End of training

README.md

@@ -22,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.8082191780821918
+      value: 0.8127853881278538
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/deit-tiny-patch16-224](https://huggingface.co/facebook/deit-tiny-patch16-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6114
-- Accuracy: 0.8082
+- Loss: 0.5154
+- Accuracy: 0.8128
 
 ## Model description
 

all_results.json

@@ -0,0 +1,18 @@
+{
+    "epoch": 19.51,
+    "eval_accuracy": 0.8127853881278538,
+    "eval_loss": 0.5153664350509644,
+    "eval_runtime": 0.8201,
+    "eval_samples_per_second": 267.033,
+    "eval_steps_per_second": 34.141,
+    "test_accuracy": 0.7534246575342466,
+    "test_loss": 0.7015746235847473,
+    "test_runtime": 1.1393,
+    "test_samples_per_second": 192.215,
+    "test_steps_per_second": 24.575,
+    "total_flos": 6.358977448272691e+16,
+    "train_loss": 0.40896380335092547,
+    "train_runtime": 131.9389,
+    "train_samples_per_second": 98.985,
+    "train_steps_per_second": 3.032
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 19.51,
+    "eval_accuracy": 0.8127853881278538,
+    "eval_loss": 0.5153664350509644,
+    "eval_runtime": 0.8201,
+    "eval_samples_per_second": 267.033,
+    "eval_steps_per_second": 34.141
+}

test_results.json

@@ -0,0 +1,7 @@
+{
+    "test_accuracy": 0.7534246575342466,
+    "test_loss": 0.7015746235847473,
+    "test_runtime": 1.1393,
+    "test_samples_per_second": 192.215,
+    "test_steps_per_second": 24.575
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 19.51,
+    "total_flos": 6.358977448272691e+16,
+    "train_loss": 0.40896380335092547,
+    "train_runtime": 131.9389,
+    "train_samples_per_second": 98.985,
+    "train_steps_per_second": 3.032
+}

trainer_state.json

@@ -0,0 +1,448 @@
+{
+  "best_metric": 0.8127853881278538,
+  "best_model_checkpoint": "deit-tiny-patch16-224-finetuned-piid/checkpoint-266",
+  "epoch": 19.51219512195122,
+  "eval_steps": 500,
+  "global_step": 400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.25e-05,
+      "loss": 1.5326,
+      "step": 10
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5e-05,
+      "loss": 1.2298,
+      "step": 20
+    },
+    {
+      "epoch": 0.98,
+      "eval_accuracy": 0.4703196347031963,
+      "eval_loss": 1.1137923002243042,
+      "eval_runtime": 0.7498,
+      "eval_samples_per_second": 292.079,
+      "eval_steps_per_second": 37.343,
+      "step": 20
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.9126,
+      "step": 30
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.7642,
+      "step": 40
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5981735159817352,
+      "eval_loss": 0.9123256206512451,
+      "eval_runtime": 0.7291,
+      "eval_samples_per_second": 300.354,
+      "eval_steps_per_second": 38.401,
+      "step": 41
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 0.6623,
+      "step": 50
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.6167,
+      "step": 60
+    },
+    {
+      "epoch": 2.98,
+      "eval_accuracy": 0.684931506849315,
+      "eval_loss": 0.6735855937004089,
+      "eval_runtime": 1.0942,
+      "eval_samples_per_second": 200.146,
+      "eval_steps_per_second": 25.589,
+      "step": 61
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 0.5708,
+      "step": 70
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5628,
+      "step": 80
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7031963470319634,
+      "eval_loss": 0.6615465879440308,
+      "eval_runtime": 0.7444,
+      "eval_samples_per_second": 294.203,
+      "eval_steps_per_second": 37.615,
+      "step": 82
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.305555555555556e-05,
+      "loss": 0.6059,
+      "step": 90
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.5616,
+      "step": 100
+    },
+    {
+      "epoch": 4.98,
+      "eval_accuracy": 0.7351598173515982,
+      "eval_loss": 0.5984886288642883,
+      "eval_runtime": 0.9744,
+      "eval_samples_per_second": 224.752,
+      "eval_steps_per_second": 28.735,
+      "step": 102
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 4.027777777777778e-05,
+      "loss": 0.5088,
+      "step": 110
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.4742,
+      "step": 120
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.7853881278538812,
+      "eval_loss": 0.4980608820915222,
+      "eval_runtime": 0.7506,
+      "eval_samples_per_second": 291.763,
+      "eval_steps_per_second": 37.303,
+      "step": 123
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4212,
+      "step": 130
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.3434,
+      "step": 140
+    },
+    {
+      "epoch": 6.98,
+      "eval_accuracy": 0.7488584474885844,
+      "eval_loss": 0.5729327201843262,
+      "eval_runtime": 0.7717,
+      "eval_samples_per_second": 283.805,
+      "eval_steps_per_second": 36.286,
+      "step": 143
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 0.3907,
+      "step": 150
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3691,
+      "step": 160
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7397260273972602,
+      "eval_loss": 0.5779256820678711,
+      "eval_runtime": 0.7364,
+      "eval_samples_per_second": 297.399,
+      "eval_steps_per_second": 38.024,
+      "step": 164
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 3.194444444444444e-05,
+      "loss": 0.3043,
+      "step": 170
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.3375,
+      "step": 180
+    },
+    {
+      "epoch": 8.98,
+      "eval_accuracy": 0.7579908675799086,
+      "eval_loss": 0.541674792766571,
+      "eval_runtime": 0.7286,
+      "eval_samples_per_second": 300.584,
+      "eval_steps_per_second": 38.431,
+      "step": 184
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.2907,
+      "step": 190
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.3192,
+      "step": 200
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.7534246575342466,
+      "eval_loss": 0.5554245710372925,
+      "eval_runtime": 0.7702,
+      "eval_samples_per_second": 284.349,
+      "eval_steps_per_second": 36.355,
+      "step": 205
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 2.6388888888888892e-05,
+      "loss": 0.3006,
+      "step": 210
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 2.5e-05,
+      "loss": 0.2795,
+      "step": 220
+    },
+    {
+      "epoch": 10.98,
+      "eval_accuracy": 0.776255707762557,
+      "eval_loss": 0.5656077265739441,
+      "eval_runtime": 0.7297,
+      "eval_samples_per_second": 300.112,
+      "eval_steps_per_second": 38.37,
+      "step": 225
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 2.361111111111111e-05,
+      "loss": 0.2999,
+      "step": 230
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.242,
+      "step": 240
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.7990867579908676,
+      "eval_loss": 0.5318821668624878,
+      "eval_runtime": 0.9729,
+      "eval_samples_per_second": 225.104,
+      "eval_steps_per_second": 28.78,
+      "step": 246
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.2281,
+      "step": 250
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.2557,
+      "step": 260
+    },
+    {
+      "epoch": 12.98,
+      "eval_accuracy": 0.8127853881278538,
+      "eval_loss": 0.5153664350509644,
+      "eval_runtime": 0.7247,
+      "eval_samples_per_second": 302.183,
+      "eval_steps_per_second": 38.635,
+      "step": 266
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 1.8055555555555555e-05,
+      "loss": 0.2134,
+      "step": 270
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2465,
+      "step": 280
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.7990867579908676,
+      "eval_loss": 0.5763050317764282,
+      "eval_runtime": 0.7251,
+      "eval_samples_per_second": 302.014,
+      "eval_steps_per_second": 38.614,
+      "step": 287
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 1.527777777777778e-05,
+      "loss": 0.3016,
+      "step": 290
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.221,
+      "step": 300
+    },
+    {
+      "epoch": 14.98,
+      "eval_accuracy": 0.8036529680365296,
+      "eval_loss": 0.5682752132415771,
+      "eval_runtime": 0.7177,
+      "eval_samples_per_second": 305.151,
+      "eval_steps_per_second": 39.015,
+      "step": 307
+    },
+    {
+      "epoch": 15.12,
+      "learning_rate": 1.25e-05,
+      "loss": 0.2028,
+      "step": 310
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.2058,
+      "step": 320
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.8127853881278538,
+      "eval_loss": 0.5852413177490234,
+      "eval_runtime": 0.7065,
+      "eval_samples_per_second": 309.982,
+      "eval_steps_per_second": 39.632,
+      "step": 328
+    },
+    {
+      "epoch": 16.1,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 0.1959,
+      "step": 330
+    },
+    {
+      "epoch": 16.59,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.1809,
+      "step": 340
+    },
+    {
+      "epoch": 16.98,
+      "eval_accuracy": 0.8082191780821918,
+      "eval_loss": 0.6281502842903137,
+      "eval_runtime": 0.7065,
+      "eval_samples_per_second": 309.973,
+      "eval_steps_per_second": 39.631,
+      "step": 348
+    },
+    {
+      "epoch": 17.07,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.19,
+      "step": 350
+    },
+    {
+      "epoch": 17.56,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.1638,
+      "step": 360
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.7945205479452054,
+      "eval_loss": 0.6289016604423523,
+      "eval_runtime": 0.7099,
+      "eval_samples_per_second": 308.501,
+      "eval_steps_per_second": 39.443,
+      "step": 369
+    },
+    {
+      "epoch": 18.05,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.2779,
+      "step": 370
+    },
+    {
+      "epoch": 18.54,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.155,
+      "step": 380
+    },
+    {
+      "epoch": 18.98,
+      "eval_accuracy": 0.8036529680365296,
+      "eval_loss": 0.613444447517395,
+      "eval_runtime": 0.9887,
+      "eval_samples_per_second": 221.506,
+      "eval_steps_per_second": 28.32,
+      "step": 389
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 1.388888888888889e-06,
+      "loss": 0.2105,
+      "step": 390
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 0.0,
+      "loss": 0.2094,
+      "step": 400
+    },
+    {
+      "epoch": 19.51,
+      "eval_accuracy": 0.8082191780821918,
+      "eval_loss": 0.6113923192024231,
+      "eval_runtime": 0.6901,
+      "eval_samples_per_second": 317.344,
+      "eval_steps_per_second": 40.574,
+      "step": 400
+    },
+    {
+      "epoch": 19.51,
+      "step": 400,
+      "total_flos": 6.358977448272691e+16,
+      "train_loss": 0.40896380335092547,
+      "train_runtime": 131.9389,
+      "train_samples_per_second": 98.985,
+      "train_steps_per_second": 3.032
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 400,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "total_flos": 6.358977448272691e+16,
+  "trial_name": null,
+  "trial_params": null
+}