final commit

.gitignore

@@ -1,4 +1,6 @@
 .idea
+found-keys*.txt
+
 
 # ---> Python
 # Byte-compiled / optimized / DLL files

README.md

@@ -1,3 +1,5 @@
 # hf-key-scraper
 
-Scrape for OpenAI keys on Hugging Face
+Scrape for OpenAI keys on Hugging Face.
+
+This requires at least one Squid proxy to function. Define your proxies in `settings.py` -> `PROXIES`.

huggingface/check-keys.py

@@ -1,8 +1,9 @@
 import argparse
-import concurrent
 from concurrent.futures import ThreadPoolExecutor
 from pathlib import Path
 
+from tqdm import tqdm
+
 from keychecker.check import check_key
 
 
@@ -17,26 +18,29 @@ def process_key(key):
 
 def main():
     parser = argparse.ArgumentParser(description='Scrape data from chub.ai.')
-    parser.add_argument('input_file', help='Path to the file containing the keys.')
+    parser.add_argument('input_file', nargs='*', help='Path to the file containing the keys.')
     args = parser.parse_args()
 
-    input_file = Path(args.input_file).resolve().absolute()
-
-    if not input_file.exists():
-        print('Input file does not exist:', input_file)
-        quit(1)
-
-    with open(input_file) as f:
-        content = set(f.read().splitlines())
+    keys = set()
+    for file in args.input_file:
+        input_file = Path(file).resolve().expanduser().absolute()
+        if not input_file.exists():
+            print('Input file does not exist:', input_file)
+            quit(1)
+        data = set(input_file.read_text().splitlines())
+        keys = keys | data
 
     # content = ['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']
 
+    print('Checking', len(keys), 'keys...')
+
     gpt_4 = set()
     gpt_4_32k = set()
     gpt_3 = set()
+    pbar = tqdm(total=len(keys))
 
     with ThreadPoolExecutor(max_workers=50) as executor:
-        results = executor.map(process_key, content)
+        results = executor.map(process_key, keys)
         for result in results:
             if result is not None:
                 if result['has_gpt_4']:
@@ -45,6 +49,8 @@ def main():
                     gpt_4_32k.add(result['api_key'])
                 if result['has_only_turbo']:
                     gpt_3.add(result['api_key'])
+            pbar.update(1)
+    pbar.close()
     print('')
 
     print('GPT4 KEYS:')

huggingface/huggingface/settings.py

@@ -92,10 +92,7 @@ TWISTED_REACTOR = "twisted.internet.asyncioreactor.AsyncioSelectorReactor"
 FEED_EXPORT_ENCODING = "utf-8"
 
 PROXIES = [
-    "http://172.0.4.7:3128",
-    "http://172.0.4.8:3128",
-    "http://172.0.4.10:3128",
-    "http://172.0.4.12:3128",
+    "http://127.0.0.1:3128"
 ]
 
 LOG_LEVEL = 'CRITICAL'

huggingface/huggingface/spiders/huggingface.py

@@ -31,5 +31,5 @@ class HuggingfaceSpider(CrawlSpider):
         text = response.text
         matches = self.regex.findall(text)
         if matches:
-            self.logger.error(f"{matches}")
+            self.logger.debug(f"{matches}")
             return FoundItem(url=response.url, matches=matches)

huggingface/huggingface/spiders/search.py

@@ -4,14 +4,14 @@ import time
 
 import coloredlogs
 import scrapy
-
+from string import ascii_lowercase, ascii_uppercase
 from ..pipelines import FoundItem
 
 
 class HFSearchSpider(scrapy.Spider):
     name = "huggingface_search"
     allowed_domains = ["huggingface.co"]
-    start_urls = ['https://huggingface.co/search/full-text?q=sk-{}&limit=100&skip={}'.format(i, j * 100) for i in ['a', 'h', 'm', 'q'] for j in range(10)]
+    start_urls = ['https://huggingface.co/search/full-text?q=sk-{}&limit=100&skip={}'.format(i, j * 100) for i in [*ascii_lowercase, *ascii_uppercase, *range(10), 'etk', 'ant'] for j in range(10)]
     logger = logging.getLogger('HFSearchSpider')
     logger.setLevel(logging.DEBUG)
     coloredlogs.install(logger=logger, level=logging.DEBUG)
@@ -27,5 +27,5 @@ class HFSearchSpider(scrapy.Spider):
         for key in keys:
             found_keys.add(''.join(key))
         if len(found_keys):
-            self.logger.error(f"{found_keys}")
+            self.logger.debug(f"{found_keys}")
             yield FoundItem(url=response.url, matches=found_keys)