metadata
library_name: transformers
license: apache-2.0
base_model: bert-base-uncased
tags:
- generated_from_trainer
- legal
- India
- BERT
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: legal_ai_India_ner_results
results: []
datasets:
- opennyaiorg/InLegalNER
legal_ai_India_ner_results
This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1041
- Precision: 0.7985
- Recall: 0.8638
- F1: 0.8298
- Accuracy: 0.9697
Model description
This model classifies named entities from the Dataset of India supreme court judgements.
Intended uses & limitations
More information needed
Training and evaluation data
Training and Evaluation Data is taken from the opennyaiorg/INLegalNer dataset
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 12
- eval_batch_size: 12
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.4662 | 1.0 | 917 | 0.1419 | 0.6784 | 0.7845 | 0.7276 | 0.9533 |
| 0.1146 | 2.0 | 1834 | 0.1084 | 0.7529 | 0.8508 | 0.7988 | 0.9623 |
| 0.0711 | 3.0 | 2751 | 0.1015 | 0.8031 | 0.8627 | 0.8318 | 0.9705 |
| 0.0546 | 4.0 | 3668 | 0.1041 | 0.7985 | 0.8638 | 0.8298 | 0.9697 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0