pembot 0.1.8__tar.gz → 0.1.9__tar.gz

{pembot-0.1.8 → pembot-0.1.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pembot
-Version: 0.1.8
+Version: 0.1.9
 Summary: A Python Package to convert PEM blog content to usseful information by leveraging LLMs
 Author-email: cyto <aryan_sidhwani@protonmail.com>
 License-Expression: MIT

pembot-0.1.9/pembot/.git/COMMIT_EDITMSG

@@ -0,0 +1 @@
+added DeepSeek-OCR as a model option to use spaces

{pembot-0.1.8 → pembot-0.1.9}/pembot/.git/logs/HEAD

@@ -20,3 +20,4 @@ e6adbc3c373070269f97ef82d4f63027d7878f67 81d01e1c63d48b096c77aae83471d42272ca9fc
 81d01e1c63d48b096c77aae83471d42272ca9fce e89cb4f5af158d26dcff5eed03dba6671a818739 cyto <silverstone965@gmail.com> 1758891589 +0530	commit: cyto/made the excel convertor compatible with suffix input from both system file and with bytes + manual suffix input
 e89cb4f5af158d26dcff5eed03dba6671a818739 e3c62c141fc65ef2be0095c49b23e06263f0b734 cyto <silverstone965@gmail.com> 1758892053 +0530	commit: silly willy mistake
 e3c62c141fc65ef2be0095c49b23e06263f0b734 d440b20aae1265dabbd3ddaafb24c35e40e3ab3c cyto <silverstone965@gmail.com> 1759127664 +0530	commit: cyto/fixed the excel file input bug, in the conversion method; added exceptions where there is invalid input
+d440b20aae1265dabbd3ddaafb24c35e40e3ab3c 8d58ec13be39949ecfe7211b42c56acd2a83dc72 cyto <silverstone965@gmail.com> 1761392079 +0530	commit: added DeepSeek-OCR as a model option to use spaces

{pembot-0.1.8 → pembot-0.1.9}/pembot/.git/logs/refs/heads/main

@@ -20,3 +20,4 @@ e6adbc3c373070269f97ef82d4f63027d7878f67 81d01e1c63d48b096c77aae83471d42272ca9fc
 81d01e1c63d48b096c77aae83471d42272ca9fce e89cb4f5af158d26dcff5eed03dba6671a818739 cyto <silverstone965@gmail.com> 1758891589 +0530	commit: cyto/made the excel convertor compatible with suffix input from both system file and with bytes + manual suffix input
 e89cb4f5af158d26dcff5eed03dba6671a818739 e3c62c141fc65ef2be0095c49b23e06263f0b734 cyto <silverstone965@gmail.com> 1758892053 +0530	commit: silly willy mistake
 e3c62c141fc65ef2be0095c49b23e06263f0b734 d440b20aae1265dabbd3ddaafb24c35e40e3ab3c cyto <silverstone965@gmail.com> 1759127664 +0530	commit: cyto/fixed the excel file input bug, in the conversion method; added exceptions where there is invalid input
+d440b20aae1265dabbd3ddaafb24c35e40e3ab3c 8d58ec13be39949ecfe7211b42c56acd2a83dc72 cyto <silverstone965@gmail.com> 1761392079 +0530	commit: added DeepSeek-OCR as a model option to use spaces

{pembot-0.1.8 → pembot-0.1.9}/pembot/.git/logs/refs/remotes/origin/main

@@ -19,3 +19,4 @@ e6adbc3c373070269f97ef82d4f63027d7878f67 81d01e1c63d48b096c77aae83471d42272ca9fc
 81d01e1c63d48b096c77aae83471d42272ca9fce e89cb4f5af158d26dcff5eed03dba6671a818739 cyto <silverstone965@gmail.com> 1758891600 +0530	update by push
 e89cb4f5af158d26dcff5eed03dba6671a818739 e3c62c141fc65ef2be0095c49b23e06263f0b734 cyto <silverstone965@gmail.com> 1758892065 +0530	update by push
 e3c62c141fc65ef2be0095c49b23e06263f0b734 d440b20aae1265dabbd3ddaafb24c35e40e3ab3c cyto <silverstone965@gmail.com> 1759127680 +0530	update by push
+d440b20aae1265dabbd3ddaafb24c35e40e3ab3c 8d58ec13be39949ecfe7211b42c56acd2a83dc72 cyto <silverstone965@gmail.com> 1761392127 +0530	update by push

pembot-0.1.9/pembot/.git/objects/8d/58ec13be39949ecfe7211b42c56acd2a83dc72

@@ -0,0 +1 @@
+x
��Mj�0@�u�ٗ�ѯc(����	F3��Բ��z���~��q-e���}�*�L�<%o��!���O�40KD<{�f�]�f�Dj]�B9�!���TOٳ�[����Wxm��{�u�1���B��µ������a�g�͡�`�RC"*�}����>.@

pembot-0.1.9/pembot/.git/refs/heads/main

@@ -0,0 +1 @@
+8d58ec13be39949ecfe7211b42c56acd2a83dc72

pembot-0.1.9/pembot/.git/refs/remotes/origin/main

@@ -0,0 +1 @@
+8d58ec13be39949ecfe7211b42c56acd2a83dc72

{pembot-0.1.8 → pembot-0.1.9}/pembot/__init__.py

@@ -1,6 +1,6 @@
 """
 A Python Package to convert PEM blog content to usseful information by leveraging LLMs
 """
-__version__ = '0.1.8'
+__version__ = '0.1.9'
 from .main import save_to_json_file, make_query
 __all__ = ["save_to_json_file", "make_query"]

{pembot-0.1.8 → pembot-0.1.9}/pembot/config/config.yaml

@@ -2,4 +2,4 @@ OUTPUT_DIR: /home/cyto/dev/pembotdir
 PAGE_DELIMITER: ___________________________ NEXT PAGE ___________________________
 app:
   name: pembot
-  version: 0.1.8
+  version: 0.1.9

{pembot-0.1.8 → pembot-0.1.9}/pembot/pdf2markdown/extract.py

@@ -88,6 +88,9 @@ class MarkdownPDFExtractor(PDFExtractor):
 
             # zerogpu public
             self.nclient= Client("deepak-mehta/ocr-simplify", hf_token= os.getenv('HF_TOKEN', ''))
+        elif "DeepSeek-OCR" in self.MODEL_NAME:
+            # zerogpu private
+            self.dclient= Client("deepak-mehta/deepseek-ocr", hf_token= os.getenv('HF_TOKEN', ''))
 
 
         self.markdown_content= ""
@@ -150,7 +153,7 @@ class MarkdownPDFExtractor(PDFExtractor):
             )
             # print("response :", response)
             return response.text
-        elif 'nanonet' in model_name:
+        elif 'nanonet' in model_name or 'DeepSeek-OCR' in model_name:
 
             result= ""
             try:
@@ -159,24 +162,33 @@ class MarkdownPDFExtractor(PDFExtractor):
                     print("file name: ", temp_file.name)
                     gr_image= handle_file(temp_file.name)
                     print("gr image : ", gr_image)
-                    result = self.nclient.predict(
-                        # model_name="Nanonets-OCR-s",
-                        # text= prompt,
-                        gr_image,
-                        # max_new_tokens=max_new_tokens,
-                        # temperature=0.6,
-                        # top_p=0.9,
-                        # top_k=50,
-                        # repetition_penalty=1.2,
-
-                        # prithiv model
-                        # api_name="/generate_image"
-
-                        max_new_tokens,
-
-                        # spaces zerogpu
-                        api_name="/predict"
-                    )
+                    if 'nanonet' in model_name:
+                        result = self.nclient.predict(
+                            # model_name="Nanonets-OCR-s",
+                            # text= prompt,
+                            gr_image,
+                            # max_new_tokens=max_new_tokens,
+                            # temperature=0.6,
+                            # top_p=0.9,
+                            # top_k=50,
+                            # repetition_penalty=1.2,
+
+                            # prithiv model
+                            # api_name="/generate_image"
+
+                            max_new_tokens,
+
+                            # spaces zerogpu
+                            api_name="/predict"
+                        )
+                    else:
+                        result = self.dclient.predict(
+                            file_input=gr_image,
+                            prompt_type= 'markdown',
+
+                            # spaces zerogpu
+                            api_name="/predict"
+                        )
                     print("ocr'd: ", result[:100] + "...")
             except Exception as e:
                 print("Error during nanonet inference", e)

{pembot-0.1.8 → pembot-0.1.9}/pembot/requirements.txt

@@ -45,7 +45,7 @@ pandas==2.3.0
 pathlib==1.0.1
 pdfminer.six==20250506
 pdfplumber==0.11.7
-pembot==0.1.8
+pembot==0.1.9
 pillow==11.3.0
 primp==0.15.0
 pyasn1==0.6.1

pembot-0.1.8/pembot/.git/COMMIT_EDITMSG

@@ -1 +0,0 @@
-cyto/fixed the excel file input bug, in the conversion method; added exceptions where there is invalid input

pembot-0.1.8/pembot/.git/refs/heads/main

@@ -1 +0,0 @@
-d440b20aae1265dabbd3ddaafb24c35e40e3ab3c

pembot-0.1.8/pembot/.git/refs/remotes/origin/main

@@ -1 +0,0 @@
-d440b20aae1265dabbd3ddaafb24c35e40e3ab3c