End of training

README.md

@@ -18,8 +18,8 @@ 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 an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7838
-- Accuracy: 0.75
+- Loss: 0.7948
+- Accuracy: 0.7841
 
 ## Model description
 

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 33.38095238095238,
+    "eval_accuracy": 0.7840909090909091,
+    "eval_loss": 0.7947837710380554,
+    "eval_runtime": 1.51,
+    "eval_samples_per_second": 58.278,
+    "eval_steps_per_second": 1.987,
+    "total_flos": 7.202076007838515e+17,
+    "train_loss": 0.4182109731435776,
+    "train_runtime": 479.8944,
+    "train_samples_per_second": 55.262,
+    "train_steps_per_second": 0.417
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 33.38095238095238,
+    "eval_accuracy": 0.7840909090909091,
+    "eval_loss": 0.7947837710380554,
+    "eval_runtime": 1.51,
+    "eval_samples_per_second": 58.278,
+    "eval_steps_per_second": 1.987
+}

runs/Feb09_22-53-44_e3629502e642/events.out.tfevents.1739142140.e3629502e642.7979.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ef7e37ac16641cda3db727045943ccd5229fb24cd740e56773a5902b407665a
+size 411

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 33.38095238095238,
+    "total_flos": 7.202076007838515e+17,
+    "train_loss": 0.4182109731435776,
+    "train_runtime": 479.8944,
+    "train_samples_per_second": 55.262,
+    "train_steps_per_second": 0.417
+}

trainer_state.json

@@ -0,0 +1,460 @@
+{
+  "best_metric": 0.7840909090909091,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-humeda-DAV52/checkpoint-120",
+  "epoch": 33.38095238095238,
+  "eval_steps": 500,
+  "global_step": 200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.20454545454545456,
+      "eval_loss": 1.6161998510360718,
+      "eval_runtime": 1.6353,
+      "eval_samples_per_second": 53.813,
+      "eval_steps_per_second": 1.835,
+      "step": 6
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.476341962814331,
+      "learning_rate": 3e-05,
+      "loss": 1.3978,
+      "step": 12
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.18181818181818182,
+      "eval_loss": 1.6140469312667847,
+      "eval_runtime": 1.5147,
+      "eval_samples_per_second": 58.098,
+      "eval_steps_per_second": 1.981,
+      "step": 12
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.4318181818181818,
+      "eval_loss": 1.4028598070144653,
+      "eval_runtime": 1.504,
+      "eval_samples_per_second": 58.51,
+      "eval_steps_per_second": 1.995,
+      "step": 18
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.7945642471313477,
+      "learning_rate": 4.993910125649561e-05,
+      "loss": 1.0539,
+      "step": 24
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5454545454545454,
+      "eval_loss": 1.2503002882003784,
+      "eval_runtime": 1.5054,
+      "eval_samples_per_second": 58.457,
+      "eval_steps_per_second": 1.993,
+      "step": 24
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.625,
+      "eval_loss": 1.001364827156067,
+      "eval_runtime": 1.4946,
+      "eval_samples_per_second": 58.879,
+      "eval_steps_per_second": 2.007,
+      "step": 30
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.5080366134643555,
+      "learning_rate": 4.9031542398457974e-05,
+      "loss": 0.7171,
+      "step": 36
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6363636363636364,
+      "eval_loss": 0.9539236426353455,
+      "eval_runtime": 1.5148,
+      "eval_samples_per_second": 58.092,
+      "eval_steps_per_second": 1.98,
+      "step": 36
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6136363636363636,
+      "eval_loss": 0.9957922101020813,
+      "eval_runtime": 1.5318,
+      "eval_samples_per_second": 57.447,
+      "eval_steps_per_second": 1.958,
+      "step": 42
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.426722764968872,
+      "learning_rate": 4.707368982147318e-05,
+      "loss": 0.5557,
+      "step": 48
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8232616186141968,
+      "eval_runtime": 1.5714,
+      "eval_samples_per_second": 56.0,
+      "eval_steps_per_second": 1.909,
+      "step": 48
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6136363636363636,
+      "eval_loss": 0.8813073635101318,
+      "eval_runtime": 2.2484,
+      "eval_samples_per_second": 39.139,
+      "eval_steps_per_second": 1.334,
+      "step": 54
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 3.06756329536438,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.4942,
+      "step": 60
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.7159090909090909,
+      "eval_loss": 0.8385125994682312,
+      "eval_runtime": 1.9967,
+      "eval_samples_per_second": 44.073,
+      "eval_steps_per_second": 1.503,
+      "step": 60
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7913870215415955,
+      "eval_runtime": 2.2117,
+      "eval_samples_per_second": 39.789,
+      "eval_steps_per_second": 1.356,
+      "step": 66
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 6.963029384613037,
+      "learning_rate": 4.039153688314145e-05,
+      "loss": 0.3957,
+      "step": 72
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.7272727272727273,
+      "eval_loss": 0.7741739153862,
+      "eval_runtime": 2.2523,
+      "eval_samples_per_second": 39.071,
+      "eval_steps_per_second": 1.332,
+      "step": 72
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.7045454545454546,
+      "eval_loss": 0.8121969699859619,
+      "eval_runtime": 2.1918,
+      "eval_samples_per_second": 40.15,
+      "eval_steps_per_second": 1.369,
+      "step": 78
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 5.263124942779541,
+      "learning_rate": 3.5959278669726935e-05,
+      "loss": 0.3664,
+      "step": 84
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7980570197105408,
+      "eval_runtime": 1.7544,
+      "eval_samples_per_second": 50.159,
+      "eval_steps_per_second": 1.71,
+      "step": 84
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.7159090909090909,
+      "eval_loss": 0.7852334976196289,
+      "eval_runtime": 2.0825,
+      "eval_samples_per_second": 42.257,
+      "eval_steps_per_second": 1.441,
+      "step": 90
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 4.2380452156066895,
+      "learning_rate": 3.104804738999169e-05,
+      "loss": 0.3042,
+      "step": 96
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.7159090909090909,
+      "eval_loss": 0.8828800320625305,
+      "eval_runtime": 2.2568,
+      "eval_samples_per_second": 38.994,
+      "eval_steps_per_second": 1.329,
+      "step": 96
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7629781365394592,
+      "eval_runtime": 2.2046,
+      "eval_samples_per_second": 39.916,
+      "eval_steps_per_second": 1.361,
+      "step": 102
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 5.526077747344971,
+      "learning_rate": 2.587248741756253e-05,
+      "loss": 0.2673,
+      "step": 108
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.793631911277771,
+      "eval_runtime": 2.2537,
+      "eval_samples_per_second": 39.047,
+      "eval_steps_per_second": 1.331,
+      "step": 108
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7491422295570374,
+      "eval_runtime": 2.2335,
+      "eval_samples_per_second": 39.4,
+      "eval_steps_per_second": 1.343,
+      "step": 114
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 3.8078646659851074,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.2308,
+      "step": 120
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.7947837710380554,
+      "eval_runtime": 2.1548,
+      "eval_samples_per_second": 40.84,
+      "eval_steps_per_second": 1.392,
+      "step": 120
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.7797514200210571,
+      "eval_runtime": 2.1593,
+      "eval_samples_per_second": 40.754,
+      "eval_steps_per_second": 1.389,
+      "step": 126
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 2.2906363010406494,
+      "learning_rate": 1.56348351646022e-05,
+      "loss": 0.2113,
+      "step": 132
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7635252475738525,
+      "eval_runtime": 2.2092,
+      "eval_samples_per_second": 39.833,
+      "eval_steps_per_second": 1.358,
+      "step": 132
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.7159090909090909,
+      "eval_loss": 0.8520774841308594,
+      "eval_runtime": 2.1989,
+      "eval_samples_per_second": 40.019,
+      "eval_steps_per_second": 1.364,
+      "step": 138
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 5.189899444580078,
+      "learning_rate": 1.1020177413231334e-05,
+      "loss": 0.1852,
+      "step": 144
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.866014838218689,
+      "eval_runtime": 2.17,
+      "eval_samples_per_second": 40.553,
+      "eval_steps_per_second": 1.382,
+      "step": 144
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7984264492988586,
+      "eval_runtime": 2.1065,
+      "eval_samples_per_second": 41.775,
+      "eval_steps_per_second": 1.424,
+      "step": 150
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 2.9657599925994873,
+      "learning_rate": 7.016504991533726e-06,
+      "loss": 0.1765,
+      "step": 156
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7750325798988342,
+      "eval_runtime": 1.8283,
+      "eval_samples_per_second": 48.132,
+      "eval_steps_per_second": 1.641,
+      "step": 156
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7935478687286377,
+      "eval_runtime": 1.7247,
+      "eval_samples_per_second": 51.024,
+      "eval_steps_per_second": 1.739,
+      "step": 162
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 2.2592532634735107,
+      "learning_rate": 3.798797596089351e-06,
+      "loss": 0.1969,
+      "step": 168
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7956154346466064,
+      "eval_runtime": 1.5304,
+      "eval_samples_per_second": 57.5,
+      "eval_steps_per_second": 1.96,
+      "step": 168
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7901678681373596,
+      "eval_runtime": 1.5307,
+      "eval_samples_per_second": 57.49,
+      "eval_steps_per_second": 1.96,
+      "step": 174
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 3.4742448329925537,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 0.1502,
+      "step": 180
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7867553234100342,
+      "eval_runtime": 1.5285,
+      "eval_samples_per_second": 57.571,
+      "eval_steps_per_second": 1.963,
+      "step": 180
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7842350006103516,
+      "eval_runtime": 1.6168,
+      "eval_samples_per_second": 54.429,
+      "eval_steps_per_second": 1.856,
+      "step": 186
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 1.7514631748199463,
+      "learning_rate": 2.4329828146074095e-07,
+      "loss": 0.1621,
+      "step": 192
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7835842370986938,
+      "eval_runtime": 2.342,
+      "eval_samples_per_second": 37.575,
+      "eval_steps_per_second": 1.281,
+      "step": 192
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7837024331092834,
+      "eval_runtime": 3.8794,
+      "eval_samples_per_second": 22.684,
+      "eval_steps_per_second": 0.773,
+      "step": 198
+    },
+    {
+      "epoch": 33.38095238095238,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7837730050086975,
+      "eval_runtime": 2.3047,
+      "eval_samples_per_second": 38.183,
+      "eval_steps_per_second": 1.302,
+      "step": 200
+    },
+    {
+      "epoch": 33.38095238095238,
+      "step": 200,
+      "total_flos": 7.202076007838515e+17,
+      "train_loss": 0.4182109731435776,
+      "train_runtime": 479.8944,
+      "train_samples_per_second": 55.262,
+      "train_steps_per_second": 0.417
+    }
+  ],
+  "logging_steps": 12,
+  "max_steps": 200,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 40,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.202076007838515e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}