109 lines
3.8 KiB
Python
109 lines
3.8 KiB
Python
# MBM's Music Visualizer: Prompt Sequence
|
|
# Allows for building of a sequence of prompts.
|
|
|
|
# Imports
|
|
import torch
|
|
from typing import Union
|
|
|
|
from .mbmPrompt import MbmPrompt
|
|
from .mbmPromptSequenceData import PromptSequenceData
|
|
|
|
# Classes
|
|
class PromptSequenceBuilder:
|
|
"""
|
|
Allows for building of a sequence of prompts.
|
|
"""
|
|
def __init__(self):
|
|
pass
|
|
|
|
@classmethod
|
|
def INPUT_TYPES(s):
|
|
return {
|
|
"required": {
|
|
"positive_1": ("CONDITIONING", ),
|
|
"negative_1": ("CONDITIONING", )
|
|
},
|
|
"optional": {
|
|
"positive_2": ("CONDITIONING", ),
|
|
"negative_2": ("CONDITIONING", ),
|
|
"prompts": ("PROMPT_SEQ", )
|
|
}
|
|
}
|
|
|
|
RETURN_TYPES = ("PROMPT_SEQ", )
|
|
RETURN_NAMES = ("PROMPTS", )
|
|
FUNCTION = "process"
|
|
CATEGORY = "MBMnodes/Prompts"
|
|
|
|
def process(self,
|
|
positive_1: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]],
|
|
negative_1: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]],
|
|
positive_2: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]] = None,
|
|
negative_2: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]] = None,
|
|
prompts: list[MbmPrompt] = None
|
|
) -> list[MbmPrompt]:
|
|
"""
|
|
Returns a list of MbmPrompt objects representing the prompt sequence.
|
|
"""
|
|
# Create a new prompt sequence
|
|
if (positive_2 is None) or (negative_2 is None):
|
|
# Include just the required one
|
|
promptsOut = [
|
|
MbmPrompt.fromComfyUiPrompts(positive_1, negative_1)
|
|
]
|
|
else:
|
|
# Include both
|
|
promptsOut = [
|
|
MbmPrompt.fromComfyUiPrompts(positive_1, negative_1),
|
|
MbmPrompt.fromComfyUiPrompts(positive_2, negative_2)
|
|
]
|
|
|
|
# Add to the given prompt sequence if provided
|
|
if prompts is not None:
|
|
promptsOut = (prompts + promptsOut)
|
|
|
|
return (promptsOut, )
|
|
|
|
class PromptSequenceBuilderAdvanced(PromptSequenceBuilder):
|
|
"""
|
|
Allows for building prompts with additional information required for more fine control of prompt sequences.
|
|
"""
|
|
@classmethod
|
|
def INPUT_TYPES(s):
|
|
inputTypes = super().INPUT_TYPES()
|
|
|
|
inputTypes["required"]["timecode_1"] = ("FLOAT", {"default": -1.0, "min": -1.0, "max": 0xffffffffffffffff})
|
|
inputTypes["optional"]["timecode_2"] = ("FLOAT", {"default": -1.0, "min": -1.0, "max": 0xffffffffffffffff})
|
|
|
|
return inputTypes
|
|
|
|
def process(self,
|
|
positive_1: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]],
|
|
negative_1: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]],
|
|
timecode_1: float,
|
|
positive_2: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]] = None,
|
|
negative_2: list[list[Union[torch.Tensor, dict[str, torch.Tensor]]]] = None,
|
|
timecode_2: float = 0.0,
|
|
prompts: list[MbmPrompt] = None
|
|
):
|
|
"""
|
|
Returns a list of MbmPrompt objects representing the prompt sequence.
|
|
"""
|
|
# Validation
|
|
if (timecode_1 < 0) or (timecode_2 < 0):
|
|
raise ValueError("Timecodes must be >= 0.0.")
|
|
|
|
# Run the super
|
|
promptsOut = super().process(positive_1, negative_1, positive_2, negative_2, prompts)
|
|
|
|
# Add the sequence data
|
|
if (positive_2 is None):
|
|
# Add data to only the last prompt
|
|
promptsOut[0][-1].data[PromptSequenceData.DATA_KEY] = PromptSequenceData(timecode_1)
|
|
else:
|
|
# Add data to the proper prompts
|
|
promptsOut[0][-2].data[PromptSequenceData.DATA_KEY] = PromptSequenceData(timecode_1)
|
|
promptsOut[0][-1].data[PromptSequenceData.DATA_KEY] = PromptSequenceData(timecode_2)
|
|
|
|
return promptsOut
|