🍻 cheers

README.md

@@ -3,6 +3,7 @@ library_name: transformers
 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
+- image-classification
 - generated_from_trainer
 metrics:
 - accuracy
@@ -19,13 +20,13 @@ should probably proofread and complete it, then remove this comment. -->
 
 # beans-leaf-disease-detection
 
-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: 0.0472
-- Accuracy: 0.9850
-- Precision: 0.9856
-- Recall: 0.9850
-- F1: 0.9850
+- Loss: 0.0404
+- Accuracy: 0.9925
+- Precision: 0.9926
+- Recall: 0.9925
+- F1: 0.9925
 
 ## Model description
 

all_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 1.0,
-    "eval_f1": 1.0,
-    "eval_loss": 0.014331846497952938,
-    "eval_precision": 1.0,
-    "eval_recall": 1.0,
-    "eval_runtime": 3.7971,
-    "eval_samples_per_second": 35.027,
-    "eval_steps_per_second": 4.477,
+    "eval_accuracy": 0.9924812030075187,
+    "eval_f1": 0.9924802319630927,
+    "eval_loss": 0.04042869806289673,
+    "eval_precision": 0.9926482873851294,
+    "eval_recall": 0.9924812030075187,
+    "eval_runtime": 1.4609,
+    "eval_samples_per_second": 91.039,
+    "eval_steps_per_second": 11.637,
     "total_flos": 3.205097416476426e+17,
-    "train_loss": 0.11858639149711682,
-    "train_runtime": 104.5608,
-    "train_samples_per_second": 39.556,
-    "train_steps_per_second": 2.487
+    "train_loss": 0.04095530418249277,
+    "train_runtime": 294.0083,
+    "train_samples_per_second": 14.068,
+    "train_steps_per_second": 0.884
 }

eval_results.json

@@ -1,11 +1,11 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 1.0,
-    "eval_f1": 1.0,
-    "eval_loss": 0.014331846497952938,
-    "eval_precision": 1.0,
-    "eval_recall": 1.0,
-    "eval_runtime": 3.7971,
-    "eval_samples_per_second": 35.027,
-    "eval_steps_per_second": 4.477
+    "eval_accuracy": 0.9924812030075187,
+    "eval_f1": 0.9924802319630927,
+    "eval_loss": 0.04042869806289673,
+    "eval_precision": 0.9926482873851294,
+    "eval_recall": 0.9924812030075187,
+    "eval_runtime": 1.4609,
+    "eval_samples_per_second": 91.039,
+    "eval_steps_per_second": 11.637
 }

runs/Apr08_12-13-15_8daea7f679cf/events.out.tfevents.1744114795.8daea7f679cf.15724.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7873338ee427a9cc0fb70bac41060bf2149b7af890a521615b19272e3efab3a6
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
     "total_flos": 3.205097416476426e+17,
-    "train_loss": 0.11858639149711682,
-    "train_runtime": 104.5608,
-    "train_samples_per_second": 39.556,
-    "train_steps_per_second": 2.487
+    "train_loss": 0.04095530418249277,
+    "train_runtime": 294.0083,
+    "train_samples_per_second": 14.068,
+    "train_steps_per_second": 0.884
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_global_step": 200,
-  "best_metric": 1.0,
+  "best_metric": 0.9924812030075187,
   "best_model_checkpoint": "./vit-base-beans-model-finetuned/checkpoint-200",
   "epoch": 4.0,
-  "eval_steps": 100,
+  "eval_steps": 50,
   "global_step": 260,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
@@ -11,225 +11,261 @@
   "log_history": [
     {
       "epoch": 0.15384615384615385,
-      "grad_norm": 1.2726824283599854,
+      "grad_norm": 0.046927113085985184,
       "learning_rate": 0.00019230769230769233,
-      "loss": 0.7734,
+      "loss": 0.0203,
       "step": 10
     },
     {
       "epoch": 0.3076923076923077,
-      "grad_norm": 2.146637201309204,
+      "grad_norm": 9.78127670288086,
       "learning_rate": 0.00018461538461538463,
-      "loss": 0.2647,
+      "loss": 0.0936,
       "step": 20
     },
     {
       "epoch": 0.46153846153846156,
-      "grad_norm": 1.4233285188674927,
+      "grad_norm": 0.15273386240005493,
       "learning_rate": 0.00017692307692307693,
-      "loss": 0.2683,
+      "loss": 0.1531,
       "step": 30
     },
     {
       "epoch": 0.6153846153846154,
-      "grad_norm": 0.6973394155502319,
+      "grad_norm": 0.059062909334897995,
       "learning_rate": 0.00016923076923076923,
-      "loss": 0.282,
+      "loss": 0.0808,
       "step": 40
     },
     {
       "epoch": 0.7692307692307693,
-      "grad_norm": 0.30112147331237793,
+      "grad_norm": 0.04636870697140694,
       "learning_rate": 0.00016153846153846155,
-      "loss": 0.1744,
+      "loss": 0.0492,
+      "step": 50
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "eval_accuracy": 0.9774436090225563,
+      "eval_f1": 0.9774379108058134,
+      "eval_loss": 0.1117313951253891,
+      "eval_precision": 0.9775956574956096,
+      "eval_recall": 0.9774436090225563,
+      "eval_runtime": 1.2431,
+      "eval_samples_per_second": 106.99,
+      "eval_steps_per_second": 13.675,
       "step": 50
     },
     {
       "epoch": 0.9230769230769231,
-      "grad_norm": 1.9841829538345337,
+      "grad_norm": 0.03537413850426674,
       "learning_rate": 0.00015384615384615385,
-      "loss": 0.1775,
+      "loss": 0.1339,
       "step": 60
     },
     {
       "epoch": 1.0769230769230769,
-      "grad_norm": 0.2848643660545349,
+      "grad_norm": 8.806337356567383,
       "learning_rate": 0.00014615384615384615,
-      "loss": 0.2133,
+      "loss": 0.0454,
       "step": 70
     },
     {
       "epoch": 1.2307692307692308,
-      "grad_norm": 0.1492587774991989,
+      "grad_norm": 10.037283897399902,
       "learning_rate": 0.00013846153846153847,
-      "loss": 0.1077,
+      "loss": 0.0988,
       "step": 80
     },
     {
       "epoch": 1.3846153846153846,
-      "grad_norm": 0.22872798144817352,
+      "grad_norm": 0.030385395511984825,
       "learning_rate": 0.00013076923076923077,
-      "loss": 0.0501,
+      "loss": 0.0539,
       "step": 90
     },
     {
       "epoch": 1.5384615384615383,
-      "grad_norm": 3.7855725288391113,
+      "grad_norm": 0.03275859355926514,
       "learning_rate": 0.0001230769230769231,
-      "loss": 0.1638,
+      "loss": 0.0432,
       "step": 100
     },
     {
       "epoch": 1.5384615384615383,
-      "eval_accuracy": 0.9473684210526315,
-      "eval_f1": 0.9468339901115093,
-      "eval_loss": 0.14574794471263885,
-      "eval_precision": 0.9514619883040936,
-      "eval_recall": 0.9473684210526315,
-      "eval_runtime": 2.8734,
-      "eval_samples_per_second": 46.286,
-      "eval_steps_per_second": 5.916,
+      "eval_accuracy": 0.9624060150375939,
+      "eval_f1": 0.9619558041418667,
+      "eval_loss": 0.14281876385211945,
+      "eval_precision": 0.9650282168005047,
+      "eval_recall": 0.9624060150375939,
+      "eval_runtime": 1.6279,
+      "eval_samples_per_second": 81.699,
+      "eval_steps_per_second": 10.443,
       "step": 100
     },
     {
       "epoch": 1.6923076923076923,
-      "grad_norm": 0.12178239226341248,
+      "grad_norm": 0.030820081010460854,
       "learning_rate": 0.00011538461538461538,
-      "loss": 0.1062,
+      "loss": 0.0186,
       "step": 110
     },
     {
       "epoch": 1.8461538461538463,
-      "grad_norm": 0.16127511858940125,
+      "grad_norm": 0.027867253869771957,
       "learning_rate": 0.0001076923076923077,
-      "loss": 0.082,
+      "loss": 0.0827,
       "step": 120
     },
     {
       "epoch": 2.0,
-      "grad_norm": 0.09819109737873077,
+      "grad_norm": 0.03597959503531456,
       "learning_rate": 0.0001,
-      "loss": 0.107,
+      "loss": 0.0319,
       "step": 130
     },
     {
       "epoch": 2.1538461538461537,
-      "grad_norm": 6.641988754272461,
+      "grad_norm": 0.12349524348974228,
       "learning_rate": 9.230769230769232e-05,
-      "loss": 0.0531,
+      "loss": 0.0368,
       "step": 140
     },
     {
       "epoch": 2.3076923076923075,
-      "grad_norm": 0.6329468488693237,
+      "grad_norm": 0.3923969566822052,
       "learning_rate": 8.461538461538461e-05,
-      "loss": 0.0407,
+      "loss": 0.0119,
+      "step": 150
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "eval_accuracy": 0.9849624060150376,
+      "eval_f1": 0.9849605066094567,
+      "eval_loss": 0.07355950772762299,
+      "eval_precision": 0.9851258581235698,
+      "eval_recall": 0.9849624060150376,
+      "eval_runtime": 1.2553,
+      "eval_samples_per_second": 105.955,
+      "eval_steps_per_second": 13.543,
       "step": 150
     },
     {
       "epoch": 2.4615384615384617,
-      "grad_norm": 0.08334755897521973,
+      "grad_norm": 5.8568878173828125,
       "learning_rate": 7.692307692307693e-05,
-      "loss": 0.0296,
+      "loss": 0.0635,
       "step": 160
     },
     {
       "epoch": 2.6153846153846154,
-      "grad_norm": 0.4335107207298279,
+      "grad_norm": 0.03230699151754379,
       "learning_rate": 6.923076923076924e-05,
-      "loss": 0.0151,
+      "loss": 0.0067,
       "step": 170
     },
     {
       "epoch": 2.769230769230769,
-      "grad_norm": 0.07147972285747528,
+      "grad_norm": 0.02458355203270912,
       "learning_rate": 6.153846153846155e-05,
-      "loss": 0.0445,
+      "loss": 0.0102,
       "step": 180
     },
     {
       "epoch": 2.9230769230769234,
-      "grad_norm": 1.928112268447876,
+      "grad_norm": 0.017980379983782768,
       "learning_rate": 5.384615384615385e-05,
-      "loss": 0.0447,
+      "loss": 0.0055,
       "step": 190
     },
     {
       "epoch": 3.076923076923077,
-      "grad_norm": 0.06142687052488327,
+      "grad_norm": 0.017595401033759117,
       "learning_rate": 4.615384615384616e-05,
-      "loss": 0.0156,
+      "loss": 0.0038,
       "step": 200
     },
     {
       "epoch": 3.076923076923077,
-      "eval_accuracy": 1.0,
-      "eval_f1": 1.0,
-      "eval_loss": 0.014331846497952938,
-      "eval_precision": 1.0,
-      "eval_recall": 1.0,
-      "eval_runtime": 2.2915,
-      "eval_samples_per_second": 58.041,
-      "eval_steps_per_second": 7.419,
+      "eval_accuracy": 0.9924812030075187,
+      "eval_f1": 0.9924802319630927,
+      "eval_loss": 0.04042869806289673,
+      "eval_precision": 0.9926482873851294,
+      "eval_recall": 0.9924812030075187,
+      "eval_runtime": 1.2785,
+      "eval_samples_per_second": 104.031,
+      "eval_steps_per_second": 13.297,
       "step": 200
     },
     {
       "epoch": 3.230769230769231,
-      "grad_norm": 0.06270013004541397,
+      "grad_norm": 0.016184426844120026,
       "learning_rate": 3.846153846153846e-05,
-      "loss": 0.0121,
+      "loss": 0.0036,
       "step": 210
     },
     {
       "epoch": 3.3846153846153846,
-      "grad_norm": 0.05996321141719818,
+      "grad_norm": 0.016925645992159843,
       "learning_rate": 3.0769230769230774e-05,
-      "loss": 0.0129,
+      "loss": 0.0037,
       "step": 220
     },
     {
       "epoch": 3.5384615384615383,
-      "grad_norm": 0.06372592598199844,
+      "grad_norm": 0.018014749512076378,
       "learning_rate": 2.307692307692308e-05,
-      "loss": 0.0112,
+      "loss": 0.003,
       "step": 230
     },
     {
       "epoch": 3.6923076923076925,
-      "grad_norm": 0.055972643196582794,
+      "grad_norm": 0.016459761187434196,
       "learning_rate": 1.5384615384615387e-05,
-      "loss": 0.0109,
+      "loss": 0.003,
       "step": 240
     },
     {
       "epoch": 3.8461538461538463,
-      "grad_norm": 0.05908067896962166,
+      "grad_norm": 0.015728654339909554,
       "learning_rate": 7.692307692307694e-06,
-      "loss": 0.011,
+      "loss": 0.0046,
+      "step": 250
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "eval_accuracy": 0.9849624060150376,
+      "eval_f1": 0.9850440384831379,
+      "eval_loss": 0.04719630256295204,
+      "eval_precision": 0.9856162144491664,
+      "eval_recall": 0.9849624060150376,
+      "eval_runtime": 1.2361,
+      "eval_samples_per_second": 107.601,
+      "eval_steps_per_second": 13.753,
       "step": 250
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.05994662642478943,
+      "grad_norm": 0.015585985034704208,
       "learning_rate": 0.0,
-      "loss": 0.0112,
+      "loss": 0.0029,
       "step": 260
     },
     {
       "epoch": 4.0,
       "step": 260,
       "total_flos": 3.205097416476426e+17,
-      "train_loss": 0.11858639149711682,
-      "train_runtime": 104.5608,
-      "train_samples_per_second": 39.556,
-      "train_steps_per_second": 2.487
+      "train_loss": 0.04095530418249277,
+      "train_runtime": 294.0083,
+      "train_samples_per_second": 14.068,
+      "train_steps_per_second": 0.884
     }
   ],
   "logging_steps": 10,
   "max_steps": 260,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
-  "save_steps": 100,
+  "save_steps": 50,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {