nllb-200-distilled-600M-finetuned_augmented_synthetic_cleaned_ar-to-en

This model is a fine-tuned version of facebook/nllb-200-distilled-600M on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7548
• Bleu: 55.0936
• Gen Len: 71.309

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: 8
• eval_batch_size: 8
• 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	Bleu	Gen Len
1.0702	1.0	2085	0.9530	45.7652	72.802
0.8945	2.0	4170	0.8525	51.7618	70.845
0.8323	3.0	6255	0.8107	52.8262	71.099
0.7552	4.0	8340	0.7867	54.4572	71.243
0.7133	5.0	10425	0.7717	55.0749	71.354
0.6844	6.0	12510	0.7648	55.5344	71.477
0.6439	7.0	14595	0.7595	55.6968	70.883
0.6338	8.0	16680	0.7546	54.7998	71.414
0.6098	9.0	18765	0.7544	54.7933	71.567
0.6066	10.0	20850	0.7548	55.0936	71.309

Framework versions

• Transformers 4.33.0
• Pytorch 1.13.1
• Datasets 2.14.4
• Tokenizers 0.13.3