e4cb84029119bc7f0a41ab739916352d

This model is a fine-tuned version of google-t5/t5-3b on the Helsinki-NLP/opus_books [de-en] dataset. It achieves the following results on the evaluation set:

• Loss: 1.6326
• Data Size: 1.0
• Epoch Runtime: 593.0467
• Bleu: 12.0167

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	2.5439	0	38.2277	6.0746
No log	1	1286	1.9440	0.0078	44.1900	8.8838
0.0415	2	2572	1.8445	0.0156	52.4201	9.8276
0.0537	3	3858	1.7678	0.0312	65.9054	9.4914
0.079	4	5144	1.7053	0.0625	84.4355	10.0735
1.7414	5	6430	1.6472	0.125	123.3373	10.8341
1.672	6	7716	1.5838	0.25	182.3106	10.8763
1.5756	7	9002	1.5212	0.5	326.3003	11.2567
1.4593	8.0	10288	1.4629	1.0	585.0266	11.9885
1.284	9.0	11574	1.4518	1.0	585.1319	12.3096
1.1162	10.0	12860	1.4714	1.0	594.7481	12.2024
0.9664	11.0	14146	1.5170	1.0	605.3794	12.2927
0.9025	12.0	15432	1.5649	1.0	599.8586	12.1782
0.7602	13.0	16718	1.6326	1.0	593.0467	12.0167

Framework versions

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