End of training

all_results.json

@@ -1,8 +1,13 @@
 {
-    "epoch": 2.89,
-    "total_flos": 1.7986296196553933e+17,
-    "train_loss": 0.9899872813308448,
-    "train_runtime": 160.5331,
-    "train_samples_per_second": 46.682,
-    "train_steps_per_second": 0.355
+    "epoch": 11.54,
+    "eval_accuracy": 0.9100719424460432,
+    "eval_loss": 0.263301283121109,
+    "eval_runtime": 5.2084,
+    "eval_samples_per_second": 53.376,
+    "eval_steps_per_second": 1.728,
+    "total_flos": 7.17214746345173e+17,
+    "train_loss": 0.48171327511469525,
+    "train_runtime": 539.5354,
+    "train_samples_per_second": 55.559,
+    "train_steps_per_second": 0.423
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 11.54,
+    "eval_accuracy": 0.9100719424460432,
+    "eval_loss": 0.263301283121109,
+    "eval_runtime": 5.2084,
+    "eval_samples_per_second": 53.376,
+    "eval_steps_per_second": 1.728
+}

runs/Apr06_13-30-37_615cc53a6797/events.out.tfevents.1712410808.615cc53a6797.172.6

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cbc579aee87251cff3126f3303e0ad9bc6677cdc243b96f680dfb7322872d53
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.89,
-    "total_flos": 1.7986296196553933e+17,
-    "train_loss": 0.9899872813308448,
-    "train_runtime": 160.5331,
-    "train_samples_per_second": 46.682,
-    "train_steps_per_second": 0.355
+    "epoch": 11.54,
+    "total_flos": 7.17214746345173e+17,
+    "train_loss": 0.48171327511469525,
+    "train_runtime": 539.5354,
+    "train_samples_per_second": 55.559,
+    "train_steps_per_second": 0.423
 }

trainer_state.json

@@ -1,91 +1,291 @@
 {
-  "best_metric": 0.5359712230215827,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-57",
-  "epoch": 2.8860759493670884,
+  "best_metric": 0.9100719424460432,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-228",
+  "epoch": 11.544303797468354,
   "eval_steps": 500,
-  "global_step": 57,
+  "global_step": 228,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.51,
-      "grad_norm": 3.057455539703369,
-      "learning_rate": 4.607843137254902e-05,
-      "loss": 1.0905,
+      "grad_norm": 4.082145690917969,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.8026,
       "step": 10
     },
     {
       "epoch": 0.96,
-      "eval_accuracy": 0.43884892086330934,
-      "eval_loss": 1.0363357067108154,
-      "eval_runtime": 7.9543,
-      "eval_samples_per_second": 34.95,
-      "eval_steps_per_second": 1.131,
+      "eval_accuracy": 0.5611510791366906,
+      "eval_loss": 0.9313456416130066,
+      "eval_runtime": 5.2218,
+      "eval_samples_per_second": 53.238,
+      "eval_steps_per_second": 1.724,
       "step": 19
     },
     {
       "epoch": 1.01,
-      "grad_norm": 8.071553230285645,
-      "learning_rate": 3.627450980392157e-05,
-      "loss": 1.0368,
+      "grad_norm": 7.958559513092041,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.8133,
       "step": 20
     },
     {
       "epoch": 1.52,
-      "grad_norm": 3.5114073753356934,
-      "learning_rate": 2.647058823529412e-05,
-      "loss": 0.9897,
+      "grad_norm": 8.814253807067871,
+      "learning_rate": 4.829268292682927e-05,
+      "loss": 0.7571,
       "step": 30
     },
     {
       "epoch": 1.97,
-      "eval_accuracy": 0.5071942446043165,
-      "eval_loss": 0.9945195913314819,
-      "eval_runtime": 5.7967,
-      "eval_samples_per_second": 47.958,
-      "eval_steps_per_second": 1.553,
+      "eval_accuracy": 0.5755395683453237,
+      "eval_loss": 0.8835484981536865,
+      "eval_runtime": 5.8943,
+      "eval_samples_per_second": 47.164,
+      "eval_steps_per_second": 1.527,
       "step": 39
     },
     {
       "epoch": 2.03,
-      "grad_norm": 6.494540214538574,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.957,
+      "grad_norm": 7.337099552154541,
+      "learning_rate": 4.585365853658537e-05,
+      "loss": 0.7116,
       "step": 40
     },
     {
       "epoch": 2.53,
-      "grad_norm": 3.738604784011841,
-      "learning_rate": 6.862745098039216e-06,
-      "loss": 0.9218,
+      "grad_norm": 6.751507759094238,
+      "learning_rate": 4.3414634146341465e-05,
+      "loss": 0.7061,
       "step": 50
     },
     {
-      "epoch": 2.89,
-      "eval_accuracy": 0.5359712230215827,
-      "eval_loss": 0.9638464450836182,
-      "eval_runtime": 5.7057,
-      "eval_samples_per_second": 48.723,
-      "eval_steps_per_second": 1.577,
-      "step": 57
+      "epoch": 2.99,
+      "eval_accuracy": 0.6870503597122302,
+      "eval_loss": 0.7589301466941833,
+      "eval_runtime": 5.3529,
+      "eval_samples_per_second": 51.935,
+      "eval_steps_per_second": 1.681,
+      "step": 59
     },
     {
-      "epoch": 2.89,
-      "step": 57,
-      "total_flos": 1.7986296196553933e+17,
-      "train_loss": 0.9899872813308448,
-      "train_runtime": 160.5331,
-      "train_samples_per_second": 46.682,
-      "train_steps_per_second": 0.355
+      "epoch": 3.04,
+      "grad_norm": 14.623546600341797,
+      "learning_rate": 4.097560975609756e-05,
+      "loss": 0.6659,
+      "step": 60
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 6.873902320861816,
+      "learning_rate": 3.853658536585366e-05,
+      "loss": 0.5911,
+      "step": 70
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7482014388489209,
+      "eval_loss": 0.6329088807106018,
+      "eval_runtime": 5.0758,
+      "eval_samples_per_second": 54.769,
+      "eval_steps_per_second": 1.773,
+      "step": 79
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 7.94564151763916,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.5374,
+      "step": 80
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 9.940902709960938,
+      "learning_rate": 3.365853658536586e-05,
+      "loss": 0.5194,
+      "step": 90
+    },
+    {
+      "epoch": 4.96,
+      "eval_accuracy": 0.7697841726618705,
+      "eval_loss": 0.5634069442749023,
+      "eval_runtime": 5.7171,
+      "eval_samples_per_second": 48.626,
+      "eval_steps_per_second": 1.574,
+      "step": 98
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 9.237541198730469,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.5044,
+      "step": 100
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 7.870697021484375,
+      "learning_rate": 2.8780487804878046e-05,
+      "loss": 0.4471,
+      "step": 110
+    },
+    {
+      "epoch": 5.97,
+      "eval_accuracy": 0.8165467625899281,
+      "eval_loss": 0.45517149567604065,
+      "eval_runtime": 9.6217,
+      "eval_samples_per_second": 28.893,
+      "eval_steps_per_second": 0.935,
+      "step": 118
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 8.180870056152344,
+      "learning_rate": 2.6341463414634148e-05,
+      "loss": 0.4194,
+      "step": 120
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 10.097016334533691,
+      "learning_rate": 2.3902439024390243e-05,
+      "loss": 0.3743,
+      "step": 130
+    },
+    {
+      "epoch": 6.99,
+      "eval_accuracy": 0.8525179856115108,
+      "eval_loss": 0.3760392963886261,
+      "eval_runtime": 5.0392,
+      "eval_samples_per_second": 55.167,
+      "eval_steps_per_second": 1.786,
+      "step": 138
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 14.651969909667969,
+      "learning_rate": 2.146341463414634e-05,
+      "loss": 0.392,
+      "step": 140
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 7.655238628387451,
+      "learning_rate": 1.902439024390244e-05,
+      "loss": 0.3686,
+      "step": 150
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8705035971223022,
+      "eval_loss": 0.3233107626438141,
+      "eval_runtime": 5.8587,
+      "eval_samples_per_second": 47.451,
+      "eval_steps_per_second": 1.536,
+      "step": 158
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 6.040632724761963,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.3411,
+      "step": 160
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 8.26773738861084,
+      "learning_rate": 1.4146341463414633e-05,
+      "loss": 0.318,
+      "step": 170
+    },
+    {
+      "epoch": 8.96,
+      "eval_accuracy": 0.8776978417266187,
+      "eval_loss": 0.3140920102596283,
+      "eval_runtime": 5.5768,
+      "eval_samples_per_second": 49.85,
+      "eval_steps_per_second": 1.614,
+      "step": 177
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 9.864940643310547,
+      "learning_rate": 1.1707317073170733e-05,
+      "loss": 0.3118,
+      "step": 180
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 6.075065612792969,
+      "learning_rate": 9.26829268292683e-06,
+      "loss": 0.3163,
+      "step": 190
+    },
+    {
+      "epoch": 9.97,
+      "eval_accuracy": 0.8992805755395683,
+      "eval_loss": 0.27717697620391846,
+      "eval_runtime": 6.2589,
+      "eval_samples_per_second": 44.417,
+      "eval_steps_per_second": 1.438,
+      "step": 197
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 5.928242206573486,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.2931,
+      "step": 200
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 5.408517837524414,
+      "learning_rate": 4.390243902439024e-06,
+      "loss": 0.2871,
+      "step": 210
+    },
+    {
+      "epoch": 10.99,
+      "eval_accuracy": 0.9028776978417267,
+      "eval_loss": 0.27071893215179443,
+      "eval_runtime": 5.0727,
+      "eval_samples_per_second": 54.803,
+      "eval_steps_per_second": 1.774,
+      "step": 217
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 4.428467273712158,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.2909,
+      "step": 220
+    },
+    {
+      "epoch": 11.54,
+      "eval_accuracy": 0.9100719424460432,
+      "eval_loss": 0.263301283121109,
+      "eval_runtime": 5.2589,
+      "eval_samples_per_second": 52.863,
+      "eval_steps_per_second": 1.711,
+      "step": 228
+    },
+    {
+      "epoch": 11.54,
+      "step": 228,
+      "total_flos": 7.17214746345173e+17,
+      "train_loss": 0.48171327511469525,
+      "train_runtime": 539.5354,
+      "train_samples_per_second": 55.559,
+      "train_steps_per_second": 0.423
     }
   ],
   "logging_steps": 10,
-  "max_steps": 57,
+  "max_steps": 228,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 12,
   "save_steps": 500,
-  "total_flos": 1.7986296196553933e+17,
+  "total_flos": 7.17214746345173e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null