t5-small_corrector_15

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

• Loss: 1.3416
• Rouge1: 34.7998
• Rouge2: 9.0842
• Rougel: 27.8188
• Rougelsum: 27.839
• Gen Len: 18.5561

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: 2e-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: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
4.2274	1.0	2365	2.9386	10.1244	1.0024	9.1029	9.1104	18.5377
2.7936	2.0	4730	2.0196	17.7168	3.0899	15.1305	15.1353	18.8883
2.2678	3.0	7095	1.7072	26.8501	5.7804	22.0034	22.0213	18.839
1.9029	4.0	9460	1.5254	32.9484	7.8531	26.4538	26.4749	18.502
1.5936	5.0	11825	1.3416	34.7998	9.0842	27.8188	27.839	18.5561

Framework versions

• Transformers 4.21.2
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1