End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 2.99,
-    "eval_accuracy": 0.9959183673469387,
-    "eval_loss": 0.02241336554288864,
-    "eval_runtime": 11.9967,
-    "eval_samples_per_second": 61.267,
-    "eval_steps_per_second": 7.669,
-    "total_flos": 4.9150232620631654e+17,
-    "train_loss": 0.0970435019064104,
-    "train_runtime": 528.4446,
-    "train_samples_per_second": 37.508,
-    "train_steps_per_second": 1.169
+    "eval_accuracy": 0.7149341806627326,
+    "eval_loss": 1.1311845779418945,
+    "eval_runtime": 48.1534,
+    "eval_samples_per_second": 45.75,
+    "eval_steps_per_second": 5.732,
+    "total_flos": 3.815394309328896e+17,
+    "train_loss": 0.09319638077480098,
+    "train_runtime": 491.2433,
+    "train_samples_per_second": 31.384,
+    "train_steps_per_second": 0.977
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.99,
-    "eval_accuracy": 0.9959183673469387,
-    "eval_loss": 0.02241336554288864,
-    "eval_runtime": 11.9967,
-    "eval_samples_per_second": 61.267,
-    "eval_steps_per_second": 7.669
+    "eval_accuracy": 0.7149341806627326,
+    "eval_loss": 1.1311845779418945,
+    "eval_runtime": 48.1534,
+    "eval_samples_per_second": 45.75,
+    "eval_steps_per_second": 5.732
 }

runs/Nov30_01-45-27_08b01d83a514/events.out.tfevents.1701309282.08b01d83a514.20755.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4aceb9125343f6442e7fd2c42aa8a2551882702524febc6d8a040dd98ee6997
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.99,
-    "total_flos": 4.9150232620631654e+17,
-    "train_loss": 0.0970435019064104,
-    "train_runtime": 528.4446,
-    "train_samples_per_second": 37.508,
-    "train_steps_per_second": 1.169
+    "total_flos": 3.815394309328896e+17,
+    "train_loss": 0.09319638077480098,
+    "train_runtime": 491.2433,
+    "train_samples_per_second": 31.384,
+    "train_steps_per_second": 0.977
 }

trainer_state.json

@@ -1,421 +1,343 @@
 {
-  "best_metric": 0.9959183673469387,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-413",
-  "epoch": 2.9927360774818403,
+  "best_metric": 0.7149341806627326,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-480",
+  "epoch": 2.9860031104199067,
   "eval_steps": 500,
-  "global_step": 618,
+  "global_step": 480,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.05,
-      "learning_rate": 8.064516129032258e-06,
-      "loss": 0.6891,
+      "epoch": 0.06,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.6441,
       "step": 10
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 1.6129032258064517e-05,
-      "loss": 0.6032,
+      "epoch": 0.12,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.5084,
       "step": 20
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 2.4193548387096777e-05,
-      "loss": 0.452,
+      "epoch": 0.19,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3659,
       "step": 30
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 3.2258064516129034e-05,
-      "loss": 0.3304,
+      "epoch": 0.25,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.2162,
       "step": 40
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.032258064516129e-05,
-      "loss": 0.2486,
+      "epoch": 0.31,
+      "learning_rate": 4.976851851851852e-05,
+      "loss": 0.1387,
       "step": 50
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.8387096774193554e-05,
-      "loss": 0.1912,
+      "epoch": 0.37,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 0.1187,
       "step": 60
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9280575539568345e-05,
-      "loss": 0.1454,
+      "epoch": 0.44,
+      "learning_rate": 4.745370370370371e-05,
+      "loss": 0.1003,
       "step": 70
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.838129496402878e-05,
-      "loss": 0.1398,
+      "epoch": 0.5,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.104,
       "step": 80
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.748201438848921e-05,
-      "loss": 0.1292,
+      "epoch": 0.56,
+      "learning_rate": 4.5138888888888894e-05,
+      "loss": 0.1448,
       "step": 90
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.658273381294964e-05,
-      "loss": 0.1324,
+      "epoch": 0.62,
+      "learning_rate": 4.3981481481481486e-05,
+      "loss": 0.0851,
       "step": 100
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.5683453237410076e-05,
-      "loss": 0.077,
+      "epoch": 0.68,
+      "learning_rate": 4.282407407407408e-05,
+      "loss": 0.0664,
       "step": 110
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.47841726618705e-05,
-      "loss": 0.0785,
+      "epoch": 0.75,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.0775,
       "step": 120
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.3884892086330935e-05,
-      "loss": 0.1228,
+      "epoch": 0.81,
+      "learning_rate": 4.0509259259259265e-05,
+      "loss": 0.0996,
       "step": 130
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.298561151079137e-05,
-      "loss": 0.054,
+      "epoch": 0.87,
+      "learning_rate": 3.935185185185186e-05,
+      "loss": 0.0929,
       "step": 140
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.20863309352518e-05,
-      "loss": 0.1124,
+      "epoch": 0.93,
+      "learning_rate": 3.8194444444444444e-05,
+      "loss": 0.0983,
       "step": 150
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.1187050359712234e-05,
-      "loss": 0.0875,
+      "epoch": 1.0,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.1091,
+      "step": 160
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.634135270086246,
+      "eval_loss": 1.134523868560791,
+      "eval_runtime": 48.9231,
+      "eval_samples_per_second": 45.03,
+      "eval_steps_per_second": 5.642,
       "step": 160
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.0287769784172666e-05,
-      "loss": 0.1152,
+      "epoch": 1.06,
+      "learning_rate": 3.587962962962963e-05,
+      "loss": 0.1446,
       "step": 170
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 3.938848920863309e-05,
-      "loss": 0.107,
+      "epoch": 1.12,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 0.0648,
       "step": 180
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 3.8489208633093525e-05,
-      "loss": 0.0887,
+      "epoch": 1.18,
+      "learning_rate": 3.3564814814814815e-05,
+      "loss": 0.0704,
       "step": 190
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 3.758992805755396e-05,
-      "loss": 0.0826,
+      "epoch": 1.24,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 0.1206,
       "step": 200
     },
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.9918367346938776,
-      "eval_loss": 0.03725149855017662,
-      "eval_runtime": 12.9921,
-      "eval_samples_per_second": 56.573,
-      "eval_steps_per_second": 7.081,
-      "step": 206
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 3.669064748201439e-05,
-      "loss": 0.0654,
+      "epoch": 1.31,
+      "learning_rate": 3.125e-05,
+      "loss": 0.0428,
       "step": 210
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 3.5791366906474824e-05,
-      "loss": 0.0904,
+      "epoch": 1.37,
+      "learning_rate": 3.0092592592592593e-05,
+      "loss": 0.0271,
       "step": 220
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 3.489208633093525e-05,
-      "loss": 0.0879,
+      "epoch": 1.43,
+      "learning_rate": 2.8935185185185186e-05,
+      "loss": 0.0786,
       "step": 230
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 3.399280575539568e-05,
-      "loss": 0.0731,
+      "epoch": 1.49,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.0233,
       "step": 240
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 3.3093525179856116e-05,
-      "loss": 0.0715,
+      "epoch": 1.56,
+      "learning_rate": 2.6620370370370372e-05,
+      "loss": 0.01,
       "step": 250
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 3.219424460431655e-05,
-      "loss": 0.0513,
+      "epoch": 1.62,
+      "learning_rate": 2.5462962962962965e-05,
+      "loss": 0.0297,
       "step": 260
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 3.129496402877698e-05,
-      "loss": 0.0458,
+      "epoch": 1.68,
+      "learning_rate": 2.4305555555555558e-05,
+      "loss": 0.0598,
       "step": 270
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 3.039568345323741e-05,
-      "loss": 0.0795,
+      "epoch": 1.74,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 0.0714,
       "step": 280
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 2.9496402877697844e-05,
-      "loss": 0.0743,
+      "epoch": 1.8,
+      "learning_rate": 2.1990740740740743e-05,
+      "loss": 0.1358,
       "step": 290
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 2.8597122302158273e-05,
-      "loss": 0.0943,
+      "epoch": 1.87,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.0164,
       "step": 300
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 2.7697841726618706e-05,
-      "loss": 0.0961,
+      "epoch": 1.93,
+      "learning_rate": 1.967592592592593e-05,
+      "loss": 0.0072,
       "step": 310
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 2.679856115107914e-05,
-      "loss": 0.0387,
+      "epoch": 1.99,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.0426,
       "step": 320
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 2.589928057553957e-05,
-      "loss": 0.0843,
+      "epoch": 2.0,
+      "eval_accuracy": 0.6627326373127553,
+      "eval_loss": 1.377668857574463,
+      "eval_runtime": 47.5961,
+      "eval_samples_per_second": 46.285,
+      "eval_steps_per_second": 5.799,
+      "step": 321
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.736111111111111e-05,
+      "loss": 0.0439,
       "step": 330
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 2.5e-05,
-      "loss": 0.0816,
+      "epoch": 2.12,
+      "learning_rate": 1.6203703703703704e-05,
+      "loss": 0.0253,
       "step": 340
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 2.4100719424460434e-05,
-      "loss": 0.0263,
+      "epoch": 2.18,
+      "learning_rate": 1.5046296296296297e-05,
+      "loss": 0.0197,
       "step": 350
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 2.3201438848920864e-05,
-      "loss": 0.0794,
+      "epoch": 2.24,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.0112,
       "step": 360
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 2.2302158273381296e-05,
-      "loss": 0.0321,
+      "epoch": 2.3,
+      "learning_rate": 1.2731481481481482e-05,
+      "loss": 0.0445,
       "step": 370
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 2.140287769784173e-05,
-      "loss": 0.0087,
+      "epoch": 2.36,
+      "learning_rate": 1.1574074074074075e-05,
+      "loss": 0.0338,
       "step": 380
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 2.050359712230216e-05,
-      "loss": 0.0621,
+      "epoch": 2.43,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.0344,
       "step": 390
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 1.960431654676259e-05,
-      "loss": 0.0501,
+      "epoch": 2.49,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.0226,
       "step": 400
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 1.8705035971223024e-05,
-      "loss": 0.0658,
+      "epoch": 2.55,
+      "learning_rate": 8.101851851851852e-06,
+      "loss": 0.0748,
       "step": 410
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.9959183673469387,
-      "eval_loss": 0.02241336554288864,
-      "eval_runtime": 14.7711,
-      "eval_samples_per_second": 49.759,
-      "eval_steps_per_second": 6.228,
-      "step": 413
-    },
-    {
-      "epoch": 2.03,
-      "learning_rate": 1.7805755395683454e-05,
-      "loss": 0.0063,
+      "epoch": 2.61,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.0099,
       "step": 420
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 1.6906474820143887e-05,
-      "loss": 0.0207,
+      "epoch": 2.67,
+      "learning_rate": 5.787037037037038e-06,
+      "loss": 0.0558,
       "step": 430
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 1.6007194244604316e-05,
-      "loss": 0.0129,
+      "epoch": 2.74,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 0.0243,
       "step": 440
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 1.5107913669064749e-05,
-      "loss": 0.0116,
+      "epoch": 2.8,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.0239,
       "step": 450
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 1.420863309352518e-05,
-      "loss": 0.0049,
+      "epoch": 2.86,
+      "learning_rate": 2.3148148148148148e-06,
+      "loss": 0.0645,
       "step": 460
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 1.3309352517985613e-05,
-      "loss": 0.0311,
+      "epoch": 2.92,
+      "learning_rate": 1.1574074074074074e-06,
+      "loss": 0.0547,
       "step": 470
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 1.2410071942446044e-05,
-      "loss": 0.037,
+      "epoch": 2.99,
+      "learning_rate": 0.0,
+      "loss": 0.0151,
       "step": 480
     },
-    {
-      "epoch": 2.37,
-      "learning_rate": 1.1510791366906475e-05,
-      "loss": 0.0572,
-      "step": 490
-    },
-    {
-      "epoch": 2.42,
-      "learning_rate": 1.0611510791366908e-05,
-      "loss": 0.0038,
-      "step": 500
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 9.71223021582734e-06,
-      "loss": 0.0551,
-      "step": 510
-    },
-    {
-      "epoch": 2.52,
-      "learning_rate": 8.812949640287769e-06,
-      "loss": 0.1015,
-      "step": 520
-    },
-    {
-      "epoch": 2.57,
-      "learning_rate": 7.913669064748202e-06,
-      "loss": 0.0154,
-      "step": 530
-    },
-    {
-      "epoch": 2.62,
-      "learning_rate": 7.014388489208633e-06,
-      "loss": 0.0531,
-      "step": 540
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 6.115107913669065e-06,
-      "loss": 0.0558,
-      "step": 550
-    },
-    {
-      "epoch": 2.71,
-      "learning_rate": 5.215827338129497e-06,
-      "loss": 0.035,
-      "step": 560
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 4.316546762589929e-06,
-      "loss": 0.0335,
-      "step": 570
-    },
-    {
-      "epoch": 2.81,
-      "learning_rate": 3.41726618705036e-06,
-      "loss": 0.0457,
-      "step": 580
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 2.5179856115107916e-06,
-      "loss": 0.016,
-      "step": 590
-    },
-    {
-      "epoch": 2.91,
-      "learning_rate": 1.618705035971223e-06,
-      "loss": 0.0124,
-      "step": 600
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 7.194244604316547e-07,
-      "loss": 0.012,
-      "step": 610
-    },
     {
       "epoch": 2.99,
-      "eval_accuracy": 0.9931972789115646,
-      "eval_loss": 0.027184385806322098,
-      "eval_runtime": 13.1253,
-      "eval_samples_per_second": 55.999,
-      "eval_steps_per_second": 7.009,
-      "step": 618
+      "eval_accuracy": 0.7149341806627326,
+      "eval_loss": 1.1311845779418945,
+      "eval_runtime": 51.1784,
+      "eval_samples_per_second": 43.045,
+      "eval_steps_per_second": 5.393,
+      "step": 480
     },
     {
       "epoch": 2.99,
-      "step": 618,
-      "total_flos": 4.9150232620631654e+17,
-      "train_loss": 0.0970435019064104,
-      "train_runtime": 528.4446,
-      "train_samples_per_second": 37.508,
-      "train_steps_per_second": 1.169
+      "step": 480,
+      "total_flos": 3.815394309328896e+17,
+      "train_loss": 0.09319638077480098,
+      "train_runtime": 491.2433,
+      "train_samples_per_second": 31.384,
+      "train_steps_per_second": 0.977
     }
   ],
   "logging_steps": 10,
-  "max_steps": 618,
+  "max_steps": 480,
   "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 4.9150232620631654e+17,
+  "total_flos": 3.815394309328896e+17,
   "trial_name": null,
   "trial_params": null
 }