Update README.md

README.md

@@ -15,25 +15,24 @@ probably proofread and complete it, then remove this comment. -->
 
 # Sultannn/bert-base-ft-ner-xtreme-id
 
-This model is a fine-tuned version of [bert-base-multilingual-uncased](https://huggingface.co/bert-base-multilingual-uncased) on an unknown dataset.
-It achieves the following results on the evaluation set:
-- Train Loss: 0.0260
-- Validation Loss: 0.1174
-- Epoch: 3
+This model is a fine-tuned version of [bert-base-multilingual-uncased](https://huggingface.co/bert-base-multilingual-uncased) on an [xtreme PAN-X.id](https://huggingface.co/datasets/xtreme/viewer/PAN-X.id/test) for **NER** downstream task.
 
-## Model description
+## Details of the downstream task (NER) - Dataset
 
-More information needed
+| Dataset                | # Examples |
+| ---------------------- | ----- |
+| Train                  | 35 K |
+| Validation             | 5 K |
 
-## Intended uses & limitations
+## Metrics on evaluation set
 
-More information needed
+| Metrics                | Score  |
+| ---------------------- | ----- |
+| Accuracy               | **97.18** |
+| F1                     | **93.26** |
+| Precision              | **92.36** |
+| Recall                 | **94.18** |
 
-## Training and evaluation data
-
-More information needed
-
-## Training procedure
 
 ### Training hyperparameters
 
@@ -41,15 +40,6 @@ The following hyperparameters were used during training:
 - optimizer: {'inner_optimizer': {'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'WarmUp', 'config': {'initial_learning_rate': 4e-05, 'decay_schedule_fn': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 4e-05, 'decay_steps': 8248, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, '__passive_serialization__': True}, 'warmup_steps': 500, 'power': 1.0, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
 - training_precision: mixed_float16
 
-### Training results
-
-| Train Loss | Validation Loss | Epoch |
-|:----------:|:---------------:|:-----:|
-| 0.2788     | 0.1341          | 0     |
-| 0.1023     | 0.1159          | 1     |
-| 0.0549     | 0.1133          | 2     |
-| 0.0260     | 0.1174          | 3     |
-
 
 ### Framework versions
 
@@ -57,3 +47,6 @@ The following hyperparameters were used during training:
 - TensorFlow 2.8.0
 - Datasets 2.1.0
 - Tokenizers 0.12.1
+
+- [Fine-tune on NER script provided by @stefan-it](https://raw.githubusercontent.com/stefan-it/fine-tuned-berts-seq/master/scripts/preprocess.py)
+