augmented_t5_pictos

This model is a fine-tuned version of vgaraujov/t5-base-spanish on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2659
• Rouge1: 45.6149
• Rouge2: 30.0038
• Rougel: 44.5481
• Rougelsum: 44.5275
• Gen Len: 7.7863

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.4139	1.0	527	1.9162	36.3092	19.4412	35.3042	35.2912	6.6688
1.8071	2.0	1054	1.6680	42.3979	26.0796	41.4105	41.3967	7.0363
1.5205	3.0	1581	1.5088	43.033	27.2079	42.0521	42.0524	7.6261
1.3732	4.0	2108	1.4220	44.7554	28.8592	43.8614	43.8616	7.3504
1.2345	5.0	2635	1.3596	45.1026	29.8953	44.055	44.0191	7.5791
1.1474	6.0	3162	1.3102	44.9068	29.1063	43.7844	43.7561	7.7521
1.0654	7.0	3689	1.2925	45.5775	29.9282	44.5345	44.5368	7.6090
0.9925	8.0	4216	1.2732	45.3677	29.7696	44.3156	44.2848	7.7585
0.9616	9.0	4743	1.2663	45.6594	30.2393	44.6785	44.7119	7.8419
0.9296	10.0	5270	1.2659	45.6149	30.0038	44.5481	44.5275	7.7863

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1