metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- wikiann
model-index:
- name: ner_marathi_bert
results: []
ner_marathi_bert
This model is a fine-tuned version of bert-base-multilingual-cased on the wikiann dataset. It achieves the following results on the evaluation set:
- Loss: 0.3606
- Overall Precision: 0.8939
- Overall Recall: 0.9030
- Overall F1: 0.8984
- Overall Accuracy: 0.9347
- Loc F1: 0.8823
- Org F1: 0.8555
- Per F1: 0.9435
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
Training results
| Training Loss | Epoch | Step | Validation Loss | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy | Loc F1 | Org F1 | Per F1 |
|---|---|---|---|---|---|---|---|---|---|---|
| 0.2961 | 3.19 | 1000 | 0.3496 | 0.8720 | 0.8841 | 0.8780 | 0.9229 | 0.8599 | 0.8210 | 0.9343 |
| 0.0613 | 6.39 | 2000 | 0.3606 | 0.8939 | 0.9030 | 0.8984 | 0.9347 | 0.8823 | 0.8555 | 0.9435 |
Framework versions
- Transformers 4.21.0
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1