From 87343ec7fdfdf19aae3467efc00cf8ff5eea2b0d Mon Sep 17 00:00:00 2001
From: Emmett McFaralne <staminacode@gmail.com>
Date: Sat, 20 Jul 2024 21:10:35 -0400
Subject: [PATCH] imcrease default max pages to use max allocation of gpus

---
 thepipe/scraper.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/thepipe/scraper.py b/thepipe/scraper.py
index 0394430..3093d6b 100644
--- a/thepipe/scraper.py
+++ b/thepipe/scraper.py
@@ -149,7 +149,7 @@ def scrape_zip(file_path: str, include_regex: Optional[str] = None, verbose: boo
 def scrape_pdf(file_path: str, ai_extraction: bool = False, text_only: bool = False, verbose: bool = False) -> List[Chunk]:
     chunks = []
     BATCH_SIZE = 16
-    MAX_PAGES = 100
+    MAX_PAGES = 128
 
     if ai_extraction:
         # if using AI extraction, for each page, generate markdown and cropped figures