c35a0c5546f057046f21c4fef3ac43ce

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

• Loss: 1.6160
• Data Size: 1.0
• Epoch Runtime: 62.2285
• Bleu: 6.4307

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	3.4307	0	4.3675	0.3637
No log	1	87	2.8840	0.0078	5.0694	0.8080
No log	2	174	2.5492	0.0156	10.9795	1.7591
No log	3	261	2.4002	0.0312	17.3669	2.1632
No log	4	348	2.2924	0.0625	24.7733	2.0369
0.1289	5	435	2.1789	0.125	25.2324	1.9628
0.64	6	522	2.0423	0.25	29.4944	2.2622
0.8468	7	609	1.8889	0.5	44.7825	2.9427
1.2505	8.0	696	1.7489	1.0	73.0264	3.7961
1.8788	9.0	783	1.6637	1.0	61.2944	4.3343
1.7018	10.0	870	1.6046	1.0	58.4812	4.6722
1.5517	11.0	957	1.5724	1.0	58.7256	5.2372
1.4407	12.0	1044	1.5539	1.0	62.1892	5.4603
1.302	13.0	1131	1.5568	1.0	58.5583	5.7801
1.2045	14.0	1218	1.5325	1.0	59.9434	5.9372
1.0864	15.0	1305	1.5407	1.0	60.4190	6.0244
1.0056	16.0	1392	1.5531	1.0	58.5635	6.1148
0.9228	17.0	1479	1.5869	1.0	59.6514	6.2292
0.8397	18.0	1566	1.6160	1.0	62.2285	6.4307

Framework versions

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