masrawy-english-arabic-translator

This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-ar on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.3696

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: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.2491	1.0	1198	2.9310
2.8555	2.0	2396	2.7350
2.6079	3.0	3594	2.6380
2.5038	4.0	4792	2.5691
2.3559	5.0	5990	2.5271
2.2868	6.0	7188	2.4843
2.1985	7.0	8386	2.4646
2.1517	8.0	9584	2.4430
2.0742	9.0	10782	2.4295
2.0326	10.0	11980	2.4169
1.9601	11.0	13178	2.4055
1.9361	12.0	14376	2.4044
1.9269	13.0	15574	2.3900
1.876	14.0	16772	2.3922
1.8449	15.0	17970	2.3801
1.7911	16.0	19168	2.3793
1.7514	17.0	20366	2.3745
1.7356	18.0	21564	2.3725
1.717	19.0	22762	2.3729
1.7136	20.0	23960	2.3731
1.6919	21.0	25158	2.3699
1.6693	22.0	26356	2.3720
1.6447	23.0	27554	2.3699
1.6159	24.0	28752	2.3684
1.6053	25.0	29950	2.3705
1.6269	26.0	31148	2.3705
1.6141	27.0	32346	2.3712
1.6109	28.0	33544	2.3704
1.6066	29.0	34742	2.3699
1.5936	30.0	35940	2.3696

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.15.1