Uploaded filtered 500k tagging train set.

README.md

@@ -5,7 +5,7 @@ task_categories:
 tags:
 - not-for-all-audiences
 size_categories:
-- 10K<n<100K
+- 100K<n<1M
 ---
 # Anime Collection
 A repo containing scripts to scrape booru sites and images scraped from it.

balance_tags.py

@@ -7,39 +7,6 @@ import random
 import argparse
 from constants import *
 
-def get_model_tags(model_tags_path):
-    if not os.path.isfile(model_tags_path):
-        raise FileNotFoundError(f"\"{model_tags_path}\" is not a file, please place one there!")
-    index_tag_dict = {}
-    with open(model_tags_path, "r", encoding="utf8") as model_tags_file:
-        for line in model_tags_file:
-            line = line.split()
-            if len(line) != 2:
-                continue
-            index_tag_dict[int(line[0])] = line[1]
-    if len(index_tag_dict) <= 0:
-        return []
-    sorted_index_tag_tuple_list = sorted(index_tag_dict.items(), key=lambda x: x[0])
-    if len(sorted_index_tag_tuple_list) != sorted_index_tag_tuple_list[-1][0] + 1:
-        raise ValueError(f"The index specified in \"{model_tags_path}\" is not continuous!")
-    return [tag for _, tag in sorted_index_tag_tuple_list]
-
-def get_tags_set(tags_path):
-    if not os.path.isfile(tags_path):
-        raise FileNotFoundError(f"\"{tags_path}\" is not a file!")
-    with open(tags_path, "r", encoding="utf8") as tags_file:
-        tags_text = tags_file.read()
-    tags_set = set()
-    for tag in tags_text.split(","):
-        tag = tag.strip()
-        if tag:
-            tag = tag.replace(" ", "_")
-            if tag == "nsfw": tag = "rating:explicit"
-            elif tag == "qfw": tag = "rating:questionable"
-            elif tag == "sfw": tag = "rating:safe"
-            tags_set.add(tag)
-    return tags_set
-
 def parse_args():
     parser = argparse.ArgumentParser(description="Balance the dataset based on tag frequency.")
     parser.add_argument("-c", "--count", type=int, help="The target selection count, must be an integer greater than 0")
@@ -64,7 +31,7 @@ def parse_args():
 def main():
     args = parse_args()
     print("Starting...\nGetting model tags...")
-    model_tags = get_model_tags(MODEL_TAGS_PATH)
+    model_tags = utils.get_model_tags(MODEL_TAGS_PATH)
     print("Getting paths...")
     image_id_image_tags_path_tuple_tuple_list = sorted(utils.get_image_id_image_tags_path_tuple_dict(IMAGE_DIR).items(), key=lambda x: x[0])
     print("Got", len(image_id_image_tags_path_tuple_tuple_list), "images.\nShuffling paths...")
@@ -75,7 +42,7 @@ def main():
     buckets = {tag: [] for tag in model_tags}
     for image_id_image_tags_path_tuple_tuple in tqdm.tqdm(image_id_image_tags_path_tuple_tuple_list, desc="Making buckets"):
         did_append = False
-        for tag in get_tags_set(image_id_image_tags_path_tuple_tuple[1][1]):
+        for tag in utils.get_tags_set(image_id_image_tags_path_tuple_tuple[1][1]):
             bucket = buckets.get(tag)
             if bucket is None:
                 continue

make_model_tags.py

@@ -0,0 +1,61 @@
+import sys
+import tqdm
+import utils
+import argparse
+from constants import *
+from collections import defaultdict
+
+def is_tag_unneeded(tag):
+    if tag.endswith("_request"):
+        return True
+    if tag.startswith("bad_") and tag.endswith("id"):
+        return True
+    if tag.endswith("_mismatch"):
+        return True
+    match tag:
+        case "tagme":
+            return True
+        case "bad_link":
+            return True
+        case "protected_link":
+            return True
+        case "inactive_account":
+            return True
+        case "bad_source":
+            return True
+        case "resized":
+            return True
+    return False
+
+def parse_args():
+    parser = argparse.ArgumentParser(description="Create model tags based on tag frequency.")
+    parser.add_argument("-i", "--min-images", type=int, default=0, help="Filter out tags with less than the specified amount of images, default to 0")
+    args = parser.parse_args()
+    if args.min_images < 0:
+        print("Minimum images must be greater than or equal to 0!")
+        sys.exit(1)
+    return args
+
+def main():
+    args = parse_args()
+    print("Starting...\nGetting paths...")
+    image_id_image_tags_path_tuple_dict = utils.get_image_id_image_tags_path_tuple_dict(IMAGE_DIR)
+    print("Got", len(image_id_image_tags_path_tuple_dict), "images.\nMaking buckets...")
+    buckets = defaultdict(int)
+    for _, tags_path in tqdm.tqdm(image_id_image_tags_path_tuple_dict.values(), desc="Making buckets"):
+        for tag in utils.get_tags_set(tags_path):
+            buckets[tag] += 1
+    print("Filtering out tags with less than", args.min_images, "images and sorting...")
+    buckets = sorted(bucket for bucket in buckets.items() if not is_tag_unneeded(bucket[0]) and bucket[1] >= args.min_images)
+    print("The new model tags list contains", len(buckets), "tags.\nSaving the result...")
+    with open(MODEL_TAGS_PATH, "w", encoding="utf8") as file:
+        for i, bucket in enumerate(buckets):
+            file.write(f"{i} {bucket[0]}\n")
+    print("Finished.")
+
+if __name__ == "__main__":
+    try:
+        main()
+    except KeyboardInterrupt:
+        print("\nScript interrupted by user, exiting...")
+        sys.exit(1)

scrape_gel.py

@@ -15,7 +15,7 @@ from constants import *
 from bs4 import BeautifulSoup
 
 async def process_link(scrape_args, scrape_state):
-    image_id = re.search("id=(\d+)", scrape_args.target).group(1)
+    image_id = re.search(r"id=(\d+)", scrape_args.target).group(1)
     image_id_int = int(image_id)
     scrape_state.last_reached_image_id = image_id_int
     image_id_already_exists = image_id in scrape_state.existing_image_ids

tagging_train/{chunk_0.tar → balanced_500k.tar}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5071eed8cbcf55e133823580778f54674b24eba60eabe20e1b1e1a3d3e0d2b58
-size 6841774080
+oid sha256:6cdcc52833adbebe1c482e424a8336ce383adb1d5f3abdaf92b3710dc0cd17a6
+size 11393341440

tagging_train/full_1m.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:454f382118a06d2ee22814cc1e64fccbe7f86fecca7bbd3a8013368052ff9aa9
+size 22291906560

utils/utils.py

@@ -70,3 +70,36 @@ def get_session(timeout=None, cookies=None):
     if timeout is not None:
         kwargs["timeout"] = aiohttp.ClientTimeout(total=timeout)
     return aiohttp.ClientSession(**kwargs)
+
+def get_model_tags(model_tags_path):
+    if not os.path.isfile(model_tags_path):
+        raise FileNotFoundError(f"\"{model_tags_path}\" is not a file, please place one there!")
+    index_tag_dict = {}
+    with open(model_tags_path, "r", encoding="utf8") as model_tags_file:
+        for line in model_tags_file:
+            line = line.split()
+            if len(line) != 2:
+                continue
+            index_tag_dict[int(line[0])] = line[1]
+    if len(index_tag_dict) <= 0:
+        return []
+    sorted_index_tag_tuple_list = sorted(index_tag_dict.items(), key=lambda x: x[0])
+    if len(sorted_index_tag_tuple_list) != sorted_index_tag_tuple_list[-1][0] + 1:
+        raise ValueError(f"The index specified in \"{model_tags_path}\" is not continuous!")
+    return [tag for _, tag in sorted_index_tag_tuple_list]
+
+def get_tags_set(tags_path):
+    if not os.path.isfile(tags_path):
+        raise FileNotFoundError(f"\"{tags_path}\" is not a file!")
+    with open(tags_path, "r", encoding="utf8") as tags_file:
+        tags_text = tags_file.read()
+    tags_set = set()
+    for tag in tags_text.split(","):
+        tag = tag.strip()
+        if tag:
+            tag = tag.replace(" ", "_")
+            if tag == "nsfw": tag = "rating:explicit"
+            elif tag == "qfw": tag = "rating:questionable"
+            elif tag == "sfw": tag = "rating:safe"
+            tags_set.add(tag)
+    return tags_set