End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 1.98,
-    "eval_accuracy": 0.9166666666666666,
-    "eval_loss": 0.25919345021247864,
-    "eval_runtime": 4.8825,
-    "eval_samples_per_second": 154.838,
-    "eval_steps_per_second": 4.916,
-    "total_flos": 2.1241704283294925e+17,
-    "train_loss": 0.7265663851391185,
-    "train_runtime": 142.6799,
-    "train_samples_per_second": 60.555,
-    "train_steps_per_second": 0.463
+    "epoch": 4.98,
+    "eval_accuracy": 0.9775132275132276,
+    "eval_loss": 0.09184322506189346,
+    "eval_runtime": 4.8893,
+    "eval_samples_per_second": 154.623,
+    "eval_steps_per_second": 4.909,
+    "total_flos": 5.346226696020296e+17,
+    "train_loss": 0.37407838611891775,
+    "train_runtime": 386.1594,
+    "train_samples_per_second": 55.935,
+    "train_steps_per_second": 0.427
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.98,
-    "eval_accuracy": 0.9166666666666666,
-    "eval_loss": 0.25919345021247864,
-    "eval_runtime": 4.8825,
-    "eval_samples_per_second": 154.838,
-    "eval_steps_per_second": 4.916
+    "epoch": 4.98,
+    "eval_accuracy": 0.9775132275132276,
+    "eval_loss": 0.09184322506189346,
+    "eval_runtime": 4.8893,
+    "eval_samples_per_second": 154.623,
+    "eval_steps_per_second": 4.909
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.98,
-    "total_flos": 2.1241704283294925e+17,
-    "train_loss": 0.7265663851391185,
-    "train_runtime": 142.6799,
-    "train_samples_per_second": 60.555,
-    "train_steps_per_second": 0.463
+    "epoch": 4.98,
+    "total_flos": 5.346226696020296e+17,
+    "train_loss": 0.37407838611891775,
+    "train_runtime": 386.1594,
+    "train_samples_per_second": 55.935,
+    "train_steps_per_second": 0.427
 }

trainer_state.json

@@ -1,79 +1,166 @@
 {
-  "best_metric": 0.9166666666666666,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-eurosat/checkpoint-66",
-  "epoch": 1.9777777777777779,
-  "global_step": 66,
+  "best_metric": 0.9775132275132276,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-eurosat/checkpoint-165",
+  "epoch": 4.977777777777778,
+  "global_step": 165,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.3,
-      "learning_rate": 4.745762711864407e-05,
-      "loss": 2.1648,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 2.2557,
       "step": 10
     },
     {
       "epoch": 0.59,
-      "learning_rate": 3.898305084745763e-05,
-      "loss": 1.1607,
+      "learning_rate": 4.8986486486486486e-05,
+      "loss": 1.5384,
       "step": 20
     },
     {
       "epoch": 0.89,
-      "learning_rate": 3.050847457627119e-05,
-      "loss": 0.4904,
+      "learning_rate": 4.560810810810811e-05,
+      "loss": 0.6233,
       "step": 30
     },
     {
       "epoch": 0.98,
-      "eval_accuracy": 0.8783068783068783,
-      "eval_loss": 0.3856896162033081,
-      "eval_runtime": 4.8675,
-      "eval_samples_per_second": 155.315,
-      "eval_steps_per_second": 4.931,
+      "eval_accuracy": 0.91005291005291,
+      "eval_loss": 0.30006203055381775,
+      "eval_runtime": 4.7638,
+      "eval_samples_per_second": 158.696,
+      "eval_steps_per_second": 5.038,
       "step": 33
     },
     {
       "epoch": 1.21,
-      "learning_rate": 2.2033898305084748e-05,
-      "loss": 0.3647,
+      "learning_rate": 4.222972972972973e-05,
+      "loss": 0.3829,
       "step": 40
     },
     {
       "epoch": 1.5,
-      "learning_rate": 1.3559322033898305e-05,
-      "loss": 0.2628,
+      "learning_rate": 3.885135135135135e-05,
+      "loss": 0.2118,
       "step": 50
     },
     {
       "epoch": 1.8,
-      "learning_rate": 5.084745762711865e-06,
-      "loss": 0.2188,
+      "learning_rate": 3.547297297297297e-05,
+      "loss": 0.1958,
       "step": 60
     },
     {
       "epoch": 1.98,
-      "eval_accuracy": 0.9166666666666666,
-      "eval_loss": 0.25919345021247864,
-      "eval_runtime": 4.9632,
-      "eval_samples_per_second": 152.321,
-      "eval_steps_per_second": 4.836,
+      "eval_accuracy": 0.9642857142857143,
+      "eval_loss": 0.12865345180034637,
+      "eval_runtime": 4.9076,
+      "eval_samples_per_second": 154.047,
+      "eval_steps_per_second": 4.89,
       "step": 66
     },
     {
-      "epoch": 1.98,
-      "step": 66,
-      "total_flos": 2.1241704283294925e+17,
-      "train_loss": 0.7265663851391185,
-      "train_runtime": 142.6799,
-      "train_samples_per_second": 60.555,
-      "train_steps_per_second": 0.463
+      "epoch": 2.12,
+      "learning_rate": 3.20945945945946e-05,
+      "loss": 0.1691,
+      "step": 70
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.8716216216216217e-05,
+      "loss": 0.1239,
+      "step": 80
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.533783783783784e-05,
+      "loss": 0.1212,
+      "step": 90
+    },
+    {
+      "epoch": 2.98,
+      "eval_accuracy": 0.9708994708994709,
+      "eval_loss": 0.11089407652616501,
+      "eval_runtime": 4.7941,
+      "eval_samples_per_second": 157.693,
+      "eval_steps_per_second": 5.006,
+      "step": 99
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.195945945945946e-05,
+      "loss": 0.1071,
+      "step": 100
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.8581081081081082e-05,
+      "loss": 0.0814,
+      "step": 110
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.5202702702702704e-05,
+      "loss": 0.0962,
+      "step": 120
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.1824324324324325e-05,
+      "loss": 0.0734,
+      "step": 130
+    },
+    {
+      "epoch": 3.98,
+      "eval_accuracy": 0.9642857142857143,
+      "eval_loss": 0.11790694296360016,
+      "eval_runtime": 4.8978,
+      "eval_samples_per_second": 154.355,
+      "eval_steps_per_second": 4.9,
+      "step": 132
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.445945945945946e-06,
+      "loss": 0.0679,
+      "step": 140
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 5.067567567567568e-06,
+      "loss": 0.0516,
+      "step": 150
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.6891891891891894e-06,
+      "loss": 0.0457,
+      "step": 160
+    },
+    {
+      "epoch": 4.98,
+      "eval_accuracy": 0.9775132275132276,
+      "eval_loss": 0.09184322506189346,
+      "eval_runtime": 4.7828,
+      "eval_samples_per_second": 158.066,
+      "eval_steps_per_second": 5.018,
+      "step": 165
+    },
+    {
+      "epoch": 4.98,
+      "step": 165,
+      "total_flos": 5.346226696020296e+17,
+      "train_loss": 0.37407838611891775,
+      "train_runtime": 386.1594,
+      "train_samples_per_second": 55.935,
+      "train_steps_per_second": 0.427
     }
   ],
-  "max_steps": 66,
-  "num_train_epochs": 2,
-  "total_flos": 2.1241704283294925e+17,
+  "max_steps": 165,
+  "num_train_epochs": 5,
+  "total_flos": 5.346226696020296e+17,
   "trial_name": null,
   "trial_params": null
 }