t5-small-finetuned-cnn

This model is a fine-tuned version of t5-small on the cnn_dailymail dataset. It achieves the following results on the evaluation set:

• Loss: 1.8436
• Rouge1: 33.2082
• Rouge2: 16.798
• Rougel: 28.9573
• Rougelsum: 31.1044

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
2.3793	1.0	359	1.8885	33.0321	16.7798	28.9367	30.9509
2.1432	2.0	718	1.8481	33.1559	16.8557	29.015	31.1122
2.0571	3.0	1077	1.8391	32.99	16.716	28.8118	30.9178
2.0001	4.0	1436	1.8357	33.0543	16.6731	28.8375	30.9604
1.9609	5.0	1795	1.8437	33.1019	16.7576	28.8669	31.001
1.925	6.0	2154	1.8402	33.1388	16.7539	28.8887	31.0262
1.9036	7.0	2513	1.8423	33.1825	16.759	28.9154	31.0656
1.8821	8.0	2872	1.8436	33.2082	16.798	28.9573	31.1044

Framework versions

• Transformers 4.14.0
• Pytorch 1.5.0
• Datasets 2.3.2
• Tokenizers 0.10.3