1cbf5c3f737ed311c3f2e67d5a20e75b

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

• Loss: 1.0103
• Data Size: 1.0
• Epoch Runtime: 397.3565
• Bleu: 9.4691

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	2.9704	0	26.5204	1.0602
No log	1	1000	2.2839	0.0078	29.8123	3.7535
No log	2	2000	2.0114	0.0156	37.8883	4.2523
No log	3	3000	1.8421	0.0312	46.9027	5.0741
0.0742	4	4000	1.6950	0.0625	58.6706	5.4402
1.8031	5	5000	1.5401	0.125	79.0344	5.5322
0.0979	6	6000	1.3804	0.25	123.0371	6.4839
0.1234	7	7000	1.2383	0.5	217.1385	7.1687
1.1698	8.0	8000	1.0910	1.0	397.1570	8.0882
1.0327	9.0	9000	1.0292	1.0	395.2226	8.7435
0.9093	10.0	10000	0.9886	1.0	399.2346	9.0022
0.8222	11.0	11000	0.9719	1.0	396.7249	9.1993
0.7442	12.0	12000	0.9670	1.0	396.7806	9.2106
0.6816	13.0	13000	0.9702	1.0	397.0271	9.2864
0.6165	14.0	14000	0.9716	1.0	398.6560	9.2914
0.5612	15.0	15000	0.9906	1.0	394.1021	9.4142
0.5104	16.0	16000	1.0103	1.0	397.3565	9.4691

Framework versions

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