acf9a452775ce9271aa747df63fc2ec8

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

• Loss: 1.5966
• Data Size: 1.0
• Epoch Runtime: 224.3470
• Bleu: 6.6874

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	3.2743	0	16.1680	0.1736
No log	1	367	2.4480	0.0078	18.2738	1.5991
No log	2	734	2.3301	0.0156	23.0401	2.4161
No log	3	1101	2.2460	0.0312	28.6740	2.7313
No log	4	1468	2.1495	0.0625	38.5878	3.2180
0.1411	5	1835	2.0702	0.125	52.8776	3.0964
2.4192	6	2202	1.9694	0.25	75.1343	3.4377
2.2196	7	2569	1.8563	0.5	128.0018	3.9203
1.9811	8.0	2936	1.7373	1.0	226.8562	4.8269
1.7948	9.0	3303	1.6520	1.0	223.0565	5.2420
1.6481	10.0	3670	1.6027	1.0	221.8362	5.6566
1.5408	11.0	4037	1.5787	1.0	224.3946	5.8434
1.4469	12.0	4404	1.5492	1.0	223.8401	6.0809
1.3758	13.0	4771	1.5371	1.0	225.0378	6.2751
1.2637	14.0	5138	1.5275	1.0	223.8217	6.3437
1.1992	15.0	5505	1.5255	1.0	223.4315	6.4837
1.1194	16.0	5872	1.5503	1.0	222.3271	6.5577
1.0565	17.0	6239	1.5695	1.0	223.3507	6.5637
0.987	18.0	6606	1.5856	1.0	223.3244	6.6690
0.9318	19.0	6973	1.5966	1.0	224.3470	6.6874

Framework versions

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