Instructions to run the code

Download the datasets

• Dialogue Datasets: https://www.dropbox.com/s/kojje169drighei/avec_iemocap_meld.zip?dl=0
  • Containing AVEC, IEMOCAP and MELD datasets
  • Datasets come from https://github.com/SenticNet/conv-emotion
    • iemocap_data.pkl, meld_data_original.pkl and AVEC/* are original datasets in the github
    • iemocap_data_act.pkl and meld_data_act.pkl are the datasets with dialogue act annotations from https://github.com/bothe/EDAs
    • meld_data.pkl is the MELD dataset with dialogue act annotations and pretrained 300d visual features for each utterance

Do A Single Run (train/valid/test)

1. Set up the configurations in config/run.ini
2. python run.py -config config/run.ini

Configuration setup

• Dialogue
  • mode = run
  • pickle_dir_path = /path/to/datasets/. The absolute path of the folder storing the datasets.
  • wordvec_path = /path/to/glove_embedding/. The path of the glove embedding file.
  • dataset_name in {'iemocap','meld','avec'}. Name of the dataset.
  • features in {'acoustic','visual','textual'}. Multiple modality names should be joined by ','.
  • label = emotion. Only emotion recognition is supported for the time being.
  • dialogue_context in {'True','False'}. Whether the data is converted to context + current utterance format (the context of a certain length is extracted for each utterance in the dialogue).
  • context_len. Length of the context. Requires dialogue_context = True.
  • seed. The random seed for the experiment.
  • load_model_from_dir in {'True','False'}. Whether the model is loaded from a saved file.
  • dir_name. The directory storing the model configurations and model parameters. Requires load_model_from_dir = True.
  • fine_tune in `{'True','False'}. Whether you want to train the model with the data.
  • model specific parameters. For running a model on the dataset, uncomment the respective area of the model and comment the areas for the other models. Please refer to the model implementations in /models/dialogue/ for the meaning of each model specific parameter.
    • supported models include but are not limited to:
      • BC-LSTM
      • CMN
      • ICON
      • DialogueRNN
      • DialogueGCN
      • Multimodal-Transformer
      • EF-LSTM
      • LF-LSTM
      • TFN
      • MFN
      • MARN
      • RMFN
      • LMF
      • Vanilla-LSTHM
      • QMN (QMNN)
    • Each model is only compatible with dialogue_context = True or dialogue_context = False. For this information, please see the respective block for each model.

Grid Search for the Best Parameters

1. Set up the configurations in config/grid_search.ini. Tweak a couple of fields in the single run configurations, as instructed below.
2. Write up the hyperparameter pool in config/grid_parameters/.
3. python run.py -config config/grid_search.ini

Configuration setup

• mode = run_grid_search
• grid_parameters_file. The name of file storing the parameters to be searched, under the folder /config/grid_parameters.
  • the format of a file is:
    • [COMMON]
    • var_1 = val_1;val_2;val_3
    • var_2 = val_1;val_2;val_3
• search_times. The number of times the program searches in the pool of parameters.
• output_file. The file storing the performances for each search in the pool of parameters. By default, it is eval/grid_search_{dataset_name}_{network_type}.csv