ac33d969d5f4596b96be48358fb3cb01

This model is a fine-tuned version of google-t5/t5-large on the Helsinki-NLP/opus_books [en-no] dataset. It achieves the following results on the evaluation set:

• Loss: 1.7156
• Data Size: 1.0
• Epoch Runtime: 44.4268
• Bleu: 7.0834

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	3.2643	0	3.7504	0.7254
No log	1	87	3.1822	0.0078	4.9347	0.7667
No log	2	174	3.0019	0.0156	5.8045	0.8808
No log	3	261	2.9132	0.0312	7.4471	0.9530
No log	4	348	2.7458	0.0625	9.6213	1.2998
0.1289	5	435	2.6183	0.125	12.3563	1.9020
0.677	6	522	2.4657	0.25	18.6116	2.2297
0.9161	7	609	2.2862	0.5	27.3908	2.7272
1.3792	8.0	696	2.0849	1.0	44.1645	3.6993
2.1299	9.0	783	1.9653	1.0	41.7885	4.3272
1.9646	10.0	870	1.8814	1.0	41.1856	4.9248
1.8343	11.0	957	1.8326	1.0	41.0896	5.3031
1.7423	12.0	1044	1.7886	1.0	42.1944	5.5034
1.6241	13.0	1131	1.7495	1.0	41.1808	5.8722
1.5412	14.0	1218	1.7138	1.0	41.5002	6.1606
1.4457	15.0	1305	1.7102	1.0	41.8326	6.3771
1.3794	16.0	1392	1.6885	1.0	41.0939	6.5643
1.3143	17.0	1479	1.6865	1.0	41.5576	6.6450
1.2585	18.0	1566	1.6766	1.0	43.3475	6.8959
1.2121	19.0	1653	1.6929	1.0	41.0873	6.8166
1.1498	20.0	1740	1.6992	1.0	42.6784	7.0438
1.0807	21.0	1827	1.7004	1.0	43.9162	6.9865
1.0593	22.0	1914	1.7156	1.0	44.4268	7.0834

Framework versions

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