53a17720563edbb21d3523333ee51677

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

• Loss: 2.0252
• Data Size: 1.0
• Epoch Runtime: 61.2643
• Bleu: 2.9307

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	3.7982	0	4.5605	0.1044
No log	1	91	3.3185	0.0078	5.2863	0.1612
No log	2	182	3.1639	0.0156	8.8117	0.1731
No log	3	273	3.0493	0.0312	13.1207	0.3087
No log	4	364	2.9257	0.0625	18.4668	0.4274
No log	5	455	2.8185	0.125	22.4691	0.5549
No log	6	546	2.6653	0.25	28.1361	0.9521
0.31	7	637	2.5022	0.5	40.0774	1.4670
2.6852	8.0	728	2.3300	1.0	70.2840	1.6240
2.4354	9.0	819	2.2218	1.0	58.7037	2.0172
2.265	10.0	910	2.1360	1.0	58.6665	2.0617
2.0976	11.0	1001	2.0811	1.0	61.1931	2.3377
1.9945	12.0	1092	2.0387	1.0	59.7506	2.4748
1.873	13.0	1183	2.0055	1.0	62.9365	2.6466
1.746	14.0	1274	2.0099	1.0	60.5703	2.8270
1.6196	15.0	1365	1.9820	1.0	59.5274	2.7930
1.5166	16.0	1456	1.9899	1.0	64.7478	2.8834
1.4091	17.0	1547	1.9990	1.0	60.6692	2.9923
1.2957	18.0	1638	2.0107	1.0	58.5272	2.9677
1.2042	19.0	1729	2.0252	1.0	61.2643	2.9307

Framework versions

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