Upload TFLayoutLMForTokenClassification

README.md

@@ -1,8 +1,8 @@
 ---
 license: mit
+base_model: cor-c/layoutlm-funsd-tf
 tags:
 - generated_from_keras_callback
-base_model: microsoft/layoutlm-base-uncased
 model-index:
 - name: layoutlm-funsd-tf
   results: []
@@ -13,10 +13,14 @@ probably proofread and complete it, then remove this comment. -->
 
 # layoutlm-funsd-tf
 
-This model is a fine-tuned version of [microsoft/layoutlm-base-uncased](https://huggingface.co/microsoft/layoutlm-base-uncased) on an unknown dataset.
+This model is a fine-tuned version of [cor-c/layoutlm-funsd-tf](https://huggingface.co/cor-c/layoutlm-funsd-tf) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Train Loss: 0.2352
-- Validation Loss: 0.6944
+- Train Loss: 0.0632
+- Validation Loss: 0.8795
+- Train Overall Precision: 0.7424
+- Train Overall Recall: 0.8038
+- Train Overall F1: 0.7719
+- Train Overall Accuracy: 0.8103
 - Epoch: 7
 
 ## Model description
@@ -36,21 +40,21 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': 3e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}
+- optimizer: {'inner_optimizer': {'module': 'transformers.optimization_tf', 'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': 2.9999999242136255e-05, 'decay': 0.0, 'beta_1': 0.8999999761581421, 'beta_2': 0.9990000128746033, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}, 'registered_name': 'AdamWeightDecay'}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
 - training_precision: mixed_float16
 
 ### Training results
 
-| Train Loss | Validation Loss | Epoch |
-|:----------:|:---------------:|:-----:|
-| 1.6827     | 1.3457          | 0     |
-| 1.1110     | 0.8420          | 1     |
-| 0.7425     | 0.6871          | 2     |
-| 0.5572     | 0.6354          | 3     |
-| 0.4218     | 0.6126          | 4     |
-| 0.3556     | 0.6430          | 5     |
-| 0.2905     | 0.6746          | 6     |
-| 0.2352     | 0.6944          | 7     |
+| Train Loss | Validation Loss | Train Overall Precision | Train Overall Recall | Train Overall F1 | Train Overall Accuracy | Epoch |
+|:----------:|:---------------:|:-----------------------:|:--------------------:|:----------------:|:----------------------:|:-----:|
+| 0.2119     | 0.7340          | 0.7292                  | 0.8053               | 0.7654           | 0.8046                 | 0     |
+| 0.1948     | 0.7521          | 0.7406                  | 0.7963               | 0.7674           | 0.8027                 | 1     |
+| 0.1485     | 0.7879          | 0.7256                  | 0.7988               | 0.7604           | 0.8019                 | 2     |
+| 0.1220     | 0.7861          | 0.7403                  | 0.7983               | 0.7682           | 0.8073                 | 3     |
+| 0.1003     | 0.8253          | 0.7495                  | 0.8018               | 0.7748           | 0.8087                 | 4     |
+| 0.0825     | 0.8617          | 0.7491                  | 0.7968               | 0.7722           | 0.8048                 | 5     |
+| 0.0676     | 0.8938          | 0.7503                  | 0.8128               | 0.7803           | 0.8062                 | 6     |
+| 0.0632     | 0.8795          | 0.7424                  | 0.8038               | 0.7719           | 0.8103                 | 7     |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "microsoft/layoutlm-base-uncased",
+  "_name_or_path": "cor-c/layoutlm-funsd-tf",
   "architectures": [
     "LayoutLMForTokenClassification"
   ],

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:18c9b880fe507a2f107a170cb5587b4a2c8f71ef5b308cd81ac56c3f6900df6d
+oid sha256:2f3f596b3bb043bd8f54859085f4038520a75fa50b18221889ee050a769e25ad
 size 450829256