fc553a27224de2e75635f62fa2816c59

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

• Loss: 1.5436
• Data Size: 1.0
• Epoch Runtime: 155.7047
• Bleu: 9.0658

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	3.7924	0	11.2305	1.8251
No log	1	367	2.8608	0.0078	13.7472	3.8329
No log	2	734	2.2088	0.0156	19.7492	5.3682
No log	3	1101	2.1027	0.0312	26.3032	6.6280
No log	4	1468	1.9772	0.0625	32.5196	8.8348
0.1183	5	1835	1.8719	0.125	45.8259	8.8600
2.0144	6	2202	1.7722	0.25	62.9738	8.6559
1.8439	7	2569	1.6538	0.5	107.4832	8.9071
1.65	8.0	2936	1.5537	1.0	182.0428	10.8758
1.4817	9.0	3303	1.4975	1.0	177.1992	10.2254
1.3351	10.0	3670	1.4718	1.0	176.8617	9.7556
1.2168	11.0	4037	1.4645	1.0	156.7527	9.5061
1.1492	12.0	4404	1.4568	1.0	155.0943	9.5538
1.0733	13.0	4771	1.4667	1.0	155.3934	9.5295
0.9613	14.0	5138	1.4939	1.0	156.2864	9.4018
0.8947	15.0	5505	1.4940	1.0	155.0916	9.4032
0.8269	16.0	5872	1.5436	1.0	155.7047	9.0658

Framework versions

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