0510345fef2342e223c7542d42a0c295

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

• Loss: 1.8762
• Data Size: 1.0
• Epoch Runtime: 42.6125
• Bleu: 3.5060

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	3.2263	0	2.8764	0.5320
No log	1	75	3.1037	0.0078	3.3154	0.4920
No log	2	150	2.8440	0.0156	6.7074	0.9489
No log	3	225	2.6345	0.0312	12.5980	1.3955
No log	4	300	2.4005	0.0625	17.0841	2.1635
No log	5	375	2.3272	0.125	22.2598	2.5225
No log	6	450	2.2119	0.25	23.4510	2.3499
No log	7	525	2.0890	0.5	29.1007	2.6035
2.0697	8.0	600	1.9600	1.0	47.8344	2.8781
1.8829	9.0	675	1.9012	1.0	49.2374	2.9863
1.7315	10.0	750	1.8658	1.0	40.5932	3.1221
1.5875	11.0	825	1.8354	1.0	43.3602	3.1060
1.4564	12.0	900	1.8228	1.0	41.2780	3.4887
1.3443	13.0	975	1.8243	1.0	43.7222	3.4529
1.2526	14.0	1050	1.8467	1.0	46.8332	3.5566
1.1513	15.0	1125	1.8695	1.0	40.4732	3.4651
1.0671	16.0	1200	1.8762	1.0	42.6125	3.5060

Framework versions

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