54f633d38de113a9040adb5e3419b747

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

• Loss: 1.2703
• Data Size: 1.0
• Epoch Runtime: 375.4680
• Bleu: 10.7514

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	2.3337	0	23.6958	3.1125
No log	1	808	2.0149	0.0078	28.0358	4.3477
No log	2	1616	1.8919	0.0156	37.1064	7.8153
No log	3	2424	1.7965	0.0312	45.5738	9.2887
0.0614	4	3232	1.7039	0.0625	58.2102	8.3308
1.8659	5	4040	1.6046	0.125	79.0933	7.6671
1.6797	6	4848	1.4961	0.25	119.6548	8.5864
1.4894	7	5656	1.3899	0.5	204.9930	9.4981
1.3789	8.0	6464	1.2833	1.0	384.9791	9.9560
1.1927	9.0	7272	1.2345	1.0	377.5094	10.7986
1.0686	10.0	8080	1.2111	1.0	379.6636	11.0554
0.9863	11.0	8888	1.2079	1.0	377.5173	10.7092
0.8754	12.0	9696	1.2174	1.0	387.5942	10.7165
0.8075	13.0	10504	1.2298	1.0	378.0924	10.8300
0.7293	14.0	11312	1.2392	1.0	391.2719	10.7792
0.6733	15.0	12120	1.2703	1.0	375.4680	10.7514

Framework versions

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