7198e9b30dbf83cff937fbc97da502d3

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

• Loss: 1.9199
• Data Size: 1.0
• Epoch Runtime: 302.2049
• Bleu: 5.7067

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	24.4950	0	21.9985	0.0125
No log	1	721	18.1813	0.0078	24.6922	0.0148
No log	2	1442	15.2718	0.0156	29.5449	0.0154
0.3753	3	2163	7.1474	0.0312	35.2999	0.0206
0.7372	4	2884	3.8456	0.0625	44.7050	0.0775
3.8083	5	3605	2.5613	0.125	64.8942	0.4858
2.8626	6	4326	2.2701	0.25	96.1871	3.6801
2.6322	7	5047	2.1435	0.5	163.1007	3.9694
2.4161	8.0	5768	2.0519	1.0	297.9778	4.7795
2.2607	9.0	6489	2.0093	1.0	295.5396	4.9885
2.188	10.0	7210	1.9728	1.0	293.4393	5.4166
2.0695	11.0	7931	1.9473	1.0	293.8989	5.4776
2.0372	12.0	8652	1.9321	1.0	295.2344	5.5985
1.9585	13.0	9373	1.9260	1.0	294.9270	5.5794
1.8729	14.0	10094	1.9120	1.0	292.9617	5.4891
1.8454	15.0	10815	1.9037	1.0	294.3007	5.5290
1.7838	16.0	11536	1.9053	1.0	297.3617	5.6152
1.7403	17.0	12257	1.9003	1.0	293.2119	5.6926
1.6988	18.0	12978	1.9057	1.0	294.2738	5.7481
1.6392	19.0	13699	1.9160	1.0	298.1034	5.7223
1.6021	20.0	14420	1.9116	1.0	292.8241	5.7237
1.5743	21.0	15141	1.9199	1.0	302.2049	5.7067

Framework versions

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