egyptian-to-fusha-9k-v1

This model is a fine-tuned version of UBC-NLP/AraT5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.3471

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 16
• 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: 15

Training results

Training Loss	Epoch	Step	Validation Loss
6.4935	1.0	458	5.5704
4.0821	2.0	916	2.9774
3.0241	3.0	1374	2.3383
2.4691	4.0	1832	2.1540
1.8886	5.0	2290	1.9994
1.5816	6.0	2748	2.0091
1.4247	7.0	3206	1.9657
1.0824	8.0	3664	2.0082
0.9926	9.0	4122	2.0718
0.9148	10.0	4580	2.1133
0.7256	11.0	5038	2.1661
0.6846	12.0	5496	2.2511
0.5831	13.0	5954	2.2889
0.5783	14.0	6412	2.3135
0.4017	15.0	6870	2.3471

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.4.1
• Tokenizers 0.22.1