t5-base-translation-spa-pbb

This model is a fine-tuned version of t5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.2411
• Bleu: 0.608
• Gen Len: 8.108

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
• lr_scheduler_warmup_steps: 10
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
2.6692	1.0	304	2.9825	0.8944	6.2582
2.6593	2.0	608	2.7422	0.0	6.9895
2.5452	3.0	912	2.6276	0.0	7.1924
2.5998	4.0	1216	2.5437	0.0	7.3347
3.0987	5.0	1520	2.4819	0.0	7.5204
2.3259	6.0	1824	2.4409	0.0	7.4466
3.2006	7.0	2128	2.3988	0.6694	7.4058
1.989	8.0	2432	2.3669	0.6097	8.1383
2.3702	9.0	2736	2.3464	0.9537	8.1542
2.3841	10.0	3040	2.3434	0.9045	7.7852
2.2193	11.0	3344	2.3119	0.9082	8.22
2.4414	12.0	3648	2.2997	0.791	8.2569
1.8003	13.0	3952	2.2848	1.0315	8.2055
1.9862	14.0	4256	2.2756	0.6622	8.2134
2.3814	15.0	4560	2.2678	0.6688	8.1634
2.145	16.0	4864	2.2606	0.8214	8.2754
2.1513	17.0	5168	2.2605	1.0985	8.2635
2.249	18.0	5472	2.2506	1.0695	8.1726
2.3972	19.0	5776	2.2477	0.663	8.22
2.1375	20.0	6080	2.2458	0.612	8.1515
2.4343	21.0	6384	2.2451	0.6825	8.1871
2.9682	22.0	6688	2.2361	0.6095	8.2306
1.8138	23.0	6992	2.2411	0.608	8.108

Framework versions

• Transformers 4.35.2
• Pytorch 1.13.1+cu117
• Datasets 2.15.0
• Tokenizers 0.15.0