helsinki-fr-ar-finetuned

This model is a fine-tuned version of Helsinki-NLP/opus-mt-fr-ar on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0986
• Bleu: 34.97

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: 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
• lr_scheduler_warmup_steps: 500
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu
0.3254	1.0	334	0.2252	8.69
0.2137	2.0	668	0.1577	18.3
0.1707	3.0	1002	0.1336	21.55
0.1421	4.0	1336	0.1213	21.92
0.1353	5.0	1670	0.1120	27.37
0.1179	6.0	2004	0.1070	28.74
0.1079	7.0	2338	0.1028	30.37
0.1066	8.0	2672	0.0999	32.1
0.1016	9.0	3006	0.0986	34.97
0.105	10.0	3340	0.0978	33.64

Framework versions

• Transformers 4.56.2
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1