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: 0.5798
-- Accuracy: 0.7594
+- Loss: 0.5370
+- Accuracy: 0.7895
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 10.0,
-    "eval_accuracy": 0.3383458646616541,
-    "eval_loss": 2.689824104309082,
-    "eval_runtime": 0.9993,
-    "eval_samples_per_second": 133.095,
-    "eval_steps_per_second": 17.012,
+    "eval_accuracy": 0.7894736842105263,
+    "eval_loss": 0.5369813442230225,
+    "eval_runtime": 1.1887,
+    "eval_samples_per_second": 111.89,
+    "eval_steps_per_second": 14.302,
     "total_flos": 8.772706474360013e+17,
-    "train_loss": 0.3764252471923828,
-    "train_runtime": 205.4123,
-    "train_samples_per_second": 50.338,
-    "train_steps_per_second": 6.329
+    "train_loss": 0.7462210376446063,
+    "train_runtime": 205.5107,
+    "train_samples_per_second": 50.314,
+    "train_steps_per_second": 6.326
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
-    "eval_accuracy": 0.3383458646616541,
-    "eval_loss": 2.689824104309082,
-    "eval_runtime": 0.9993,
-    "eval_samples_per_second": 133.095,
-    "eval_steps_per_second": 17.012
+    "eval_accuracy": 0.7894736842105263,
+    "eval_loss": 0.5369813442230225,
+    "eval_runtime": 1.1887,
+    "eval_samples_per_second": 111.89,
+    "eval_steps_per_second": 14.302
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
     "total_flos": 8.772706474360013e+17,
-    "train_loss": 0.3764252471923828,
-    "train_runtime": 205.4123,
-    "train_samples_per_second": 50.338,
-    "train_steps_per_second": 6.329
+    "train_loss": 0.7462210376446063,
+    "train_runtime": 205.5107,
+    "train_samples_per_second": 50.314,
+    "train_steps_per_second": 6.326
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.0367523655295372,
-  "best_model_checkpoint": "./beans_outputs/checkpoint-390",
+  "best_metric": 0.5369813442230225,
+  "best_model_checkpoint": "./beans_outputs/checkpoint-1170",
   "epoch": 10.0,
   "eval_steps": 500,
   "global_step": 1300,
@@ -10,193 +10,193 @@
   "log_history": [
     {
       "epoch": 0.7692307692307693,
-      "grad_norm": 4.539805889129639,
-      "learning_rate": 0.0005,
-      "loss": 0.4713,
+      "grad_norm": 3.450658082962036,
+      "learning_rate": 5e-05,
+      "loss": 1.0967,
       "step": 100
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9548872180451128,
-      "eval_loss": 0.09497568756341934,
-      "eval_runtime": 1.0331,
-      "eval_samples_per_second": 128.734,
-      "eval_steps_per_second": 16.455,
+      "eval_accuracy": 0.5413533834586466,
+      "eval_loss": 0.9973052740097046,
+      "eval_runtime": 0.9887,
+      "eval_samples_per_second": 134.519,
+      "eval_steps_per_second": 17.194,
       "step": 130
     },
     {
       "epoch": 1.5384615384615383,
-      "grad_norm": 0.8214866518974304,
-      "learning_rate": 0.0005,
-      "loss": 0.1717,
+      "grad_norm": 1.2139726877212524,
+      "learning_rate": 5e-05,
+      "loss": 1.0161,
       "step": 200
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.08848864585161209,
-      "eval_runtime": 0.9725,
-      "eval_samples_per_second": 136.757,
-      "eval_steps_per_second": 17.48,
+      "eval_accuracy": 0.5789473684210527,
+      "eval_loss": 0.9221953749656677,
+      "eval_runtime": 1.1584,
+      "eval_samples_per_second": 114.811,
+      "eval_steps_per_second": 14.675,
       "step": 260
     },
     {
       "epoch": 2.3076923076923075,
-      "grad_norm": 5.174630165100098,
-      "learning_rate": 0.0005,
-      "loss": 0.1941,
+      "grad_norm": 1.6421459913253784,
+      "learning_rate": 5e-05,
+      "loss": 0.9472,
       "step": 300
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9924812030075187,
-      "eval_loss": 0.0367523655295372,
-      "eval_runtime": 1.0358,
-      "eval_samples_per_second": 128.405,
-      "eval_steps_per_second": 16.413,
+      "eval_accuracy": 0.6390977443609023,
+      "eval_loss": 0.8625620007514954,
+      "eval_runtime": 1.1687,
+      "eval_samples_per_second": 113.798,
+      "eval_steps_per_second": 14.546,
       "step": 390
     },
     {
       "epoch": 3.076923076923077,
-      "grad_norm": 6.90311336517334,
-      "learning_rate": 0.0005,
-      "loss": 0.178,
+      "grad_norm": 2.6952731609344482,
+      "learning_rate": 5e-05,
+      "loss": 0.8456,
       "step": 400
     },
     {
       "epoch": 3.8461538461538463,
-      "grad_norm": 0.0417233482003212,
-      "learning_rate": 0.0005,
-      "loss": 0.1484,
+      "grad_norm": 1.875662088394165,
+      "learning_rate": 5e-05,
+      "loss": 0.7532,
       "step": 500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.10054013133049011,
-      "eval_runtime": 1.0366,
-      "eval_samples_per_second": 128.308,
-      "eval_steps_per_second": 16.4,
+      "eval_accuracy": 0.7142857142857143,
+      "eval_loss": 0.7298586368560791,
+      "eval_runtime": 1.176,
+      "eval_samples_per_second": 113.091,
+      "eval_steps_per_second": 14.455,
       "step": 520
     },
     {
       "epoch": 4.615384615384615,
-      "grad_norm": 3.222952365875244,
-      "learning_rate": 0.0005,
-      "loss": 0.1574,
+      "grad_norm": 2.776125192642212,
+      "learning_rate": 5e-05,
+      "loss": 0.6792,
       "step": 600
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9774436090225563,
-      "eval_loss": 0.06399582326412201,
-      "eval_runtime": 1.1232,
-      "eval_samples_per_second": 118.415,
-      "eval_steps_per_second": 15.136,
+      "eval_accuracy": 0.706766917293233,
+      "eval_loss": 0.6639184951782227,
+      "eval_runtime": 1.2266,
+      "eval_samples_per_second": 108.429,
+      "eval_steps_per_second": 13.859,
       "step": 650
     },
     {
       "epoch": 5.384615384615385,
-      "grad_norm": 0.5326167345046997,
-      "learning_rate": 0.0005,
-      "loss": 0.1948,
+      "grad_norm": 2.474877119064331,
+      "learning_rate": 5e-05,
+      "loss": 0.6645,
       "step": 700
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9172932330827067,
-      "eval_loss": 0.18919076025485992,
-      "eval_runtime": 1.1717,
-      "eval_samples_per_second": 113.51,
-      "eval_steps_per_second": 14.509,
+      "eval_accuracy": 0.7218045112781954,
+      "eval_loss": 0.6237967610359192,
+      "eval_runtime": 1.15,
+      "eval_samples_per_second": 115.652,
+      "eval_steps_per_second": 14.783,
       "step": 780
     },
     {
       "epoch": 6.153846153846154,
-      "grad_norm": 2.2912113666534424,
-      "learning_rate": 0.0005,
-      "loss": 0.2541,
+      "grad_norm": 2.956679344177246,
+      "learning_rate": 5e-05,
+      "loss": 0.6473,
       "step": 800
     },
     {
       "epoch": 6.923076923076923,
-      "grad_norm": 4.357283592224121,
-      "learning_rate": 0.0005,
-      "loss": 0.4159,
+      "grad_norm": 9.533896446228027,
+      "learning_rate": 5e-05,
+      "loss": 0.6056,
       "step": 900
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.9699248120300752,
-      "eval_loss": 0.15302829444408417,
-      "eval_runtime": 1.1496,
-      "eval_samples_per_second": 115.692,
-      "eval_steps_per_second": 14.788,
+      "eval_accuracy": 0.7443609022556391,
+      "eval_loss": 0.6903008818626404,
+      "eval_runtime": 1.1553,
+      "eval_samples_per_second": 115.117,
+      "eval_steps_per_second": 14.714,
       "step": 910
     },
     {
       "epoch": 7.6923076923076925,
-      "grad_norm": 2.716820240020752,
-      "learning_rate": 0.0005,
-      "loss": 0.4951,
+      "grad_norm": 4.010281085968018,
+      "learning_rate": 5e-05,
+      "loss": 0.614,
       "step": 1000
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.6842105263157895,
-      "eval_loss": 0.7176167368888855,
-      "eval_runtime": 0.9764,
-      "eval_samples_per_second": 136.21,
-      "eval_steps_per_second": 17.41,
+      "eval_accuracy": 0.706766917293233,
+      "eval_loss": 0.6278253793716431,
+      "eval_runtime": 1.1805,
+      "eval_samples_per_second": 112.66,
+      "eval_steps_per_second": 14.4,
       "step": 1040
     },
     {
       "epoch": 8.461538461538462,
-      "grad_norm": 1.9368160963058472,
-      "learning_rate": 0.0005,
-      "loss": 0.6818,
+      "grad_norm": 2.6526243686676025,
+      "learning_rate": 5e-05,
+      "loss": 0.6298,
       "step": 1100
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.42857142857142855,
-      "eval_loss": 1.4557244777679443,
-      "eval_runtime": 1.1951,
-      "eval_samples_per_second": 111.285,
-      "eval_steps_per_second": 14.224,
+      "eval_accuracy": 0.7894736842105263,
+      "eval_loss": 0.5369813442230225,
+      "eval_runtime": 0.995,
+      "eval_samples_per_second": 133.666,
+      "eval_steps_per_second": 17.085,
       "step": 1170
     },
     {
       "epoch": 9.23076923076923,
-      "grad_norm": 1.1992454528808594,
-      "learning_rate": 0.0005,
-      "loss": 0.7561,
+      "grad_norm": 4.809407711029053,
+      "learning_rate": 5e-05,
+      "loss": 0.6149,
       "step": 1200
     },
     {
       "epoch": 10.0,
-      "grad_norm": 3.8898727893829346,
-      "learning_rate": 0.0005,
-      "loss": 0.7749,
+      "grad_norm": 9.877020835876465,
+      "learning_rate": 5e-05,
+      "loss": 0.5867,
       "step": 1300
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.7293233082706767,
-      "eval_loss": 0.7070097923278809,
-      "eval_runtime": 1.1878,
-      "eval_samples_per_second": 111.974,
-      "eval_steps_per_second": 14.312,
+      "eval_accuracy": 0.7593984962406015,
+      "eval_loss": 0.579809308052063,
+      "eval_runtime": 1.1918,
+      "eval_samples_per_second": 111.6,
+      "eval_steps_per_second": 14.265,
       "step": 1300
     },
     {
       "epoch": 10.0,
       "step": 1300,
       "total_flos": 8.772706474360013e+17,
-      "train_loss": 0.3764252471923828,
-      "train_runtime": 205.4123,
-      "train_samples_per_second": 50.338,
-      "train_steps_per_second": 6.329
+      "train_loss": 0.7462210376446063,
+      "train_runtime": 205.5107,
+      "train_samples_per_second": 50.314,
+      "train_steps_per_second": 6.326
     }
   ],
   "logging_steps": 100,