mt5-small-finetuned-news-summary-kaggle

This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.6907
• Rouge1: 26.6547
• Rouge2: 10.1
• Rougel: 24.0137
• Rougelsum: 23.9999

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

https://www.kaggle.com/datasets/sunnysai12345/news-summary

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5.6e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
No log	1.0	220	3.9956	14.9021	3.3744	13.4763	13.499
8.3183	2.0	440	3.1550	17.9472	5.9671	16.6974	16.6959
8.3183	3.0	660	2.8950	21.2665	7.4266	19.5041	19.4837
4.0457	4.0	880	2.8087	25.063	9.4484	22.746	22.7351
4.0457	5.0	1100	2.7375	25.5269	9.4299	23.0623	23.0075
3.6505	6.0	1320	2.7091	25.8308	9.3392	23.2001	23.1586
3.6505	7.0	1540	2.6949	26.2177	9.8536	23.5946	23.6358
3.5175	8.0	1760	2.6907	26.6547	10.1	24.0137	23.9999

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2