Model save

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 an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3163
-- Accuracy: 0.9293
+- Loss: 0.3535
+- Accuracy: 0.9346
 
 ## Model description
 
@@ -51,10 +51,10 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.2997        | 1.0   | 48   | 0.7613          | 0.8168   |
-| 0.1109        | 2.0   | 96   | 0.3247          | 0.9084   |
-| 0.0943        | 3.0   | 144  | 0.3684          | 0.9162   |
-| 0.0368        | 4.0   | 192  | 0.3163          | 0.9293   |
+| 0.1493        | 1.0   | 48   | 0.7434          | 0.8534   |
+| 0.1631        | 2.0   | 96   | 0.6245          | 0.8743   |
+| 0.0489        | 3.0   | 144  | 0.4446          | 0.9110   |
+| 0.0285        | 4.0   | 192  | 0.3535          | 0.9346   |
 
 
 ### Framework versions

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
     "total_flos": 4.72760450039808e+17,
-    "train_loss": 0.09667936308930318,
-    "train_runtime": 211.3464,
-    "train_samples_per_second": 28.863,
-    "train_steps_per_second": 0.908
+    "train_loss": 0.15895162567418689,
+    "train_runtime": 211.1483,
+    "train_samples_per_second": 28.89,
+    "train_steps_per_second": 0.909
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:316234cc42f7c08e247f35119eb9198b2c07c0906f0cd78fb835a067d3e63c07
+oid sha256:873e4869a48ed79b6e85ec8a79b9cafb9ba8d11205ad1863c2f17448005e5b04
 size 343267040

runs/Dec09_11-26-01_85cc5ef1c5bf/events.out.tfevents.1733744577.85cc5ef1c5bf.23.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:951b644c6ed0168bd4ee3157f0c00a0ab05056f3bfefcaaaed0d0ed22a670a2b
+size 11335

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
     "total_flos": 4.72760450039808e+17,
-    "train_loss": 0.09667936308930318,
-    "train_runtime": 211.3464,
-    "train_samples_per_second": 28.863,
-    "train_steps_per_second": 0.908
+    "train_loss": 0.15895162567418689,
+    "train_runtime": 211.1483,
+    "train_samples_per_second": 28.89,
+    "train_steps_per_second": 0.909
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.3182551860809326,
+  "best_metric": 0.31634724140167236,
   "best_model_checkpoint": "finetuned-bangladeshi-traditional-food/checkpoint-192",
   "epoch": 4.0,
   "eval_steps": 500,
@@ -10,181 +10,181 @@
   "log_history": [
     {
       "epoch": 0.20833333333333334,
-      "grad_norm": 119921.5,
+      "grad_norm": 733914.25,
       "learning_rate": 0.00018958333333333332,
-      "loss": 0.2202,
+      "loss": 0.0903,
       "step": 10
     },
     {
       "epoch": 0.4166666666666667,
-      "grad_norm": 331074.28125,
+      "grad_norm": 165444.109375,
       "learning_rate": 0.0001791666666666667,
-      "loss": 0.1853,
+      "loss": 0.1926,
       "step": 20
     },
     {
       "epoch": 0.625,
-      "grad_norm": 264337.0,
+      "grad_norm": 491501.1875,
       "learning_rate": 0.00016875,
-      "loss": 0.2269,
+      "loss": 0.498,
       "step": 30
     },
     {
       "epoch": 0.8333333333333334,
-      "grad_norm": 103972.0625,
+      "grad_norm": 181531.828125,
       "learning_rate": 0.00015833333333333332,
-      "loss": 0.176,
+      "loss": 0.2997,
       "step": 40
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8900523560209425,
-      "eval_loss": 0.4411921799182892,
-      "eval_runtime": 6.5407,
-      "eval_samples_per_second": 58.403,
-      "eval_steps_per_second": 3.669,
+      "eval_accuracy": 0.8167539267015707,
+      "eval_loss": 0.7613248825073242,
+      "eval_runtime": 6.5218,
+      "eval_samples_per_second": 58.573,
+      "eval_steps_per_second": 3.68,
       "step": 48
     },
     {
       "epoch": 1.0416666666666667,
-      "grad_norm": 7157.15185546875,
+      "grad_norm": 390060.03125,
       "learning_rate": 0.0001479166666666667,
-      "loss": 0.1162,
+      "loss": 0.3007,
       "step": 50
     },
     {
       "epoch": 1.25,
-      "grad_norm": 5605.81591796875,
+      "grad_norm": 374682.71875,
       "learning_rate": 0.0001375,
-      "loss": 0.0597,
+      "loss": 0.2778,
       "step": 60
     },
     {
       "epoch": 1.4583333333333333,
-      "grad_norm": 210989.8125,
+      "grad_norm": 104593.75,
       "learning_rate": 0.00012708333333333332,
-      "loss": 0.0551,
+      "loss": 0.2456,
       "step": 70
     },
     {
       "epoch": 1.6666666666666665,
-      "grad_norm": 338559.40625,
+      "grad_norm": 156420.5625,
       "learning_rate": 0.00011666666666666668,
-      "loss": 0.1035,
+      "loss": 0.1309,
       "step": 80
     },
     {
       "epoch": 1.875,
-      "grad_norm": 382570.28125,
+      "grad_norm": 30851.041015625,
       "learning_rate": 0.00010625000000000001,
-      "loss": 0.0726,
+      "loss": 0.1109,
       "step": 90
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9267015706806283,
-      "eval_loss": 0.34653016924858093,
-      "eval_runtime": 6.5723,
-      "eval_samples_per_second": 58.123,
-      "eval_steps_per_second": 3.652,
+      "eval_accuracy": 0.9083769633507853,
+      "eval_loss": 0.3246798515319824,
+      "eval_runtime": 6.5686,
+      "eval_samples_per_second": 58.155,
+      "eval_steps_per_second": 3.654,
       "step": 96
     },
     {
       "epoch": 2.0833333333333335,
-      "grad_norm": 2441.15869140625,
+      "grad_norm": 2772.370361328125,
       "learning_rate": 9.583333333333334e-05,
-      "loss": 0.0405,
+      "loss": 0.1209,
       "step": 100
     },
     {
       "epoch": 2.2916666666666665,
-      "grad_norm": 240823.875,
+      "grad_norm": 162168.65625,
       "learning_rate": 8.541666666666666e-05,
-      "loss": 0.0559,
+      "loss": 0.1226,
       "step": 110
     },
     {
       "epoch": 2.5,
-      "grad_norm": 384660.9375,
+      "grad_norm": 229507.6875,
       "learning_rate": 7.500000000000001e-05,
-      "loss": 0.0461,
+      "loss": 0.1504,
       "step": 120
     },
     {
       "epoch": 2.7083333333333335,
-      "grad_norm": 727094.5625,
+      "grad_norm": 226514.609375,
       "learning_rate": 6.458333333333334e-05,
-      "loss": 0.0519,
+      "loss": 0.1271,
       "step": 130
     },
     {
       "epoch": 2.9166666666666665,
-      "grad_norm": 2871.189453125,
+      "grad_norm": 7009.35693359375,
       "learning_rate": 5.4166666666666664e-05,
-      "loss": 0.0356,
+      "loss": 0.0943,
       "step": 140
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9214659685863874,
-      "eval_loss": 0.3606209456920624,
-      "eval_runtime": 6.8485,
-      "eval_samples_per_second": 55.779,
-      "eval_steps_per_second": 3.504,
+      "eval_accuracy": 0.9162303664921466,
+      "eval_loss": 0.3683938980102539,
+      "eval_runtime": 6.5149,
+      "eval_samples_per_second": 58.635,
+      "eval_steps_per_second": 3.684,
       "step": 144
     },
     {
       "epoch": 3.125,
-      "grad_norm": 1312.48974609375,
+      "grad_norm": 25972.21875,
       "learning_rate": 4.375e-05,
-      "loss": 0.0605,
+      "loss": 0.0418,
       "step": 150
     },
     {
       "epoch": 3.3333333333333335,
-      "grad_norm": 251352.375,
+      "grad_norm": 133353.5625,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.0467,
+      "loss": 0.0497,
       "step": 160
     },
     {
       "epoch": 3.5416666666666665,
-      "grad_norm": 2361.97216796875,
+      "grad_norm": 229467.96875,
       "learning_rate": 2.2916666666666667e-05,
-      "loss": 0.1208,
+      "loss": 0.0402,
       "step": 170
     },
     {
       "epoch": 3.75,
-      "grad_norm": 1161.3221435546875,
+      "grad_norm": 85779.9921875,
       "learning_rate": 1.25e-05,
-      "loss": 0.078,
+      "loss": 0.1172,
       "step": 180
     },
     {
       "epoch": 3.9583333333333335,
-      "grad_norm": 66789.09375,
+      "grad_norm": 188882.28125,
       "learning_rate": 2.0833333333333334e-06,
-      "loss": 0.0649,
+      "loss": 0.0368,
       "step": 190
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9397905759162304,
-      "eval_loss": 0.3182551860809326,
-      "eval_runtime": 6.6427,
-      "eval_samples_per_second": 57.507,
-      "eval_steps_per_second": 3.613,
+      "eval_accuracy": 0.9293193717277487,
+      "eval_loss": 0.31634724140167236,
+      "eval_runtime": 6.6575,
+      "eval_samples_per_second": 57.379,
+      "eval_steps_per_second": 3.605,
       "step": 192
     },
     {
       "epoch": 4.0,
       "step": 192,
       "total_flos": 4.72760450039808e+17,
-      "train_loss": 0.09667936308930318,
-      "train_runtime": 211.3464,
-      "train_samples_per_second": 28.863,
-      "train_steps_per_second": 0.908
+      "train_loss": 0.15895162567418689,
+      "train_runtime": 211.1483,
+      "train_samples_per_second": 28.89,
+      "train_steps_per_second": 0.909
     }
   ],
   "logging_steps": 10,