0a387033a4d88069da63a9ee6eac827c

This model is a fine-tuned version of google-t5/t5-3b on the Helsinki-NLP/opus_books [de-es] dataset. It achieves the following results on the evaluation set:

• Loss: 1.3979
• Data Size: 1.0
• Epoch Runtime: 315.0832
• Bleu: 7.9170

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Bleu
No log	0	0	2.6860	0	20.6344	1.3640
No log	1	688	2.1519	0.0078	23.8429	5.2944
No log	2	1376	1.9147	0.0156	31.9751	6.5657
No log	3	2064	1.8024	0.0312	42.1712	7.0616
0.0736	4	2752	1.7196	0.0625	53.3232	7.6174
0.1468	5	3440	1.6418	0.125	68.0928	8.4165
1.722	6	4128	1.5686	0.25	98.5165	8.2005
1.5958	7	4816	1.4875	0.5	170.8546	9.0824
1.4745	8.0	5504	1.4052	1.0	319.8372	8.5661
1.3383	9.0	6192	1.3645	1.0	323.0955	8.4099
1.2184	10.0	6880	1.3487	1.0	329.5231	8.2980
1.1016	11.0	7568	1.3498	1.0	326.4803	8.4330
1.0111	12.0	8256	1.3560	1.0	316.7683	8.5156
0.9364	13.0	8944	1.3770	1.0	312.7931	8.0993
0.8706	14.0	9632	1.3979	1.0	315.0832	7.9170

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1