End of training

README.md

@@ -1,4 +1,5 @@
 ---
+library_name: transformers
 base_model: layoutlmv3
 tags:
 - generated_from_trainer
@@ -21,16 +22,16 @@ model-index:
     metrics:
     - name: Precision
       type: precision
-      value: 0.9511823035850496
+      value: 0.9587155963302753
     - name: Recall
       type: recall
-      value: 0.968167701863354
+      value: 0.9736024844720497
     - name: F1
       type: f1
-      value: 0.9595998460946518
+      value: 0.9661016949152543
     - name: Accuracy
       type: accuracy
-      value: 0.9596774193548387
+      value: 0.9656196943972836
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -40,11 +41,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [layoutlmv3](https://huggingface.co/layoutlmv3) on the mp-02/cord dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2305
-- Precision: 0.9512
-- Recall: 0.9682
-- F1: 0.9596
-- Accuracy: 0.9597
+- Loss: 0.2067
+- Precision: 0.9587
+- Recall: 0.9736
+- F1: 0.9661
+- Accuracy: 0.9656
 
 ## Model description
 
@@ -64,8 +65,8 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 10
-- eval_batch_size: 10
+- train_batch_size: 12
+- eval_batch_size: 12
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
@@ -73,23 +74,23 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
-|:-------------:|:------:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
-| No log        | 3.125  | 250  | 0.6703          | 0.8163    | 0.8626 | 0.8388 | 0.8502   |
-| 1.0387        | 6.25   | 500  | 0.3617          | 0.8935    | 0.9317 | 0.9122 | 0.9253   |
-| 1.0387        | 9.375  | 750  | 0.2860          | 0.9320    | 0.9581 | 0.9449 | 0.9423   |
-| 0.2107        | 12.5   | 1000 | 0.2416          | 0.9429    | 0.9620 | 0.9523 | 0.9525   |
-| 0.2107        | 15.625 | 1250 | 0.2120          | 0.9550    | 0.9713 | 0.9630 | 0.9631   |
-| 0.0887        | 18.75  | 1500 | 0.2211          | 0.9474    | 0.9658 | 0.9566 | 0.9588   |
-| 0.0887        | 21.875 | 1750 | 0.2304          | 0.9474    | 0.9658 | 0.9566 | 0.9576   |
-| 0.0515        | 25.0   | 2000 | 0.2344          | 0.9526    | 0.9682 | 0.9603 | 0.9584   |
-| 0.0515        | 28.125 | 2250 | 0.2300          | 0.9490    | 0.9674 | 0.9581 | 0.9601   |
-| 0.0366        | 31.25  | 2500 | 0.2305          | 0.9512    | 0.9682 | 0.9596 | 0.9597   |
+| Training Loss | Epoch   | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
+|:-------------:|:-------:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
+| No log        | 3.7313  | 250  | 0.6302          | 0.8384    | 0.8742 | 0.8559 | 0.8633   |
+| 1.0061        | 7.4627  | 500  | 0.3168          | 0.9144    | 0.9457 | 0.9298 | 0.9372   |
+| 1.0061        | 11.1940 | 750  | 0.2460          | 0.9444    | 0.9620 | 0.9531 | 0.9533   |
+| 0.1885        | 14.9254 | 1000 | 0.2109          | 0.9534    | 0.9689 | 0.9611 | 0.9622   |
+| 0.1885        | 18.6567 | 1250 | 0.2030          | 0.9571    | 0.9705 | 0.9638 | 0.9618   |
+| 0.0721        | 22.3881 | 1500 | 0.1985          | 0.9562    | 0.9666 | 0.9614 | 0.9626   |
+| 0.0721        | 26.1194 | 1750 | 0.2002          | 0.9557    | 0.9705 | 0.9630 | 0.9656   |
+| 0.04          | 29.8507 | 2000 | 0.2086          | 0.9579    | 0.9720 | 0.9649 | 0.9643   |
+| 0.04          | 33.5821 | 2250 | 0.2032          | 0.9587    | 0.9736 | 0.9661 | 0.9660   |
+| 0.0313        | 37.3134 | 2500 | 0.2067          | 0.9587    | 0.9736 | 0.9661 | 0.9656   |
 
 
 ### Framework versions
 
-- Transformers 4.42.4
+- Transformers 4.44.2
 - Pytorch 2.4.0+cu118
 - Datasets 2.21.0
 - Tokenizers 0.19.1