pipeline.py

from diffusers import DiffusionPipeline
import torch
import os
import yaml
from audioldm_train.utilities.tools import build_dataset_json_from_list
from infer_mos5 import infer  # Importing the infer function

class MOSDiffusionPipeline(DiffusionPipeline):

    def __init__(self, config_yaml, list_inference, reload_from_ckpt=None):
        """
        Initialize the MOS Diffusion pipeline.

        Args:
            config_yaml (str): Path to the YAML configuration file.
            list_inference (str): Path to the file containing inference prompts.
            reload_from_ckpt (str, optional): Checkpoint path to reload from.
        """
        super().__init__()

        # we load and process the yaml config
        self.config_yaml = config_yaml
        self.list_inference = list_inference
        self.reload_from_ckpt = reload_from_ckpt

        # we load the yaml config
        config_yaml_path = os.path.join(self.config_yaml)
        self.configs = yaml.load(open(config_yaml_path, "r"), Loader=yaml.FullLoader)

        # override checkpoint if provided--
        if self.reload_from_ckpt is not None:
            self.configs["reload_from_ckpt"] = self.reload_from_ckpt

        self.dataset_key = build_dataset_json_from_list(self.list_inference)
        self.exp_name = os.path.basename(self.config_yaml.split(".")[0])
        self.exp_group_name = os.path.basename(os.path.dirname(self.config_yaml))

    @torch.no_grad()
    def __call__(self, *args, **kwargs):
        """
        Run the MOS Diffusion Pipeline. This method calls the infer function from infer_mos5.py.

        Args:
            *args: Additional arguments.
            **kwargs: Keyword arguments that may contain overrides for configurations.

        Returns:
            None. Inference is performed and samples are generated.
        """
        # here call the infer function to perform the inference
        infer(
            dataset_key=self.dataset_key, 
            configs=self.configs, 
            config_yaml_path=self.config_yaml, 
            exp_group_name=self.exp_group_name, 
            exp_name=self.exp_name
        )

# # This is an example of how to use the pipeline
# if __name__ == "__main__":
#     pipeline = MOSDiffusionPipeline(
#         config_yaml="audioldm_train/config/mos_as_token/qa_mdt.yaml",
#         list_inference="/content/qa-mdt/test_prompts/good_prompts_1.lst",
#         reload_from_ckpt="/content/qa-mdt/checkpoint_389999.ckpt"
#     )

#     # Run the pipeline
#     pipeline()