End of training

README.md

@@ -18,7 +18,7 @@ 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 an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.2547
+- Loss: 1.2286
 - Accuracy: 0.5192
 
 ## Model description

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 30.0,
+    "eval_accuracy": 0.5192307692307693,
+    "eval_loss": 1.2286193370819092,
+    "eval_runtime": 0.9797,
+    "eval_samples_per_second": 53.078,
+    "eval_steps_per_second": 2.041,
+    "total_flos": 7.115852825454182e+17,
+    "train_loss": 3.1241997612847223,
+    "train_runtime": 594.7221,
+    "train_samples_per_second": 36.773,
+    "train_steps_per_second": 0.151
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 30.0,
+    "eval_accuracy": 0.5192307692307693,
+    "eval_loss": 1.2286193370819092,
+    "eval_runtime": 0.9797,
+    "eval_samples_per_second": 53.078,
+    "eval_steps_per_second": 2.041
+}

runs/Jan20_18-12-26_2dbac0f7b711/events.out.tfevents.1737397395.2dbac0f7b711.1282.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0836714fb25cbbf1c347312f26e530d48a5240d6b19d82199a73312e0c0ecaa
+size 405

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 30.0,
+    "total_flos": 7.115852825454182e+17,
+    "train_loss": 3.1241997612847223,
+    "train_runtime": 594.7221,
+    "train_samples_per_second": 36.773,
+    "train_steps_per_second": 0.151
+}

trainer_state.json

@@ -0,0 +1,375 @@
+{
+  "best_metric": 0.5192307692307693,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-humeda-DAV9/checkpoint-63",
+  "epoch": 30.0,
+  "eval_steps": 500,
+  "global_step": 90,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.23076923076923078,
+      "eval_loss": 1.5944099426269531,
+      "eval_runtime": 0.986,
+      "eval_samples_per_second": 52.74,
+      "eval_steps_per_second": 2.028,
+      "step": 3
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.2692307692307692,
+      "eval_loss": 1.5511040687561035,
+      "eval_runtime": 0.9229,
+      "eval_samples_per_second": 56.344,
+      "eval_steps_per_second": 2.167,
+      "step": 6
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.34615384615384615,
+      "eval_loss": 1.491486668586731,
+      "eval_runtime": 1.2511,
+      "eval_samples_per_second": 41.564,
+      "eval_steps_per_second": 1.599,
+      "step": 9
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 9.355170249938965,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 6.2974,
+      "step": 10
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.46153846153846156,
+      "eval_loss": 1.4388242959976196,
+      "eval_runtime": 0.9203,
+      "eval_samples_per_second": 56.504,
+      "eval_steps_per_second": 2.173,
+      "step": 12
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.46153846153846156,
+      "eval_loss": 1.3927408456802368,
+      "eval_runtime": 0.9499,
+      "eval_samples_per_second": 54.741,
+      "eval_steps_per_second": 2.105,
+      "step": 15
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.4423076923076923,
+      "eval_loss": 1.3393510580062866,
+      "eval_runtime": 0.9523,
+      "eval_samples_per_second": 54.605,
+      "eval_steps_per_second": 2.1,
+      "step": 18
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 14.36314582824707,
+      "learning_rate": 2.9942920471376185e-05,
+      "loss": 5.3611,
+      "step": 20
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.4423076923076923,
+      "eval_loss": 1.3107666969299316,
+      "eval_runtime": 0.9443,
+      "eval_samples_per_second": 55.069,
+      "eval_steps_per_second": 2.118,
+      "step": 21
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.34615384615384615,
+      "eval_loss": 1.3680166006088257,
+      "eval_runtime": 1.2909,
+      "eval_samples_per_second": 40.281,
+      "eval_steps_per_second": 1.549,
+      "step": 24
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.40384615384615385,
+      "eval_loss": 1.2717851400375366,
+      "eval_runtime": 0.9305,
+      "eval_samples_per_second": 55.883,
+      "eval_steps_per_second": 2.149,
+      "step": 27
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 21.04581642150879,
+      "learning_rate": 2.7990381056766583e-05,
+      "loss": 3.7205,
+      "step": 30
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.4230769230769231,
+      "eval_loss": 1.2678520679473877,
+      "eval_runtime": 0.9355,
+      "eval_samples_per_second": 55.588,
+      "eval_steps_per_second": 2.138,
+      "step": 30
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.40384615384615385,
+      "eval_loss": 1.3009899854660034,
+      "eval_runtime": 0.9418,
+      "eval_samples_per_second": 55.212,
+      "eval_steps_per_second": 2.124,
+      "step": 33
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.4230769230769231,
+      "eval_loss": 1.2598135471343994,
+      "eval_runtime": 0.9337,
+      "eval_samples_per_second": 55.694,
+      "eval_steps_per_second": 2.142,
+      "step": 36
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.4230769230769231,
+      "eval_loss": 1.2016476392745972,
+      "eval_runtime": 0.9331,
+      "eval_samples_per_second": 55.726,
+      "eval_steps_per_second": 2.143,
+      "step": 39
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 22.220523834228516,
+      "learning_rate": 2.360364654526569e-05,
+      "loss": 2.8178,
+      "step": 40
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.4423076923076923,
+      "eval_loss": 1.1934146881103516,
+      "eval_runtime": 0.938,
+      "eval_samples_per_second": 55.435,
+      "eval_steps_per_second": 2.132,
+      "step": 42
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.4807692307692308,
+      "eval_loss": 1.1842211484909058,
+      "eval_runtime": 1.3187,
+      "eval_samples_per_second": 39.433,
+      "eval_steps_per_second": 1.517,
+      "step": 45
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.15390944480896,
+      "eval_runtime": 0.9603,
+      "eval_samples_per_second": 54.152,
+      "eval_steps_per_second": 2.083,
+      "step": 48
+    },
+    {
+      "epoch": 16.666666666666668,
+      "grad_norm": 24.42645263671875,
+      "learning_rate": 1.760472266500396e-05,
+      "loss": 2.4001,
+      "step": 50
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.4807692307692308,
+      "eval_loss": 1.1308224201202393,
+      "eval_runtime": 0.9273,
+      "eval_samples_per_second": 56.077,
+      "eval_steps_per_second": 2.157,
+      "step": 51
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.46153846153846156,
+      "eval_loss": 1.217308521270752,
+      "eval_runtime": 0.926,
+      "eval_samples_per_second": 56.154,
+      "eval_steps_per_second": 2.16,
+      "step": 54
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.1669617891311646,
+      "eval_runtime": 0.9426,
+      "eval_samples_per_second": 55.165,
+      "eval_steps_per_second": 2.122,
+      "step": 57
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 23.83008575439453,
+      "learning_rate": 1.1117714323462188e-05,
+      "loss": 2.081,
+      "step": 60
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.1792218685150146,
+      "eval_runtime": 0.9338,
+      "eval_samples_per_second": 55.688,
+      "eval_steps_per_second": 2.142,
+      "step": 60
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.5192307692307693,
+      "eval_loss": 1.2286193370819092,
+      "eval_runtime": 0.9342,
+      "eval_samples_per_second": 55.66,
+      "eval_steps_per_second": 2.141,
+      "step": 63
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.26326322555542,
+      "eval_runtime": 1.1302,
+      "eval_samples_per_second": 46.008,
+      "eval_steps_per_second": 1.77,
+      "step": 66
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.2379720211029053,
+      "eval_runtime": 0.9486,
+      "eval_samples_per_second": 54.817,
+      "eval_steps_per_second": 2.108,
+      "step": 69
+    },
+    {
+      "epoch": 23.333333333333332,
+      "grad_norm": 18.731117248535156,
+      "learning_rate": 5.3581858547019095e-06,
+      "loss": 1.8588,
+      "step": 70
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.4807692307692308,
+      "eval_loss": 1.24978768825531,
+      "eval_runtime": 1.2983,
+      "eval_samples_per_second": 40.053,
+      "eval_steps_per_second": 1.54,
+      "step": 72
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.2591068744659424,
+      "eval_runtime": 0.9309,
+      "eval_samples_per_second": 55.857,
+      "eval_steps_per_second": 2.148,
+      "step": 75
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.2652957439422607,
+      "eval_runtime": 0.9419,
+      "eval_samples_per_second": 55.21,
+      "eval_steps_per_second": 2.123,
+      "step": 78
+    },
+    {
+      "epoch": 26.666666666666668,
+      "grad_norm": 19.568700790405273,
+      "learning_rate": 1.4053831944502509e-06,
+      "loss": 1.7634,
+      "step": 80
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.259857177734375,
+      "eval_runtime": 0.9424,
+      "eval_samples_per_second": 55.178,
+      "eval_steps_per_second": 2.122,
+      "step": 81
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.2548739910125732,
+      "eval_runtime": 0.9289,
+      "eval_samples_per_second": 55.978,
+      "eval_steps_per_second": 2.153,
+      "step": 84
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.5192307692307693,
+      "eval_loss": 1.2544516324996948,
+      "eval_runtime": 1.0558,
+      "eval_samples_per_second": 49.25,
+      "eval_steps_per_second": 1.894,
+      "step": 87
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 25.50164222717285,
+      "learning_rate": 0.0,
+      "loss": 1.8177,
+      "step": 90
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.5192307692307693,
+      "eval_loss": 1.2546559572219849,
+      "eval_runtime": 1.2806,
+      "eval_samples_per_second": 40.606,
+      "eval_steps_per_second": 1.562,
+      "step": 90
+    },
+    {
+      "epoch": 30.0,
+      "step": 90,
+      "total_flos": 7.115852825454182e+17,
+      "train_loss": 3.1241997612847223,
+      "train_runtime": 594.7221,
+      "train_samples_per_second": 36.773,
+      "train_steps_per_second": 0.151
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 90,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 30,
+  "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": 7.115852825454182e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}