🍻 cheers

README.md

@@ -18,8 +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-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the beans dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0395
-- Accuracy: 0.9850
+- Loss: 0.0442
+- Accuracy: 0.9925
 
 ## Model description
 
@@ -51,8 +51,8 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss | Accuracy |
 |:-------------:|:------:|:----:|:---------------:|:--------:|
-| 0.0879        | 1.5385 | 100  | 0.0950          | 0.9774   |
-| 0.0243        | 3.0769 | 200  | 0.0395          | 0.9850   |
+| 0.059         | 1.5385 | 100  | 0.0442          | 0.9925   |
+| 0.0359        | 3.0769 | 200  | 0.0564          | 0.9850   |
 
 
 ### Framework versions

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 0.9849624060150376,
-    "eval_loss": 0.039455611258745193,
-    "eval_runtime": 1.3415,
-    "eval_samples_per_second": 99.143,
-    "eval_steps_per_second": 12.672,
+    "eval_accuracy": 0.9924812030075187,
+    "eval_loss": 0.04420160874724388,
+    "eval_runtime": 1.4571,
+    "eval_samples_per_second": 91.278,
+    "eval_steps_per_second": 11.667,
     "total_flos": 3.205097416476426e+17,
-    "train_loss": 0.12025929483083578,
-    "train_runtime": 94.3031,
-    "train_samples_per_second": 43.859,
-    "train_steps_per_second": 2.757
+    "train_loss": 0.10883640790214906,
+    "train_runtime": 98.1266,
+    "train_samples_per_second": 42.15,
+    "train_steps_per_second": 2.65
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
-    "eval_accuracy": 0.9849624060150376,
-    "eval_loss": 0.039455611258745193,
-    "eval_runtime": 1.3415,
-    "eval_samples_per_second": 99.143,
-    "eval_steps_per_second": 12.672
+    "eval_accuracy": 0.9924812030075187,
+    "eval_loss": 0.04420160874724388,
+    "eval_runtime": 1.4571,
+    "eval_samples_per_second": 91.278,
+    "eval_steps_per_second": 11.667
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:da23118451465ef10341a98aa766e4137db222474ee4f6002366b8a02de2d1d4
+oid sha256:7ae565612793076afedc56601e30174a8547c043284cacc77c3a5a48a76492b8
 size 343227052

runs/Jun28_08-52-04_98d0e026ed58/events.out.tfevents.1719564731.98d0e026ed58.442.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fe243b730cce285a21db431e637e91c5ad670f6cc2fdbba763a41ae725d42b1
+size 11323

runs/Jun28_08-52-04_98d0e026ed58/events.out.tfevents.1719564869.98d0e026ed58.442.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68c8e585624332599748c0f2dbba7ce38570c9f66161a07bb411404d0538c3a9
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
     "total_flos": 3.205097416476426e+17,
-    "train_loss": 0.12025929483083578,
-    "train_runtime": 94.3031,
-    "train_samples_per_second": 43.859,
-    "train_steps_per_second": 2.757
+    "train_loss": 0.10883640790214906,
+    "train_runtime": 98.1266,
+    "train_samples_per_second": 42.15,
+    "train_steps_per_second": 2.65
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.039455611258745193,
-  "best_model_checkpoint": "./vit-base-beans-demo-v5/checkpoint-200",
+  "best_metric": 0.04420160874724388,
+  "best_model_checkpoint": "./vit-base-beans-demo-v5/checkpoint-100",
   "epoch": 4.0,
   "eval_steps": 100,
   "global_step": 260,
@@ -10,212 +10,212 @@
   "log_history": [
     {
       "epoch": 0.15384615384615385,
-      "grad_norm": 1.5127050876617432,
+      "grad_norm": 1.251067042350769,
       "learning_rate": 0.00019230769230769233,
-      "loss": 0.832,
+      "loss": 0.7455,
       "step": 10
     },
     {
       "epoch": 0.3076923076923077,
-      "grad_norm": 2.9642956256866455,
+      "grad_norm": 2.8409852981567383,
       "learning_rate": 0.00018461538461538463,
-      "loss": 0.3281,
+      "loss": 0.2967,
       "step": 20
     },
     {
       "epoch": 0.46153846153846156,
-      "grad_norm": 1.479327917098999,
+      "grad_norm": 2.9720468521118164,
       "learning_rate": 0.00017692307692307693,
-      "loss": 0.1852,
+      "loss": 0.1919,
       "step": 30
     },
     {
       "epoch": 0.6153846153846154,
-      "grad_norm": 4.251507759094238,
+      "grad_norm": 0.5581620335578918,
       "learning_rate": 0.00016923076923076923,
-      "loss": 0.2552,
+      "loss": 0.2157,
       "step": 40
     },
     {
       "epoch": 0.7692307692307693,
-      "grad_norm": 2.2233493328094482,
+      "grad_norm": 1.9146449565887451,
       "learning_rate": 0.00016153846153846155,
-      "loss": 0.173,
+      "loss": 0.206,
       "step": 50
     },
     {
       "epoch": 0.9230769230769231,
-      "grad_norm": 1.1962614059448242,
+      "grad_norm": 0.5213171243667603,
       "learning_rate": 0.00015384615384615385,
-      "loss": 0.2716,
+      "loss": 0.2305,
       "step": 60
     },
     {
       "epoch": 1.0769230769230769,
-      "grad_norm": 2.2528934478759766,
+      "grad_norm": 0.17810313403606415,
       "learning_rate": 0.00014615384615384615,
-      "loss": 0.3024,
+      "loss": 0.1452,
       "step": 70
     },
     {
       "epoch": 1.2307692307692308,
-      "grad_norm": 0.2029009759426117,
+      "grad_norm": 0.20280644297599792,
       "learning_rate": 0.00013846153846153847,
-      "loss": 0.1413,
+      "loss": 0.0821,
       "step": 80
     },
     {
       "epoch": 1.3846153846153846,
-      "grad_norm": 0.16266554594039917,
+      "grad_norm": 0.3441774845123291,
       "learning_rate": 0.00013076923076923077,
-      "loss": 0.0459,
+      "loss": 0.0305,
       "step": 90
     },
     {
       "epoch": 1.5384615384615383,
-      "grad_norm": 3.4238033294677734,
+      "grad_norm": 4.757013320922852,
       "learning_rate": 0.0001230769230769231,
-      "loss": 0.0879,
+      "loss": 0.059,
       "step": 100
     },
     {
       "epoch": 1.5384615384615383,
-      "eval_accuracy": 0.9774436090225563,
-      "eval_loss": 0.09501739591360092,
-      "eval_runtime": 1.6492,
-      "eval_samples_per_second": 80.643,
-      "eval_steps_per_second": 10.308,
+      "eval_accuracy": 0.9924812030075187,
+      "eval_loss": 0.04420160874724388,
+      "eval_runtime": 1.2347,
+      "eval_samples_per_second": 107.718,
+      "eval_steps_per_second": 13.768,
       "step": 100
     },
     {
       "epoch": 1.6923076923076923,
-      "grad_norm": 4.021958827972412,
+      "grad_norm": 0.09952733665704727,
       "learning_rate": 0.00011538461538461538,
-      "loss": 0.0663,
+      "loss": 0.0634,
       "step": 110
     },
     {
       "epoch": 1.8461538461538463,
-      "grad_norm": 0.8424084186553955,
+      "grad_norm": 0.8956493735313416,
       "learning_rate": 0.0001076923076923077,
-      "loss": 0.0762,
+      "loss": 0.1432,
       "step": 120
     },
     {
       "epoch": 2.0,
-      "grad_norm": 0.11821229010820389,
+      "grad_norm": 0.09242899715900421,
       "learning_rate": 0.0001,
-      "loss": 0.068,
+      "loss": 0.0325,
       "step": 130
     },
     {
       "epoch": 2.1538461538461537,
-      "grad_norm": 0.11869572848081589,
+      "grad_norm": 6.364304542541504,
       "learning_rate": 9.230769230769232e-05,
-      "loss": 0.0463,
+      "loss": 0.1169,
       "step": 140
     },
     {
       "epoch": 2.3076923076923075,
-      "grad_norm": 5.829068660736084,
+      "grad_norm": 4.19351863861084,
       "learning_rate": 8.461538461538461e-05,
-      "loss": 0.0358,
+      "loss": 0.0268,
       "step": 150
     },
     {
       "epoch": 2.4615384615384617,
-      "grad_norm": 0.09531725943088531,
+      "grad_norm": 0.12382465600967407,
       "learning_rate": 7.692307692307693e-05,
-      "loss": 0.0459,
+      "loss": 0.0341,
       "step": 160
     },
     {
       "epoch": 2.6153846153846154,
-      "grad_norm": 0.13866978883743286,
+      "grad_norm": 0.09067798405885696,
       "learning_rate": 6.923076923076924e-05,
-      "loss": 0.0309,
+      "loss": 0.019,
       "step": 170
     },
     {
       "epoch": 2.769230769230769,
-      "grad_norm": 0.18827354907989502,
+      "grad_norm": 3.0015292167663574,
       "learning_rate": 6.153846153846155e-05,
-      "loss": 0.0156,
+      "loss": 0.0641,
       "step": 180
     },
     {
       "epoch": 2.9230769230769234,
-      "grad_norm": 0.07095403224229813,
+      "grad_norm": 0.06810443103313446,
       "learning_rate": 5.384615384615385e-05,
-      "loss": 0.0169,
+      "loss": 0.0138,
       "step": 190
     },
     {
       "epoch": 3.076923076923077,
-      "grad_norm": 0.06902670115232468,
+      "grad_norm": 0.07230933755636215,
       "learning_rate": 4.615384615384616e-05,
-      "loss": 0.0243,
+      "loss": 0.0359,
       "step": 200
     },
     {
       "epoch": 3.076923076923077,
       "eval_accuracy": 0.9849624060150376,
-      "eval_loss": 0.039455611258745193,
-      "eval_runtime": 1.1868,
-      "eval_samples_per_second": 112.071,
-      "eval_steps_per_second": 14.325,
+      "eval_loss": 0.05636923760175705,
+      "eval_runtime": 1.2039,
+      "eval_samples_per_second": 110.472,
+      "eval_steps_per_second": 14.12,
       "step": 200
     },
     {
       "epoch": 3.230769230769231,
-      "grad_norm": 0.6385956406593323,
+      "grad_norm": 0.07314398884773254,
       "learning_rate": 3.846153846153846e-05,
-      "loss": 0.0133,
+      "loss": 0.0143,
       "step": 210
     },
     {
       "epoch": 3.3846153846153846,
-      "grad_norm": 0.0793854370713234,
+      "grad_norm": 0.0658462792634964,
       "learning_rate": 3.0769230769230774e-05,
-      "loss": 0.0124,
+      "loss": 0.0122,
       "step": 220
     },
     {
       "epoch": 3.5384615384615383,
-      "grad_norm": 0.06871400028467178,
+      "grad_norm": 0.08664824813604355,
       "learning_rate": 2.307692307692308e-05,
-      "loss": 0.0122,
+      "loss": 0.0152,
       "step": 230
     },
     {
       "epoch": 3.6923076923076925,
-      "grad_norm": 0.10652589797973633,
+      "grad_norm": 0.06487639248371124,
       "learning_rate": 1.5384615384615387e-05,
-      "loss": 0.0124,
+      "loss": 0.0117,
       "step": 240
     },
     {
       "epoch": 3.8461538461538463,
-      "grad_norm": 0.059241101145744324,
+      "grad_norm": 0.06605548411607742,
       "learning_rate": 7.692307692307694e-06,
-      "loss": 0.0118,
+      "loss": 0.0114,
       "step": 250
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.1098334938287735,
+      "grad_norm": 0.05876839533448219,
       "learning_rate": 0.0,
-      "loss": 0.0156,
+      "loss": 0.0122,
       "step": 260
     },
     {
       "epoch": 4.0,
       "step": 260,
       "total_flos": 3.205097416476426e+17,
-      "train_loss": 0.12025929483083578,
-      "train_runtime": 94.3031,
-      "train_samples_per_second": 43.859,
-      "train_steps_per_second": 2.757
+      "train_loss": 0.10883640790214906,
+      "train_runtime": 98.1266,
+      "train_samples_per_second": 42.15,
+      "train_steps_per_second": 2.65
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8baf81fe96f98069659da52bd60eddef3e6c99293239ad3b6032cfe3637bc378
+oid sha256:ca3afcee552df97ba46efe95b3759aa31467154c2d8d1354970d418da0d0e997
 size 5112