ConvLab/t5-small-dst-multiwoz21

t5-small-dst-multiwoz21

This model is a fine-tuned version of t5-small on MultiWOZ 2.1.

Refer to ConvLab-3 for model description and usage.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• —learning_rate: 0.001
• —trainbatchsize: 64
• —evalbatchsize: 64
• —seed: 42
• —gradientaccumulationsteps: 2
• —totaltrainbatch_size: 128
• —optimizer: Adafactor
• —lrschedulertype: linear
• —num_epochs: 10.0

Framework versions

• —Transformers 4.20.1
• —Pytorch 1.11.0+cu113
• —Datasets 2.3.2
• —Tokenizers 0.12.1