Bring mgds to OneTrainer code standards

src/mgds/MGDS.py

@@ -1,7 +1,7 @@
 import random
 
 import torch
-from torch.utils.data import DataLoader, Dataset, IterableDataset
+from torch.utils.data import DataLoader, IterableDataset
 
 from mgds.ConceptPipelineModule import ConceptPipelineModule
 from mgds.LoadingPipeline import LoadingPipeline
@@ -19,7 +19,7 @@ def __init__(
             device: torch.device,
             concepts: list[dict],
             settings: dict,
-            definition: [PipelineModule],
+            definition: list[PipelineModule],
             batch_size: int, #local batch size
             state: PipelineState,
             seed: int = 42,

src/mgds/PipelineModule.py

@@ -4,13 +4,17 @@
 from concurrent import futures
 from contextlib import ExitStack
 from random import Random
+from typing import TYPE_CHECKING
 
 import torch
 
 from mgds.pipelineModuleTypes.RandomAccessPipelineModule import RandomAccessPipelineModule
 from mgds.pipelineModuleTypes.SerialPipelineModule import SerialPipelineModule
 from mgds.pipelineModuleTypes.SingleVariationRandomAccessPipelineModule import SingleVariationRandomAccessPipelineModule
 
+if TYPE_CHECKING:
+    from mgds.LoadingPipeline import LoadingPipeline
+
 
 class PipelineState:
     """Container for state shared amongst all pipeline modules in a pipeline.
@@ -142,7 +146,6 @@ def _get_previous_item(self, variation: int, name: str, index: int):
     def _get_previous_length(self, name: str):
         split_name = name.split('.')
         item_name = split_name[0]
-        path_names = split_name[1::]
 
         for previous_module_index in range(self.__module_index - 1, -1, -1):
             module = self.pipeline.modules[previous_module_index]

src/mgds/pipelineModules/DropTags.py

@@ -144,7 +144,7 @@ def get_item(self, variation: int, index: int, requested_name: str = None) -> di
                 for s in dropout_tags:
                     if special_tag_mode == "WHITELIST" and s in special_tags_list:
                         pruned_tags.append(s)
-                    elif special_tag_mode == "BLACKLIST" and not(s in special_tags_list):
+                    elif special_tag_mode == "BLACKLIST" and s not in special_tags_list:
                         pruned_tags.append(s)
             elif dropout_mode.startswith("RANDOM"):     
                 #iterate through dropout_tags and add to pruned_tags if random > probability
@@ -156,7 +156,7 @@ def get_item(self, variation: int, index: int, requested_name: str = None) -> di
                 elif special_tag_mode == "BLACKLIST":
                     for i, s in enumerate(dropout_tags):
                         if rand.random() > self.probability_weighted(probability, dropout_mode, i, len(dropout_tags)) \
-                        or not(s in special_tags_list):
+                        or s not in special_tags_list:
                             pruned_tags.append(s)
                 else:   #NONE or any other unexpected values
                     for i, s in enumerate(dropout_tags):

src/mgds/pipelineModules/ImageToVideo.py

@@ -1,5 +1,3 @@
-from transformers import CLIPTokenizer, T5Tokenizer, T5TokenizerFast, GemmaTokenizer, LlamaTokenizer
-
 from mgds.PipelineModule import PipelineModule
 from mgds.pipelineModuleTypes.RandomAccessPipelineModule import RandomAccessPipelineModule
 

src/mgds/pipelineModules/LoadVideo.py

@@ -123,7 +123,7 @@ def get_item(self, variation: int, index: int, requested_name: str = None) -> di
 
             except FileNotFoundError:
                 video_tensor = None
-            except Exception as e:
+            except Exception:
                 print("could not load video, it might be corrupted: " + path)
                 raise
         else: