removing the tokens concept.

agent/api.py

@@ -163,6 +163,8 @@ async def embedd_text(embedding: EmbeddTextRequest, llm_backend: LLMBackend) ->
 
     """
     logger.info("Embedding Text")
+
+    # TODO: REWORK THE TOKEN
     token = validate_token(token=llm_backend.token, llm_backend=llm_backend, aleph_alpha_key=ALEPH_ALPHA_API_KEY, openai_key=OPENAI_API_KEY)
 
     service = LLMContext(LLMStrategyFactory.get_strategy(strategy_type=llm_backend.llm_provider, token=token, collection_name=llm_backend.collection_name))
@@ -195,7 +197,6 @@ def search(search: SearchParams, llm_backend: LLMBackend) -> list[SearchResponse
 
     """
     logger.info("Searching for Documents")
-    llm_backend.token = validate_token(token=llm_backend.token, llm_backend=llm_backend, aleph_alpha_key=ALEPH_ALPHA_API_KEY, openai_key=OPENAI_API_KEY)
 
     service = LLMContext(LLMStrategyFactory.get_strategy(strategy_type=llm_backend.llm_provider, token=llm_backend.token, collection_name=llm_backend.collection_name))
 

agent/backend/LLMBase.py

@@ -9,9 +9,8 @@ class LLMBase(ABC):
     """The LLM Base Strategy."""
 
     @abstractmethod
-    def __init__(self, token: str | None, collection_name: str | None) -> None:
+    def __init__(self, collection_name: str | None) -> None:
         """Init the LLM Base."""
-        self.token = token
         self.collection_name = collection_name
 
     @abstractmethod

agent/backend/LLMStrategy.py

@@ -66,9 +66,9 @@ def __init__(self, llm: LLMBase) -> None:
         """Init the Context."""
         self.llm = llm
 
-    def change_strategy(self, strategy_type: str, token: str, collection_name: str) -> None:
+    def change_strategy(self, strategy_type: str, collection_name: str) -> None:
         """Changes the strategy using the Factory."""
-        self.llm = LLMStrategyFactory.get_strategy(strategy_type=strategy_type, token=token, collection_name=collection_name)
+        self.llm = LLMStrategyFactory.get_strategy(strategy_type=strategy_type, collection_name=collection_name)
 
     def search(self, search: SearchParams) -> list:
         """Wrapper for the search."""

agent/backend/aleph_alpha_service.py

@@ -45,13 +45,10 @@ class AlephAlphaService(LLMBase):
     """Aleph Alpha Strategy implementation."""
 
     @load_config(location="config/main.yml")
-    def __init__(self, cfg: DictConfig, collection_name: str, token: str) -> None:
+    def __init__(self, cfg: DictConfig, collection_name: str) -> None:
         """Initialize the Aleph Alpha Service."""
-        super().__init__(token=token, collection_name=collection_name)
+        super().__init__(collection_name=collection_name)
         """Initialize the Aleph Alpha Service."""
-        if token:
-            os.environ["ALEPH_ALPHA_API_KEY"] = token
-
         self.cfg = cfg
 
         if collection_name:
@@ -61,7 +58,6 @@ def __init__(self, cfg: DictConfig, collection_name: str, token: str) -> None:
 
         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,
         )
@@ -72,7 +68,7 @@ def __init__(self, cfg: DictConfig, collection_name: str, token: str) -> None:
 
     def get_tokenizer(self) -> None:
         """Initialize the tokenizer."""
-        client = Client(token=self.aleph_alpha_token)
+        client = Client(token=os.getenv("ALEPH_ALPHA_API_KEY"))
         self.tokenizer = client.tokenizer("luminous-base")
 
     def count_tokens(self, text: str) -> int:
@@ -115,7 +111,7 @@ def summarize_text(self, text: str) -> str:
 
         """
         # TODO: rewrite because deprecated.
-        client = Client(token=self.aleph_alpha_token)
+        client = Client(token=os.getenv("ALEPH_ALPHA_API_KEY"))
         document = Document.from_text(text=text)
         request = SummarizationRequest(document=document)
         response = client.summarize(request=request)

agent/backend/cohere_service.py

@@ -1,5 +1,4 @@
 """Cohere Backend."""
-import os
 
 from dotenv import load_dotenv
 from langchain_cohere import ChatCohere, CohereEmbeddings
@@ -30,12 +29,9 @@ class CohereService(LLMBase):
     """Wrapper for cohere llms."""
 
     @load_config(location="config/main.yml")
-    def __init__(self, cfg: DictConfig, collection_name: str | None, token: str | None) -> None:
+    def __init__(self, cfg: DictConfig, collection_name: str | None) -> None:
         """Init the Cohere Service."""
-        super().__init__(token=token, collection_name=collection_name)
-
-        if token:
-            os.environ["COHERE_API_KEY"] = token
+        super().__init__(collection_name=collection_name)
 
         self.cfg = cfg
 

agent/backend/gpt4all_service.py

@@ -30,15 +30,11 @@ class GPT4AllService(LLMBase):
     """GPT4ALL Backend Service."""
 
     @load_config(location="config/main.yml")
-    def __init__(self, cfg: DictConfig, collection_name: str, token: str | None) -> None:
+    def __init__(self, cfg: DictConfig, collection_name: str) -> None:
         """Init the GPT4ALL Service."""
-        self.cfg = cfg
-        self.token = token
+        super().__init__(collection_name=collection_name)
 
-        if collection_name:
-            self.collection_name = collection_name
-        else:
-            self.collection_name = self.cfg.qdrant.collection_name_gpt4all
+        self.cfg = cfg
 
         embedding = GPT4AllEmbeddings(model_name="nomic-embed-text-v1.5.f16.gguf")
 

agent/backend/ollama_service.py

@@ -30,9 +30,9 @@ class OllamaService(LLMBase):
     """Wrapper for Ollama llms."""
 
     @load_config(location="config/main.yml")
-    def __init__(self, cfg: DictConfig, collection_name: str | None, token: str | None) -> None:
+    def __init__(self, cfg: DictConfig, collection_name: str | None) -> None:
         """Init the Ollama Service."""
-        super().__init__(token=token, collection_name=collection_name)
+        super().__init__(collection_name=collection_name)
 
         self.cfg = cfg
 
@@ -122,7 +122,7 @@ def summarize_text(self, text: str) -> str:
 if __name__ == "__main__":
     query = "Was ist Attention?"
 
-    Ollama_service = OllamaService(collection_name="", token="")
+    Ollama_service = OllamaService(collection_name="")
 
     Ollama_service.embed_documents(directory="tests/resources/")
 

agent/backend/open_ai_service.py

@@ -29,14 +29,11 @@ class OpenAIService(LLMBase):
     """OpenAI Backend Service."""
 
     @load_config(location="config/main.yml")
-    def __init__(self, cfg: DictConfig, collection_name: str, token: str) -> None:
+    def __init__(self, cfg: DictConfig, collection_name: str) -> None:
         """Init the OpenAI Service."""
-        super().__init__(token=token, collection_name=collection_name)
+        super().__init__(collection_name=collection_name)
 
         """Openai Service."""
-        if token:
-            os.environ["COHERE_API_KEY"] = token
-
         self.cfg = cfg
 
         if collection_name:

agent/utils/utility.py

@@ -8,7 +8,6 @@
 from loguru import logger
 
 from agent.data_model.internal_model import RetrievalResults
-from agent.data_model.request_data_model import LLMProvider
 
 # add new languages to detect here
 languages = [Language.ENGLISH, Language.GERMAN]
@@ -132,63 +131,6 @@ def load_prompt_template(prompt_name: str, task: str) -> PromptTemplate:
     return prompt_template
 
 
-def get_token(token: str | None, llm_provider: str | LLMProvider | None, aleph_alpha_key: str | None, openai_key: str | None) -> str:
-    """Get the token from the environment variables or the parameter.
-
-    Args:
-    ----
-        token (str, optional): Token from the REST service.
-        llm_provider (Union[str, LLMProvider], optional): LLM provider. Defaults to "openai".
-        aleph_alpha_key (str, optional): Key from the .env file. Defaults to None.
-        openai_key (str, optional): Key from the .env file. Defaults to None.
-
-    Returns:
-    -------
-        str: Token for the LLM Provider of choice.
-
-    Raises:
-    ------
-        ValueError: If no token is provided.
-
-    """
-    if isinstance(llm_provider, str):
-        llm_provider = LLMProvider.normalize(llm_provider)
-
-    if token in ("string", ""):
-        token = None
-
-    if token:
-        return token
-
-    env_token = aleph_alpha_key if llm_provider == LLMProvider.ALEPH_ALPHA else openai_key
-    if not env_token and not token:
-        msg = "No token provided."
-        raise ValueError(msg)
-    return env_token
-
-
-def validate_token(token: str | None, llm_backend: str | LLMProvider, aleph_alpha_key: str | None, openai_key: str | None) -> str:
-    """Test if a token is available, and raise an error if it is missing when needed.
-
-    Args:
-    ----
-        token (str): Token from the request
-        llm_backend (str): Backend from the request
-        aleph_alpha_key (str): Key from the .env file
-        openai_key (str): Key from the .env file
-
-    Raises:
-    ------
-        ValueError: If the llm backend is AA or OpenAI and there is no token.
-
-    Returns:
-    -------
-        str: Token
-
-    """
-    return get_token(token, llm_backend.llm_provider, aleph_alpha_key, openai_key) if llm_backend != "gpt4all" else "gpt4all"
-
-
 def convert_qdrant_result_to_retrieval_results(docs: list) -> list[RetrievalResults]:
     """Converts the Qdrant result to a list of tuples.
 

config/main.yml

@@ -12,7 +12,7 @@ qdrant:
 # COHERE CONFIG
 cohere_embeddings:
   embedding_model_name: "embed-multilingual-v3.0"
-  size: 2048
+  size: 1024
 
 # OLLAMA CONFIG
 ollama_embeddings:

prompts/chat/aleph_alpha_chat.j2

@@ -0,0 +1,6 @@
+### Instruction:
+{{question}} If there's no answer, say "NO_ANSWER_IN_TEXT".
+### Input:
+Text:{{context}}
+Question:{{question}}
+### Response:

tests/unit_tests/test_utility.py

@@ -1,5 +1,5 @@
 """Tests for the utility functions."""
-from agent.utils.utility import generate_prompt, validate_token
+from agent.utils.utility import generate_prompt, set_token
 
 
 def test_generate_prompt() -> None:
@@ -37,30 +37,30 @@ def test_combine_text_from_list() -> None:
 
 def test_validate_token() -> None:
     """Test that validate_token returns the correct token."""
-    token = validate_token(token="example_token", llm_backend="openai", aleph_alpha_key="example_key_a", openai_key="example_key_o")
+    token = set_token(token="example_token", llm_backend="openai", aleph_alpha_key="example_key_a", openai_key="example_key_o")
 
     assert token == "example_token"
 
-    token = validate_token(token="", llm_backend="aleph-alpha", aleph_alpha_key="example_key_a", openai_key="example_key_o")
+    token = set_token(token="", llm_backend="aleph-alpha", aleph_alpha_key="example_key_a", openai_key="example_key_o")
 
     assert token == "example_key_a"
 
-    token = validate_token(token="", llm_backend="openai", aleph_alpha_key="example_key_a", openai_key="example_key_o")
+    token = set_token(token="", llm_backend="openai", aleph_alpha_key="example_key_a", openai_key="example_key_o")
 
     assert token == "example_key_o"
 
-    token = validate_token(token=None, llm_backend="openai", aleph_alpha_key="example_key_a", openai_key="example_key_o")
+    token = set_token(token=None, llm_backend="openai", aleph_alpha_key="example_key_a", openai_key="example_key_o")
 
     assert token == "example_key_o"
 
-    token = validate_token(token="", llm_backend="gpt4all", aleph_alpha_key="example_key_a", openai_key="example_key_o")
+    token = set_token(token="", llm_backend="gpt4all", aleph_alpha_key="example_key_a", openai_key="example_key_o")
 
     assert token == "gpt4all"
 
     from agent.data_model.request_data_model import LLMProvider
 
     backend = LLMProvider.ALEPH_ALPHA
 
-    token = validate_token(token="", llm_backend=backend, aleph_alpha_key="example_key_a", openai_key="example_key_o")
+    token = set_token(token="", llm_backend=backend, aleph_alpha_key="example_key_a", openai_key="example_key_o")
 
     assert token == "example_key_a"