End of training

README.md

@@ -2,6 +2,8 @@
 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
+- image-classification
+- vision
 - generated_from_trainer
 metrics:
 - accuracy
@@ -15,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # vit-base-beans
 
-This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
+This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the beans dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.2269
 - Accuracy: 0.9699

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 7.0,
     "eval_accuracy": 0.9699248120300752,
-    "eval_loss": 0.22723796963691711,
-    "eval_runtime": 4.9743,
-    "eval_samples_per_second": 26.737,
-    "eval_steps_per_second": 0.603,
+    "eval_loss": 0.22694912552833557,
+    "eval_runtime": 7.1736,
+    "eval_samples_per_second": 18.54,
+    "eval_steps_per_second": 0.418,
     "total_flos": 5.608920478833746e+17,
-    "train_loss": 0.46667203782987193,
-    "train_runtime": 402.2426,
-    "train_samples_per_second": 17.994,
-    "train_steps_per_second": 0.296
+    "train_loss": 0.46666095236770244,
+    "train_runtime": 510.5808,
+    "train_samples_per_second": 14.176,
+    "train_steps_per_second": 0.233
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 7.0,
     "eval_accuracy": 0.9699248120300752,
-    "eval_loss": 0.22723796963691711,
-    "eval_runtime": 4.9743,
-    "eval_samples_per_second": 26.737,
-    "eval_steps_per_second": 0.603
+    "eval_loss": 0.22694912552833557,
+    "eval_runtime": 7.1736,
+    "eval_samples_per_second": 18.54,
+    "eval_steps_per_second": 0.418
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 7.0,
     "total_flos": 5.608920478833746e+17,
-    "train_loss": 0.46667203782987193,
-    "train_runtime": 402.2426,
-    "train_samples_per_second": 17.994,
-    "train_steps_per_second": 0.296
+    "train_loss": 0.46666095236770244,
+    "train_runtime": 510.5808,
+    "train_samples_per_second": 14.176,
+    "train_steps_per_second": 0.233
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.22723796963691711,
+  "best_metric": 0.22694912552833557,
   "best_model_checkpoint": "./beans_outputs/checkpoint-119",
   "epoch": 7.0,
   "eval_steps": 500,
@@ -10,7 +10,7 @@
   "log_history": [
     {
       "epoch": 0.5882352941176471,
-      "grad_norm": 1.3763134479522705,
+      "grad_norm": 1.3620986938476562,
       "learning_rate": 1.831932773109244e-05,
       "loss": 0.9859,
       "step": 10
@@ -18,144 +18,144 @@
     {
       "epoch": 1.0,
       "eval_accuracy": 0.9323308270676691,
-      "eval_loss": 0.7491587996482849,
-      "eval_runtime": 5.0252,
-      "eval_samples_per_second": 26.467,
-      "eval_steps_per_second": 0.597,
+      "eval_loss": 0.7491994500160217,
+      "eval_runtime": 7.0157,
+      "eval_samples_per_second": 18.958,
+      "eval_steps_per_second": 0.428,
       "step": 17
     },
     {
       "epoch": 1.1764705882352942,
-      "grad_norm": 1.1917294263839722,
+      "grad_norm": 1.1871144771575928,
       "learning_rate": 1.6638655462184877e-05,
-      "loss": 0.8079,
+      "loss": 0.8078,
       "step": 20
     },
     {
       "epoch": 1.7647058823529411,
-      "grad_norm": 1.144236445426941,
+      "grad_norm": 1.1515430212020874,
       "learning_rate": 1.4957983193277313e-05,
-      "loss": 0.6764,
+      "loss": 0.6759,
       "step": 30
     },
     {
       "epoch": 2.0,
       "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.5274316668510437,
-      "eval_runtime": 5.1613,
-      "eval_samples_per_second": 25.769,
-      "eval_steps_per_second": 0.581,
+      "eval_loss": 0.5281645059585571,
+      "eval_runtime": 7.0064,
+      "eval_samples_per_second": 18.983,
+      "eval_steps_per_second": 0.428,
       "step": 34
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 1.2157258987426758,
+      "grad_norm": 1.2202690839767456,
       "learning_rate": 1.3277310924369749e-05,
-      "loss": 0.5491,
+      "loss": 0.5495,
       "step": 40
     },
     {
       "epoch": 2.9411764705882355,
-      "grad_norm": 1.031990885734558,
+      "grad_norm": 1.0186351537704468,
       "learning_rate": 1.1596638655462186e-05,
-      "loss": 0.4601,
+      "loss": 0.461,
       "step": 50
     },
     {
       "epoch": 3.0,
       "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.37144508957862854,
-      "eval_runtime": 5.4329,
-      "eval_samples_per_second": 24.48,
-      "eval_steps_per_second": 0.552,
+      "eval_loss": 0.37340047955513,
+      "eval_runtime": 7.3765,
+      "eval_samples_per_second": 18.03,
+      "eval_steps_per_second": 0.407,
       "step": 51
     },
     {
       "epoch": 3.5294117647058822,
-      "grad_norm": 1.2084527015686035,
+      "grad_norm": 1.1645742654800415,
       "learning_rate": 9.915966386554622e-06,
-      "loss": 0.4047,
+      "loss": 0.4048,
       "step": 60
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.2976459562778473,
-      "eval_runtime": 5.0095,
-      "eval_samples_per_second": 26.549,
-      "eval_steps_per_second": 0.599,
+      "eval_loss": 0.2986946403980255,
+      "eval_runtime": 7.0493,
+      "eval_samples_per_second": 18.867,
+      "eval_steps_per_second": 0.426,
       "step": 68
     },
     {
       "epoch": 4.117647058823529,
-      "grad_norm": 1.0828431844711304,
+      "grad_norm": 1.0221604108810425,
       "learning_rate": 8.23529411764706e-06,
-      "loss": 0.3269,
+      "loss": 0.3271,
       "step": 70
     },
     {
       "epoch": 4.705882352941177,
-      "grad_norm": 1.3986183404922485,
+      "grad_norm": 1.340811014175415,
       "learning_rate": 6.5546218487394966e-06,
-      "loss": 0.3172,
+      "loss": 0.3174,
       "step": 80
     },
     {
       "epoch": 5.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.25337785482406616,
-      "eval_runtime": 5.2266,
-      "eval_samples_per_second": 25.447,
-      "eval_steps_per_second": 0.574,
+      "eval_loss": 0.25426048040390015,
+      "eval_runtime": 7.1157,
+      "eval_samples_per_second": 18.691,
+      "eval_steps_per_second": 0.422,
       "step": 85
     },
     {
       "epoch": 5.294117647058823,
-      "grad_norm": 1.0007528066635132,
+      "grad_norm": 1.0393356084823608,
       "learning_rate": 4.873949579831933e-06,
-      "loss": 0.2775,
+      "loss": 0.2772,
       "step": 90
     },
     {
       "epoch": 5.882352941176471,
-      "grad_norm": 0.8358347415924072,
+      "grad_norm": 0.8360141515731812,
       "learning_rate": 3.1932773109243696e-06,
-      "loss": 0.2545,
+      "loss": 0.2542,
       "step": 100
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.9774436090225563,
-      "eval_loss": 0.22837479412555695,
-      "eval_runtime": 5.7431,
-      "eval_samples_per_second": 23.158,
-      "eval_steps_per_second": 0.522,
+      "eval_loss": 0.22818955779075623,
+      "eval_runtime": 7.1818,
+      "eval_samples_per_second": 18.519,
+      "eval_steps_per_second": 0.418,
       "step": 102
     },
     {
       "epoch": 6.470588235294118,
-      "grad_norm": 1.0932303667068481,
+      "grad_norm": 1.0879504680633545,
       "learning_rate": 1.5126050420168068e-06,
-      "loss": 0.2632,
+      "loss": 0.2627,
       "step": 110
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.22723796963691711,
-      "eval_runtime": 5.0242,
-      "eval_samples_per_second": 26.472,
-      "eval_steps_per_second": 0.597,
+      "eval_loss": 0.22694912552833557,
+      "eval_runtime": 7.2426,
+      "eval_samples_per_second": 18.364,
+      "eval_steps_per_second": 0.414,
       "step": 119
     },
     {
       "epoch": 7.0,
       "step": 119,
       "total_flos": 5.608920478833746e+17,
-      "train_loss": 0.46667203782987193,
-      "train_runtime": 402.2426,
-      "train_samples_per_second": 17.994,
-      "train_steps_per_second": 0.296
+      "train_loss": 0.46666095236770244,
+      "train_runtime": 510.5808,
+      "train_samples_per_second": 14.176,
+      "train_steps_per_second": 0.233
     }
   ],
   "logging_steps": 10,