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: 1.2066
+- Loss: 8.3393
 - Accuracy: 0.4565
 
 ## Model description

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 34.29,
+    "eval_accuracy": 0.45652173913043476,
+    "eval_loss": 8.339312553405762,
+    "eval_runtime": 0.9027,
+    "eval_samples_per_second": 50.958,
+    "eval_steps_per_second": 3.323,
+    "train_loss": 2.8776248772939046,
+    "train_runtime": 206.7662,
+    "train_samples_per_second": 41.206,
+    "train_steps_per_second": 0.58
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 34.29,
+    "eval_accuracy": 0.45652173913043476,
+    "eval_loss": 8.339312553405762,
+    "eval_runtime": 0.9027,
+    "eval_samples_per_second": 50.958,
+    "eval_steps_per_second": 3.323
+}

runs/Dec02_15-35-36_DESKTOP-SKBE9FB/events.out.tfevents.1733175887.DESKTOP-SKBE9FB.18648.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:490a4ae5b2ef3b5abb2c5a9b7a392200f661feb2e79e63504cc6ce2665d20163
+size 405

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 34.29,
+    "train_loss": 2.8776248772939046,
+    "train_runtime": 206.7662,
+    "train_samples_per_second": 41.206,
+    "train_steps_per_second": 0.58
+}

trainer_state.json

@@ -0,0 +1,417 @@
+{
+  "best_metric": 0.45652173913043476,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-DMAE-5e-2\\checkpoint-10",
+  "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": 0.32608695652173914,
+      "eval_loss": 22.410741806030273,
+      "eval_runtime": 0.7553,
+      "eval_samples_per_second": 60.901,
+      "eval_steps_per_second": 3.972,
+      "step": 3
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 11.26321792602539,
+      "eval_runtime": 0.7728,
+      "eval_samples_per_second": 59.527,
+      "eval_steps_per_second": 3.882,
+      "step": 7
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.04583333333333334,
+      "loss": 15.4384,
+      "step": 10
+    },
+    {
+      "epoch": 2.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 8.339312553405762,
+      "eval_runtime": 0.7917,
+      "eval_samples_per_second": 58.104,
+      "eval_steps_per_second": 3.789,
+      "step": 10
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.10869565217391304,
+      "eval_loss": 2.280815839767456,
+      "eval_runtime": 0.7452,
+      "eval_samples_per_second": 61.731,
+      "eval_steps_per_second": 4.026,
+      "step": 14
+    },
+    {
+      "epoch": 4.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 2.978447198867798,
+      "eval_runtime": 0.7387,
+      "eval_samples_per_second": 62.273,
+      "eval_steps_per_second": 4.061,
+      "step": 17
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 0.04166666666666667,
+      "loss": 4.9288,
+      "step": 20
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 2.475663900375366,
+      "eval_runtime": 0.7452,
+      "eval_samples_per_second": 61.731,
+      "eval_steps_per_second": 4.026,
+      "step": 21
+    },
+    {
+      "epoch": 6.86,
+      "eval_accuracy": 0.10869565217391304,
+      "eval_loss": 2.5933523178100586,
+      "eval_runtime": 0.7372,
+      "eval_samples_per_second": 62.399,
+      "eval_steps_per_second": 4.07,
+      "step": 24
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 2.3187241554260254,
+      "eval_runtime": 0.7272,
+      "eval_samples_per_second": 63.258,
+      "eval_steps_per_second": 4.126,
+      "step": 28
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 0.037500000000000006,
+      "loss": 2.426,
+      "step": 30
+    },
+    {
+      "epoch": 8.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 1.7270220518112183,
+      "eval_runtime": 0.7982,
+      "eval_samples_per_second": 57.631,
+      "eval_steps_per_second": 3.759,
+      "step": 31
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.4868004322052002,
+      "eval_runtime": 0.7112,
+      "eval_samples_per_second": 64.683,
+      "eval_steps_per_second": 4.218,
+      "step": 35
+    },
+    {
+      "epoch": 10.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 1.3056219816207886,
+      "eval_runtime": 0.7972,
+      "eval_samples_per_second": 57.704,
+      "eval_steps_per_second": 3.763,
+      "step": 38
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 0.03333333333333333,
+      "loss": 1.5653,
+      "step": 40
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2236616611480713,
+      "eval_runtime": 0.7612,
+      "eval_samples_per_second": 60.433,
+      "eval_steps_per_second": 3.941,
+      "step": 42
+    },
+    {
+      "epoch": 12.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.4644097089767456,
+      "eval_runtime": 0.7272,
+      "eval_samples_per_second": 63.26,
+      "eval_steps_per_second": 4.126,
+      "step": 45
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2427252531051636,
+      "eval_runtime": 0.7122,
+      "eval_samples_per_second": 64.592,
+      "eval_steps_per_second": 4.213,
+      "step": 49
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 0.02916666666666667,
+      "loss": 1.4578,
+      "step": 50
+    },
+    {
+      "epoch": 14.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 1.247389554977417,
+      "eval_runtime": 0.7527,
+      "eval_samples_per_second": 61.115,
+      "eval_steps_per_second": 3.986,
+      "step": 52
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.3440945148468018,
+      "eval_runtime": 0.7337,
+      "eval_samples_per_second": 62.698,
+      "eval_steps_per_second": 4.089,
+      "step": 56
+    },
+    {
+      "epoch": 16.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 1.3480498790740967,
+      "eval_runtime": 0.8027,
+      "eval_samples_per_second": 57.306,
+      "eval_steps_per_second": 3.737,
+      "step": 59
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 0.025,
+      "loss": 1.3492,
+      "step": 60
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2835181951522827,
+      "eval_runtime": 0.7937,
+      "eval_samples_per_second": 57.957,
+      "eval_steps_per_second": 3.78,
+      "step": 63
+    },
+    {
+      "epoch": 18.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.3153202533721924,
+      "eval_runtime": 0.7487,
+      "eval_samples_per_second": 61.442,
+      "eval_steps_per_second": 4.007,
+      "step": 66
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 0.020833333333333336,
+      "loss": 1.3089,
+      "step": 70
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2387202978134155,
+      "eval_runtime": 0.8447,
+      "eval_samples_per_second": 54.457,
+      "eval_steps_per_second": 3.552,
+      "step": 70
+    },
+    {
+      "epoch": 20.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 1.2538148164749146,
+      "eval_runtime": 0.7427,
+      "eval_samples_per_second": 61.938,
+      "eval_steps_per_second": 4.039,
+      "step": 73
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2386040687561035,
+      "eval_runtime": 0.8212,
+      "eval_samples_per_second": 56.017,
+      "eval_steps_per_second": 3.653,
+      "step": 77
+    },
+    {
+      "epoch": 22.86,
+      "learning_rate": 0.016666666666666666,
+      "loss": 1.2345,
+      "step": 80
+    },
+    {
+      "epoch": 22.86,
+      "eval_accuracy": 0.32608695652173914,
+      "eval_loss": 1.261245608329773,
+      "eval_runtime": 0.8717,
+      "eval_samples_per_second": 52.77,
+      "eval_steps_per_second": 3.442,
+      "step": 80
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2124842405319214,
+      "eval_runtime": 0.8932,
+      "eval_samples_per_second": 51.5,
+      "eval_steps_per_second": 3.359,
+      "step": 84
+    },
+    {
+      "epoch": 24.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2078908681869507,
+      "eval_runtime": 0.7602,
+      "eval_samples_per_second": 60.513,
+      "eval_steps_per_second": 3.946,
+      "step": 87
+    },
+    {
+      "epoch": 25.71,
+      "learning_rate": 0.0125,
+      "loss": 1.2105,
+      "step": 90
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2138991355895996,
+      "eval_runtime": 0.7567,
+      "eval_samples_per_second": 60.792,
+      "eval_steps_per_second": 3.965,
+      "step": 91
+    },
+    {
+      "epoch": 26.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2171320915222168,
+      "eval_runtime": 0.8692,
+      "eval_samples_per_second": 52.921,
+      "eval_steps_per_second": 3.451,
+      "step": 94
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.218336820602417,
+      "eval_runtime": 0.8997,
+      "eval_samples_per_second": 51.128,
+      "eval_steps_per_second": 3.334,
+      "step": 98
+    },
+    {
+      "epoch": 28.57,
+      "learning_rate": 0.008333333333333333,
+      "loss": 1.2095,
+      "step": 100
+    },
+    {
+      "epoch": 28.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2093594074249268,
+      "eval_runtime": 0.7922,
+      "eval_samples_per_second": 58.066,
+      "eval_steps_per_second": 3.787,
+      "step": 101
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.206130862236023,
+      "eval_runtime": 0.8217,
+      "eval_samples_per_second": 55.981,
+      "eval_steps_per_second": 3.651,
+      "step": 105
+    },
+    {
+      "epoch": 30.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2068984508514404,
+      "eval_runtime": 0.7457,
+      "eval_samples_per_second": 61.688,
+      "eval_steps_per_second": 4.023,
+      "step": 108
+    },
+    {
+      "epoch": 31.43,
+      "learning_rate": 0.004166666666666667,
+      "loss": 1.203,
+      "step": 110
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2073019742965698,
+      "eval_runtime": 0.7547,
+      "eval_samples_per_second": 60.953,
+      "eval_steps_per_second": 3.975,
+      "step": 112
+    },
+    {
+      "epoch": 32.86,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.207156777381897,
+      "eval_runtime": 0.8097,
+      "eval_samples_per_second": 56.811,
+      "eval_steps_per_second": 3.705,
+      "step": 115
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2066789865493774,
+      "eval_runtime": 0.7932,
+      "eval_samples_per_second": 57.993,
+      "eval_steps_per_second": 3.782,
+      "step": 119
+    },
+    {
+      "epoch": 34.29,
+      "learning_rate": 0.0,
+      "loss": 1.1996,
+      "step": 120
+    },
+    {
+      "epoch": 34.29,
+      "eval_accuracy": 0.45652173913043476,
+      "eval_loss": 1.2066394090652466,
+      "eval_runtime": 0.7892,
+      "eval_samples_per_second": 58.289,
+      "eval_steps_per_second": 3.801,
+      "step": 120
+    },
+    {
+      "epoch": 34.29,
+      "step": 120,
+      "total_flos": 2.3770905934823424e+17,
+      "train_loss": 2.8776248772939046,
+      "train_runtime": 206.7662,
+      "train_samples_per_second": 41.206,
+      "train_steps_per_second": 0.58
+    }
+  ],
+  "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
+}