a50e0479df31a6b42995051c03abb9ef

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

• Loss: 1.6243
• Data Size: 1.0
• Epoch Runtime: 26.5925
• Bleu: 5.2474

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	3.1227	0	1.8720	0.3497
No log	1	27	2.8097	0.0078	2.7344	0.4528
No log	2	54	2.5536	0.0156	8.0554	0.6786
No log	3	81	2.3421	0.0312	11.1991	1.2236
No log	4	108	2.1753	0.0625	15.5740	2.1698
No log	5	135	2.0890	0.125	20.8313	2.1033
No log	6	162	2.0273	0.25	27.9351	1.9696
No log	7	189	1.9091	0.5	27.0546	2.1619
0.5362	8.0	216	1.7893	1.0	31.7981	2.8972
0.5362	9.0	243	1.7106	1.0	31.8682	3.5717
2.0484	10.0	270	1.6700	1.0	34.2794	4.1177
2.0484	11.0	297	1.6465	1.0	28.6308	4.3165
1.7421	12.0	324	1.6207	1.0	28.6074	4.4859
1.5418	13.0	351	1.6082	1.0	31.8599	4.6081
1.5418	14.0	378	1.5989	1.0	27.9289	4.7485
1.3658	15.0	405	1.6066	1.0	28.3605	5.0321
1.3658	16.0	432	1.6011	1.0	31.4451	5.1330
1.2282	17.0	459	1.6088	1.0	33.0436	5.1568
1.2282	18.0	486	1.6243	1.0	26.5925	5.2474

Framework versions

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