mt5-small-finetuned

This model is a fine-tuned version of google/mt5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.3994
• Rouge1: 20.69
• Rouge2: 6.09
• Rougel: 20.15

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: 0.0001
• train_batch_size: 9
• eval_batch_size: 9
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel
4.7204	1.45	500	2.6053	16.93	4.91	16.71
3.1289	2.9	1000	2.4878	18.05	5.26	17.79
2.8862	4.35	1500	2.4109	17.45	5.06	17.04
2.7669	5.8	2000	2.4006	18.61	5.28	18.12
2.6433	7.25	2500	2.4017	18.81	5.67	18.5
2.5514	8.7	3000	2.3917	19.5	5.88	19.09
2.4947	10.14	3500	2.3994	20.69	6.09	20.15
2.3995	11.59	4000	2.3608	20.2	6.51	19.67
2.3798	13.04	4500	2.3251	20.1	6.25	19.71
2.3029	14.49	5000	2.3387	19.75	6.11	19.37
2.2563	15.94	5500	2.3372	20.28	6.32	19.74
2.2109	17.39	6000	2.3410	20.67	6.38	20.13

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2