56db742cbb869d57e899fa6d448778f7

This model is a fine-tuned version of google/mt5-large on the Helsinki-NLP/opus_books [es-nl] dataset. It achieves the following results on the evaluation set:

• Loss: 1.6348
• Data Size: 1.0
• Epoch Runtime: 326.7068
• Bleu: 8.7659

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	26.5351	0	24.4325	0.0114
No log	1	806	25.8299	0.0078	27.2925	0.0109
No log	2	1612	22.7544	0.0156	32.5647	0.0132
No log	3	2418	17.6702	0.0312	39.0016	0.0214
0.6336	4	3224	4.5612	0.0625	48.1287	0.0732
3.8694	5	4030	2.5615	0.125	68.4718	0.6566
2.9035	6	4836	2.2192	0.25	106.9074	4.5923
2.5495	7	5642	2.0284	0.5	179.4702	5.4765
2.2949	8.0	6448	1.8873	1.0	325.1696	6.3413
2.132	9.0	7254	1.8147	1.0	321.5022	6.7938
2.015	10.0	8060	1.7655	1.0	322.6876	7.2142
1.946	11.0	8866	1.7264	1.0	333.2729	7.4862
1.8469	12.0	9672	1.7044	1.0	336.3128	7.7866
1.7509	13.0	10478	1.6751	1.0	326.8451	7.9545
1.711	14.0	11284	1.6633	1.0	322.8929	8.1829
1.6198	15.0	12090	1.6473	1.0	321.3832	8.2777
1.5941	16.0	12896	1.6367	1.0	329.6941	8.3974
1.5335	17.0	13702	1.6337	1.0	331.9570	8.4875
1.4438	18.0	14508	1.6348	1.0	340.2806	8.5504
1.4625	19.0	15314	1.6388	1.0	326.7978	8.6558
1.3825	20.0	16120	1.6359	1.0	331.6755	8.6934
1.3631	21.0	16926	1.6348	1.0	326.7068	8.7659

Framework versions

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