Merge branch 'update' of github.com:mfmezger/conversational-agent-lan…

…gchain into update

agent/api.py

@@ -27,15 +27,12 @@
     QAResponse,
     SearchResponse,
 )
-from agent.utils.vdb import (
+from agent.utils.utility import (
     combine_text_from_list,
     create_tmp_folder,
-    initialize_aleph_alpha_vector_db,
-    initialize_gpt4all_vector_db,
-    initialize_open_ai_vector_db,
-    load_vec_db_conn,
     validate_token,
 )
+from agent.utils.vdb import initialize_aleph_alpha_vector_db, initialize_gpt4all_vector_db, initialize_open_ai_vector_db, load_vec_db_conn
 
 # add file logger for loguru
 # logger.add("logs/file_{time}.log", backtrace=False, diagnose=False)

agent/backend/aleph_alpha_service.py

@@ -17,7 +17,6 @@
 from langchain.text_splitter import NLTKTextSplitter
 from langchain_community.document_loaders import DirectoryLoader, PyPDFium2Loader, TextLoader
 from langchain_community.embeddings import AlephAlphaAsymmetricSemanticEmbedding
-from langchain_community.vectorstores import Qdrant
 from loguru import logger
 from omegaconf import DictConfig
 from ultra_simple_config import load_config
@@ -28,12 +27,8 @@
     RAGRequest,
     SearchRequest,
 )
-from agent.utils.utility import (
-    convert_qdrant_result_to_retrieval_results,
-    generate_collection_aleph_alpha,
-    generate_prompt,
-)
-from agent.utils.vdb import init_vdb
+from agent.utils.utility import convert_qdrant_result_to_retrieval_results, generate_prompt
+from agent.utils.vdb import generate_collection_aleph_alpha, init_vdb
 
 nltk.download("punkt")  # This needs to be installed for the tokenizer to work.
 load_dotenv()
@@ -67,7 +62,13 @@ def __init__(self, cfg: DictConfig, collection_name: str, token: str) -> None:
         else:
             self.collection_name = self.cfg.qdrant.collection_name_aa
 
-        self.vector_db = self.get_db_connection(self.collection_name)
+        embedding = AlephAlphaAsymmetricSemanticEmbedding(
+            model=self.cfg.aleph_alpha_embeddings.model_name,
+            aleph_alpha_api_key=self.aleph_alpha_token,
+            normalize=self.cfg.aleph_alpha_embeddings.normalize,
+            compress_to_size=self.cfg.aleph_alpha_embeddings.compress_to_size,
+        )
+        self.vector_db = init_vdb(cfg=self.cfg, collection_name=collection_name, embedding=embedding)
 
     def get_tokenizer(self) -> None:
         """Initialize the tokenizer."""
@@ -84,30 +85,11 @@ def count_tokens(self, text: str) -> int:
         Returns:
         -------
             int: Number of tokens.
+
         """
         tokens = self.tokenizer.encode(text)
         return len(tokens)
 
-    def get_db_connection(self, collection_name: str) -> Qdrant:
-        """Initializes a connection to the Qdrant DB.
-
-        Args:
-        ----
-            collection_name (str): The name of the collection in the Qdrant DB.
-
-        Returns:
-        -------
-            Qdrant: The Qdrant DB connection.
-        """
-        embedding = AlephAlphaAsymmetricSemanticEmbedding(
-            model=self.cfg.aleph_alpha_embeddings.model_name,
-            aleph_alpha_api_key=self.aleph_alpha_token,
-            normalize=self.cfg.aleph_alpha_embeddings.normalize,
-            compress_to_size=self.cfg.aleph_alpha_embeddings.compress_to_size,
-        )
-
-        return init_vdb(self.cfg, collection_name, embedding)
-
     def create_collection(self, name: str) -> bool:
         """Create a new collection in the Qdrant DB.
 
@@ -130,6 +112,7 @@ def summarize_text(self, text: str) -> str:
         Returns:
         -------
             str: The summary of the text.
+
         """
         # TODO: rewrite because deprecated.
         client = Client(token=self.aleph_alpha_token)
@@ -154,6 +137,7 @@ def generate(self, text: str) -> str:
         Raises:
         ------
             ValueError: If the text or token is None or empty, or if the response or completion is empty.
+
         """
         if not text:
             msg = "Text cannot be None or empty."
@@ -196,6 +180,7 @@ def embed_documents(self, directory: str, file_ending: str = ".pdf") -> None:
         Returns:
         -------
             None
+
         """
         if file_ending == ".pdf":
             loader = DirectoryLoader(directory, glob="*" + file_ending, loader_cls=PyPDFium2Loader)
@@ -235,6 +220,7 @@ def search(self, search: SearchRequest, filtering: Filtering) -> list[tuple[Lang
         Returns:
         -------
             List[Tuple[Document, float]]: A list of tuples containing the documents and their similarity scores.
+
         """
         docs = self.vector_db.similarity_search_with_score(query=search.query, k=search.amount, score_threshold=filtering.threshold)
         logger.info(f"SUCCESS: {len(docs)} Documents found.")
@@ -253,6 +239,7 @@ def rag(self, rag: RAGRequest, search: SearchRequest, filtering: Filtering) -> t
         Returns:
         -------
             Tuple[str, str, List[RetrievalResults]]: The answer, the prompt and the metadata.
+
         """
         documents = self.search(search=search, filtering=filtering)
         if search.amount == 0:
@@ -323,6 +310,7 @@ def process_documents_aleph_alpha(self, folder: str, processing_type: str) -> li
         Raises:
         ------
             ValueError: If the type is not one of 'qa', 'summarization', or 'invoice'.
+
         """
         # load the documents
         loader = DirectoryLoader(folder, glob="*.pdf", loader_cls=PyPDFium2Loader)

agent/backend/cohere_service.py

@@ -1 +1,34 @@
 """Cohere Backend."""
+from dotenv import load_dotenv
+
+from agent.backend.LLMBase import LLMBackend, LLMBase
+from agent.data_model.request_data_model import (
+    Filtering,
+    RAGRequest,
+    SearchRequest,
+)
+
+load_dotenv()
+
+
+class CohereService(LLMBase):
+
+    """Wrapper for cohere llms."""
+
+    def embed_documents(self, directory: str, llm_backend: LLMBackend) -> None:
+        """Embedd new docments in the Qdrant DB."""
+
+    def create_collection(self, name: str) -> bool:
+        """Create a new collection in the Vector Database."""
+
+    def search(self, search: SearchRequest, filtering: Filtering) -> list:
+        """Searches the documents in the Qdrant DB with semantic search."""
+
+    def generate(self, prompt: str) -> str:
+        """Generate text from a prompt."""
+
+    def rag(self, rag: RAGRequest, search: SearchRequest, filtering: Filtering) -> tuple:
+        """Retrieval Augmented Generation."""
+
+    def summarize_text(self, text: str) -> str:
+        """Summarize text."""

agent/backend/gpt4all_service.py

@@ -5,7 +5,6 @@
 from langchain.text_splitter import NLTKTextSplitter
 from langchain_community.document_loaders import DirectoryLoader, PyPDFium2Loader, TextLoader
 from langchain_community.embeddings import GPT4AllEmbeddings
-from langchain_community.vectorstores.qdrant import Qdrant
 from loguru import logger
 from omegaconf import DictConfig
 from ultra_simple_config import load_config
@@ -17,12 +16,10 @@
     RAGRequest,
     SearchRequest,
 )
-from agent.utils.utility import (
-    convert_qdrant_result_to_retrieval_results,
-    generate_collection_gpt4all,
-    generate_prompt,
-)
-from agent.utils.vdb import init_vdb
+from agent.utils.utility import convert_qdrant_result_to_retrieval_results, generate_prompt
+from agent.utils.vdb import generate_collection_gpt4all, init_vdb
+
+# nltk.download("punkt")
 
 load_dotenv()
 
@@ -42,24 +39,24 @@ def __init__(self, cfg: DictConfig, collection_name: str, token: str | None) ->
         else:
             self.collection_name = self.cfg.qdrant.collection_name_gpt4all
 
-        self.vector_db = self.get_db_connection()
+        embedding = GPT4AllEmbeddings()
+        self.vector_db = init_vdb(cfg=self.cfg, collection_name=collection_name, embedding=embedding)
 
-    def get_db_connection(self) -> Qdrant:
-        """Initializes a connection to the Qdrant DB.
+        # create retriever from the vector database
 
-        Args:
-        ----
-            cfg (DictConfig): The configuration file loaded via OmegaConf.
-            aleph_alpha_token (str): The Aleph Alpha API token.
-
-        Returns:
-        -------
-            Qdrant: The Qdrant DB connection.
+        # query = "Was ist Attention?"
+        # results = retriever.invoke(query=query)
+        # print(results)
+        # self.chain =
 
-        """
-        embedding = GPT4AllEmbeddings()
+    def create_search_chain(self, search_kwargs: dict[str, any] | None = None):
+        if search_kwargs is None:
+            search_kwargs = {}
+        return self.vector_db.as_retriever(search_kwargs=search_kwargs)
 
-        return init_vdb(self.cfg, self.collection_name, embedding)
+    def create_rag_chain(self, search_chain):
+        llm = GPT4All(self.cfg.gpt4all_completion.completion_model)
+        return search_chain | llm
 
     def create_collection(self, name: str) -> bool:
         """Create a new collection in the Vector Database."""
@@ -186,23 +183,31 @@ def rag(self, rag_request: RAGRequest, search: SearchRequest, filtering: Filteri
 
 
 if __name__ == "__main__":
+    query = "Was ist Attention?"
+
     gpt4all_service = GPT4AllService(collection_name="gpt4all", token="")
 
-    gpt4all_service.embed_documents(directory="tests/resources/")
+    # gpt4all_service.embed_documents(directory="tests/resources/")
+
+    retriever = gpt4all_service.create_search_chain(search_kwargs={"k": 3})
+
+    results = (retriever.invoke(query),)  # config={'callbacks': [ConsoleCallbackHandler()]})
+
+    rag_chain = gpt4all_service.create_rag_chain(search_chain=retriever)
 
-    docs = gpt4all_service.search(SearchRequest(query="Was ist Attention?", amount=3), Filtering(threshold=0.0, collection_name="gpt4all"))
+    # docs = gpt4all_service.search(SearchRequest(query, amount=3), Filtering(threshold=0.0, collection_name="gpt4all"))
 
-    logger.info(f"Documents: {docs}")
+    # logger.info(f"Documents: {docs}")
 
-    answer, prompt, meta_data = gpt4all_service.rag(
-        RAGRequest(language="detect", history={}),
-        SearchRequest(
-            query="Was ist Attention?",
-            amount=3,
-        ),
-        Filtering(threshold=0.0, collection_name="gpt4all"),
-    )
+    # answer, prompt, meta_data = gpt4all_service.rag(
+    #     RAGRequest(language="detect", history={}),
+    #     SearchRequest(
+    #         query=query,
+    #         amount=3,
+    #     ),
+    #     Filtering(threshold=0.0, collection_name="gpt4all"),
+    # )
 
-    logger.info(f"Answer: {answer}")
-    logger.info(f"Prompt: {prompt}")
-    logger.info(f"Metadata: {meta_data}")
+    # logger.info(f"Answer: {answer}")
+    # logger.info(f"Prompt: {prompt}")
+    # logger.info(f"Metadata: {meta_data}")