SeyedAli's picture
Update README.md
777b1f7
metadata
base_model: HooshvareLab/bert-base-parsbert-uncased
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - accuracy
model-index:
  - name: output
    results: []
language:
  - fa

Persian Text Emotion Detection

This model is a fine-tuned version of HooshvareLab/bert-base-parsbert-uncased on a custom dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2551
  • Precision: 0.9362
  • Recall: 0.9360
  • Fscore: 0.9359
  • Accuracy: 0.9360

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: 10

Training results

Training Loss Epoch Step Validation Loss Precision Recall Fscore Accuracy
No log 1.0 348 0.3054 0.9166 0.9144 0.9136 0.9144
0.5158 2.0 696 0.2551 0.9362 0.9360 0.9359 0.9360
0.1469 3.0 1044 0.3670 0.9283 0.9259 0.9245 0.9259
0.1469 4.0 1392 0.3833 0.9331 0.9317 0.9307 0.9317
0.0453 5.0 1740 0.4241 0.9356 0.9345 0.9342 0.9345
0.0237 6.0 2088 0.3750 0.9441 0.9439 0.9437 0.9439
0.0237 7.0 2436 0.3986 0.9389 0.9388 0.9385 0.9388
0.009 8.0 2784 0.4100 0.9407 0.9403 0.9397 0.9403
0.0053 9.0 3132 0.4005 0.9403 0.9403 0.9401 0.9403
0.0053 10.0 3480 0.3986 0.9410 0.9410 0.9408 0.9410

Framework versions

  • Transformers 4.33.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.5
  • Tokenizers 0.13.3