mt5-small-finetuned-amazon-en-es

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

• Loss: 2.4413
• Rouge1: 22.6804
• Rouge2: 8.3299
• Rougel: 17.9992
• Rougelsum: 20.7342

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: 5.6e-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: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
7.77	1.0	240	2.7230	17.25	5.629	14.0381	15.8959
3.7586	2.0	480	2.5949	19.4577	6.9354	15.772	17.8773
3.4314	3.0	720	2.5355	20.0511	7.6417	16.0889	18.4551
3.2892	4.0	960	2.4845	20.3951	7.88	16.601	19.0048
3.1954	5.0	1200	2.4612	20.1806	7.2656	16.2658	18.6222
3.1128	6.0	1440	2.4544	22.5647	8.0899	17.8057	20.487
3.103	7.0	1680	2.4498	22.7048	8.384	17.978	20.6871
3.0708	8.0	1920	2.4413	22.6804	8.3299	17.9992	20.7342

Framework versions

• Transformers 4.20.1
• Pytorch 1.12.0+cu113
• Datasets 2.3.2
• Tokenizers 0.12.1