metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- xtreme
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: bert-multilingual-finetuned-xtreme-tamil-ner
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: xtreme
type: xtreme
config: PAN-X.ta
split: train
args: PAN-X.ta
metrics:
- name: Precision
type: precision
value: 0.746268656716418
- name: Recall
type: recall
value: 0.819672131147541
- name: F1
type: f1
value: 0.7812500000000001
- name: Accuracy
type: accuracy
value: 0.9236328484625299
bert-multilingual-finetuned-xtreme-tamil-ner
This model is a fine-tuned version of bert-base-multilingual-uncased on the xtreme dataset. It achieves the following results on the evaluation set:
- Loss: 0.2338
- Precision: 0.7463
- Recall: 0.8197
- F1: 0.7813
- Accuracy: 0.9236
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.3899 | 1.0 | 469 | 0.2517 | 0.6893 | 0.7893 | 0.7360 | 0.9143 |
| 0.2093 | 2.0 | 938 | 0.2338 | 0.7463 | 0.8197 | 0.7813 | 0.9236 |
Framework versions
- Transformers 4.22.1
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1