64e2108fc55edcc92327e54cb544ba55

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

• Loss: 1.1715
• Data Size: 1.0
• Epoch Runtime: 455.8688
• Bleu: 8.7940

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	2.6616	0	28.2970	1.7281
No log	1	966	2.2039	0.0078	32.5855	2.7291
No log	2	1932	2.0539	0.0156	43.6727	3.8064
0.0505	3	2898	1.9161	0.0312	48.5563	4.2945
2.0261	4	3864	1.7818	0.0625	63.5191	4.3479
1.8616	5	4830	1.6349	0.125	87.2131	5.1001
1.6977	6	5796	1.4992	0.25	139.6249	5.8257
1.4842	7	6762	1.3557	0.5	240.0775	6.6527
1.2894	8.0	7728	1.2250	1.0	446.9474	7.6431
1.1553	9.0	8694	1.1645	1.0	447.2167	8.0025
1.0527	10.0	9660	1.1327	1.0	454.2894	8.3299
0.9251	11.0	10626	1.1213	1.0	455.9962	8.4725
0.8492	12.0	11592	1.1249	1.0	434.8186	8.5219
0.7611	13.0	12558	1.1337	1.0	458.0648	8.6364
0.7174	14.0	13524	1.1428	1.0	457.5247	8.7487
0.6356	15.0	14490	1.1715	1.0	455.8688	8.7940

Framework versions

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