arabic-eou-model-v1

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02-twitter on the xmjo/arabic-eou-dataset-v1 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4713
  • Accuracy: 0.7922
  • Precision: 0.7790
  • Recall: 0.8378
  • F1: 0.8073

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: 1.5e-05
  • train_batch_size: 128
  • eval_batch_size: 128
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 3
  • label_smoothing_factor: 0.015

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1
No log 0.4651 40 0.5364 0.7518 0.7183 0.8597 0.7827
No log 0.9302 80 0.4941 0.7709 0.7511 0.8364 0.7915
No log 1.3953 120 0.4806 0.7793 0.7479 0.8681 0.8035
No log 1.8605 160 0.4746 0.7841 0.7573 0.8604 0.8055
No log 2.3256 200 0.4721 0.7878 0.7681 0.8477 0.8059
No log 2.7907 240 0.4713 0.7922 0.7790 0.8378 0.8073

Framework versions

  • Transformers 4.55.4
  • Pytorch 2.9.1+cu128
  • Datasets 4.4.1
  • Tokenizers 0.21.4
Downloads last month
-
Safetensors
Model size
0.1B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Space using xmjo/arabic-eou-model-v1 1