1fb838bc28638006ccb886623eaa9890

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

• Loss: 1.3220
• Data Size: 1.0
• Epoch Runtime: 365.0267
• Bleu: 10.0696

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	2.0995	0	26.0721	4.5287
No log	1	872	1.9112	0.0078	29.1658	5.4509
No log	2	1744	1.8442	0.0156	34.2727	5.9087
0.0323	3	2616	1.7786	0.0312	39.9419	6.4910
0.1232	4	3488	1.7221	0.0625	52.6335	6.0637
1.876	5	4360	1.6597	0.125	72.9768	6.4588
1.774	6	5232	1.5870	0.25	115.1484	7.0494
1.6322	7	6104	1.5060	0.5	209.8697	7.7572
1.5164	8.0	6976	1.4202	1.0	381.3269	8.6993
1.407	9.0	7848	1.3725	1.0	371.6459	9.2602
1.3461	10.0	8720	1.3451	1.0	376.9327	9.5082
1.2717	11.0	9592	1.3261	1.0	376.4887	9.4906
1.1858	12.0	10464	1.3127	1.0	378.0147	9.6528
1.1458	13.0	11336	1.3025	1.0	370.0599	9.8180
1.067	14.0	12208	1.3077	1.0	375.3955	9.9657
1.0357	15.0	13080	1.3096	1.0	366.5659	10.0176
0.9849	16.0	13952	1.3130	1.0	378.3173	10.0986
0.9144	17.0	14824	1.3220	1.0	365.0267	10.0696

Framework versions

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