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: 0.8094
-- Accuracy: 0.7386
+- Loss: 0.7753
+- Accuracy: 0.75
 
 ## Model description
 

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 33.38095238095238,
+    "eval_accuracy": 0.75,
+    "eval_loss": 0.7753319144248962,
+    "eval_runtime": 1.5195,
+    "eval_samples_per_second": 57.914,
+    "eval_steps_per_second": 1.974,
+    "total_flos": 7.202076007838515e+17,
+    "train_loss": 0.5288270854949951,
+    "train_runtime": 627.2045,
+    "train_samples_per_second": 52.854,
+    "train_steps_per_second": 0.159
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 33.38095238095238,
+    "eval_accuracy": 0.75,
+    "eval_loss": 0.7753319144248962,
+    "eval_runtime": 1.5195,
+    "eval_samples_per_second": 57.914,
+    "eval_steps_per_second": 1.974
+}

runs/Feb10_00-33-52_e3629502e642/events.out.tfevents.1739148302.e3629502e642.30696.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d06d9efbfd8042d769faf6cbb900c57a9ad28059f56029046deb61d8fc17123
+size 405

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 33.38095238095238,
+    "total_flos": 7.202076007838515e+17,
+    "train_loss": 0.5288270854949951,
+    "train_runtime": 627.2045,
+    "train_samples_per_second": 52.854,
+    "train_steps_per_second": 0.159
+}

trainer_state.json

@@ -0,0 +1,418 @@
+{
+  "best_metric": 0.75,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-humeda-DAV54/checkpoint-60",
+  "epoch": 33.38095238095238,
+  "eval_steps": 500,
+  "global_step": 100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.19318181818181818,
+      "eval_loss": 1.585249423980713,
+      "eval_runtime": 1.4912,
+      "eval_samples_per_second": 59.014,
+      "eval_steps_per_second": 2.012,
+      "step": 3
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.3181818181818182,
+      "eval_loss": 1.578413486480713,
+      "eval_runtime": 1.5431,
+      "eval_samples_per_second": 57.028,
+      "eval_steps_per_second": 1.944,
+      "step": 6
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.4318181818181818,
+      "eval_loss": 1.537353515625,
+      "eval_runtime": 1.5969,
+      "eval_samples_per_second": 55.107,
+      "eval_steps_per_second": 1.879,
+      "step": 9
+    },
+    {
+      "epoch": 3.380952380952381,
+      "grad_norm": 3.083775520324707,
+      "learning_rate": 5e-05,
+      "loss": 1.3768,
+      "step": 10
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.4090909090909091,
+      "eval_loss": 1.4629433155059814,
+      "eval_runtime": 1.5346,
+      "eval_samples_per_second": 57.345,
+      "eval_steps_per_second": 1.955,
+      "step": 12
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5340909090909091,
+      "eval_loss": 1.2221930027008057,
+      "eval_runtime": 2.5597,
+      "eval_samples_per_second": 34.379,
+      "eval_steps_per_second": 1.172,
+      "step": 15
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5454545454545454,
+      "eval_loss": 1.2436859607696533,
+      "eval_runtime": 1.5399,
+      "eval_samples_per_second": 57.146,
+      "eval_steps_per_second": 1.948,
+      "step": 18
+    },
+    {
+      "epoch": 6.761904761904762,
+      "grad_norm": 4.605942249298096,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.942,
+      "step": 20
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5340909090909091,
+      "eval_loss": 1.242801308631897,
+      "eval_runtime": 1.5797,
+      "eval_samples_per_second": 55.707,
+      "eval_steps_per_second": 1.899,
+      "step": 21
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5340909090909091,
+      "eval_loss": 1.1750913858413696,
+      "eval_runtime": 2.5156,
+      "eval_samples_per_second": 34.982,
+      "eval_steps_per_second": 1.193,
+      "step": 24
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5795454545454546,
+      "eval_loss": 1.1278696060180664,
+      "eval_runtime": 2.3024,
+      "eval_samples_per_second": 38.222,
+      "eval_steps_per_second": 1.303,
+      "step": 27
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 6.71976900100708,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.6265,
+      "step": 30
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6477272727272727,
+      "eval_loss": 0.9868423342704773,
+      "eval_runtime": 2.2207,
+      "eval_samples_per_second": 39.627,
+      "eval_steps_per_second": 1.351,
+      "step": 30
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.6363636363636364,
+      "eval_loss": 0.9661000967025757,
+      "eval_runtime": 2.178,
+      "eval_samples_per_second": 40.404,
+      "eval_steps_per_second": 1.377,
+      "step": 33
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.6136363636363636,
+      "eval_loss": 0.9892013669013977,
+      "eval_runtime": 2.2016,
+      "eval_samples_per_second": 39.97,
+      "eval_steps_per_second": 1.363,
+      "step": 36
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.6818181818181818,
+      "eval_loss": 0.87156081199646,
+      "eval_runtime": 1.6274,
+      "eval_samples_per_second": 54.073,
+      "eval_steps_per_second": 1.843,
+      "step": 39
+    },
+    {
+      "epoch": 13.380952380952381,
+      "grad_norm": 6.5638837814331055,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.5106,
+      "step": 40
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.7272727272727273,
+      "eval_loss": 0.8009552359580994,
+      "eval_runtime": 1.5443,
+      "eval_samples_per_second": 56.984,
+      "eval_steps_per_second": 1.943,
+      "step": 42
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.6818181818181818,
+      "eval_loss": 0.8596124053001404,
+      "eval_runtime": 1.5414,
+      "eval_samples_per_second": 57.091,
+      "eval_steps_per_second": 1.946,
+      "step": 45
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.6931818181818182,
+      "eval_loss": 0.8257151246070862,
+      "eval_runtime": 2.2212,
+      "eval_samples_per_second": 39.619,
+      "eval_steps_per_second": 1.351,
+      "step": 48
+    },
+    {
+      "epoch": 16.761904761904763,
+      "grad_norm": 5.321930885314941,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.4183,
+      "step": 50
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.7045454545454546,
+      "eval_loss": 0.8189888000488281,
+      "eval_runtime": 1.5383,
+      "eval_samples_per_second": 57.207,
+      "eval_steps_per_second": 1.95,
+      "step": 51
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.7272727272727273,
+      "eval_loss": 0.7627521753311157,
+      "eval_runtime": 2.2402,
+      "eval_samples_per_second": 39.283,
+      "eval_steps_per_second": 1.339,
+      "step": 54
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.7159090909090909,
+      "eval_loss": 0.7801702618598938,
+      "eval_runtime": 2.2768,
+      "eval_samples_per_second": 38.651,
+      "eval_steps_per_second": 1.318,
+      "step": 57
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 3.8304524421691895,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.3267,
+      "step": 60
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7753319144248962,
+      "eval_runtime": 2.3302,
+      "eval_samples_per_second": 37.766,
+      "eval_steps_per_second": 1.287,
+      "step": 60
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7770728468894958,
+      "eval_runtime": 2.2672,
+      "eval_samples_per_second": 38.815,
+      "eval_steps_per_second": 1.323,
+      "step": 63
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7770181894302368,
+      "eval_runtime": 2.3268,
+      "eval_samples_per_second": 37.821,
+      "eval_steps_per_second": 1.289,
+      "step": 66
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.7272727272727273,
+      "eval_loss": 0.7940635681152344,
+      "eval_runtime": 2.2104,
+      "eval_samples_per_second": 39.812,
+      "eval_steps_per_second": 1.357,
+      "step": 69
+    },
+    {
+      "epoch": 23.38095238095238,
+      "grad_norm": 5.505289077758789,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.3008,
+      "step": 70
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.7272727272727273,
+      "eval_loss": 0.7921419739723206,
+      "eval_runtime": 2.1726,
+      "eval_samples_per_second": 40.505,
+      "eval_steps_per_second": 1.381,
+      "step": 72
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7899277210235596,
+      "eval_runtime": 2.2223,
+      "eval_samples_per_second": 39.599,
+      "eval_steps_per_second": 1.35,
+      "step": 75
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7849180698394775,
+      "eval_runtime": 2.1648,
+      "eval_samples_per_second": 40.65,
+      "eval_steps_per_second": 1.386,
+      "step": 78
+    },
+    {
+      "epoch": 26.761904761904763,
+      "grad_norm": 5.414725303649902,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.2795,
+      "step": 80
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7891153693199158,
+      "eval_runtime": 2.2542,
+      "eval_samples_per_second": 39.038,
+      "eval_steps_per_second": 1.331,
+      "step": 81
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7973006963729858,
+      "eval_runtime": 2.205,
+      "eval_samples_per_second": 39.909,
+      "eval_steps_per_second": 1.361,
+      "step": 84
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8067693710327148,
+      "eval_runtime": 2.1816,
+      "eval_samples_per_second": 40.338,
+      "eval_steps_per_second": 1.375,
+      "step": 87
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 4.411905765533447,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 0.2526,
+      "step": 90
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8087642788887024,
+      "eval_runtime": 1.5036,
+      "eval_samples_per_second": 58.526,
+      "eval_steps_per_second": 1.995,
+      "step": 90
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8097642660140991,
+      "eval_runtime": 1.8153,
+      "eval_samples_per_second": 48.476,
+      "eval_steps_per_second": 1.653,
+      "step": 93
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8095622658729553,
+      "eval_runtime": 1.82,
+      "eval_samples_per_second": 48.352,
+      "eval_steps_per_second": 1.648,
+      "step": 96
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8095269799232483,
+      "eval_runtime": 1.7559,
+      "eval_samples_per_second": 50.118,
+      "eval_steps_per_second": 1.709,
+      "step": 99
+    },
+    {
+      "epoch": 33.38095238095238,
+      "grad_norm": 5.03811502456665,
+      "learning_rate": 0.0,
+      "loss": 0.2544,
+      "step": 100
+    },
+    {
+      "epoch": 33.38095238095238,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.8093982338905334,
+      "eval_runtime": 1.7511,
+      "eval_samples_per_second": 50.253,
+      "eval_steps_per_second": 1.713,
+      "step": 100
+    },
+    {
+      "epoch": 33.38095238095238,
+      "step": 100,
+      "total_flos": 7.202076007838515e+17,
+      "train_loss": 0.5288270854949951,
+      "train_runtime": 627.2045,
+      "train_samples_per_second": 52.854,
+      "train_steps_per_second": 0.159
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "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
+}