langroid 0.47.0__tar.gz → 0.47.2__tar.gz

{langroid-0.47.0 → langroid-0.47.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: langroid
-Version: 0.47.0
+Version: 0.47.2
 Summary: Harness LLMs with Multi-Agent Programming
 Author-email: Prasad Chalasani <pchalasani@gmail.com>
 License: MIT

{langroid-0.47.0 → langroid-0.47.2}/langroid/agent/special/doc_chat_agent.py

@@ -237,9 +237,7 @@ class DocChatAgent(ChatAgent):
         self.chunked_docs: List[Document] = []
         self.chunked_docs_clean: List[Document] = []
         self.response: None | Document = None
-
-        if len(config.doc_paths) > 0:
-            self.ingest()
+        self.ingest()
 
     def clear(self) -> None:
         """Clear the document collection and the specific collection in vecdb"""
@@ -275,8 +273,9 @@ class DocChatAgent(ChatAgent):
             # But let's get all the chunked docs so we can
             # do keyword and other non-vector searches
             if self.vecdb is None:
-                raise ValueError("VecDB not set")
-            self.setup_documents(filter=self.config.filter)
+                logger.warning("VecDB not set: cannot ingest docs.")
+            else:
+                self.setup_documents(filter=self.config.filter)
             return
         self.ingest_doc_paths(self.config.doc_paths)  # type: ignore
 

{langroid-0.47.0 → langroid-0.47.2}/langroid/agent/special/retriever_agent.py

@@ -1,5 +1,5 @@
 """
-Deprecated: use DocChatAgent instead, with DocChatAgentConfig.retrieve_only=True,
+DEPRECATED: use DocChatAgent instead, with DocChatAgentConfig.retrieve_only=True,
 and if you want to retrieve FULL relevant doc-contents rather than just extracts,
 then set DocChatAgentConfig.extraction_granularity=-1
 
@@ -47,10 +47,12 @@ class RetrieverAgent(DocChatAgent):
         )
 
     def get_records(self) -> Sequence[Document]:
-        raise NotImplementedError
+        # subclasses should override
+        return []
 
     def ingest(self) -> None:
         records = self.get_records()
         if self.vecdb is None:
-            raise ValueError("No vector store specified")
-        self.vecdb.add_documents(records)
+            logger.warning("Vector store not configured. Cannot ingest records.")
+        else:
+            self.vecdb.add_documents(records)

{langroid-0.47.0 → langroid-0.47.2}/langroid/language_models/openai_gpt.py

@@ -48,7 +48,6 @@ from langroid.language_models.base import (
 from langroid.language_models.config import HFPromptFormatterConfig
 from langroid.language_models.model_info import (
     DeepSeekModel,
-    GeminiModel,
     OpenAI_API_ParamInfo,
 )
 from langroid.language_models.model_info import (
@@ -672,10 +671,8 @@ class OpenAIGPT(LanguageModel):
         return self.config.completion_model in openai_completion_models
 
     def is_gemini_model(self) -> bool:
-        gemini_models = [e.value for e in GeminiModel]
-        return self.chat_model_orig in gemini_models or self.chat_model_orig.startswith(
-            "gemini/"
-        )
+        """Are we using the gemini OpenAI-compatible API?"""
+        return self.chat_model_orig.startswith("gemini/")
 
     def is_deepseek_model(self) -> bool:
         deepseek_models = [e.value for e in DeepSeekModel]

{langroid-0.47.0 → langroid-0.47.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "langroid"
-version = "0.47.0"
+version = "0.47.2"
 authors = [
     {name = "Prasad Chalasani", email = "pchalasani@gmail.com"},
 ]