch": 0.84,
       "learning_rate": 3.996101364522417e-05,
-      "loss": 0.2496,
+      "loss": 0.2651,
       "step": 160
     },
     {
       "epoch": 0.89,
       "learning_rate": 3.898635477582846e-05,
-      "loss": 0.3115,
+      "loss": 0.2923,
       "step": 170
     },
     {
       "epoch": 0.95,
       "learning_rate": 3.8011695906432746e-05,
-      "loss": 0.2363,
+      "loss": 0.2435,
       "step": 180
     },
     {
       "epoch": 1.0,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.2471,
+      "loss": 0.2409,
       "step": 190
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9485185185185185,
-      "eval_loss": 0.14287887513637543,
-      "eval_runtime": 5.6127,
-      "eval_samples_per_second": 481.053,
-      "eval_steps_per_second": 15.144,
+      "eval_accuracy": 0.9666666666666667,
+      "eval_loss": 0.11242207884788513,
+      "eval_runtime": 5.1679,
+      "eval_samples_per_second": 522.456,
+      "eval_steps_per_second": 16.448,
       "step": 190
     },
     {
       "epoch": 1.05,
       "learning_rate": 3.606237816764133e-05,
-      "loss": 0.2516,
+      "loss": 0.236,
       "step": 200
     },
     {
       "epoch": 1.11,
       "learning_rate": 3.508771929824561e-05,
-      "loss": 0.2287,
+      "loss": 0.216,
       "step": 210
     },
     {
       "epoch": 1.16,
       "learning_rate": 3.41130604288499e-05,
-      "loss": 0.2086,
+      "loss": 0.2762,
       "step": 220
     },
     {
       "epoch": 1.21,
       "learning_rate": 3.313840155945419e-05,
-      "loss": 0.2055,
+      "loss": 0.2386,
       "step": 230
     },
     {
       "epoch": 1.26,
       "learning_rate": 3.216374269005848e-05,
-      "loss": 0.2444,
+      "loss": 0.229,
       "step": 240
     },
     {
       "epoch": 1.32,
       "learning_rate": 3.118908382066277e-05,
-      "loss": 0.2222,
+      "loss": 0.2486,
       "step": 250
     },
     {
       "epoch": 1.37,
       "learning_rate": 3.0214424951267055e-05,
-      "loss": 0.2201,
+      "loss": 0.2034,
       "step": 260
     },
     {
       "epoch": 1.42,
       "learning_rate": 2.9239766081871346e-05,
-      "loss": 0.2105,
+      "loss": 0.2117,
       "step": 270
     },
     {
       "epoch": 1.47,
       "learning_rate": 2.8265107212475634e-05,
-      "loss": 0.2093,
+      "loss": 0.2064,
       "step": 280
     },
     {
       "epoch": 1.53,
       "learning_rate": 2.729044834307992e-05,
-      "loss": 0.1814,
+      "loss": 0.1925,
       "step": 290
     },
     {
       "epoch": 1.58,
       "learning_rate": 2.6315789473684212e-05,
-      "loss": 0.2059,
+      "loss": 0.2298,
       "step": 300
     },
     {
       "epoch": 1.63,
       "learning_rate": 2.53411306042885e-05,
-      "loss": 0.2003,
+      "loss": 0.1805,
       "step": 310
     },
     {
       "epoch": 1.68,
       "learning_rate": 2.4366471734892787e-05,
-      "loss": 0.2221,
+      "loss": 0.1766,
       "step": 320
     },
     {
       "epoch": 1.74,
       "learning_rate": 2.3391812865497074e-05,
-      "loss": 0.1465,
+      "loss": 0.1411,
       "step": 330
     },
     {
       "epoch": 1.79,
       "learning_rate": 2.2417153996101365e-05,
-      "loss": 0.2161,
+      "loss": 0.1523,
       "step": 340
     },
     {
       "epoch": 1.84,
       "learning_rate": 2.1442495126705653e-05,
-      "loss": 0.2064,
+      "loss": 0.1804,
       "step": 350
     },
     {
       "epoch": 1.89,
       "learning_rate": 2.046783625730994e-05,
-      "loss": 0.2124,
+      "loss": 0.1744,
       "step": 360
     },
     {
       "epoch": 1.95,
       "learning_rate": 1.949317738791423e-05,
-      "loss": 0.1777,
+      "loss": 0.2013,
       "step": 370
     },
     {
       "epoch": 2.0,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.1417,
+      "loss": 0.2039,
       "step": 380
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9737037037037037,
-      "eval_loss": 0.08577410876750946,
-      "eval_runtime": 4.8219,
-      "eval_samples_per_second": 559.942,
-      "eval_steps_per_second": 17.628,
+      "eval_accuracy": 0.9685185185185186,
+      "eval_loss": 0.09637241065502167,
+      "eval_runtime": 4.8698,
+      "eval_samples_per_second": 554.433,
+      "eval_steps_per_second": 17.454,
       "step": 380
     },
     {
       "epoch": 2.05,
       "learning_rate": 1.7543859649122806e-05,
-      "loss": 0.1367,
+      "loss": 0.2079,
       "step": 390
     },
     {
       "epoch": 2.11,
       "learning_rate": 1.6569200779727097e-05,
-      "loss": 0.1907,
+      "loss": 0.1616,
       "step": 400
     },
     {
       "epoch": 2.16,
       "learning_rate": 1.5594541910331384e-05,
-      "loss": 0.1684,
+      "loss": 0.1533,
       "step": 410
     },
     {
       "epoch": 2.21,
       "learning_rate": 1.4619883040935673e-05,
-      "loss": 0.1499,
+      "loss": 0.1751,
       "step": 420
     },
     {
       "epoch": 2.26,
       "learning_rate": 1.364522417153996e-05,
-      "loss": 0.1723,
+      "loss": 0.1646,
       "step": 430
     },
     {
       "epoch": 2.32,
       "learning_rate": 1.267056530214425e-05,
-      "loss": 0.1611,
+      "loss": 0.155,
       "step": 440
     },
     {
       "epoch": 2.37,
       "learning_rate": 1.1695906432748537e-05,
-      "loss": 0.1398,
+      "loss": 0.1811,
       "step": 450
     },
     {
       "epoch": 2.42,
       "learning_rate": 1.0721247563352826e-05,
-      "loss": 0.1364,
+      "loss": 0.1466,
       "step": 460
     },
     {
       "epoch": 2.47,
       "learning_rate": 9.746588693957115e-06,
-      "loss": 0.1461,
+      "loss": 0.1605,
       "step": 470
     },
     {
       "epoch": 2.53,
       "learning_rate": 8.771929824561403e-06,
-      "loss": 0.1528,
+      "loss": 0.1341,
       "step": 480
     },
     {
       "epoch": 2.58,
       "learning_rate": 7.797270955165692e-06,
-      "loss": 0.193,
+      "loss": 0.1258,
       "step": 490
     },
     {
       "epoch": 2.63,
       "learning_rate": 6.82261208576998e-06,
-      "loss": 0.1531,
+      "loss": 0.1462,
       "step": 500
     },
     {
       "epoch": 2.68,
       "learning_rate": 5.8479532163742686e-06,
-      "loss": 0.141,
+      "loss": 0.1665,
       "step": 510
     },
     {
       "epoch": 2.74,
       "learning_rate": 4.873294346978558e-06,
-      "loss": 0.1675,
+      "loss": 0.1508,
       "step": 520
     },
     {
       "epoch": 2.79,
       "learning_rate": 3.898635477582846e-06,
-      "loss": 0.1464,
+      "loss": 0.1402,
       "step": 530
     },
     {
       "epoch": 2.84,
       "learning_rate": 2.9239766081871343e-06,
-      "loss": 0.1444,
+      "loss": 0.1448,
       "step": 540
     },
     {
       "epoch": 2.89,
       "learning_rate": 1.949317738791423e-06,
-      "loss": 0.1194,
+      "loss": 0.1066,
       "step": 550
     },
     {
       "epoch": 2.95,
       "learning_rate": 9.746588693957115e-07,
-      "loss": 0.1461,
+      "loss": 0.1478,
       "step": 560
     },
     {
       "epoch": 3.0,
       "learning_rate": 0.0,
-      "loss": 0.1534,
+      "loss": 0.1254,
       "step": 570
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9751851851851852,
-      "eval_loss": 0.07488196343183517,
-      "eval_runtime": 4.8556,
-      "eval_samples_per_second": 556.06,
-      "eval_steps_per_second": 17.506,
+      "eval_accuracy": 0.9748148148148148,
+      "eval_loss": 0.0716383084654808,
+      "eval_runtime": 4.8486,
+      "eval_samples_per_second": 556.865,
+      "eval_steps_per_second": 17.531,
       "step": 570
     },
     {
       "epoch": 3.0,
       "step": 570,
       "total_flos": 1.8124066505760768e+18,
-      "train_loss": 0.34760786274023225,
-      "train_runtime": 263.5996,
-      "train_samples_per_second": 276.556,
-      "train_steps_per_second": 2.162
+      "train_loss": 0.3320666231607136,
+      "train_runtime": 260.3624,
+      "train_samples_per_second": 279.994,
+      "train_steps_per_second": 2.189
     }
   ],
   "logging_steps": 10,