20a3a431ad0fd5416dbfc44dba8f6a27

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

• Loss: 1.4599
• Data Size: 1.0
• Epoch Runtime: 183.0574
• Bleu: 7.8718

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	2.7828	0	11.5270	1.3929
No log	1	390	2.3682	0.0078	13.9453	2.5667
No log	2	780	2.2389	0.0156	20.9544	3.3957
No log	3	1170	2.1298	0.0312	25.2727	3.7033
No log	4	1560	2.0076	0.0625	31.7153	4.4871
0.1265	5	1950	1.8912	0.125	43.8931	4.6747
0.3017	6	2340	1.7544	0.25	65.0467	5.5113
1.7994	7	2730	1.6127	0.5	107.6365	6.3217
1.5789	8.0	3120	1.4697	1.0	190.0869	7.1668
1.3958	9.0	3510	1.4051	1.0	189.1154	7.3471
1.2664	10.0	3900	1.3679	1.0	183.5182	7.6866
1.1608	11.0	4290	1.3542	1.0	183.0621	7.8722
1.0498	12.0	4680	1.3473	1.0	180.0907	7.9511
0.975	13.0	5070	1.3466	1.0	183.5270	7.9492
0.896	14.0	5460	1.3628	1.0	179.8843	7.9660
0.7922	15.0	5850	1.4089	1.0	182.2905	8.0350
0.7378	16.0	6240	1.4297	1.0	191.0450	7.9710
0.6759	17.0	6630	1.4599	1.0	183.0574	7.8718

Framework versions

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