9b119aadb047a403da687c37c5c13ece

This model is a fine-tuned version of google/long-t5-tglobal-xl on the Helsinki-NLP/opus_books [es-fi] dataset. It achieves the following results on the evaluation set:

• Loss: 2.1560
• Data Size: 1.0
• Epoch Runtime: 54.9502
• Bleu: 0.9746

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	3.7815	0	4.0760	0.0666
No log	1	83	3.3776	0.0078	4.6796	0.0690
No log	2	166	3.1356	0.0156	10.4660	0.1529
No log	3	249	3.0205	0.0312	16.8049	0.1821
0.1101	4	332	2.8977	0.0625	21.9118	0.1338
0.1101	5	415	2.7921	0.125	24.6784	0.1492
0.1101	6	498	2.6768	0.25	28.2243	0.1895
0.5116	7	581	2.5291	0.5	41.9877	0.3425
2.7389	8.0	664	2.3740	1.0	63.2199	0.6335
2.5966	9.0	747	2.2830	1.0	59.8315	0.6210
2.3991	10.0	830	2.2124	1.0	55.5377	0.7248
2.2481	11.0	913	2.1609	1.0	58.3829	0.7245
2.18	12.0	996	2.1386	1.0	56.0157	0.7657
2.0321	13.0	1079	2.1092	1.0	58.9364	0.7355
1.9481	14.0	1162	2.0960	1.0	61.8901	0.8235
1.8677	15.0	1245	2.0781	1.0	56.5170	0.8296
1.7545	16.0	1328	2.0780	1.0	55.1804	0.8630
1.6672	17.0	1411	2.1007	1.0	57.5731	0.9432
1.602	18.0	1494	2.0955	1.0	54.8193	0.8842
1.4741	19.0	1577	2.1109	1.0	57.8128	0.9085
1.4093	20.0	1660	2.1560	1.0	54.9502	0.9746

Framework versions

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