2596874123624287b083a24be30fb73e

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

• Loss: 1.8436
• Data Size: 1.0
• Epoch Runtime: 40.6215
• Bleu: 4.1192

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	3.9880	0	3.0576	0.4132
No log	1	74	3.6240	0.0078	3.6180	0.6141
No log	2	148	2.8719	0.0156	6.8153	1.5338
0.1209	3	222	2.5144	0.0312	11.1008	2.0916
0.1209	4	296	2.3940	0.0625	15.4705	2.7176
0.1796	5	370	2.2923	0.125	20.9756	3.5691
0.1796	6	444	2.1907	0.25	25.1332	3.2682
0.4975	7	518	2.0551	0.5	29.4209	2.5545
1.3846	8.0	592	1.9280	1.0	47.4893	2.9849
1.8909	9.0	666	1.8553	1.0	48.1310	3.3759
1.7741	10.0	740	1.8038	1.0	39.2973	3.5026
1.5797	11.0	814	1.7633	1.0	43.6140	4.0478
1.4863	12.0	888	1.7477	1.0	39.8674	3.9655
1.3169	13.0	962	1.7410	1.0	42.4716	4.1988
1.2517	14.0	1036	1.7376	1.0	45.3696	4.2779
1.115	15.0	1110	1.7625	1.0	40.7063	4.1594
1.0376	16.0	1184	1.7984	1.0	42.2548	4.0825
0.9407	17.0	1258	1.8033	1.0	45.4371	4.1904
0.8875	18.0	1332	1.8436	1.0	40.6215	4.1192

Framework versions

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