diff --git a/convlab/dst/trippy/README.md~ b/convlab/dst/trippy/README.md~ deleted file mode 100644 index b5f35120505ecb27793db102c4f2d52bb2760ad8..0000000000000000000000000000000000000000 --- a/convlab/dst/trippy/README.md~ +++ /dev/null @@ -1,62 +0,0 @@ -# Introduction - -This is the TripPy DST module for ConvLab-3. - -## Supported encoders - -* RoBERTa -* BERT (full support w.i.p.) -* ELECTRA (full support w.i.p.) - -## Supported datasets - -* MultiWOZ 2.X -* Unified Data Format - -## Requirements - -transformers (tested: 4.18.0) -torch (tested: 1.8.0) - -# Parameters - -``` -model_type # Default: "roberta", Type of the model (Supported: "roberta", "bert", "electra") -model_name # Default: "roberta-base", Name of the model (Use -h to print a list of names) -model_path # Path to a model checkpoint -dataset_name # Default: "multiwoz21", Name of the dataset the model was trained on and/or is being applied to -local_files_only # Default: False, Set to True to load local files only. Useful for offline systems -nlu_usr_config # Path to a NLU config file. Only needed for internal evaluation -nlu_sys_config # Path to a NLU config file. Only needed for internal evaluation -nlu_usr_path # Path to a NLU model file. Only needed for internal evaluation -nlu_sys_path # Path to a NLU model file. Only needed for internal evaluation -no_eval # Default: True, Set to True if internal evaluation should be conducted -no_history # Default: False, Set to True if dialogue history should be omitted during inference -``` - -# Training - -TripPy can easily be trained for the abovementioned supported datasets using the original code in the official [TripPy repository](https://gitlab.cs.uni-duesseldorf.de/general/dsml/trippy-public). Simply clone the code and run the appropriate DO.* script to train a TripPy DST. After training, set model_path to the preferred checkpoint to use TripPy in ConvLab-3. - -# Training and evaluation with PPO policy - -Switch to the directory: -``` -cd ../../policy/ppo -``` - -Edit trippy_config.json and trippy_config_eval.json accordingly, e.g., edit paths to model checkpoints. - -For training, run -``` -train.py --path trippy_config.json -``` - -For evaluation, run -``` -train.py --path trippy_config_eval.json -``` - -# Paper - -[TripPy: A Triple Copy Strategy for Value Independent Neural Dialog State Tracking](https://aclanthology.org/2020.sigdial-1.4/)