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: 0.7712
-- Accuracy: 0.7727
+- Loss: 0.7678
+- Accuracy: 0.8068
 
 ## Model description
 

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 41.76190476190476,
+    "eval_accuracy": 0.8068181818181818,
+    "eval_loss": 0.7677886486053467,
+    "eval_runtime": 2.5988,
+    "eval_samples_per_second": 33.861,
+    "eval_steps_per_second": 1.154,
+    "total_flos": 9.011135985411686e+17,
+    "train_loss": 0.39907308292388916,
+    "train_runtime": 596.2283,
+    "train_samples_per_second": 55.6,
+    "train_steps_per_second": 0.419
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 41.76190476190476,
+    "eval_accuracy": 0.8068181818181818,
+    "eval_loss": 0.7677886486053467,
+    "eval_runtime": 2.5988,
+    "eval_samples_per_second": 33.861,
+    "eval_steps_per_second": 1.154
+}

runs/Feb10_01-45-09_e3629502e642/events.out.tfevents.1739152558.e3629502e642.51423.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2dd1a8d021cdb62878f5e0874270b2f96904a53820df5edabdb526dc4af5e6b3
+size 411

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 41.76190476190476,
+    "total_flos": 9.011135985411686e+17,
+    "train_loss": 0.39907308292388916,
+    "train_runtime": 596.2283,
+    "train_samples_per_second": 55.6,
+    "train_steps_per_second": 0.419
+}

trainer_state.json

@@ -0,0 +1,560 @@
+{
+  "best_metric": 0.8068181818181818,
+  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-humeda-DAV55/checkpoint-102",
+  "epoch": 41.76190476190476,
+  "eval_steps": 500,
+  "global_step": 250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.06818181818181818,
+      "eval_loss": 1.715864658355713,
+      "eval_runtime": 1.5297,
+      "eval_samples_per_second": 57.528,
+      "eval_steps_per_second": 1.961,
+      "step": 6
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5547009706497192,
+      "learning_rate": 2.4e-05,
+      "loss": 1.4016,
+      "step": 12
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.125,
+      "eval_loss": 1.6803200244903564,
+      "eval_runtime": 2.2184,
+      "eval_samples_per_second": 39.669,
+      "eval_steps_per_second": 1.352,
+      "step": 12
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.29545454545454547,
+      "eval_loss": 1.5241920948028564,
+      "eval_runtime": 2.1977,
+      "eval_samples_per_second": 40.042,
+      "eval_steps_per_second": 1.365,
+      "step": 18
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.8719537258148193,
+      "learning_rate": 4.8e-05,
+      "loss": 1.105,
+      "step": 24
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.4318181818181818,
+      "eval_loss": 1.2325453758239746,
+      "eval_runtime": 2.3307,
+      "eval_samples_per_second": 37.757,
+      "eval_steps_per_second": 1.287,
+      "step": 24
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5113636363636364,
+      "eval_loss": 1.2531273365020752,
+      "eval_runtime": 1.8518,
+      "eval_samples_per_second": 47.521,
+      "eval_steps_per_second": 1.62,
+      "step": 30
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 5.038722991943359,
+      "learning_rate": 4.9705709536163824e-05,
+      "loss": 0.8077,
+      "step": 36
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5,
+      "eval_loss": 1.1025725603103638,
+      "eval_runtime": 1.5547,
+      "eval_samples_per_second": 56.604,
+      "eval_steps_per_second": 1.93,
+      "step": 36
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5909090909090909,
+      "eval_loss": 0.9851853847503662,
+      "eval_runtime": 1.5272,
+      "eval_samples_per_second": 57.623,
+      "eval_steps_per_second": 1.964,
+      "step": 42
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.3146438598632812,
+      "learning_rate": 4.8721900291112415e-05,
+      "loss": 0.6405,
+      "step": 48
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7045454545454546,
+      "eval_loss": 0.8992521166801453,
+      "eval_runtime": 1.5232,
+      "eval_samples_per_second": 57.773,
+      "eval_steps_per_second": 1.97,
+      "step": 48
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6363636363636364,
+      "eval_loss": 0.9155292510986328,
+      "eval_runtime": 1.5814,
+      "eval_samples_per_second": 55.646,
+      "eval_steps_per_second": 1.897,
+      "step": 54
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 4.260365009307861,
+      "learning_rate": 4.707368982147318e-05,
+      "loss": 0.5807,
+      "step": 60
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.5568181818181818,
+      "eval_loss": 1.044386863708496,
+      "eval_runtime": 1.5257,
+      "eval_samples_per_second": 57.677,
+      "eval_steps_per_second": 1.966,
+      "step": 60
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7695552706718445,
+      "eval_runtime": 1.5142,
+      "eval_samples_per_second": 58.118,
+      "eval_steps_per_second": 1.981,
+      "step": 66
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 8.558947563171387,
+      "learning_rate": 4.480724108387977e-05,
+      "loss": 0.511,
+      "step": 72
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.7159090909090909,
+      "eval_loss": 0.9045647978782654,
+      "eval_runtime": 1.569,
+      "eval_samples_per_second": 56.087,
+      "eval_steps_per_second": 1.912,
+      "step": 72
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7709473967552185,
+      "eval_runtime": 1.5249,
+      "eval_samples_per_second": 57.71,
+      "eval_steps_per_second": 1.967,
+      "step": 78
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 6.5993757247924805,
+      "learning_rate": 4.198603260653792e-05,
+      "loss": 0.4578,
+      "step": 84
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.7272727272727273,
+      "eval_loss": 0.8067413568496704,
+      "eval_runtime": 2.3833,
+      "eval_samples_per_second": 36.923,
+      "eval_steps_per_second": 1.259,
+      "step": 84
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.8393685817718506,
+      "eval_runtime": 2.2836,
+      "eval_samples_per_second": 38.536,
+      "eval_steps_per_second": 1.314,
+      "step": 90
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 8.13113021850586,
+      "learning_rate": 3.868908058731376e-05,
+      "loss": 0.357,
+      "step": 96
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7472593188285828,
+      "eval_runtime": 2.2018,
+      "eval_samples_per_second": 39.967,
+      "eval_steps_per_second": 1.363,
+      "step": 96
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.8068181818181818,
+      "eval_loss": 0.7677886486053467,
+      "eval_runtime": 2.2228,
+      "eval_samples_per_second": 39.591,
+      "eval_steps_per_second": 1.35,
+      "step": 102
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 5.903696060180664,
+      "learning_rate": 3.5008725813922386e-05,
+      "loss": 0.313,
+      "step": 108
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7593179941177368,
+      "eval_runtime": 2.1703,
+      "eval_samples_per_second": 40.547,
+      "eval_steps_per_second": 1.382,
+      "step": 108
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.759705126285553,
+      "eval_runtime": 2.1997,
+      "eval_samples_per_second": 40.006,
+      "eval_steps_per_second": 1.364,
+      "step": 114
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 2.582517147064209,
+      "learning_rate": 3.104804738999169e-05,
+      "loss": 0.2741,
+      "step": 120
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7861429452896118,
+      "eval_runtime": 2.2726,
+      "eval_samples_per_second": 38.723,
+      "eval_steps_per_second": 1.32,
+      "step": 120
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7381357550621033,
+      "eval_runtime": 2.2359,
+      "eval_samples_per_second": 39.358,
+      "eval_steps_per_second": 1.342,
+      "step": 126
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 4.310458660125732,
+      "learning_rate": 2.6917975703170466e-05,
+      "loss": 0.2508,
+      "step": 132
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.7954545454545454,
+      "eval_loss": 0.7421664595603943,
+      "eval_runtime": 2.1769,
+      "eval_samples_per_second": 40.424,
+      "eval_steps_per_second": 1.378,
+      "step": 132
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.7751468420028687,
+      "eval_runtime": 2.2291,
+      "eval_samples_per_second": 39.478,
+      "eval_steps_per_second": 1.346,
+      "step": 138
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 2.611931562423706,
+      "learning_rate": 2.2734185495055503e-05,
+      "loss": 0.1992,
+      "step": 144
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.7386363636363636,
+      "eval_loss": 0.7757924795150757,
+      "eval_runtime": 2.2227,
+      "eval_samples_per_second": 39.592,
+      "eval_steps_per_second": 1.35,
+      "step": 144
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.7272288799285889,
+      "eval_runtime": 2.1971,
+      "eval_samples_per_second": 40.052,
+      "eval_steps_per_second": 1.365,
+      "step": 150
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 1.5732656717300415,
+      "learning_rate": 1.8613856051605243e-05,
+      "loss": 0.1897,
+      "step": 156
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.7842504978179932,
+      "eval_runtime": 2.1412,
+      "eval_samples_per_second": 41.099,
+      "eval_steps_per_second": 1.401,
+      "step": 156
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7606151700019836,
+      "eval_runtime": 2.0615,
+      "eval_samples_per_second": 42.687,
+      "eval_steps_per_second": 1.455,
+      "step": 162
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 3.5426816940307617,
+      "learning_rate": 1.467238925438646e-05,
+      "loss": 0.2024,
+      "step": 168
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.7954545454545454,
+      "eval_loss": 0.7455670833587646,
+      "eval_runtime": 1.8782,
+      "eval_samples_per_second": 46.853,
+      "eval_steps_per_second": 1.597,
+      "step": 168
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.7954545454545454,
+      "eval_loss": 0.7653156518936157,
+      "eval_runtime": 1.985,
+      "eval_samples_per_second": 44.333,
+      "eval_steps_per_second": 1.511,
+      "step": 174
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 3.386720895767212,
+      "learning_rate": 1.1020177413231334e-05,
+      "loss": 0.172,
+      "step": 180
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7677273154258728,
+      "eval_runtime": 2.0548,
+      "eval_samples_per_second": 42.826,
+      "eval_steps_per_second": 1.46,
+      "step": 180
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7420690059661865,
+      "eval_runtime": 2.0796,
+      "eval_samples_per_second": 42.315,
+      "eval_steps_per_second": 1.443,
+      "step": 186
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 0.8537516593933105,
+      "learning_rate": 7.759511406608255e-06,
+      "loss": 0.1561,
+      "step": 192
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7326412796974182,
+      "eval_runtime": 2.0655,
+      "eval_samples_per_second": 42.605,
+      "eval_steps_per_second": 1.452,
+      "step": 192
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.7613636363636364,
+      "eval_loss": 0.7541264891624451,
+      "eval_runtime": 1.9193,
+      "eval_samples_per_second": 45.851,
+      "eval_steps_per_second": 1.563,
+      "step": 198
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 3.0382492542266846,
+      "learning_rate": 4.981715726281666e-06,
+      "loss": 0.1472,
+      "step": 204
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7634887099266052,
+      "eval_runtime": 2.3886,
+      "eval_samples_per_second": 36.842,
+      "eval_steps_per_second": 1.256,
+      "step": 204
+    },
+    {
+      "epoch": 35.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7503542304039001,
+      "eval_runtime": 1.5141,
+      "eval_samples_per_second": 58.122,
+      "eval_steps_per_second": 1.981,
+      "step": 210
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 0.8586391806602478,
+      "learning_rate": 2.764590667717562e-06,
+      "loss": 0.1402,
+      "step": 216
+    },
+    {
+      "epoch": 36.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7600700855255127,
+      "eval_runtime": 1.4959,
+      "eval_samples_per_second": 58.826,
+      "eval_steps_per_second": 2.005,
+      "step": 216
+    },
+    {
+      "epoch": 37.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.7682745456695557,
+      "eval_runtime": 1.5095,
+      "eval_samples_per_second": 58.296,
+      "eval_steps_per_second": 1.987,
+      "step": 222
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 0.7802270650863647,
+      "learning_rate": 1.170233305176327e-06,
+      "loss": 0.1414,
+      "step": 228
+    },
+    {
+      "epoch": 38.0,
+      "eval_accuracy": 0.7840909090909091,
+      "eval_loss": 0.7706876993179321,
+      "eval_runtime": 1.5074,
+      "eval_samples_per_second": 58.378,
+      "eval_steps_per_second": 1.99,
+      "step": 228
+    },
+    {
+      "epoch": 39.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7727323174476624,
+      "eval_runtime": 1.5866,
+      "eval_samples_per_second": 55.463,
+      "eval_steps_per_second": 1.891,
+      "step": 234
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 2.9806997776031494,
+      "learning_rate": 2.4329828146074095e-07,
+      "loss": 0.1344,
+      "step": 240
+    },
+    {
+      "epoch": 40.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7721273303031921,
+      "eval_runtime": 1.5038,
+      "eval_samples_per_second": 58.517,
+      "eval_steps_per_second": 1.995,
+      "step": 240
+    },
+    {
+      "epoch": 41.0,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.7714928984642029,
+      "eval_runtime": 1.5205,
+      "eval_samples_per_second": 57.876,
+      "eval_steps_per_second": 1.973,
+      "step": 246
+    },
+    {
+      "epoch": 41.76190476190476,
+      "eval_accuracy": 0.7727272727272727,
+      "eval_loss": 0.771232008934021,
+      "eval_runtime": 1.8203,
+      "eval_samples_per_second": 48.344,
+      "eval_steps_per_second": 1.648,
+      "step": 250
+    },
+    {
+      "epoch": 41.76190476190476,
+      "step": 250,
+      "total_flos": 9.011135985411686e+17,
+      "train_loss": 0.39907308292388916,
+      "train_runtime": 596.2283,
+      "train_samples_per_second": 55.6,
+      "train_steps_per_second": 0.419
+    }
+  ],
+  "logging_steps": 12,
+  "max_steps": 250,
+  "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": 9.011135985411686e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}