Model save

README.md

@@ -4,6 +4,8 @@ license: apache-2.0
 base_model: google/vit-base-patch16-224
 tags:
 - generated_from_trainer
+metrics:
+- accuracy
 model-index:
 - name: vit-base-patch16-224_rice-leaf-disease-augmented-v2_fft
   results: []
@@ -16,13 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- epoch: 18.0
-- eval_accuracy: 0.9196
-- eval_loss: 0.3620
-- eval_runtime: 11.0713
-- eval_samples_per_second: 30.349
-- eval_steps_per_second: 0.542
-- step: 2250
+- Accuracy: 0.9196
+- Loss: 0.3620
 
 ## Model description
 
@@ -51,6 +48,30 @@ The following hyperparameters were used during training:
 - num_epochs: 15
 - mixed_precision_training: Native AMP
 
+### Training results
+
+| Training Loss | Epoch | Step | Accuracy | Validation Loss |
+|:-------------:|:-----:|:----:|:--------:|:---------------:|
+| 1.9482        | 1.0   | 125  | 0.5685   | 1.5012          |
+| 0.9894        | 2.0   | 250  | 0.7976   | 0.6444          |
+| 0.3321        | 3.0   | 375  | 0.8958   | 0.3859          |
+| 0.1115        | 4.0   | 500  | 0.9107   | 0.3081          |
+| 0.0387        | 5.0   | 625  | 0.9137   | 0.2980          |
+| 0.0204        | 6.0   | 750  | 0.9137   | 0.2936          |
+| 0.0169        | 7.0   | 875  | 0.9196   | 0.2953          |
+| 0.0078        | 8.0   | 1000 | 0.9226   | 0.3067          |
+| 0.0034        | 9.0   | 1125 | 0.9286   | 0.3087          |
+| 0.0025        | 10.0  | 1250 | 0.9196   | 0.3139          |
+| 0.0023        | 11.0  | 1375 | 0.9196   | 0.3142          |
+| 0.0019        | 12.0  | 1500 | 0.9196   | 0.3288          |
+| 0.0013        | 13.0  | 1625 | 0.9196   | 0.3359          |
+| 0.001         | 14.0  | 1750 | 0.9226   | 0.3413          |
+| 0.0009        | 15.0  | 1875 | 0.9226   | 0.3425          |
+| 0.0009        | 16.0  | 2000 | 0.9226   | 0.3481          |
+| 0.0007        | 17.0  | 2125 | 0.9226   | 0.3571          |
+| 0.0006        | 18.0  | 2250 | 0.9196   | 0.3620          |
+
+
 ### Framework versions
 
 - Transformers 4.48.3

logs/events.out.tfevents.1740444395.4a8a76f3e68a.1955.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa79a1d4ec4bca93aaf1b6810cbf2efcad36d9606487ce3e1c783c4f1bf89cc2
-size 88
+oid sha256:665349651650707fd38cbf97bd56fc049951be360fcfe9126e5e5ef8a0229fa4
+size 6021

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7ccfeb9fcbfcb69bb1dd3f91ab957cedb45851be3c8d25ba88c6576bc370a613
+oid sha256:3e45518fd16a0cfc3f3db29de59fd516f8807a340ff671f0c2bd29aa81adca17
 size 343242432

trainer_state.json

@@ -1,17 +1,306 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 0,
+  "best_metric": 0.29358136653900146,
+  "best_model_checkpoint": "./drive/Shareddrives/CS198-Drones/training_output/vit-base-patch16-224_rice-leaf-disease-augmented-v2_fft/checkpoint-750",
+  "epoch": 18.0,
   "eval_steps": 500,
-  "global_step": 0,
+  "global_step": 2250,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
-  "log_history": [],
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8496286869049072,
+      "learning_rate": 5e-06,
+      "loss": 1.9482,
+      "step": 125
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5684523809523809,
+      "eval_loss": 1.5012105703353882,
+      "eval_runtime": 11.0652,
+      "eval_samples_per_second": 30.365,
+      "eval_steps_per_second": 0.542,
+      "step": 125
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.772235155105591,
+      "learning_rate": 1e-05,
+      "loss": 0.9894,
+      "step": 250
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7976190476190477,
+      "eval_loss": 0.6444453001022339,
+      "eval_runtime": 11.0129,
+      "eval_samples_per_second": 30.51,
+      "eval_steps_per_second": 0.545,
+      "step": 250
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4745649099349976,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.3321,
+      "step": 375
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8958333333333334,
+      "eval_loss": 0.3858731687068939,
+      "eval_runtime": 10.958,
+      "eval_samples_per_second": 30.663,
+      "eval_steps_per_second": 0.548,
+      "step": 375
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.1253288984298706,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.1115,
+      "step": 500
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9107142857142857,
+      "eval_loss": 0.3080737888813019,
+      "eval_runtime": 10.9888,
+      "eval_samples_per_second": 30.577,
+      "eval_steps_per_second": 0.546,
+      "step": 500
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.3589422404766083,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0387,
+      "step": 625
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9136904761904762,
+      "eval_loss": 0.2979942262172699,
+      "eval_runtime": 11.1535,
+      "eval_samples_per_second": 30.125,
+      "eval_steps_per_second": 0.538,
+      "step": 625
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.2502361536026001,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.0204,
+      "step": 750
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9136904761904762,
+      "eval_loss": 0.29358136653900146,
+      "eval_runtime": 11.1371,
+      "eval_samples_per_second": 30.169,
+      "eval_steps_per_second": 0.539,
+      "step": 750
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.5141955018043518,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.0169,
+      "step": 875
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9196428571428571,
+      "eval_loss": 0.2952845096588135,
+      "eval_runtime": 11.0115,
+      "eval_samples_per_second": 30.513,
+      "eval_steps_per_second": 0.545,
+      "step": 875
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.1391526609659195,
+      "learning_rate": 7.500000000000003e-06,
+      "loss": 0.0078,
+      "step": 1000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9226190476190477,
+      "eval_loss": 0.3067362606525421,
+      "eval_runtime": 11.2875,
+      "eval_samples_per_second": 29.767,
+      "eval_steps_per_second": 0.532,
+      "step": 1000
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.06930585205554962,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.0034,
+      "step": 1125
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9285714285714286,
+      "eval_loss": 0.3087221682071686,
+      "eval_runtime": 10.9679,
+      "eval_samples_per_second": 30.635,
+      "eval_steps_per_second": 0.547,
+      "step": 1125
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.09097771346569061,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.0025,
+      "step": 1250
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9196428571428571,
+      "eval_loss": 0.3138602674007416,
+      "eval_runtime": 10.3394,
+      "eval_samples_per_second": 32.497,
+      "eval_steps_per_second": 0.58,
+      "step": 1250
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.06938016414642334,
+      "learning_rate": 1e-05,
+      "loss": 0.0023,
+      "step": 1375
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.9196428571428571,
+      "eval_loss": 0.3141930401325226,
+      "eval_runtime": 10.7534,
+      "eval_samples_per_second": 31.246,
+      "eval_steps_per_second": 0.558,
+      "step": 1375
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.04166670888662338,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.0019,
+      "step": 1500
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.9196428571428571,
+      "eval_loss": 0.3288079798221588,
+      "eval_runtime": 10.8781,
+      "eval_samples_per_second": 30.888,
+      "eval_steps_per_second": 0.552,
+      "step": 1500
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.04881810024380684,
+      "learning_rate": 5.86824088833465e-06,
+      "loss": 0.0013,
+      "step": 1625
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.9196428571428571,
+      "eval_loss": 0.33594179153442383,
+      "eval_runtime": 11.2397,
+      "eval_samples_per_second": 29.894,
+      "eval_steps_per_second": 0.534,
+      "step": 1625
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.02085023745894432,
+      "learning_rate": 2.5000000000000036e-06,
+      "loss": 0.001,
+      "step": 1750
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.9226190476190477,
+      "eval_loss": 0.34129947423934937,
+      "eval_runtime": 11.0362,
+      "eval_samples_per_second": 30.445,
+      "eval_steps_per_second": 0.544,
+      "step": 1750
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.028393927961587906,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.0009,
+      "step": 1875
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.9226190476190477,
+      "eval_loss": 0.3424765169620514,
+      "eval_runtime": 10.9571,
+      "eval_samples_per_second": 30.665,
+      "eval_steps_per_second": 0.548,
+      "step": 1875
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.018229400739073753,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.0009,
+      "step": 2000
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.9226190476190477,
+      "eval_loss": 0.3481498956680298,
+      "eval_runtime": 11.0764,
+      "eval_samples_per_second": 30.335,
+      "eval_steps_per_second": 0.542,
+      "step": 2000
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.018818996846675873,
+      "learning_rate": 7.499999999999999e-06,
+      "loss": 0.0007,
+      "step": 2125
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.9226190476190477,
+      "eval_loss": 0.35712891817092896,
+      "eval_runtime": 10.3576,
+      "eval_samples_per_second": 32.44,
+      "eval_steps_per_second": 0.579,
+      "step": 2125
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.021823769435286522,
+      "learning_rate": 4.131759111665352e-06,
+      "loss": 0.0006,
+      "step": 2250
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.9196428571428571,
+      "eval_loss": 0.36197924613952637,
+      "eval_runtime": 11.0713,
+      "eval_samples_per_second": 30.349,
+      "eval_steps_per_second": 0.542,
+      "step": 2250
+    }
+  ],
   "logging_steps": 500,
-  "max_steps": 2500,
+  "max_steps": 1875,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 20,
+  "num_train_epochs": 15,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -19,13 +308,13 @@
         "should_epoch_stop": false,
         "should_evaluate": false,
         "should_log": false,
-        "should_save": false,
+        "should_save": true,
         "should_training_stop": false
       },
       "attributes": {}
     }
   },
-  "total_flos": 0,
+  "total_flos": 1.1159446583771136e+19,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null