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: 0.5646
 - Accuracy: 0.9699

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9624060150375939,
-    "eval_loss": 0.34422725439071655,
-    "eval_runtime": 8.5078,
-    "eval_samples_per_second": 15.633,
-    "eval_steps_per_second": 0.353,
-    "total_flos": 4.006371770595533e+17,
-    "train_loss": 0.5861698781742769,
-    "train_runtime": 731.5711,
-    "train_samples_per_second": 7.067,
-    "train_steps_per_second": 0.116
+    "epoch": 3.0,
+    "eval_accuracy": 0.9699248120300752,
+    "eval_loss": 0.564608097076416,
+    "eval_runtime": 17.4903,
+    "eval_samples_per_second": 7.604,
+    "eval_steps_per_second": 0.172,
+    "total_flos": 2.4038230623573197e+17,
+    "train_loss": 0.7542588278359058,
+    "train_runtime": 427.2792,
+    "train_samples_per_second": 7.26,
+    "train_steps_per_second": 0.119
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9624060150375939,
-    "eval_loss": 0.34422725439071655,
-    "eval_runtime": 8.5078,
-    "eval_samples_per_second": 15.633,
-    "eval_steps_per_second": 0.353
+    "epoch": 3.0,
+    "eval_accuracy": 0.9699248120300752,
+    "eval_loss": 0.564608097076416,
+    "eval_runtime": 17.4903,
+    "eval_samples_per_second": 7.604,
+    "eval_steps_per_second": 0.172
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "total_flos": 4.006371770595533e+17,
-    "train_loss": 0.5861698781742769,
-    "train_runtime": 731.5711,
-    "train_samples_per_second": 7.067,
-    "train_steps_per_second": 0.116
+    "epoch": 3.0,
+    "total_flos": 2.4038230623573197e+17,
+    "train_loss": 0.7542588278359058,
+    "train_runtime": 427.2792,
+    "train_samples_per_second": 7.26,
+    "train_steps_per_second": 0.119
 }

trainer_state.json

@@ -1,128 +1,89 @@
 {
-  "best_metric": 0.34422725439071655,
-  "best_model_checkpoint": "./beans_outputs/checkpoint-85",
-  "epoch": 5.0,
+  "best_metric": 0.564608097076416,
+  "best_model_checkpoint": "./beans_outputs/checkpoint-51",
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 85,
+  "global_step": 51,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.5882352941176471,
-      "grad_norm": 1.339193344116211,
-      "learning_rate": 1.7647058823529414e-05,
-      "loss": 0.9867,
+      "grad_norm": 1.342671275138855,
+      "learning_rate": 1.607843137254902e-05,
+      "loss": 0.9886,
       "step": 10
     },
     {
       "epoch": 1.0,
       "eval_accuracy": 0.924812030075188,
-      "eval_loss": 0.7571204304695129,
-      "eval_runtime": 17.907,
-      "eval_samples_per_second": 7.427,
-      "eval_steps_per_second": 0.168,
+      "eval_loss": 0.7758036851882935,
+      "eval_runtime": 13.3932,
+      "eval_samples_per_second": 9.93,
+      "eval_steps_per_second": 0.224,
       "step": 17
     },
     {
       "epoch": 1.1764705882352942,
-      "grad_norm": 1.186477780342102,
-      "learning_rate": 1.5294117647058822e-05,
-      "loss": 0.8134,
+      "grad_norm": 1.2002716064453125,
+      "learning_rate": 1.215686274509804e-05,
+      "loss": 0.8266,
       "step": 20
     },
     {
       "epoch": 1.7647058823529411,
-      "grad_norm": 1.7653396129608154,
-      "learning_rate": 1.2941176470588238e-05,
-      "loss": 0.6897,
+      "grad_norm": 1.1757010221481323,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.7226,
       "step": 30
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.5546792149543762,
-      "eval_runtime": 18.1491,
-      "eval_samples_per_second": 7.328,
-      "eval_steps_per_second": 0.165,
+      "eval_accuracy": 0.9548872180451128,
+      "eval_loss": 0.6197953820228577,
+      "eval_runtime": 12.3405,
+      "eval_samples_per_second": 10.778,
+      "eval_steps_per_second": 0.243,
       "step": 34
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 1.223302960395813,
-      "learning_rate": 1.0588235294117648e-05,
-      "loss": 0.5764,
+      "grad_norm": 1.2492908239364624,
+      "learning_rate": 4.313725490196079e-06,
+      "loss": 0.6423,
       "step": 40
     },
     {
       "epoch": 2.9411764705882355,
-      "grad_norm": 1.0631064176559448,
-      "learning_rate": 8.23529411764706e-06,
-      "loss": 0.4993,
+      "grad_norm": 1.134194254875183,
+      "learning_rate": 3.921568627450981e-07,
+      "loss": 0.6041,
       "step": 50
     },
     {
       "epoch": 3.0,
       "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.42078566551208496,
-      "eval_runtime": 18.1595,
-      "eval_samples_per_second": 7.324,
-      "eval_steps_per_second": 0.165,
+      "eval_loss": 0.564608097076416,
+      "eval_runtime": 12.5805,
+      "eval_samples_per_second": 10.572,
+      "eval_steps_per_second": 0.238,
       "step": 51
     },
     {
-      "epoch": 3.5294117647058822,
-      "grad_norm": 1.1994314193725586,
-      "learning_rate": 5.882352941176471e-06,
-      "loss": 0.455,
-      "step": 60
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.3615361154079437,
-      "eval_runtime": 18.2494,
-      "eval_samples_per_second": 7.288,
-      "eval_steps_per_second": 0.164,
-      "step": 68
-    },
-    {
-      "epoch": 4.117647058823529,
-      "grad_norm": 1.0398813486099243,
-      "learning_rate": 3.529411764705883e-06,
-      "loss": 0.3916,
-      "step": 70
-    },
-    {
-      "epoch": 4.705882352941177,
-      "grad_norm": 1.11141836643219,
-      "learning_rate": 1.1764705882352942e-06,
-      "loss": 0.3931,
-      "step": 80
-    },
-    {
-      "epoch": 5.0,
-      "eval_accuracy": 0.9624060150375939,
-      "eval_loss": 0.34422725439071655,
-      "eval_runtime": 10.0226,
-      "eval_samples_per_second": 13.27,
-      "eval_steps_per_second": 0.299,
-      "step": 85
-    },
-    {
-      "epoch": 5.0,
-      "step": 85,
-      "total_flos": 4.006371770595533e+17,
-      "train_loss": 0.5861698781742769,
-      "train_runtime": 731.5711,
-      "train_samples_per_second": 7.067,
-      "train_steps_per_second": 0.116
+      "epoch": 3.0,
+      "step": 51,
+      "total_flos": 2.4038230623573197e+17,
+      "train_loss": 0.7542588278359058,
+      "train_runtime": 427.2792,
+      "train_samples_per_second": 7.26,
+      "train_steps_per_second": 0.119
     }
   ],
   "logging_steps": 10,
-  "max_steps": 85,
+  "max_steps": 51,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -136,7 +97,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.006371770595533e+17,
+  "total_flos": 2.4038230623573197e+17,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null