---
language:
- en
license: apache-2.0
tags:
- roberta
- classification
- dialog state tracking
- conversational system
- task-oriented dialog
datasets:
- ConvLab/tm1
- Convlab/tm2
- Convlab/tm3
metrics:
- Joint Goal Accuracy
- Slot F1

model-index:
- name: setsumbt-dst-tm123
  results:
  - task:
      type: classification
      name: dialog state tracking
    dataset:
      type: ConvLab/tm1
      name: TM1+TM2+TM3
      split: test
    metrics:
      - type: Joint Goal Accuracy
        value: 24.9
        name: JGA
      - type: Slot F1
        value: 65.5
        name: Slot F1

---

# SetSUMBT-dst-tm1-tm2-tm3

This model is a fine-tuned version [SetSUMBT](https://github.com/ConvLab/ConvLab-3/tree/master/convlab/dst/setsumbt) of [roberta-base](https://huggingface.co/roberta-base) on [Taskmaster1](https://huggingface.co/datasets/ConvLab/tm1), [Taskmaster2](https://huggingface.co/datasets/ConvLab/tm2) and [Taskmaster3](https://huggingface.co/datasets/ConvLab/tm3).

Refer to [ConvLab-3](https://github.com/ConvLab/ConvLab-3) for model description and usage.

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.00001
- train_batch_size: 2
- eval_batch_size: 8
- seed: 0
- gradient_accumulation_steps: 1
- optimizer: AdamW
- lr_scheduler_type: linear
- num_epochs: 50.0

### Framework versions

- Transformers 4.17.0
- Pytorch 1.8.0+cu110
- Datasets 2.3.2
- Tokenizers 0.12.1