93dab9f25502fb3dde0b0f1a467d67cc

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

• Loss: 0.9916
• Data Size: 1.0
• Epoch Runtime: 22.6504
• Bleu: 18.4079

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	2.0942	0	2.0070	4.1931
No log	1	35	2.0200	0.0078	2.6001	4.0017
No log	2	70	1.8950	0.0156	3.2940	4.4829
No log	3	105	1.7346	0.0312	5.4302	5.1878
No log	4	140	1.6203	0.0625	7.8263	6.0648
No log	5	175	1.5167	0.125	9.0388	8.2982
No log	6	210	1.4429	0.25	11.1934	9.8345
No log	7	245	1.3368	0.5	15.2762	11.3725
0.3179	8.0	280	1.1997	1.0	21.9345	12.9820
1.4161	9.0	315	1.1268	1.0	19.8059	13.4222
1.2184	10.0	350	1.0754	1.0	22.1018	14.6163
1.2184	11.0	385	1.0473	1.0	21.2029	14.3354
1.0558	12.0	420	1.0314	1.0	21.5731	15.6313
0.9325	13.0	455	1.0132	1.0	21.2095	16.5828
0.9325	14.0	490	1.0068	1.0	20.3950	16.2022
0.8349	15.0	525	0.9867	1.0	20.8965	17.0375
0.7558	16.0	560	0.9902	1.0	20.8363	17.0010
0.7558	17.0	595	0.9942	1.0	20.8230	16.6790
0.6726	18.0	630	0.9893	1.0	20.6770	16.9605
0.5971	19.0	665	0.9916	1.0	22.6504	18.4079

Framework versions

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