Merge pull request #54 from ai-forever/dev

release 1.0.0

DPF/filters/images/aesthetic_improved_filter.py

@@ -73,7 +73,22 @@ def get_improved_aesthetic_model(cache_folder: str) -> MLP:
 
 class ImprovedAestheticFilter(ImageFilter):
     """
-    ImprovedAestheticFilter class
+    Filter for improved aesthetic score calculating with LAION model. This repository is used:
+    https://github.com/christophschuhmann/improved-aesthetic-predictor
+
+    Parameters
+    ----------
+    weights_folder: str
+        Path to the folder where the weights are located.
+        If there are no weights, they will be downloaded automatically
+    device: str = "cuda:0"
+        Device to use
+    workers: int = 16
+        Number of processes to use for reading data and calculating flow scores
+    batch_size: int = 64
+        Batch size for model
+    pbar: bool = True
+        Whether to use a progress bar
     """
 
     def __init__(

DPF/filters/images/hash_filters.py

@@ -30,7 +30,16 @@ def get_phash(pil_img: Image.Image, hash_size: int = 8, highfreq_factor: int = 4
 
 class PHashFilter(ImageFilter):
     """
-    PHashFilter class
+    Filter for calculating PHash (perceptual hash) for images
+
+    Parameters
+    ----------
+    sim_hash_size: int = 8
+        Hash size for PHash
+    workers: int = 16
+        Number of processes to use for reading data and calculating flow scores
+    pbar: bool = True
+        Whether to use a progress bar
     """
 
     def __init__(

DPF/filters/images/info_filter.py

@@ -50,7 +50,14 @@ def get_image_info(img_bytes: bytes, data: dict[str, Any], key_column: str) -> I
 
 class ImageInfoFilter(ImageFilter):
     """
-    ImageInfoFilter class
+    Filter for gathering basic info about images (width, height, number of channels)
+
+    Parameters
+    ----------
+    workers: int = 16
+        Number of parallel dataloader workers
+    pbar: bool = True
+        Whether to show progress bar
     """
 
     def __init__(self, workers: int = 16, pbar: bool = True, _pbar_position: int = 0):

DPF/filters/images/nsfw_filter.py

@@ -26,8 +26,6 @@
 
 
 def load_safety_model(clip_model: str, cache_folder: str, device: Union[str, torch.device]) -> Any:
-    """load the safety model"""
-
     gpus = tf.config.list_physical_devices("GPU")
     if gpus:
         try:
@@ -38,7 +36,7 @@ def load_safety_model(clip_model: str, cache_folder: str, device: Union[str, tor
             pass
 
     if clip_model == "ViT-L/14":
-        model_dir = cache_folder + "/clip_autokeras_binary_nsfw"
+        model_dir = os.path.join(cache_folder, "clip_autokeras_binary_nsfw")
         url_model = (
             "https://raw.githubusercontent.com/LAION-AI/"
             "CLIP-based-NSFW-Detector/main/clip_autokeras_binary_nsfw.zip"
@@ -48,12 +46,13 @@ def load_safety_model(clip_model: str, cache_folder: str, device: Union[str, tor
 
     if not os.path.exists(model_dir):
         os.makedirs(cache_folder, exist_ok=True)
-        path_to_zip_file = cache_folder + "/clip_autokeras_binary_nsfw.zip"
+        path_to_zip_file = os.path.join(cache_folder, "clip_autokeras_binary_nsfw.zip")
         urlretrieve(url_model, path_to_zip_file)
         with zipfile.ZipFile(path_to_zip_file, "r") as zip_ref:
             zip_ref.extractall(cache_folder)
 
     with tf.device(device):
+        print(model_dir)
         loaded_model = load_model(model_dir, custom_objects=ak.CUSTOM_OBJECTS)
 
     return loaded_model
@@ -72,7 +71,6 @@ class NSFWFilter(ImageFilter):
 
     def __init__(
         self,
-        clip_model: str,
         weights_folder: str,
         workers: int = 16,
         batch_size: int = 64,
@@ -81,7 +79,7 @@ def __init__(
         _pbar_position: int = 0
     ):
         super().__init__(pbar, _pbar_position)
-
+        clip_model = "ViT-L/14"
         self.num_workers = workers
         self.batch_size = batch_size
         self.device = device

DPF/filters/images/ocr_filter.py

@@ -23,6 +23,7 @@ def __init__(
         self,
         weights_path: str,
         model_name: Optional[str] = None,
+        text_box_col: str = "text_boxes",
         device: str = "cuda:0",
         workers: int = 16,
         pad: int = 5,
@@ -73,7 +74,7 @@ def __init__(
 
         self.AlignCollate = AlignCollate(imgH=self.opt.imgH, imgW=self.opt.imgW, keep_ratio_with_pad=self.opt.PAD)
         #
-        self.text_box_col = "text_boxes"
+        self.text_box_col = text_box_col
         self.ocr_col = f"OCR_{self.model_name}"
 
     @property

DPF/filters/images/ocr_model/model.py

@@ -1,19 +1,3 @@
-"""
-Copyright (c) 2019-present NAVER Corp.
-
-Licensed under the Apache License, Version 2.0 (the "License");
-you may not use this file except in compliance with the License.
-You may obtain a copy of the License at
-
-    http://www.apache.org/licenses/LICENSE-2.0
-
-Unless required by applicable law or agreed to in writing, software
-distributed under the License is distributed on an "AS IS" BASIS,
-WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-See the License for the specific language governing permissions and
-limitations under the License.
-"""
-
 import torch.nn as nn
 
 from .modules.feature_extraction import (
@@ -34,14 +18,12 @@ def __init__(self, opt):
         self.stages = {'Trans': opt.Transformation, 'Feat': opt.FeatureExtraction,
                        'Seq': opt.SequenceModeling, 'Pred': opt.Prediction}
 
-        """ Transformation """
         if opt.Transformation == 'TPS':
             self.Transformation = TPS_SpatialTransformerNetwork(
                 F=opt.num_fiducial, I_size=(opt.imgH, opt.imgW), I_r_size=(opt.imgH, opt.imgW), I_channel_num=opt.input_channel)
         else:
             print('No Transformation module specified')
 
-        """ FeatureExtraction """
         if opt.FeatureExtraction == 'VGG':
             self.FeatureExtraction = VGG_FeatureExtractor(opt.input_channel, opt.output_channel)
         elif opt.FeatureExtraction == 'RCNN':
@@ -53,7 +35,6 @@ def __init__(self, opt):
         self.FeatureExtraction_output = opt.output_channel  # int(imgH/16-1) * 512
         self.AdaptiveAvgPool = nn.AdaptiveAvgPool2d((None, 1))  # Transform final (imgH/16-1) -> 1
 
-        """ Sequence modeling"""
         if opt.SequenceModeling == 'BiLSTM':
             self.SequenceModeling = nn.Sequential(
                 BidirectionalLSTM(self.FeatureExtraction_output, opt.hidden_size, opt.hidden_size),
@@ -63,7 +44,6 @@ def __init__(self, opt):
             print('No SequenceModeling module specified')
             self.SequenceModeling_output = self.FeatureExtraction_output
 
-        """ Prediction """
         if opt.Prediction == 'CTC':
             self.Prediction = nn.Linear(self.SequenceModeling_output, opt.num_class)
         elif opt.Prediction == 'Attn':
@@ -72,22 +52,18 @@ def __init__(self, opt):
             raise Exception('Prediction is neither CTC or Attn')
 
     def forward(self, input, text, is_train=True):
-        """ Transformation stage """
         if not self.stages['Trans'] == "None":
             input = self.Transformation(input)
 
-        """ Feature extraction stage """
         visual_feature = self.FeatureExtraction(input)
         visual_feature = self.AdaptiveAvgPool(visual_feature.permute(0, 3, 1, 2))  # [b, c, h, w] -> [b, w, c, h]
         visual_feature = visual_feature.squeeze(3)
 
-        """ Sequence modeling stage """
         if self.stages['Seq'] == 'BiLSTM':
             contextual_feature = self.SequenceModeling(visual_feature)
         else:
             contextual_feature = visual_feature  # for convenience. this is NOT contextually modeled by BiLSTM
 
-        """ Prediction stage """
         if self.stages['Pred'] == 'CTC':
             prediction = self.Prediction(contextual_feature.contiguous())
         else:

DPF/filters/texts/google_translate_filter.py

@@ -42,7 +42,26 @@ def translate_batch(translator: BaseTranslator, batch: list[str], delimiter: str
 
 class GoogleTranslateFilter(ColumnFilter):
     """
-    GoogleTranslateFilter class
+    Filter for translating texts with google translate api
+
+    Parameters
+    ----------
+    text_column_name: str = "text"
+        Name of column with texts
+    source_lang: str = "auto"
+        Source language to translate from
+    target_lang: str = "en"
+        Language to translate to
+    max_symbols_in_batch: int = 3000
+        Maximum symbols in one request to API.
+    timeout: float = 1
+        Timeout between requests
+    timeout_on_error: float = 3
+        Timeout between requests if error occured
+    num_retries_per_batch: int = 1
+        Number of retries of errors occured
+    pbar: bool = True
+        Whether to use a progress bar
     """
 
     def __init__(

DPF/filters/texts/lang_filter.py

@@ -7,7 +7,16 @@
 
 class LangFilter(ColumnFilter):
     """
-    LangFilter class
+    Filter for text language detection
+
+    Parameters
+    ----------
+    text_column_name: str = "text"
+        Name of column with texts
+    workers: int = 16
+        Number of processes to use
+    pbar: bool = True
+        Whether to use a progress bar
     """
 
     def __init__(