End of training

README.md

@@ -26,16 +26,16 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.9619047619047619
+      value: 0.9714285714285714
     - name: Precision
       type: precision
-      value: 0.9583926593893372
+      value: 0.9696825396825397
     - name: Recall
       type: recall
-      value: 0.9619047619047619
+      value: 0.9714285714285714
     - name: F1
       type: f1
-      value: 0.9587351719211175
+      value: 0.9695078031212484
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -45,11 +45,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0711
-- Accuracy: 0.9619
-- Precision: 0.9584
-- Recall: 0.9619
-- F1: 0.9587
+- Loss: 0.0814
+- Accuracy: 0.9714
+- Precision: 0.9697
+- Recall: 0.9714
+- F1: 0.9695
 
 ## Model description
 

all_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 4.615384615384615,
-    "eval_accuracy": 1.0,
-    "eval_f1": 1.0,
-    "eval_loss": 0.0053217145614326,
-    "eval_precision": 1.0,
-    "eval_recall": 1.0,
-    "eval_runtime": 1.8276,
-    "eval_samples_per_second": 24.075,
-    "eval_steps_per_second": 1.094,
-    "total_flos": 4.543674786614477e+16,
-    "train_loss": 0.02700179343422254,
-    "train_runtime": 121.8192,
-    "train_samples_per_second": 16.13,
-    "train_steps_per_second": 0.123
+    "epoch": 9.887640449438202,
+    "eval_accuracy": 0.9714285714285714,
+    "eval_f1": 0.9695078031212484,
+    "eval_loss": 0.08143213391304016,
+    "eval_precision": 0.9696825396825397,
+    "eval_recall": 0.9714285714285714,
+    "eval_runtime": 31.2051,
+    "eval_samples_per_second": 10.094,
+    "eval_steps_per_second": 0.32,
+    "total_flos": 6.968127952294871e+17,
+    "train_loss": 0.13535271666266702,
+    "train_runtime": 2769.4739,
+    "train_samples_per_second": 10.233,
+    "train_steps_per_second": 0.079
 }

eval_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 4.615384615384615,
-    "eval_accuracy": 1.0,
-    "eval_f1": 1.0,
-    "eval_loss": 0.0053217145614326,
-    "eval_precision": 1.0,
-    "eval_recall": 1.0,
-    "eval_runtime": 1.8276,
-    "eval_samples_per_second": 24.075,
-    "eval_steps_per_second": 1.094
+    "epoch": 9.887640449438202,
+    "eval_accuracy": 0.9714285714285714,
+    "eval_f1": 0.9695078031212484,
+    "eval_loss": 0.08143213391304016,
+    "eval_precision": 0.9696825396825397,
+    "eval_recall": 0.9714285714285714,
+    "eval_runtime": 31.2051,
+    "eval_samples_per_second": 10.094,
+    "eval_steps_per_second": 0.32
 }

runs/Oct25_10-28-31_254845bf780a/events.out.tfevents.1729855041.254845bf780a.231.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9eda89293f0f203d7b9913a3d33664584862a12db04180b39a15cf8dd86c9a4
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.615384615384615,
-    "total_flos": 4.543674786614477e+16,
-    "train_loss": 0.02700179343422254,
-    "train_runtime": 121.8192,
-    "train_samples_per_second": 16.13,
-    "train_steps_per_second": 0.123
+    "epoch": 9.887640449438202,
+    "total_flos": 6.968127952294871e+17,
+    "train_loss": 0.13535271666266702,
+    "train_runtime": 2769.4739,
+    "train_samples_per_second": 10.233,
+    "train_steps_per_second": 0.079
 }

trainer_state.json

@@ -1,94 +1,301 @@
 {
-  "best_metric": 1.0,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-6",
-  "epoch": 4.615384615384615,
+  "best_metric": 0.9695078031212484,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-111",
+  "epoch": 9.887640449438202,
   "eval_steps": 500,
-  "global_step": 15,
+  "global_step": 220,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.9230769230769231,
-      "eval_accuracy": 0.9318181818181818,
-      "eval_f1": 0.9303133199567835,
-      "eval_loss": 0.13062183558940887,
-      "eval_precision": 0.9386363636363636,
-      "eval_recall": 0.9318181818181818,
-      "eval_runtime": 2.2426,
-      "eval_samples_per_second": 19.62,
-      "eval_steps_per_second": 0.892,
-      "step": 3
-    },
-    {
-      "epoch": 1.8461538461538463,
-      "eval_accuracy": 1.0,
-      "eval_f1": 1.0,
-      "eval_loss": 0.0053217145614326,
-      "eval_precision": 1.0,
-      "eval_recall": 1.0,
-      "eval_runtime": 1.7899,
-      "eval_samples_per_second": 24.583,
-      "eval_steps_per_second": 1.117,
-      "step": 6
-    },
-    {
-      "epoch": 2.769230769230769,
-      "eval_accuracy": 1.0,
-      "eval_f1": 1.0,
-      "eval_loss": 0.001675223233178258,
-      "eval_precision": 1.0,
-      "eval_recall": 1.0,
-      "eval_runtime": 1.7929,
-      "eval_samples_per_second": 24.541,
-      "eval_steps_per_second": 1.115,
-      "step": 9
-    },
-    {
-      "epoch": 3.076923076923077,
-      "grad_norm": 7.759745121002197,
-      "learning_rate": 1.923076923076923e-05,
-      "loss": 0.0353,
+      "epoch": 0.449438202247191,
+      "grad_norm": 6.729045391082764,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.6016,
       "step": 10
     },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 5.153387546539307,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.2262,
+      "step": 20
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "eval_accuracy": 0.9365079365079365,
+      "eval_f1": 0.9058027582617746,
+      "eval_loss": 0.20612730085849762,
+      "eval_precision": 0.8770471151423532,
+      "eval_recall": 0.9365079365079365,
+      "eval_runtime": 29.0895,
+      "eval_samples_per_second": 10.829,
+      "eval_steps_per_second": 0.344,
+      "step": 22
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 2.6891238689422607,
+      "learning_rate": 4.797979797979798e-05,
+      "loss": 0.1932,
+      "step": 30
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 10.437292098999023,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.1666,
+      "step": 40
+    },
+    {
+      "epoch": 1.9775280898876404,
+      "eval_accuracy": 0.9333333333333333,
+      "eval_f1": 0.9042145593869733,
+      "eval_loss": 0.12735532224178314,
+      "eval_precision": 0.8768577494692145,
+      "eval_recall": 0.9333333333333333,
+      "eval_runtime": 28.9476,
+      "eval_samples_per_second": 10.882,
+      "eval_steps_per_second": 0.345,
+      "step": 44
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 13.23202896118164,
+      "learning_rate": 4.292929292929293e-05,
+      "loss": 0.1408,
+      "step": 50
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 13.037389755249023,
+      "learning_rate": 4.0404040404040405e-05,
+      "loss": 0.1168,
+      "step": 60
+    },
+    {
+      "epoch": 2.966292134831461,
+      "eval_accuracy": 0.9523809523809523,
+      "eval_f1": 0.9437855353283663,
+      "eval_loss": 0.10540811717510223,
+      "eval_precision": 0.9461043676729951,
+      "eval_recall": 0.9523809523809523,
+      "eval_runtime": 28.9757,
+      "eval_samples_per_second": 10.871,
+      "eval_steps_per_second": 0.345,
+      "step": 66
+    },
+    {
+      "epoch": 3.146067415730337,
+      "grad_norm": 13.592888832092285,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.132,
+      "step": 70
+    },
+    {
+      "epoch": 3.595505617977528,
+      "grad_norm": 4.916988849639893,
+      "learning_rate": 3.535353535353535e-05,
+      "loss": 0.0984,
+      "step": 80
+    },
     {
       "epoch": 4.0,
-      "eval_accuracy": 1.0,
-      "eval_f1": 1.0,
-      "eval_loss": 0.0020660320296883583,
-      "eval_precision": 1.0,
-      "eval_recall": 1.0,
-      "eval_runtime": 2.6393,
-      "eval_samples_per_second": 16.671,
-      "eval_steps_per_second": 0.758,
-      "step": 13
-    },
-    {
-      "epoch": 4.615384615384615,
-      "eval_accuracy": 1.0,
-      "eval_f1": 1.0,
-      "eval_loss": 0.0022493237629532814,
-      "eval_precision": 1.0,
-      "eval_recall": 1.0,
-      "eval_runtime": 2.0209,
-      "eval_samples_per_second": 21.772,
-      "eval_steps_per_second": 0.99,
-      "step": 15
-    },
-    {
-      "epoch": 4.615384615384615,
-      "step": 15,
-      "total_flos": 4.543674786614477e+16,
-      "train_loss": 0.02700179343422254,
-      "train_runtime": 121.8192,
-      "train_samples_per_second": 16.13,
-      "train_steps_per_second": 0.123
+      "eval_accuracy": 0.9619047619047619,
+      "eval_f1": 0.9599166265832932,
+      "eval_loss": 0.08242347091436386,
+      "eval_precision": 0.9590699155916548,
+      "eval_recall": 0.9619047619047619,
+      "eval_runtime": 29.2105,
+      "eval_samples_per_second": 10.784,
+      "eval_steps_per_second": 0.342,
+      "step": 89
+    },
+    {
+      "epoch": 4.044943820224719,
+      "grad_norm": 5.384190082550049,
+      "learning_rate": 3.282828282828283e-05,
+      "loss": 0.0997,
+      "step": 90
+    },
+    {
+      "epoch": 4.49438202247191,
+      "grad_norm": 6.7096028327941895,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.1319,
+      "step": 100
+    },
+    {
+      "epoch": 4.943820224719101,
+      "grad_norm": 3.9921090602874756,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.1028,
+      "step": 110
+    },
+    {
+      "epoch": 4.98876404494382,
+      "eval_accuracy": 0.9714285714285714,
+      "eval_f1": 0.9695078031212484,
+      "eval_loss": 0.08143213391304016,
+      "eval_precision": 0.9696825396825397,
+      "eval_recall": 0.9714285714285714,
+      "eval_runtime": 29.3091,
+      "eval_samples_per_second": 10.748,
+      "eval_steps_per_second": 0.341,
+      "step": 111
+    },
+    {
+      "epoch": 5.393258426966292,
+      "grad_norm": 7.567589282989502,
+      "learning_rate": 2.5252525252525256e-05,
+      "loss": 0.0806,
+      "step": 120
+    },
+    {
+      "epoch": 5.842696629213483,
+      "grad_norm": 3.001021385192871,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.1082,
+      "step": 130
+    },
+    {
+      "epoch": 5.97752808988764,
+      "eval_accuracy": 0.9492063492063492,
+      "eval_f1": 0.9329456755199329,
+      "eval_loss": 0.08350013196468353,
+      "eval_precision": 0.9518195273822283,
+      "eval_recall": 0.9492063492063492,
+      "eval_runtime": 28.3828,
+      "eval_samples_per_second": 11.098,
+      "eval_steps_per_second": 0.352,
+      "step": 133
+    },
+    {
+      "epoch": 6.292134831460674,
+      "grad_norm": 3.2272369861602783,
+      "learning_rate": 2.0202020202020203e-05,
+      "loss": 0.0914,
+      "step": 140
+    },
+    {
+      "epoch": 6.741573033707866,
+      "grad_norm": 11.393094062805176,
+      "learning_rate": 1.7676767676767676e-05,
+      "loss": 0.0962,
+      "step": 150
+    },
+    {
+      "epoch": 6.966292134831461,
+      "eval_accuracy": 0.9587301587301588,
+      "eval_f1": 0.9582359734643997,
+      "eval_loss": 0.08720001578330994,
+      "eval_precision": 0.9578027275395696,
+      "eval_recall": 0.9587301587301588,
+      "eval_runtime": 28.5518,
+      "eval_samples_per_second": 11.033,
+      "eval_steps_per_second": 0.35,
+      "step": 155
+    },
+    {
+      "epoch": 7.191011235955056,
+      "grad_norm": 3.3998546600341797,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.1013,
+      "step": 160
+    },
+    {
+      "epoch": 7.640449438202247,
+      "grad_norm": 2.875291109085083,
+      "learning_rate": 1.2626262626262628e-05,
+      "loss": 0.0799,
+      "step": 170
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9587301587301588,
+      "eval_f1": 0.9545950048462611,
+      "eval_loss": 0.08030053973197937,
+      "eval_precision": 0.9543377887748749,
+      "eval_recall": 0.9587301587301588,
+      "eval_runtime": 28.171,
+      "eval_samples_per_second": 11.182,
+      "eval_steps_per_second": 0.355,
+      "step": 178
+    },
+    {
+      "epoch": 8.089887640449438,
+      "grad_norm": 3.642427682876587,
+      "learning_rate": 1.0101010101010101e-05,
+      "loss": 0.0703,
+      "step": 180
+    },
+    {
+      "epoch": 8.539325842696629,
+      "grad_norm": 11.242568016052246,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.0869,
+      "step": 190
+    },
+    {
+      "epoch": 8.98876404494382,
+      "grad_norm": 11.542448997497559,
+      "learning_rate": 5.050505050505051e-06,
+      "loss": 0.0954,
+      "step": 200
+    },
+    {
+      "epoch": 8.98876404494382,
+      "eval_accuracy": 0.9619047619047619,
+      "eval_f1": 0.9587351719211175,
+      "eval_loss": 0.06851905584335327,
+      "eval_precision": 0.9583926593893372,
+      "eval_recall": 0.9619047619047619,
+      "eval_runtime": 28.2782,
+      "eval_samples_per_second": 11.139,
+      "eval_steps_per_second": 0.354,
+      "step": 200
+    },
+    {
+      "epoch": 9.438202247191011,
+      "grad_norm": 4.964309215545654,
+      "learning_rate": 2.5252525252525253e-06,
+      "loss": 0.0806,
+      "step": 210
+    },
+    {
+      "epoch": 9.887640449438202,
+      "grad_norm": 3.8892452716827393,
+      "learning_rate": 0.0,
+      "loss": 0.0771,
+      "step": 220
+    },
+    {
+      "epoch": 9.887640449438202,
+      "eval_accuracy": 0.9619047619047619,
+      "eval_f1": 0.9587351719211175,
+      "eval_loss": 0.07111740857362747,
+      "eval_precision": 0.9583926593893372,
+      "eval_recall": 0.9619047619047619,
+      "eval_runtime": 28.6883,
+      "eval_samples_per_second": 10.98,
+      "eval_steps_per_second": 0.349,
+      "step": 220
+    },
+    {
+      "epoch": 9.887640449438202,
+      "step": 220,
+      "total_flos": 6.968127952294871e+17,
+      "train_loss": 0.13535271666266702,
+      "train_runtime": 2769.4739,
+      "train_samples_per_second": 10.233,
+      "train_steps_per_second": 0.079
     }
   ],
   "logging_steps": 10,
-  "max_steps": 15,
+  "max_steps": 220,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 10,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -102,7 +309,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.543674786614477e+16,
+  "total_flos": 6.968127952294871e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null