pembot 0.0.8__tar.gz → 0.1.0__tar.gz

{pembot-0.0.8 → pembot-0.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pembot
-Version: 0.0.8
+Version: 0.1.0
 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-0.1.0/pembot/.git/COMMIT_EDITMSG

@@ -0,0 +1 @@
+fixed the type_info array pushing bug

{pembot-0.0.8 → pembot-0.1.0}/pembot/.git/logs/HEAD

@@ -8,3 +8,5 @@ eb75e1c49f1e5b79dca17ccdbec8067756523238 0bdb4169fc0f312b8698f1df17a258fff163aea
 0bdb4169fc0f312b8698f1df17a258fff163aeaa 9528bbccd167e3f4ad583a1ae9fac98a52620e27 cyto <silverstone965@gmail.com> 1750947488 +0530	commit: handled local llm nonexistent error properly for choice of just passing None as llm_client;
 9528bbccd167e3f4ad583a1ae9fac98a52620e27 ef0503a60244391590b16042019032e91d7cc30d cyto <silverstone965@gmail.com> 1751872559 +0530	commit: added a model_name_parameter to change models quicky
 ef0503a60244391590b16042019032e91d7cc30d af80ddb5890f062e364ea8ade2d602df4e12de8c cyto <silverstone965@gmail.com> 1751896700 +0530	commit: added model name to convertor
+af80ddb5890f062e364ea8ade2d602df4e12de8c 0d28f73897db0c9a9351ee9e64d2a0fe27db2705 cyto <silverstone965@gmail.com> 1752081786 +0530	commit: upgraded to 0.0.8; added ability to classify the document chunks into one more filterable parameter 'type'; other query related improvements..
+0d28f73897db0c9a9351ee9e64d2a0fe27db2705 a898d2c3947d30d8be64bd2bbcef68f956d5456b cyto <silverstone965@gmail.com> 1752136060 +0530	commit: fixed the type_info array pushing bug

{pembot-0.0.8 → pembot-0.1.0}/pembot/.git/logs/refs/heads/main

@@ -8,3 +8,5 @@ eb75e1c49f1e5b79dca17ccdbec8067756523238 0bdb4169fc0f312b8698f1df17a258fff163aea
 0bdb4169fc0f312b8698f1df17a258fff163aeaa 9528bbccd167e3f4ad583a1ae9fac98a52620e27 cyto <silverstone965@gmail.com> 1750947488 +0530	commit: handled local llm nonexistent error properly for choice of just passing None as llm_client;
 9528bbccd167e3f4ad583a1ae9fac98a52620e27 ef0503a60244391590b16042019032e91d7cc30d cyto <silverstone965@gmail.com> 1751872559 +0530	commit: added a model_name_parameter to change models quicky
 ef0503a60244391590b16042019032e91d7cc30d af80ddb5890f062e364ea8ade2d602df4e12de8c cyto <silverstone965@gmail.com> 1751896700 +0530	commit: added model name to convertor
+af80ddb5890f062e364ea8ade2d602df4e12de8c 0d28f73897db0c9a9351ee9e64d2a0fe27db2705 cyto <silverstone965@gmail.com> 1752081786 +0530	commit: upgraded to 0.0.8; added ability to classify the document chunks into one more filterable parameter 'type'; other query related improvements..
+0d28f73897db0c9a9351ee9e64d2a0fe27db2705 a898d2c3947d30d8be64bd2bbcef68f956d5456b cyto <silverstone965@gmail.com> 1752136060 +0530	commit: fixed the type_info array pushing bug

{pembot-0.0.8 → pembot-0.1.0}/pembot/.git/logs/refs/remotes/origin/main

@@ -7,3 +7,5 @@ eb75e1c49f1e5b79dca17ccdbec8067756523238 0bdb4169fc0f312b8698f1df17a258fff163aea
 0bdb4169fc0f312b8698f1df17a258fff163aeaa 9528bbccd167e3f4ad583a1ae9fac98a52620e27 cyto <silverstone965@gmail.com> 1750947502 +0530	update by push
 9528bbccd167e3f4ad583a1ae9fac98a52620e27 ef0503a60244391590b16042019032e91d7cc30d cyto <silverstone965@gmail.com> 1751872581 +0530	update by push
 ef0503a60244391590b16042019032e91d7cc30d af80ddb5890f062e364ea8ade2d602df4e12de8c cyto <silverstone965@gmail.com> 1751896713 +0530	update by push
+af80ddb5890f062e364ea8ade2d602df4e12de8c 0d28f73897db0c9a9351ee9e64d2a0fe27db2705 cyto <silverstone965@gmail.com> 1752081801 +0530	update by push
+0d28f73897db0c9a9351ee9e64d2a0fe27db2705 a898d2c3947d30d8be64bd2bbcef68f956d5456b cyto <silverstone965@gmail.com> 1752136119 +0530	update by push

pembot-0.1.0/pembot/.git/objects/0d/28f73897db0c9a9351ee9e64d2a0fe27db2705

@@ -0,0 +1,3 @@
+x
��An� E��)��E�B`���z�	*�!#��%W�����߷�)�
+J��Z��K%�����HZMl���4"�� cz�mXx���
+�Y�IxaVfd�H<����@l]�G]r
w�_{�o.{�+OF�<�ػ��A�� ��[�B+�5�
V��jwl��P���Eo�t�8��y�.��ꅁ�;��-���CX�D���K�M������c��Vύo�m����rBስل���������jv�

pembot-0.1.0/pembot/.git/objects/a8/98d2c3947d30d8be64bd2bbcef68f956d5456b

@@ -0,0 +1 @@
+x
��݊� @�^�s�P������2�MfR���<��~pश,�A�;�HC*:�h]��8T���	)��,ņ�V��X���Q���XE�

pembot-0.1.0/pembot/.git/refs/heads/main

@@ -0,0 +1 @@
+a898d2c3947d30d8be64bd2bbcef68f956d5456b

pembot-0.1.0/pembot/.git/refs/remotes/origin/main

@@ -0,0 +1 @@
+a898d2c3947d30d8be64bd2bbcef68f956d5456b

{pembot-0.0.8 → pembot-0.1.0}/pembot/TextEmbedder/mongodb_embedder.py

@@ -69,22 +69,27 @@ def search_within_document(
     if document_belongs_to_a_type:
         project_dict['type']= 1
 
+    vectorSearchParams= {
+        'queryVector': aggregate_query_embedding,
+        'path': 'embedding',
+
+        #number of nearest neighbors to consider
+        'numCandidates': 100,
+        'limit': limit,
+        'index': index_name,
+
+    }
+
+    #filter (if a type or docid filter is given) to search only within the specified documents search space
+    if document_name_id:
+        vectorSearchParams['filter']= (
+            { "type": {"$in": [document_belongs_to_a_type ]} } if document_belongs_to_a_type else
+            { 'docId': document_name_id }
+        )
+
     pipeline = [
         {
-            '$vectorSearch': {
-                'queryVector': aggregate_query_embedding,
-                'path': 'embedding',
-
-                #number of nearest neighbors to consider
-                'numCandidates': 100,
-                'limit': limit,
-                'index': index_name,
-
-                #filter to search only within the specified document
-                'filter':
-                    { "type": {"$in": [document_belongs_to_a_type ]} } if document_belongs_to_a_type else
-                    { 'docId': document_name_id }
-            }
+            '$vectorSearch': vectorSearchParams
         },
 
         # to exclude the MongoDB internal _id
@@ -93,6 +98,8 @@ def search_within_document(
         }
     ]
 
+
+
     # print("sesraching now:")
     results = list(embeddings_collection.aggregate(pipeline))
     # print("search results: ", results)
@@ -253,8 +260,8 @@ def process_document_and_embed(
                     {'docId': document_name_id, 'chunk_number': i + 1},
                     {
                         '$set': doc_set,
-                        '$push': {
-                            "type": type_info
+                        '$addToSet': {
+                            "type": { '$each': type_info }
                         }
                     },
                     upsert=True

{pembot-0.0.8 → pembot-0.1.0}/pembot/__init__.py

@@ -1,6 +1,6 @@
 """
 A Python Package to convert PEM blog content to usseful information by leveraging LLMs
 """
-__version__ = '0.0.8'
+__version__ = '0.1.0'
 from .main import save_to_json_file, make_query
 __all__ = ["save_to_json_file", "make_query"]

{pembot-0.0.8 → pembot-0.1.0}/pembot/config/config.yaml

@@ -2,4 +2,4 @@ OUTPUT_DIR: /home/cyto/dev/pembotdir
 PAGE_DELIMITER: ___________________________ NEXT PAGE ___________________________
 app:
   name: pembot
-  version: 0.0.8
+  version: 0.1.0

{pembot-0.0.8 → pembot-0.1.0}/pembot/query.py

@@ -82,7 +82,13 @@ def multi_embedding_average(llm_client, inference_client, descriptions, model= "
 
 
 
-def rag_query_llm(db_client, llm_client, inference_client, user_query: str, document_id: str, required_fields_descriptions: list[str], model_name: str = "deepseek-ai/DeepSeek-R1-0528-Qwen3-8B", ollama_base_url: str = "http://localhost:11434", no_of_fields= 4, embedding_model= "BAAI/bge-en-icl", llm_provider_name: PROVIDER_T= "novita", index_name: str= "test_search", embeddings_collection= "doc_chunks", document_belongs_to_a_type= ""):
+def rag_query_llm(db_client, llm_client, inference_client,
+    user_query: str, document_id: str, required_fields_descriptions: list[str],
+    model_name: str = "deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
+    ollama_base_url: str = "http://localhost:11434", no_of_fields= 4,
+    embedding_model= "BAAI/bge-en-icl", llm_provider_name: PROVIDER_T= "novita",
+    index_name: str= "test_search", embeddings_collection= "doc_chunks",
+    document_belongs_to_a_type= "", prompt_prefix= ""):
     """
     Performs a RAG (Retrieval Augmented Generation) query using a Hugging Face
     embedding model, ChromaDB for retrieval, and a local Ollama model for generation.
@@ -139,6 +145,7 @@ def rag_query_llm(db_client, llm_client, inference_client, user_query: str, docu
     If you don't know the answer, just say that you don't know, don't try to make up an answer.
 
     Context:
+    {prompt_prefix}
     {context}
 
     Question: {user_query}

{pembot-0.0.8 → pembot-0.1.0}/pembot/requirements.txt

@@ -40,7 +40,7 @@ pandas==2.3.0
 pathlib==1.0.1
 pdfminer.six==20250506
 pdfplumber==0.11.7
-pembot==0.0.6
+pembot==0.0.8
 pillow==11.3.0
 pyasn1==0.6.1
 pyasn1_modules==0.4.2

pembot-0.0.8/pembot/.git/COMMIT_EDITMSG

@@ -1 +0,0 @@
-added model name to convertor

pembot-0.0.8/pembot/.git/refs/heads/main

@@ -1 +0,0 @@
-af80ddb5890f062e364ea8ade2d602df4e12de8c

pembot-0.0.8/pembot/.git/refs/remotes/origin/main

@@ -1 +0,0 @@
-af80ddb5890f062e364ea8ade2d602df4e12de8c