Anwaarma's picture
End of training
1488c57
metadata
base_model: Ammar-alhaj-ali/arabic-MARBERT-sentiment
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: Improved-MARBERT-attempt2
    results: []

Improved-MARBERT-attempt2

This model is a fine-tuned version of Ammar-alhaj-ali/arabic-MARBERT-sentiment on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5969
  • Accuracy: 0.85

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 Accuracy
No log 0.07 50 0.3387 0.87
No log 0.14 100 0.4445 0.85
No log 0.21 150 0.6001 0.79
No log 0.27 200 1.0466 0.78
No log 0.34 250 0.5674 0.81
No log 0.41 300 0.3374 0.85
No log 0.48 350 0.7026 0.76
No log 0.55 400 0.3953 0.84
No log 0.62 450 0.2559 0.88
0.3031 0.68 500 0.2551 0.91
0.3031 0.75 550 0.5071 0.83
0.3031 0.82 600 0.5743 0.81
0.3031 0.89 650 0.3198 0.92
0.3031 0.96 700 0.3801 0.87
0.3031 1.03 750 0.3169 0.91
0.3031 1.1 800 0.4731 0.87
0.3031 1.16 850 0.4643 0.91
0.3031 1.23 900 0.5969 0.85

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.7
  • Tokenizers 0.14.1