mt5-base-finetuned-ar-to-en

This model is a fine-tuned version of google/mt5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: nan
• Bleu: 0.0111
• Gen Len: 6.732

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
163.1788	1.0	816	nan	0.0111	6.732
1.1396	2.0	1632	nan	0.0111	6.732
0.0381	3.0	2448	nan	0.0111	6.732
0.0	4.0	3264	nan	0.0111	6.732
155.5697	5.0	4080	nan	0.0111	6.732
74.9948	6.0	4896	nan	0.0111	6.732
0.116	6.13	5000	nan	0.0111	6.732

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1