End of training

README.md

@@ -5,6 +5,8 @@ license: apache-2.0
 metrics:
 - accuracy
 tags:
+- image-classification
+- vision
 - generated_from_trainer
 model-index:
 - name: only-lora-beans-vit-base-patch16-224-in21k
@@ -16,10 +18,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # only-lora-beans-vit-base-patch16-224-in21k
 
-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: 1.0935
-- Accuracy: 0.3459
+- Loss: 1.1110
+- Accuracy: 0.3308
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 10.0,
     "eval_accuracy": 0.3308270676691729,
-    "eval_loss": 1.895585536956787,
-    "eval_runtime": 0.9592,
-    "eval_samples_per_second": 138.655,
-    "eval_steps_per_second": 17.723,
+    "eval_loss": 1.1110202074050903,
+    "eval_runtime": 1.0014,
+    "eval_samples_per_second": 132.816,
+    "eval_steps_per_second": 16.976,
     "total_flos": 8.772706474360013e+17,
-    "train_loss": 0.7308059575007512,
-    "train_runtime": 175.3451,
-    "train_samples_per_second": 58.969,
-    "train_steps_per_second": 7.414
+    "train_loss": 0.6310529987628644,
+    "train_runtime": 186.1113,
+    "train_samples_per_second": 55.558,
+    "train_steps_per_second": 6.985
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
     "eval_accuracy": 0.3308270676691729,
-    "eval_loss": 1.895585536956787,
-    "eval_runtime": 0.9592,
-    "eval_samples_per_second": 138.655,
-    "eval_steps_per_second": 17.723
+    "eval_loss": 1.1110202074050903,
+    "eval_runtime": 1.0014,
+    "eval_samples_per_second": 132.816,
+    "eval_steps_per_second": 16.976
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
     "total_flos": 8.772706474360013e+17,
-    "train_loss": 0.7308059575007512,
-    "train_runtime": 175.3451,
-    "train_samples_per_second": 58.969,
-    "train_steps_per_second": 7.414
+    "train_loss": 0.6310529987628644,
+    "train_runtime": 186.1113,
+    "train_samples_per_second": 55.558,
+    "train_steps_per_second": 6.985
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.3044010102748871,
-  "best_model_checkpoint": "./beans_outputs/checkpoint-650",
+  "best_metric": 0.17903044819831848,
+  "best_model_checkpoint": "./beans_outputs/checkpoint-520",
   "epoch": 10.0,
   "eval_steps": 500,
   "global_step": 1300,
@@ -10,193 +10,193 @@
   "log_history": [
     {
       "epoch": 0.7692307692307693,
-      "grad_norm": 2.5466179847717285,
+      "grad_norm": 2.7811906337738037,
       "learning_rate": 5e-05,
-      "loss": 1.0208,
+      "loss": 0.9748,
       "step": 100
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8045112781954887,
-      "eval_loss": 0.8930914998054504,
-      "eval_runtime": 0.9307,
-      "eval_samples_per_second": 142.898,
-      "eval_steps_per_second": 18.265,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_loss": 0.7215397357940674,
+      "eval_runtime": 0.9364,
+      "eval_samples_per_second": 142.034,
+      "eval_steps_per_second": 18.155,
       "step": 130
     },
     {
       "epoch": 1.5384615384615383,
-      "grad_norm": 1.4668034315109253,
+      "grad_norm": 1.6656582355499268,
       "learning_rate": 5e-05,
-      "loss": 0.8683,
+      "loss": 0.6849,
       "step": 200
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_loss": 0.6855143308639526,
-      "eval_runtime": 0.9382,
-      "eval_samples_per_second": 141.765,
-      "eval_steps_per_second": 18.12,
+      "eval_accuracy": 0.9473684210526315,
+      "eval_loss": 0.4107130765914917,
+      "eval_runtime": 0.9628,
+      "eval_samples_per_second": 138.137,
+      "eval_steps_per_second": 17.657,
       "step": 260
     },
     {
       "epoch": 2.3076923076923075,
-      "grad_norm": 1.5885818004608154,
+      "grad_norm": 1.416643738746643,
       "learning_rate": 5e-05,
-      "loss": 0.7297,
+      "loss": 0.4838,
       "step": 300
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_loss": 0.5008501410484314,
-      "eval_runtime": 0.9428,
-      "eval_samples_per_second": 141.075,
-      "eval_steps_per_second": 18.032,
+      "eval_accuracy": 0.9473684210526315,
+      "eval_loss": 0.2423216700553894,
+      "eval_runtime": 0.9821,
+      "eval_samples_per_second": 135.426,
+      "eval_steps_per_second": 17.31,
       "step": 390
     },
     {
       "epoch": 3.076923076923077,
-      "grad_norm": 2.4792442321777344,
+      "grad_norm": 2.6065406799316406,
       "learning_rate": 5e-05,
-      "loss": 0.6084,
+      "loss": 0.3533,
       "step": 400
     },
     {
       "epoch": 3.8461538461538463,
-      "grad_norm": 1.0500737428665161,
+      "grad_norm": 1.5872589349746704,
       "learning_rate": 5e-05,
-      "loss": 0.4881,
+      "loss": 0.2594,
       "step": 500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9323308270676691,
-      "eval_loss": 0.3685053288936615,
-      "eval_runtime": 0.9444,
-      "eval_samples_per_second": 140.829,
-      "eval_steps_per_second": 18.001,
+      "eval_accuracy": 0.9624060150375939,
+      "eval_loss": 0.17903044819831848,
+      "eval_runtime": 0.9895,
+      "eval_samples_per_second": 134.417,
+      "eval_steps_per_second": 17.181,
       "step": 520
     },
     {
       "epoch": 4.615384615384615,
-      "grad_norm": 0.9487807154655457,
+      "grad_norm": 2.6757798194885254,
       "learning_rate": 5e-05,
-      "loss": 0.3733,
+      "loss": 0.2082,
       "step": 600
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_loss": 0.3044010102748871,
-      "eval_runtime": 0.9975,
-      "eval_samples_per_second": 133.327,
-      "eval_steps_per_second": 17.042,
+      "eval_accuracy": 0.9323308270676691,
+      "eval_loss": 0.19146375358104706,
+      "eval_runtime": 1.0393,
+      "eval_samples_per_second": 127.973,
+      "eval_steps_per_second": 16.357,
       "step": 650
     },
     {
       "epoch": 5.384615384615385,
-      "grad_norm": 2.811962604522705,
+      "grad_norm": 5.998991012573242,
       "learning_rate": 5e-05,
-      "loss": 0.3229,
+      "loss": 0.2159,
       "step": 700
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.8646616541353384,
-      "eval_loss": 0.3873802721500397,
-      "eval_runtime": 0.9366,
-      "eval_samples_per_second": 142.002,
-      "eval_steps_per_second": 18.151,
+      "eval_accuracy": 0.9097744360902256,
+      "eval_loss": 0.3304254412651062,
+      "eval_runtime": 0.9409,
+      "eval_samples_per_second": 141.349,
+      "eval_steps_per_second": 18.067,
       "step": 780
     },
     {
       "epoch": 6.153846153846154,
-      "grad_norm": 2.758183717727661,
+      "grad_norm": 4.108273029327393,
       "learning_rate": 5e-05,
-      "loss": 0.4189,
+      "loss": 0.339,
       "step": 800
     },
     {
       "epoch": 6.923076923076923,
-      "grad_norm": 5.297445297241211,
+      "grad_norm": 5.80717134475708,
       "learning_rate": 5e-05,
-      "loss": 0.5796,
+      "loss": 0.5954,
       "step": 900
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.5939849624060151,
-      "eval_loss": 0.7966009974479675,
-      "eval_runtime": 0.9577,
-      "eval_samples_per_second": 138.872,
-      "eval_steps_per_second": 17.75,
+      "eval_accuracy": 0.6917293233082706,
+      "eval_loss": 0.6861125230789185,
+      "eval_runtime": 1.0533,
+      "eval_samples_per_second": 126.266,
+      "eval_steps_per_second": 16.139,
       "step": 910
     },
     {
       "epoch": 7.6923076923076925,
-      "grad_norm": 28.03814697265625,
+      "grad_norm": 4.3353142738342285,
       "learning_rate": 5e-05,
-      "loss": 0.765,
+      "loss": 0.8045,
       "step": 1000
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.5263157894736842,
-      "eval_loss": 0.9421801567077637,
-      "eval_runtime": 0.9477,
-      "eval_samples_per_second": 140.335,
-      "eval_steps_per_second": 17.938,
+      "eval_accuracy": 0.48872180451127817,
+      "eval_loss": 1.0407782793045044,
+      "eval_runtime": 1.1746,
+      "eval_samples_per_second": 113.234,
+      "eval_steps_per_second": 14.474,
       "step": 1040
     },
     {
       "epoch": 8.461538461538462,
-      "grad_norm": 5.901334285736084,
+      "grad_norm": 1.6110053062438965,
       "learning_rate": 5e-05,
-      "loss": 1.0732,
+      "loss": 1.0506,
       "step": 1100
     },
     {
       "epoch": 9.0,
       "eval_accuracy": 0.3308270676691729,
-      "eval_loss": 1.0903689861297607,
-      "eval_runtime": 0.9358,
-      "eval_samples_per_second": 142.119,
-      "eval_steps_per_second": 18.166,
+      "eval_loss": 1.1016151905059814,
+      "eval_runtime": 1.2259,
+      "eval_samples_per_second": 108.493,
+      "eval_steps_per_second": 13.868,
       "step": 1170
     },
     {
       "epoch": 9.23076923076923,
-      "grad_norm": 1.3560993671417236,
+      "grad_norm": 0.6255662441253662,
       "learning_rate": 5e-05,
-      "loss": 1.0879,
+      "loss": 1.1038,
       "step": 1200
     },
     {
       "epoch": 10.0,
-      "grad_norm": 37.35905456542969,
+      "grad_norm": 6.890110969543457,
       "learning_rate": 5e-05,
-      "loss": 1.1643,
+      "loss": 1.1299,
       "step": 1300
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.41353383458646614,
-      "eval_loss": 1.0773515701293945,
-      "eval_runtime": 0.9946,
-      "eval_samples_per_second": 133.719,
-      "eval_steps_per_second": 17.092,
+      "eval_accuracy": 0.3458646616541353,
+      "eval_loss": 1.0934619903564453,
+      "eval_runtime": 1.1913,
+      "eval_samples_per_second": 111.642,
+      "eval_steps_per_second": 14.27,
       "step": 1300
     },
     {
       "epoch": 10.0,
       "step": 1300,
       "total_flos": 8.772706474360013e+17,
-      "train_loss": 0.7308059575007512,
-      "train_runtime": 175.3451,
-      "train_samples_per_second": 58.969,
-      "train_steps_per_second": 7.414
+      "train_loss": 0.6310529987628644,
+      "train_runtime": 186.1113,
+      "train_samples_per_second": 55.558,
+      "train_steps_per_second": 6.985
     }
   ],
   "logging_steps": 100,