mt5-small-finetuned-31jan-4

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: 2.5165
• Rouge1: 19.31
• Rouge2: 6.34
• Rougel: 19.06
• Rougelsum: 19.09

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: 10
• eval_batch_size: 10
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
6.4223	1.0	217	2.8162	13.51	3.46	13.13	13.26
3.4986	2.0	434	2.7158	15.95	4.09	15.66	15.77
3.2297	3.0	651	2.6552	16.82	4.3	16.4	16.52
3.0796	4.0	868	2.6526	17.99	5.02	17.6	17.79
2.969	5.0	1085	2.6005	18.05	5.22	17.78	17.79
2.8939	6.0	1302	2.5879	18.22	5.17	17.93	18.01
2.8147	7.0	1519	2.5569	18.25	5.56	18.03	18.14
2.7642	8.0	1736	2.5541	18.24	5.38	18.07	18.19
2.724	9.0	1953	2.5493	18.86	5.7	18.51	18.63
2.6962	10.0	2170	2.5320	19.12	5.72	18.93	19.01
2.6499	11.0	2387	2.5224	18.78	5.69	18.6	18.66
2.6242	12.0	2604	2.5272	19.23	5.82	18.96	18.99
2.6088	13.0	2821	2.5122	19.51	6.16	19.26	19.36
2.5976	14.0	3038	2.5218	19.06	6.23	18.82	18.87
2.5775	15.0	3255	2.5165	19.31	6.34	19.06	19.09

Framework versions

• Transformers 4.26.0
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2