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: 0.2272
 - Accuracy: 0.9699

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 7.0,
     "eval_accuracy": 0.9699248120300752,
-    "eval_loss": 0.22653304040431976,
-    "eval_runtime": 5.9117,
-    "eval_samples_per_second": 22.498,
-    "eval_steps_per_second": 0.507,
+    "eval_loss": 0.22723796963691711,
+    "eval_runtime": 4.9743,
+    "eval_samples_per_second": 26.737,
+    "eval_steps_per_second": 0.603,
     "total_flos": 5.608920478833746e+17,
-    "train_loss": 0.4664680296633424,
-    "train_runtime": 551.2895,
-    "train_samples_per_second": 13.129,
-    "train_steps_per_second": 0.216
+    "train_loss": 0.46667203782987193,
+    "train_runtime": 402.2426,
+    "train_samples_per_second": 17.994,
+    "train_steps_per_second": 0.296
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 7.0,
     "eval_accuracy": 0.9699248120300752,
-    "eval_loss": 0.22653304040431976,
-    "eval_runtime": 5.9117,
-    "eval_samples_per_second": 22.498,
-    "eval_steps_per_second": 0.507
+    "eval_loss": 0.22723796963691711,
+    "eval_runtime": 4.9743,
+    "eval_samples_per_second": 26.737,
+    "eval_steps_per_second": 0.603
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 7.0,
     "total_flos": 5.608920478833746e+17,
-    "train_loss": 0.4664680296633424,
-    "train_runtime": 551.2895,
-    "train_samples_per_second": 13.129,
-    "train_steps_per_second": 0.216
+    "train_loss": 0.46667203782987193,
+    "train_runtime": 402.2426,
+    "train_samples_per_second": 17.994,
+    "train_steps_per_second": 0.296
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.22653304040431976,
+  "best_metric": 0.22723796963691711,
   "best_model_checkpoint": "./beans_outputs/checkpoint-119",
   "epoch": 7.0,
   "eval_steps": 500,
@@ -10,7 +10,7 @@
   "log_history": [
     {
       "epoch": 0.5882352941176471,
-      "grad_norm": 1.3733673095703125,
+      "grad_norm": 1.3763134479522705,
       "learning_rate": 1.831932773109244e-05,
       "loss": 0.9859,
       "step": 10
@@ -18,144 +18,144 @@
     {
       "epoch": 1.0,
       "eval_accuracy": 0.9323308270676691,
-      "eval_loss": 0.7493444681167603,
-      "eval_runtime": 6.3127,
-      "eval_samples_per_second": 21.069,
-      "eval_steps_per_second": 0.475,
+      "eval_loss": 0.7491587996482849,
+      "eval_runtime": 5.0252,
+      "eval_samples_per_second": 26.467,
+      "eval_steps_per_second": 0.597,
       "step": 17
     },
     {
       "epoch": 1.1764705882352942,
-      "grad_norm": 1.184097409248352,
+      "grad_norm": 1.1917294263839722,
       "learning_rate": 1.6638655462184877e-05,
-      "loss": 0.808,
+      "loss": 0.8079,
       "step": 20
     },
     {
       "epoch": 1.7647058823529411,
-      "grad_norm": 1.149975299835205,
+      "grad_norm": 1.144236445426941,
       "learning_rate": 1.4957983193277313e-05,
-      "loss": 0.6766,
+      "loss": 0.6764,
       "step": 30
     },
     {
       "epoch": 2.0,
       "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.529090940952301,
-      "eval_runtime": 6.2922,
-      "eval_samples_per_second": 21.137,
-      "eval_steps_per_second": 0.477,
+      "eval_loss": 0.5274316668510437,
+      "eval_runtime": 5.1613,
+      "eval_samples_per_second": 25.769,
+      "eval_steps_per_second": 0.581,
       "step": 34
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 1.216352939605713,
+      "grad_norm": 1.2157258987426758,
       "learning_rate": 1.3277310924369749e-05,
-      "loss": 0.55,
+      "loss": 0.5491,
       "step": 40
     },
     {
       "epoch": 2.9411764705882355,
-      "grad_norm": 1.0173835754394531,
+      "grad_norm": 1.031990885734558,
       "learning_rate": 1.1596638655462186e-05,
-      "loss": 0.4604,
+      "loss": 0.4601,
       "step": 50
     },
     {
       "epoch": 3.0,
       "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.37185198068618774,
-      "eval_runtime": 6.6034,
-      "eval_samples_per_second": 20.141,
-      "eval_steps_per_second": 0.454,
+      "eval_loss": 0.37144508957862854,
+      "eval_runtime": 5.4329,
+      "eval_samples_per_second": 24.48,
+      "eval_steps_per_second": 0.552,
       "step": 51
     },
     {
       "epoch": 3.5294117647058822,
-      "grad_norm": 1.165634274482727,
+      "grad_norm": 1.2084527015686035,
       "learning_rate": 9.915966386554622e-06,
-      "loss": 0.404,
+      "loss": 0.4047,
       "step": 60
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.29681986570358276,
-      "eval_runtime": 7.3605,
-      "eval_samples_per_second": 18.069,
-      "eval_steps_per_second": 0.408,
+      "eval_loss": 0.2976459562778473,
+      "eval_runtime": 5.0095,
+      "eval_samples_per_second": 26.549,
+      "eval_steps_per_second": 0.599,
       "step": 68
     },
     {
       "epoch": 4.117647058823529,
-      "grad_norm": 1.0317189693450928,
+      "grad_norm": 1.0828431844711304,
       "learning_rate": 8.23529411764706e-06,
-      "loss": 0.326,
+      "loss": 0.3269,
       "step": 70
     },
     {
       "epoch": 4.705882352941177,
-      "grad_norm": 1.5363612174987793,
+      "grad_norm": 1.3986183404922485,
       "learning_rate": 6.5546218487394966e-06,
-      "loss": 0.3168,
+      "loss": 0.3172,
       "step": 80
     },
     {
       "epoch": 5.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.2534410059452057,
-      "eval_runtime": 6.7412,
-      "eval_samples_per_second": 19.73,
-      "eval_steps_per_second": 0.445,
+      "eval_loss": 0.25337785482406616,
+      "eval_runtime": 5.2266,
+      "eval_samples_per_second": 25.447,
+      "eval_steps_per_second": 0.574,
       "step": 85
     },
     {
       "epoch": 5.294117647058823,
-      "grad_norm": 1.0671030282974243,
+      "grad_norm": 1.0007528066635132,
       "learning_rate": 4.873949579831933e-06,
-      "loss": 0.2769,
+      "loss": 0.2775,
       "step": 90
     },
     {
       "epoch": 5.882352941176471,
-      "grad_norm": 0.8577371835708618,
+      "grad_norm": 0.8358347415924072,
       "learning_rate": 3.1932773109243696e-06,
-      "loss": 0.254,
+      "loss": 0.2545,
       "step": 100
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.9774436090225563,
-      "eval_loss": 0.22781571745872498,
-      "eval_runtime": 7.0586,
-      "eval_samples_per_second": 18.842,
-      "eval_steps_per_second": 0.425,
+      "eval_loss": 0.22837479412555695,
+      "eval_runtime": 5.7431,
+      "eval_samples_per_second": 23.158,
+      "eval_steps_per_second": 0.522,
       "step": 102
     },
     {
       "epoch": 6.470588235294118,
-      "grad_norm": 1.0884819030761719,
+      "grad_norm": 1.0932303667068481,
       "learning_rate": 1.5126050420168068e-06,
-      "loss": 0.263,
+      "loss": 0.2632,
       "step": 110
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.22653304040431976,
-      "eval_runtime": 6.2522,
-      "eval_samples_per_second": 21.272,
-      "eval_steps_per_second": 0.48,
+      "eval_loss": 0.22723796963691711,
+      "eval_runtime": 5.0242,
+      "eval_samples_per_second": 26.472,
+      "eval_steps_per_second": 0.597,
       "step": 119
     },
     {
       "epoch": 7.0,
       "step": 119,
       "total_flos": 5.608920478833746e+17,
-      "train_loss": 0.4664680296633424,
-      "train_runtime": 551.2895,
-      "train_samples_per_second": 13.129,
-      "train_steps_per_second": 0.216
+      "train_loss": 0.46667203782987193,
+      "train_runtime": 402.2426,
+      "train_samples_per_second": 17.994,
+      "train_steps_per_second": 0.296
     }
   ],
   "logging_steps": 10,