End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 4.95,
-    "eval_accuracy": 0.9738372093023255,
-    "eval_loss": 0.10162452608346939,
-    "eval_runtime": 6.5939,
-    "eval_samples_per_second": 156.509,
-    "eval_steps_per_second": 5.005,
-    "total_flos": 1.1422818298339983e+18,
-    "train_loss": 0.08449313590923944,
-    "train_runtime": 710.4711,
-    "train_samples_per_second": 65.316,
-    "train_steps_per_second": 0.507
+    "epoch": 4.99,
+    "eval_accuracy": 0.9796511627906976,
+    "eval_loss": 0.07256749272346497,
+    "eval_runtime": 7.0579,
+    "eval_samples_per_second": 146.218,
+    "eval_steps_per_second": 4.676,
+    "total_flos": 1.151826529604567e+18,
+    "train_loss": 0.13382491601838006,
+    "train_runtime": 783.9355,
+    "train_samples_per_second": 59.195,
+    "train_steps_per_second": 0.459
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.95,
-    "eval_accuracy": 0.9738372093023255,
-    "eval_loss": 0.10162452608346939,
-    "eval_runtime": 6.5939,
-    "eval_samples_per_second": 156.509,
-    "eval_steps_per_second": 5.005
+    "epoch": 4.99,
+    "eval_accuracy": 0.9796511627906976,
+    "eval_loss": 0.07256749272346497,
+    "eval_runtime": 7.0579,
+    "eval_samples_per_second": 146.218,
+    "eval_steps_per_second": 4.676
 }

runs/Dec18_11-49-22_61f348adfad0/events.out.tfevents.1702901007.61f348adfad0.741.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48270b8e93dc15081da5a1919720ac014262971b9532e657996bc8b891714e2b
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 4.95,
-    "total_flos": 1.1422818298339983e+18,
-    "train_loss": 0.08449313590923944,
-    "train_runtime": 710.4711,
-    "train_samples_per_second": 65.316,
-    "train_steps_per_second": 0.507
+    "epoch": 4.99,
+    "total_flos": 1.151826529604567e+18,
+    "train_loss": 0.13382491601838006,
+    "train_runtime": 783.9355,
+    "train_samples_per_second": 59.195,
+    "train_steps_per_second": 0.459
 }

trainer_state.json

@@ -1,8 +1,7 @@
 {
-  "best_metric": 0.9738372093023255,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-blank_img/checkpoint-360",
-  "epoch": 4.948453608247423,
-  "eval_steps": 500,
+  "best_metric": 0.9796511627906976,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-blank_img/checkpoint-288",
+  "epoch": 4.989690721649485,
   "global_step": 360,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
@@ -11,495 +10,493 @@
     {
       "epoch": 0.07,
       "learning_rate": 6.944444444444445e-06,
-      "loss": 0.0881,
+      "loss": 0.7256,
       "step": 5
     },
     {
       "epoch": 0.14,
       "learning_rate": 1.388888888888889e-05,
-      "loss": 0.0624,
+      "loss": 0.6125,
       "step": 10
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.063,
+      "loss": 0.4648,
       "step": 15
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 0.0901,
+      "loss": 0.3041,
       "step": 20
     },
     {
       "epoch": 0.34,
       "learning_rate": 3.472222222222222e-05,
-      "loss": 0.0872,
+      "loss": 0.2065,
       "step": 25
     },
     {
       "epoch": 0.41,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.1016,
+      "loss": 0.1469,
       "step": 30
     },
     {
       "epoch": 0.48,
       "learning_rate": 4.8611111111111115e-05,
-      "loss": 0.0905,
+      "loss": 0.1734,
       "step": 35
     },
     {
       "epoch": 0.55,
       "learning_rate": 4.938271604938271e-05,
-      "loss": 0.1495,
+      "loss": 0.1736,
       "step": 40
     },
     {
       "epoch": 0.62,
       "learning_rate": 4.8611111111111115e-05,
-      "loss": 0.0949,
+      "loss": 0.1559,
       "step": 45
     },
     {
       "epoch": 0.69,
       "learning_rate": 4.783950617283951e-05,
-      "loss": 0.1178,
+      "loss": 0.1355,
       "step": 50
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.70679012345679e-05,
-      "loss": 0.08,
+      "loss": 0.0896,
       "step": 55
     },
     {
       "epoch": 0.82,
       "learning_rate": 4.62962962962963e-05,
-      "loss": 0.0919,
+      "loss": 0.0994,
       "step": 60
     },
     {
       "epoch": 0.89,
       "learning_rate": 4.5524691358024696e-05,
-      "loss": 0.0773,
+      "loss": 0.127,
       "step": 65
     },
     {
       "epoch": 0.96,
       "learning_rate": 4.4753086419753084e-05,
-      "loss": 0.0502,
+      "loss": 0.1329,
       "step": 70
     },
     {
       "epoch": 0.99,
-      "eval_accuracy": 0.9651162790697675,
-      "eval_loss": 0.12999847531318665,
-      "eval_runtime": 6.6038,
-      "eval_samples_per_second": 156.273,
-      "eval_steps_per_second": 4.997,
+      "eval_accuracy": 0.9767441860465116,
+      "eval_loss": 0.08817464113235474,
+      "eval_runtime": 7.3528,
+      "eval_samples_per_second": 140.354,
+      "eval_steps_per_second": 4.488,
       "step": 72
     },
     {
-      "epoch": 1.03,
+      "epoch": 1.04,
       "learning_rate": 4.3981481481481486e-05,
-      "loss": 0.0838,
+      "loss": 0.1512,
       "step": 75
     },
     {
-      "epoch": 1.1,
+      "epoch": 1.11,
       "learning_rate": 4.3209876543209875e-05,
-      "loss": 0.067,
+      "loss": 0.1248,
       "step": 80
     },
     {
-      "epoch": 1.17,
+      "epoch": 1.18,
       "learning_rate": 4.243827160493827e-05,
-      "loss": 0.0932,
+      "loss": 0.1396,
       "step": 85
     },
     {
-      "epoch": 1.24,
+      "epoch": 1.25,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.1037,
+      "loss": 0.1231,
       "step": 90
     },
     {
-      "epoch": 1.31,
+      "epoch": 1.32,
       "learning_rate": 4.089506172839506e-05,
-      "loss": 0.1057,
+      "loss": 0.1228,
       "step": 95
     },
     {
-      "epoch": 1.37,
+      "epoch": 1.38,
       "learning_rate": 4.012345679012346e-05,
-      "loss": 0.0774,
+      "loss": 0.1277,
       "step": 100
     },
     {
-      "epoch": 1.44,
+      "epoch": 1.45,
       "learning_rate": 3.935185185185186e-05,
-      "loss": 0.1107,
+      "loss": 0.1116,
       "step": 105
     },
     {
-      "epoch": 1.51,
+      "epoch": 1.52,
       "learning_rate": 3.8580246913580246e-05,
-      "loss": 0.1109,
+      "loss": 0.1236,
       "step": 110
     },
     {
-      "epoch": 1.58,
+      "epoch": 1.59,
       "learning_rate": 3.780864197530865e-05,
-      "loss": 0.1078,
+      "loss": 0.1254,
       "step": 115
     },
     {
-      "epoch": 1.65,
+      "epoch": 1.66,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.0935,
+      "loss": 0.1004,
       "step": 120
     },
     {
-      "epoch": 1.72,
+      "epoch": 1.73,
       "learning_rate": 3.626543209876543e-05,
-      "loss": 0.0741,
+      "loss": 0.1199,
       "step": 125
     },
     {
-      "epoch": 1.79,
+      "epoch": 1.8,
       "learning_rate": 3.5493827160493834e-05,
-      "loss": 0.1031,
+      "loss": 0.1098,
       "step": 130
     },
     {
-      "epoch": 1.86,
+      "epoch": 1.87,
       "learning_rate": 3.472222222222222e-05,
-      "loss": 0.0638,
+      "loss": 0.0787,
       "step": 135
     },
     {
-      "epoch": 1.92,
+      "epoch": 1.93,
       "learning_rate": 3.395061728395062e-05,
-      "loss": 0.0915,
+      "loss": 0.1247,
       "step": 140
     },
     {
       "epoch": 1.99,
-      "learning_rate": 3.317901234567901e-05,
-      "loss": 0.1107,
-      "step": 145
+      "eval_accuracy": 0.9767441860465116,
+      "eval_loss": 0.08047417551279068,
+      "eval_runtime": 7.6368,
+      "eval_samples_per_second": 135.135,
+      "eval_steps_per_second": 4.321,
+      "step": 144
     },
     {
-      "epoch": 1.99,
-      "eval_accuracy": 0.9728682170542635,
-      "eval_loss": 0.10234300047159195,
-      "eval_runtime": 7.16,
-      "eval_samples_per_second": 144.135,
-      "eval_steps_per_second": 4.609,
+      "epoch": 2.01,
+      "learning_rate": 3.317901234567901e-05,
+      "loss": 0.1116,
       "step": 145
     },
     {
-      "epoch": 2.06,
+      "epoch": 2.08,
       "learning_rate": 3.240740740740741e-05,
-      "loss": 0.0784,
+      "loss": 0.122,
       "step": 150
     },
     {
-      "epoch": 2.13,
+      "epoch": 2.15,
       "learning_rate": 3.16358024691358e-05,
-      "loss": 0.0672,
+      "loss": 0.1109,
       "step": 155
     },
     {
-      "epoch": 2.2,
+      "epoch": 2.22,
       "learning_rate": 3.08641975308642e-05,
-      "loss": 0.0598,
+      "loss": 0.109,
       "step": 160
     },
     {
-      "epoch": 2.27,
+      "epoch": 2.29,
       "learning_rate": 3.0092592592592593e-05,
-      "loss": 0.0641,
+      "loss": 0.104,
       "step": 165
     },
     {
-      "epoch": 2.34,
+      "epoch": 2.36,
       "learning_rate": 2.9320987654320992e-05,
-      "loss": 0.0805,
+      "loss": 0.0849,
       "step": 170
     },
     {
-      "epoch": 2.41,
+      "epoch": 2.43,
       "learning_rate": 2.8549382716049384e-05,
-      "loss": 0.0864,
+      "loss": 0.1088,
       "step": 175
     },
     {
-      "epoch": 2.47,
+      "epoch": 2.49,
       "learning_rate": 2.777777777777778e-05,
-      "loss": 0.0954,
+      "loss": 0.0988,
       "step": 180
     },
     {
-      "epoch": 2.54,
+      "epoch": 2.56,
       "learning_rate": 2.700617283950617e-05,
-      "loss": 0.0717,
+      "loss": 0.1072,
       "step": 185
     },
     {
-      "epoch": 2.61,
+      "epoch": 2.63,
       "learning_rate": 2.623456790123457e-05,
-      "loss": 0.0634,
+      "loss": 0.1098,
       "step": 190
     },
     {
-      "epoch": 2.68,
+      "epoch": 2.7,
       "learning_rate": 2.5462962962962965e-05,
-      "loss": 0.1082,
+      "loss": 0.1288,
       "step": 195
     },
     {
-      "epoch": 2.75,
+      "epoch": 2.77,
       "learning_rate": 2.4691358024691357e-05,
-      "loss": 0.0748,
+      "loss": 0.1401,
       "step": 200
     },
     {
-      "epoch": 2.82,
+      "epoch": 2.84,
       "learning_rate": 2.3919753086419755e-05,
-      "loss": 0.1221,
+      "loss": 0.1237,
       "step": 205
     },
     {
-      "epoch": 2.89,
+      "epoch": 2.91,
       "learning_rate": 2.314814814814815e-05,
-      "loss": 0.0757,
+      "loss": 0.1063,
       "step": 210
     },
     {
-      "epoch": 2.96,
+      "epoch": 2.98,
       "learning_rate": 2.2376543209876542e-05,
-      "loss": 0.0917,
+      "loss": 0.0742,
       "step": 215
     },
     {
-      "epoch": 3.0,
-      "eval_accuracy": 0.9651162790697675,
-      "eval_loss": 0.12774048745632172,
-      "eval_runtime": 6.6558,
-      "eval_samples_per_second": 155.053,
-      "eval_steps_per_second": 4.958,
-      "step": 218
+      "epoch": 2.99,
+      "eval_accuracy": 0.9767441860465116,
+      "eval_loss": 0.07210300117731094,
+      "eval_runtime": 7.3216,
+      "eval_samples_per_second": 140.953,
+      "eval_steps_per_second": 4.507,
+      "step": 216
     },
     {
-      "epoch": 3.02,
+      "epoch": 3.05,
       "learning_rate": 2.1604938271604937e-05,
-      "loss": 0.0553,
+      "loss": 0.1008,
       "step": 220
     },
     {
-      "epoch": 3.09,
+      "epoch": 3.12,
       "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.0942,
+      "loss": 0.1196,
       "step": 225
     },
     {
-      "epoch": 3.16,
+      "epoch": 3.19,
       "learning_rate": 2.006172839506173e-05,
-      "loss": 0.0834,
+      "loss": 0.0712,
       "step": 230
     },
     {
-      "epoch": 3.23,
+      "epoch": 3.26,
       "learning_rate": 1.9290123456790123e-05,
-      "loss": 0.057,
+      "loss": 0.0726,
       "step": 235
     },
     {
-      "epoch": 3.3,
+      "epoch": 3.33,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.0773,
+      "loss": 0.0964,
       "step": 240
     },
     {
-      "epoch": 3.37,
+      "epoch": 3.4,
       "learning_rate": 1.7746913580246917e-05,
-      "loss": 0.0617,
+      "loss": 0.0932,
       "step": 245
     },
     {
-      "epoch": 3.44,
+      "epoch": 3.47,
       "learning_rate": 1.697530864197531e-05,
-      "loss": 0.0785,
+      "loss": 0.1022,
       "step": 250
     },
     {
-      "epoch": 3.51,
+      "epoch": 3.54,
       "learning_rate": 1.6203703703703704e-05,
-      "loss": 0.057,
+      "loss": 0.1033,
       "step": 255
     },
     {
-      "epoch": 3.57,
+      "epoch": 3.6,
       "learning_rate": 1.54320987654321e-05,
-      "loss": 0.0746,
+      "loss": 0.0913,
       "step": 260
     },
     {
-      "epoch": 3.64,
+      "epoch": 3.67,
       "learning_rate": 1.4660493827160496e-05,
-      "loss": 0.1117,
+      "loss": 0.0873,
       "step": 265
     },
     {
-      "epoch": 3.71,
+      "epoch": 3.74,
       "learning_rate": 1.388888888888889e-05,
-      "loss": 0.0923,
+      "loss": 0.1095,
       "step": 270
     },
     {
-      "epoch": 3.78,
+      "epoch": 3.81,
       "learning_rate": 1.3117283950617285e-05,
-      "loss": 0.0599,
+      "loss": 0.1047,
       "step": 275
     },
     {
-      "epoch": 3.85,
+      "epoch": 3.88,
       "learning_rate": 1.2345679012345678e-05,
-      "loss": 0.0895,
+      "loss": 0.0707,
       "step": 280
     },
     {
-      "epoch": 3.92,
+      "epoch": 3.95,
       "learning_rate": 1.1574074074074075e-05,
-      "loss": 0.0933,
+      "loss": 0.0745,
       "step": 285
     },
     {
       "epoch": 3.99,
-      "learning_rate": 1.0802469135802469e-05,
-      "loss": 0.1022,
-      "step": 290
+      "eval_accuracy": 0.9796511627906976,
+      "eval_loss": 0.07256749272346497,
+      "eval_runtime": 6.7881,
+      "eval_samples_per_second": 152.03,
+      "eval_steps_per_second": 4.861,
+      "step": 288
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.9718992248062015,
-      "eval_loss": 0.12579788267612457,
-      "eval_runtime": 7.0589,
-      "eval_samples_per_second": 146.199,
-      "eval_steps_per_second": 4.675,
-      "step": 291
+      "epoch": 4.03,
+      "learning_rate": 1.0802469135802469e-05,
+      "loss": 0.129,
+      "step": 290
     },
     {
-      "epoch": 4.05,
+      "epoch": 4.1,
       "learning_rate": 1.0030864197530866e-05,
-      "loss": 0.0889,
+      "loss": 0.1298,
       "step": 295
     },
     {
-      "epoch": 4.12,
+      "epoch": 4.16,
       "learning_rate": 9.259259259259259e-06,
-      "loss": 0.088,
+      "loss": 0.089,
       "step": 300
     },
     {
-      "epoch": 4.19,
+      "epoch": 4.23,
       "learning_rate": 8.487654320987654e-06,
-      "loss": 0.0679,
+      "loss": 0.0907,
       "step": 305
     },
     {
-      "epoch": 4.26,
+      "epoch": 4.3,
       "learning_rate": 7.71604938271605e-06,
-      "loss": 0.0632,
+      "loss": 0.0983,
       "step": 310
     },
     {
-      "epoch": 4.33,
+      "epoch": 4.37,
       "learning_rate": 6.944444444444445e-06,
-      "loss": 0.095,
+      "loss": 0.0769,
       "step": 315
     },
     {
-      "epoch": 4.4,
+      "epoch": 4.44,
       "learning_rate": 6.172839506172839e-06,
-      "loss": 0.084,
+      "loss": 0.1115,
       "step": 320
     },
     {
-      "epoch": 4.47,
+      "epoch": 4.51,
       "learning_rate": 5.401234567901234e-06,
-      "loss": 0.07,
+      "loss": 0.058,
       "step": 325
     },
     {
-      "epoch": 4.54,
+      "epoch": 4.58,
       "learning_rate": 4.6296296296296296e-06,
-      "loss": 0.0692,
+      "loss": 0.0972,
       "step": 330
     },
     {
-      "epoch": 4.6,
+      "epoch": 4.65,
       "learning_rate": 3.858024691358025e-06,
-      "loss": 0.0575,
+      "loss": 0.1036,
       "step": 335
     },
     {
-      "epoch": 4.67,
+      "epoch": 4.71,
       "learning_rate": 3.0864197530864196e-06,
-      "loss": 0.0625,
+      "loss": 0.0675,
       "step": 340
     },
     {
-      "epoch": 4.74,
+      "epoch": 4.78,
       "learning_rate": 2.3148148148148148e-06,
-      "loss": 0.0715,
+      "loss": 0.0909,
       "step": 345
     },
     {
-      "epoch": 4.81,
+      "epoch": 4.85,
       "learning_rate": 1.5432098765432098e-06,
-      "loss": 0.0651,
+      "loss": 0.1148,
       "step": 350
     },
     {
-      "epoch": 4.88,
+      "epoch": 4.92,
       "learning_rate": 7.716049382716049e-07,
-      "loss": 0.1432,
+      "loss": 0.0802,
       "step": 355
     },
     {
-      "epoch": 4.95,
+      "epoch": 4.99,
       "learning_rate": 0.0,
-      "loss": 0.0888,
+      "loss": 0.1289,
       "step": 360
     },
     {
-      "epoch": 4.95,
-      "eval_accuracy": 0.9738372093023255,
-      "eval_loss": 0.10162452608346939,
-      "eval_runtime": 6.9642,
-      "eval_samples_per_second": 148.186,
-      "eval_steps_per_second": 4.739,
+      "epoch": 4.99,
+      "eval_accuracy": 0.9728682170542635,
+      "eval_loss": 0.08483530580997467,
+      "eval_runtime": 6.8918,
+      "eval_samples_per_second": 149.743,
+      "eval_steps_per_second": 4.788,
       "step": 360
     },
     {
-      "epoch": 4.95,
+      "epoch": 4.99,
       "step": 360,
-      "total_flos": 1.1422818298339983e+18,
-      "train_loss": 0.08449313590923944,
-      "train_runtime": 710.4711,
-      "train_samples_per_second": 65.316,
-      "train_steps_per_second": 0.507
+      "total_flos": 1.151826529604567e+18,
+      "train_loss": 0.13382491601838006,
+      "train_runtime": 783.9355,
+      "train_samples_per_second": 59.195,
+      "train_steps_per_second": 0.459
     }
   ],
-  "logging_steps": 5,
   "max_steps": 360,
   "num_train_epochs": 5,
-  "save_steps": 500,
-  "total_flos": 1.1422818298339983e+18,
+  "total_flos": 1.151826529604567e+18,
   "trial_name": null,
   "trial_params": null
 }