End of training

README.md

@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8719
+- Loss: 1.0462
 - Accuracy: 0.6522
 
 ## Model description

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 34.29,
+    "eval_accuracy": 0.6521739130434783,
+    "eval_loss": 1.0461729764938354,
+    "eval_runtime": 1.0708,
+    "eval_samples_per_second": 42.958,
+    "eval_steps_per_second": 2.802,
+    "train_loss": 1.0580236832300822,
+    "train_runtime": 204.3045,
+    "train_samples_per_second": 41.702,
+    "train_steps_per_second": 0.587
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 34.29,
+    "eval_accuracy": 0.6521739130434783,
+    "eval_loss": 1.0461729764938354,
+    "eval_runtime": 1.0708,
+    "eval_samples_per_second": 42.958,
+    "eval_steps_per_second": 2.802
+}

runs/Dec03_15-25-20_DESKTOP-SKBE9FB/events.out.tfevents.1733261368.DESKTOP-SKBE9FB.13724.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84118ba598ad2e7095123416113615358c0a4aba17246a67802cb9f9f7420791
+size 405

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 34.29,
+    "train_loss": 1.0580236832300822,
+    "train_runtime": 204.3045,
+    "train_samples_per_second": 41.702,
+    "train_steps_per_second": 0.587
+}

trainer_state.json

@@ -0,0 +1,417 @@
+{
+  "best_metric": 0.6521739130434783,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-DMAE-U\\checkpoint-63",
+  "epoch": 34.285714285714285,
+  "eval_steps": 500,
+  "global_step": 120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.86,
+      "eval_accuracy": 0.4782608695652174,
+      "eval_loss": 1.3654131889343262,
+      "eval_runtime": 0.9573,
+      "eval_samples_per_second": 48.051,
+      "eval_steps_per_second": 3.134,
+      "step": 3
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.3027186393737793,
+      "eval_runtime": 0.9204,
+      "eval_samples_per_second": 49.976,
+      "eval_steps_per_second": 3.259,
+      "step": 7
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 1.3356,
+      "step": 10
+    },
+    {
+      "epoch": 2.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2579652070999146,
+      "eval_runtime": 0.9757,
+      "eval_samples_per_second": 47.144,
+      "eval_steps_per_second": 3.075,
+      "step": 10
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2156718969345093,
+      "eval_runtime": 0.7647,
+      "eval_samples_per_second": 60.155,
+      "eval_steps_per_second": 3.923,
+      "step": 14
+    },
+    {
+      "epoch": 4.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2120847702026367,
+      "eval_runtime": 0.7481,
+      "eval_samples_per_second": 61.489,
+      "eval_steps_per_second": 4.01,
+      "step": 17
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.202,
+      "step": 20
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2014096975326538,
+      "eval_runtime": 0.7414,
+      "eval_samples_per_second": 62.046,
+      "eval_steps_per_second": 4.047,
+      "step": 21
+    },
+    {
+      "epoch": 6.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2013375759124756,
+      "eval_runtime": 0.9176,
+      "eval_samples_per_second": 50.131,
+      "eval_steps_per_second": 3.269,
+      "step": 24
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.1949347257614136,
+      "eval_runtime": 0.782,
+      "eval_samples_per_second": 58.821,
+      "eval_steps_per_second": 3.836,
+      "step": 28
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.1884,
+      "step": 30
+    },
+    {
+      "epoch": 8.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.1933584213256836,
+      "eval_runtime": 0.8684,
+      "eval_samples_per_second": 52.973,
+      "eval_steps_per_second": 3.455,
+      "step": 31
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.1916431188583374,
+      "eval_runtime": 0.8147,
+      "eval_samples_per_second": 56.461,
+      "eval_steps_per_second": 3.682,
+      "step": 35
+    },
+    {
+      "epoch": 10.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.1829259395599365,
+      "eval_runtime": 0.7563,
+      "eval_samples_per_second": 60.82,
+      "eval_steps_per_second": 3.967,
+      "step": 38
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.1351,
+      "step": 40
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.1567723751068115,
+      "eval_runtime": 0.7667,
+      "eval_samples_per_second": 59.997,
+      "eval_steps_per_second": 3.913,
+      "step": 42
+    },
+    {
+      "epoch": 12.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.1370779275894165,
+      "eval_runtime": 0.7567,
+      "eval_samples_per_second": 60.789,
+      "eval_steps_per_second": 3.965,
+      "step": 45
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.4782608695652174,
+      "eval_loss": 1.1238458156585693,
+      "eval_runtime": 0.7742,
+      "eval_samples_per_second": 59.415,
+      "eval_steps_per_second": 3.875,
+      "step": 49
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 1.132,
+      "step": 50
+    },
+    {
+      "epoch": 14.86,
+      "eval_accuracy": 0.5217391304347826,
+      "eval_loss": 1.118269681930542,
+      "eval_runtime": 0.8372,
+      "eval_samples_per_second": 54.946,
+      "eval_steps_per_second": 3.583,
+      "step": 52
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.6086956521739131,
+      "eval_loss": 1.0961614847183228,
+      "eval_runtime": 0.7592,
+      "eval_samples_per_second": 60.592,
+      "eval_steps_per_second": 3.952,
+      "step": 56
+    },
+    {
+      "epoch": 16.86,
+      "eval_accuracy": 0.6086956521739131,
+      "eval_loss": 1.0737055540084839,
+      "eval_runtime": 0.7442,
+      "eval_samples_per_second": 61.814,
+      "eval_steps_per_second": 4.031,
+      "step": 59
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 2e-05,
+      "loss": 1.0659,
+      "step": 60
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 1.0461729764938354,
+      "eval_runtime": 0.7988,
+      "eval_samples_per_second": 57.588,
+      "eval_steps_per_second": 3.756,
+      "step": 63
+    },
+    {
+      "epoch": 18.86,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 1.0217252969741821,
+      "eval_runtime": 0.7604,
+      "eval_samples_per_second": 60.495,
+      "eval_steps_per_second": 3.945,
+      "step": 66
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.0299,
+      "step": 70
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.9955050349235535,
+      "eval_runtime": 0.8634,
+      "eval_samples_per_second": 53.281,
+      "eval_steps_per_second": 3.475,
+      "step": 70
+    },
+    {
+      "epoch": 20.86,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.9766696691513062,
+      "eval_runtime": 0.7632,
+      "eval_samples_per_second": 60.272,
+      "eval_steps_per_second": 3.931,
+      "step": 73
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.9494532942771912,
+      "eval_runtime": 0.7922,
+      "eval_samples_per_second": 58.064,
+      "eval_steps_per_second": 3.787,
+      "step": 77
+    },
+    {
+      "epoch": 22.86,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.9684,
+      "step": 80
+    },
+    {
+      "epoch": 22.86,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.9327569007873535,
+      "eval_runtime": 0.7722,
+      "eval_samples_per_second": 59.572,
+      "eval_steps_per_second": 3.885,
+      "step": 80
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.917550265789032,
+      "eval_runtime": 0.897,
+      "eval_samples_per_second": 51.283,
+      "eval_steps_per_second": 3.345,
+      "step": 84
+    },
+    {
+      "epoch": 24.86,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.9077643752098083,
+      "eval_runtime": 0.7482,
+      "eval_samples_per_second": 61.483,
+      "eval_steps_per_second": 4.01,
+      "step": 87
+    },
+    {
+      "epoch": 25.71,
+      "learning_rate": 1e-05,
+      "loss": 0.9301,
+      "step": 90
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.896602988243103,
+      "eval_runtime": 0.7612,
+      "eval_samples_per_second": 60.433,
+      "eval_steps_per_second": 3.941,
+      "step": 91
+    },
+    {
+      "epoch": 26.86,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.8951462507247925,
+      "eval_runtime": 0.7502,
+      "eval_samples_per_second": 61.315,
+      "eval_steps_per_second": 3.999,
+      "step": 94
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.8893879652023315,
+      "eval_runtime": 0.7989,
+      "eval_samples_per_second": 57.58,
+      "eval_steps_per_second": 3.755,
+      "step": 98
+    },
+    {
+      "epoch": 28.57,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.9258,
+      "step": 100
+    },
+    {
+      "epoch": 28.86,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.8819690346717834,
+      "eval_runtime": 0.7512,
+      "eval_samples_per_second": 61.237,
+      "eval_steps_per_second": 3.994,
+      "step": 101
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.6304347826086957,
+      "eval_loss": 0.8770626783370972,
+      "eval_runtime": 0.7942,
+      "eval_samples_per_second": 57.916,
+      "eval_steps_per_second": 3.777,
+      "step": 105
+    },
+    {
+      "epoch": 30.86,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.8775607943534851,
+      "eval_runtime": 0.7553,
+      "eval_samples_per_second": 60.904,
+      "eval_steps_per_second": 3.972,
+      "step": 108
+    },
+    {
+      "epoch": 31.43,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.8877,
+      "step": 110
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.8753806352615356,
+      "eval_runtime": 0.7237,
+      "eval_samples_per_second": 63.563,
+      "eval_steps_per_second": 4.145,
+      "step": 112
+    },
+    {
+      "epoch": 32.86,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.8731884360313416,
+      "eval_runtime": 0.7117,
+      "eval_samples_per_second": 64.63,
+      "eval_steps_per_second": 4.215,
+      "step": 115
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.8720559477806091,
+      "eval_runtime": 0.838,
+      "eval_samples_per_second": 54.891,
+      "eval_steps_per_second": 3.58,
+      "step": 119
+    },
+    {
+      "epoch": 34.29,
+      "learning_rate": 0.0,
+      "loss": 0.8953,
+      "step": 120
+    },
+    {
+      "epoch": 34.29,
+      "eval_accuracy": 0.6521739130434783,
+      "eval_loss": 0.871898889541626,
+      "eval_runtime": 0.7853,
+      "eval_samples_per_second": 58.579,
+      "eval_steps_per_second": 3.82,
+      "step": 120
+    },
+    {
+      "epoch": 34.29,
+      "step": 120,
+      "total_flos": 2.3770905934823424e+17,
+      "train_loss": 1.0580236832300822,
+      "train_runtime": 204.3045,
+      "train_samples_per_second": 41.702,
+      "train_steps_per_second": 0.587
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 120,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 40,
+  "save_steps": 500,
+  "total_flos": 2.3770905934823424e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}