langroid 0.1.262__tar.gz → 0.1.265__tar.gz

{langroid-0.1.262 → langroid-0.1.265}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: langroid
-Version: 0.1.262
+Version: 0.1.265
 Summary: Harness LLMs with Multi-Agent Programming
 License: MIT
 Author: Prasad Chalasani
@@ -51,10 +51,9 @@ Requires-Dist: grpcio (>=1.62.1,<2.0.0)
 Requires-Dist: halo (>=0.0.31,<0.0.32)
 Requires-Dist: huggingface-hub (>=0.21.2,<0.22.0) ; extra == "hf-transformers" or extra == "all" or extra == "transformers"
 Requires-Dist: jinja2 (>=3.1.2,<4.0.0)
-Requires-Dist: lancedb (>=0.6.2,<0.7.0) ; extra == "vecdbs" or extra == "all" or extra == "lancedb"
+Requires-Dist: lancedb (>=0.8.2,<0.9.0) ; extra == "vecdbs" or extra == "lancedb"
 Requires-Dist: litellm (>=1.30.1,<2.0.0) ; extra == "all" or extra == "litellm"
 Requires-Dist: lxml (>=4.9.3,<5.0.0)
-Requires-Dist: meilisearch (>=0.28.3,<0.29.0) ; extra == "meilisearch"
 Requires-Dist: meilisearch-python-sdk (>=2.2.3,<3.0.0) ; extra == "meilisearch"
 Requires-Dist: metaphor-python (>=0.1.23,<0.2.0) ; extra == "all" or extra == "metaphor"
 Requires-Dist: momento (>=1.10.2,<2.0.0) ; extra == "momento"
@@ -68,8 +67,8 @@ Requires-Dist: pdf2image (>=1.17.0,<2.0.0) ; extra == "doc-chat" or extra == "al
 Requires-Dist: pdfplumber (>=0.10.2,<0.11.0) ; extra == "doc-chat" or extra == "all" or extra == "pdf-parsers"
 Requires-Dist: prettytable (>=3.8.0,<4.0.0)
 Requires-Dist: psycopg2 (>=2.9.7,<3.0.0) ; extra == "db" or extra == "all" or extra == "postgres" or extra == "sql"
-Requires-Dist: pyarrow (==15.0.0) ; extra == "vecdbs" or extra == "all" or extra == "lancedb"
-Requires-Dist: pydantic (==1.10.13)
+Requires-Dist: pyarrow (==15.0.0) ; extra == "vecdbs" or extra == "lancedb"
+Requires-Dist: pydantic (>=1,<3)
 Requires-Dist: pygithub (>=1.58.1,<2.0.0)
 Requires-Dist: pygments (>=2.15.1,<3.0.0)
 Requires-Dist: pymupdf (>=1.23.3,<2.0.0) ; extra == "doc-chat" or extra == "all" or extra == "pdf-parsers"
@@ -91,7 +90,7 @@ Requires-Dist: rich (>=13.3.4,<14.0.0)
 Requires-Dist: scrapy (>=2.11.0,<3.0.0) ; extra == "scrapy"
 Requires-Dist: sentence-transformers (==2.2.2) ; extra == "hf-transformers" or extra == "all" or extra == "hf-embeddings"
 Requires-Dist: sqlalchemy (>=2.0.19,<3.0.0) ; extra == "db" or extra == "all" or extra == "sql"
-Requires-Dist: tantivy (>=0.21.0,<0.22.0) ; extra == "vecdbs" or extra == "all" or extra == "lancedb"
+Requires-Dist: tantivy (>=0.21.0,<0.22.0) ; extra == "vecdbs" or extra == "lancedb"
 Requires-Dist: thefuzz (>=0.20.0,<0.21.0)
 Requires-Dist: tiktoken (>=0.7.0,<0.8.0)
 Requires-Dist: torch (==2.0.0) ; extra == "hf-transformers" or extra == "all" or extra == "hf-embeddings" or extra == "transformers"
@@ -515,9 +514,10 @@ with a postgres db, you will need to:
 ### Set up environment variables (API keys, etc)
 
 To get started, all you need is an OpenAI API Key.
-If you don't have one, see [this OpenAI Page](https://help.openai.com/en/collections/3675940-getting-started-with-openai-api).
-Currently only OpenAI models are supported. Others will be added later
-(Pull Requests welcome!).
+If you don't have one, see [this OpenAI Page](https://platform.openai.com/docs/quickstart).
+(Note that while this is the simplest way to get started, Langroid works with practically any LLM, not just those from OpenAI. 
+See the guides to using [Open/Local LLMs](https://langroid.github.io/langroid/tutorials/local-llm-setup/), 
+and other [non-OpenAI](https://langroid.github.io/langroid/tutorials/non-openai-llms/) proprietary LLMs.) 
 
 In the root of the repo, copy the `.env-template` file to a new file `.env`: 
 ```bash
@@ -545,9 +545,7 @@ All of the following environment variable settings are optional, and some are on
 to use specific features (as noted below).
 
 - **Qdrant** Vector Store API Key, URL. This is only required if you want to use Qdrant cloud.
-  The default vector store in our RAG agent (`DocChatAgent`) is LanceDB which uses file storage,
-  and you do not need to set up any environment variables for that.
-  Alternatively [Chroma](https://docs.trychroma.com/) is also currently supported. 
+  Alternatively [Chroma](https://docs.trychroma.com/) or [LanceDB](https://lancedb.com/) are also currently supported. 
   We use the local-storage version of Chroma, so there is no need for an API key.
 - **Redis** Password, host, port: This is optional, and only needed to cache LLM API responses
   using Redis Cloud. Redis [offers](https://redis.com/try-free/) a free 30MB Redis account
@@ -642,11 +640,12 @@ and they are **not** complete runnable examples! For that we encourage you to
 consult the [`langroid-examples`](https://github.com/langroid/langroid-examples) 
 repository.
 
-:information_source: The various LLM prompts and instructions in Langroid
-have been tested to work well with GPT4.
-Switching to GPT3.5-Turbo is easy via a config flag
-(e.g., `cfg = OpenAIGPTConfig(chat_model=OpenAIChatModel.GPT3_5_TURBO)`),
-and may suffice for some applications, but in general you may see inferior results.
+:information_source:
+The various LLM prompts and instructions in Langroid
+have been tested to work well with GPT-4 (and to some extent GPT-4o).
+Switching to other LLMs (local/open and proprietary) is easy (see guides mentioned above),
+and may suffice for some applications, but in general you may see inferior results
+unless you adjust the prompts and/or the multi-agent setup.
 
 
 :book: Also see the
@@ -921,7 +920,7 @@ config = DocChatAgentConfig(
     "https://en.wikipedia.org/wiki/N-gram_language_model",
     "/path/to/my/notes-on-language-models.txt",
   ],
-  vecdb=lr.vector_store.LanceDBConfig(),
+  vecdb=lr.vector_store.QdrantDBConfig(),
 )
 ```
 

{langroid-0.1.262 → langroid-0.1.265}/README.md

@@ -411,9 +411,10 @@ with a postgres db, you will need to:
 ### Set up environment variables (API keys, etc)
 
 To get started, all you need is an OpenAI API Key.
-If you don't have one, see [this OpenAI Page](https://help.openai.com/en/collections/3675940-getting-started-with-openai-api).
-Currently only OpenAI models are supported. Others will be added later
-(Pull Requests welcome!).
+If you don't have one, see [this OpenAI Page](https://platform.openai.com/docs/quickstart).
+(Note that while this is the simplest way to get started, Langroid works with practically any LLM, not just those from OpenAI. 
+See the guides to using [Open/Local LLMs](https://langroid.github.io/langroid/tutorials/local-llm-setup/), 
+and other [non-OpenAI](https://langroid.github.io/langroid/tutorials/non-openai-llms/) proprietary LLMs.) 
 
 In the root of the repo, copy the `.env-template` file to a new file `.env`: 
 ```bash
@@ -441,9 +442,7 @@ All of the following environment variable settings are optional, and some are on
 to use specific features (as noted below).
 
 - **Qdrant** Vector Store API Key, URL. This is only required if you want to use Qdrant cloud.
-  The default vector store in our RAG agent (`DocChatAgent`) is LanceDB which uses file storage,
-  and you do not need to set up any environment variables for that.
-  Alternatively [Chroma](https://docs.trychroma.com/) is also currently supported. 
+  Alternatively [Chroma](https://docs.trychroma.com/) or [LanceDB](https://lancedb.com/) are also currently supported. 
   We use the local-storage version of Chroma, so there is no need for an API key.
 - **Redis** Password, host, port: This is optional, and only needed to cache LLM API responses
   using Redis Cloud. Redis [offers](https://redis.com/try-free/) a free 30MB Redis account
@@ -538,11 +537,12 @@ and they are **not** complete runnable examples! For that we encourage you to
 consult the [`langroid-examples`](https://github.com/langroid/langroid-examples) 
 repository.
 
-:information_source: The various LLM prompts and instructions in Langroid
-have been tested to work well with GPT4.
-Switching to GPT3.5-Turbo is easy via a config flag
-(e.g., `cfg = OpenAIGPTConfig(chat_model=OpenAIChatModel.GPT3_5_TURBO)`),
-and may suffice for some applications, but in general you may see inferior results.
+:information_source:
+The various LLM prompts and instructions in Langroid
+have been tested to work well with GPT-4 (and to some extent GPT-4o).
+Switching to other LLMs (local/open and proprietary) is easy (see guides mentioned above),
+and may suffice for some applications, but in general you may see inferior results
+unless you adjust the prompts and/or the multi-agent setup.
 
 
 :book: Also see the
@@ -817,7 +817,7 @@ config = DocChatAgentConfig(
     "https://en.wikipedia.org/wiki/N-gram_language_model",
     "/path/to/my/notes-on-language-models.txt",
   ],
-  vecdb=lr.vector_store.LanceDBConfig(),
+  vecdb=lr.vector_store.QdrantDBConfig(),
 )
 ```
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/base.py

@@ -20,7 +20,6 @@ from typing import (
     no_type_check,
 )
 
-from pydantic import BaseSettings, ValidationError, validator
 from rich import print
 from rich.console import Console
 from rich.markup import escape
@@ -41,6 +40,7 @@ from langroid.mytypes import Entity
 from langroid.parsing.parse_json import extract_top_level_json
 from langroid.parsing.parser import Parser, ParsingConfig
 from langroid.prompts.prompts_config import PromptsConfig
+from langroid.pydantic_v1 import BaseSettings, ValidationError, validator
 from langroid.utils.configuration import settings
 from langroid.utils.constants import NO_ANSWER
 from langroid.utils.output import status

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/callbacks/chainlit.py

@@ -7,9 +7,8 @@ import logging
 import textwrap
 from typing import Any, Callable, Dict, List, Literal, Optional, no_type_check
 
-from pydantic import BaseSettings
-
 from langroid.exceptions import LangroidImportError
+from langroid.pydantic_v1 import BaseSettings
 
 try:
     import chainlit as cl

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/chat_document.py

@@ -2,8 +2,6 @@ import json
 from enum import Enum
 from typing import List, Optional, Union
 
-from pydantic import BaseModel, Extra
-
 from langroid.agent.tool_message import ToolMessage
 from langroid.language_models.base import (
     LLMFunctionCall,
@@ -15,6 +13,7 @@ from langroid.language_models.base import (
 from langroid.mytypes import DocMetaData, Document, Entity
 from langroid.parsing.agent_chats import parse_message
 from langroid.parsing.parse_json import extract_top_level_json, top_level_json_field
+from langroid.pydantic_v1 import BaseModel, Extra
 from langroid.utils.output.printing import shorten_text
 
 
@@ -48,7 +47,7 @@ class ChatDocMetaData(DocMetaData):
     block: None | Entity = None
     sender_name: str = ""
     recipient: str = ""
-    usage: Optional[LLMTokenUsage]
+    usage: Optional[LLMTokenUsage] = None
     cached: bool = False
     displayed: bool = False
     has_citation: bool = False

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/openai_assistant.py

@@ -15,7 +15,6 @@ from openai.types.beta.assistant_update_params import (
 )
 from openai.types.beta.threads import Message, Run
 from openai.types.beta.threads.runs import RunStep
-from pydantic import BaseModel
 from rich import print
 
 from langroid.agent.chat_agent import ChatAgent, ChatAgentConfig
@@ -27,6 +26,7 @@ from langroid.language_models.openai_gpt import (
     OpenAIGPT,
     OpenAIGPTConfig,
 )
+from langroid.pydantic_v1 import BaseModel
 from langroid.utils.configuration import settings
 from langroid.utils.system import generate_user_id, update_hash
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/special/doc_chat_agent.py

@@ -1313,7 +1313,7 @@ class DocChatAgent(ChatAgent):
             meta.update(extracts[0].metadata)
             return ChatDocument(
                 content="\n\n".join([e.content for e in extracts]),
-                metadata=ChatDocMetaData(**meta),
+                metadata=ChatDocMetaData(**meta),  # type: ignore
             )
         response = self.get_summary_answer(query, extracts)
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/special/lance_rag/query_planner_agent.py

@@ -191,7 +191,7 @@ class LanceQueryPlanAgent(ChatAgent):
             # save result, to be used in query_plan_feedback()
             self.result = msg.content
             # assemble QueryPlanAnswerTool...
-            query_plan_answer_tool = QueryPlanAnswerTool(
+            query_plan_answer_tool = QueryPlanAnswerTool(  # type: ignore
                 plan=self.curr_query_plan,
                 answer=self.result,
             )

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/special/lance_tools.py

@@ -1,8 +1,7 @@
 import logging
 
-from pydantic import BaseModel
-
 from langroid.agent.tool_message import ToolMessage
+from langroid.pydantic_v1 import BaseModel
 
 logger = logging.getLogger(__name__)
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/special/neo4j/neo4j_chat_agent.py

@@ -2,11 +2,11 @@ import json
 import logging
 from typing import TYPE_CHECKING, Any, Dict, List, Optional, Union
 
-from pydantic import BaseModel, BaseSettings
 from rich import print
 from rich.console import Console
 
 from langroid.agent import ToolMessage
+from langroid.pydantic_v1 import BaseModel, BaseSettings
 
 if TYPE_CHECKING:
     import neo4j

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/special/sql/utils/tools.py

@@ -29,7 +29,7 @@ class GetTableSchemaTool(ToolMessage):
     @classmethod
     def example(cls) -> "GetTableSchemaTool":
         return cls(
-            tables=("employees", "departments", "sales"),
+            tables=["employees", "departments", "sales"],
         )
 
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/task.py

@@ -20,7 +20,6 @@ from typing import (
 )
 
 import numpy as np
-from pydantic import BaseModel
 from rich import print
 from rich.markup import escape
 
@@ -37,6 +36,7 @@ from langroid.exceptions import InfiniteLoopException
 from langroid.mytypes import Entity
 from langroid.parsing.parse_json import extract_top_level_json
 from langroid.parsing.routing import parse_addressed_message
+from langroid.pydantic_v1 import BaseModel
 from langroid.utils.configuration import settings
 from langroid.utils.constants import (
     DONE,
@@ -106,7 +106,7 @@ class Task:
     """
 
     # class variable called `cache` that is a RedisCache object
-    cache: RedisCache = RedisCache(RedisCacheConfig(fake=False))
+    _cache: RedisCache | None = None
 
     def __init__(
         self,
@@ -332,6 +332,12 @@ class Task:
             config=self.config,
         )
 
+    @classmethod
+    def cache(cls) -> RedisCache:
+        if cls._cache is None:
+            cls._cache = RedisCache(RedisCacheConfig(fake=False))
+        return cls._cache
+
     def __repr__(self) -> str:
         return f"{self.name}"
 
@@ -350,7 +356,7 @@ class Task:
         E.g. key = "kill", value = "1"
         """
         try:
-            self.cache.store(f"{self.session_id}:{key}", value)
+            self.cache().store(f"{self.session_id}:{key}", value)
         except Exception as e:
             logging.error(f"Error in Task._cache_session_store: {e}")
 
@@ -360,7 +366,7 @@ class Task:
         """
         session_id_key = f"{self.session_id}:{key}"
         try:
-            cached_val = self.cache.retrieve(session_id_key)
+            cached_val = self.cache().retrieve(session_id_key)
         except Exception as e:
             logging.error(f"Error in Task._cache_session_lookup: {e}")
             return None
@@ -384,7 +390,7 @@ class Task:
         Kill the session with the given session_id.
         """
         session_id_kill_key = f"{session_id}:kill"
-        cls.cache.store(session_id_kill_key, "1")
+        cls.cache().store(session_id_kill_key, "1")
 
     def kill(self) -> None:
         """
@@ -567,13 +573,15 @@ class Task:
         Runs asynchronously.
 
         Args:
-            msg (str|ChatDocument): initial message to process; if None,
+            msg (str|ChatDocument): initial *user-role* message to process; if None,
                 the LLM will respond to its initial `self.task_messages`
                 which set up and kick off the overall task.
                 The agent tries to achieve this goal by looping
                 over `self.step()` until the task is considered
                 done; this can involve a series of messages produced by Agent,
-                LLM or Human (User).
+                LLM or Human (User). Note that `msg`, if passed, is treated as
+                message with role `user`; a "system" role message should not be
+                passed here.
             turns (int): number of turns to run the task for;
                 default is -1, which means run until task is done.
             caller (Task|None): the calling task, if any
@@ -1059,6 +1067,9 @@ class Task:
         """
         Get result of task. This is the default behavior.
         Derived classes can override this.
+
+        Note the result of a task is returned as if it is from the User entity.
+
         Returns:
             ChatDocument: result of task
         """
@@ -1071,7 +1082,7 @@ class Task:
         fun_call = result_msg.function_call if result_msg else None
         tool_messages = result_msg.tool_messages if result_msg else []
         block = result_msg.metadata.block if result_msg else None
-        recipient = result_msg.metadata.recipient if result_msg else None
+        recipient = result_msg.metadata.recipient if result_msg else ""
         tool_ids = result_msg.metadata.tool_ids if result_msg else []
         status = result_msg.metadata.status if result_msg else None
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/agent/tool_message.py

@@ -13,9 +13,9 @@ from random import choice
 from typing import Any, Dict, List, Tuple, Type
 
 from docstring_parser import parse
-from pydantic import BaseModel
 
 from langroid.language_models.base import LLMFunctionSpec
+from langroid.pydantic_v1 import BaseModel
 from langroid.utils.pydantic_utils import (
     _recursive_purge_dict_key,
     generate_simple_schema,
@@ -73,7 +73,7 @@ class ToolMessage(ABC, BaseModel):
         - a tuple (description, ToolMessage instance), where the description is
             a natural language "thought" that leads to the tool usage,
             e.g. ("I want to find the square of 5",  SquareTool(num=5))
-            In some scenarios, ncluding such a description can significantly
+            In some scenarios, including such a description can significantly
             enhance reliability of tool use.
         Returns:
         """

{langroid-0.1.262 → langroid-0.1.265}/langroid/cachedb/base.py

@@ -1,7 +1,7 @@
 from abc import ABC, abstractmethod
 from typing import Any, Dict, List
 
-from pydantic import BaseSettings
+from langroid.pydantic_v1 import BaseSettings
 
 
 class CacheDBConfig(BaseSettings):

{langroid-0.1.262 → langroid-0.1.265}/langroid/cachedb/redis_cachedb.py

@@ -23,6 +23,8 @@ class RedisCacheConfig(CacheDBConfig):
 class RedisCache(CacheDB):
     """Redis implementation of the CacheDB."""
 
+    _warned_password: bool = False
+
     def __init__(self, config: RedisCacheConfig):
         """
         Initialize a RedisCache with the given config.
@@ -40,10 +42,12 @@ class RedisCache(CacheDB):
             redis_host = os.getenv("REDIS_HOST")
             redis_port = os.getenv("REDIS_PORT")
             if None in [redis_password, redis_host, redis_port]:
-                logger.warning(
-                    """REDIS_PASSWORD, REDIS_HOST, REDIS_PORT not set in .env file,
-                    using fake redis client"""
-                )
+                if not RedisCache._warned_password:
+                    logger.warning(
+                        """REDIS_PASSWORD, REDIS_HOST, REDIS_PORT not set in .env file,
+                        using fake redis client"""
+                    )
+                    RedisCache._warned_password = True
                 self.pool = fakeredis.FakeStrictRedis()  # type: ignore
             else:
                 self.pool = redis.ConnectionPool(  # type: ignore

{langroid-0.1.262 → langroid-0.1.265}/langroid/embedding_models/base.py

@@ -2,9 +2,9 @@ import logging
 from abc import ABC, abstractmethod
 
 import numpy as np
-from pydantic import BaseSettings
 
 from langroid.mytypes import EmbeddingFunction
+from langroid.pydantic_v1 import BaseSettings
 
 logging.getLogger("openai").setLevel(logging.ERROR)
 

{langroid-0.1.262 → langroid-0.1.265}/langroid/language_models/base.py

@@ -1,5 +1,4 @@
 import ast
-import asyncio
 import json
 import logging
 from abc import ABC, abstractmethod
@@ -7,18 +6,11 @@ from datetime import datetime
 from enum import Enum
 from typing import Any, Callable, Dict, List, Optional, Tuple, Type, Union
 
-import aiohttp
-from pydantic import BaseModel, BaseSettings, Field
-
 from langroid.cachedb.base import CacheDBConfig
-from langroid.mytypes import Document
 from langroid.parsing.agent_chats import parse_message
 from langroid.parsing.parse_json import top_level_json_field
 from langroid.prompts.dialog import collate_chat_history
-from langroid.prompts.templates import (
-    EXTRACTION_PROMPT_GPT4,
-    SUMMARY_ANSWER_PROMPT_GPT4,
-)
+from langroid.pydantic_v1 import BaseModel, BaseSettings, Field
 from langroid.utils.configuration import settings
 from langroid.utils.output.printing import show_if_debug
 
@@ -184,7 +176,7 @@ class LLMResponse(BaseModel):
     message: str
     tool_id: str = ""  # used by OpenAIAssistant
     function_call: Optional[LLMFunctionCall] = None
-    usage: Optional[LLMTokenUsage]
+    usage: Optional[LLMTokenUsage] = None
     cached: bool = False
 
     def __str__(self) -> str:
@@ -487,107 +479,6 @@ class LanguageModel(ABC):
         show_if_debug(prompt, "FOLLOWUP->STANDALONE-RESPONSE= ")
         return standalone
 
-    async def get_verbatim_extract_async(self, question: str, passage: Document) -> str:
-        """
-        Asynchronously, get verbatim extract from passage
-        that is relevant to a question.
-        Asynch allows parallel calls to the LLM API.
-        """
-        async with aiohttp.ClientSession():
-            templatized_prompt = EXTRACTION_PROMPT_GPT4
-            final_prompt = templatized_prompt.format(
-                question=question, content=passage.content
-            )
-            show_if_debug(final_prompt, "EXTRACT-PROMPT= ")
-            final_extract = await self.agenerate(prompt=final_prompt, max_tokens=1024)
-            show_if_debug(final_extract.message.strip(), "EXTRACT-RESPONSE= ")
-        return final_extract.message.strip()
-
-    async def _get_verbatim_extracts(
-        self,
-        question: str,
-        passages: List[Document],
-    ) -> List[Document]:
-        async with aiohttp.ClientSession():
-            verbatim_extracts = await asyncio.gather(
-                *(self.get_verbatim_extract_async(question, P) for P in passages)
-            )
-        metadatas = [P.metadata for P in passages]
-        # return with metadata so we can use it downstream, e.g. to cite sources
-        return [
-            Document(content=e, metadata=m)
-            for e, m in zip(verbatim_extracts, metadatas)
-        ]
-
-    def get_verbatim_extracts(
-        self, question: str, passages: List[Document]
-    ) -> List[Document]:
-        """
-        From each passage, extract verbatim text that is relevant to a question,
-        using concurrent API calls to the LLM.
-        Args:
-            question: question to be answered
-            passages: list of passages from which to extract relevant verbatim text
-            LLM: LanguageModel to use for generating the prompt and extract
-        Returns:
-            list of verbatim extracts from passages that are relevant to question
-        """
-        docs = asyncio.run(self._get_verbatim_extracts(question, passages))
-        return docs
-
-    def get_summary_answer(self, question: str, passages: List[Document]) -> Document:
-        """
-        Given a question and a list of (possibly) doc snippets,
-        generate an answer if possible
-        Args:
-            question: question to answer
-            passages: list of `Document` objects each containing a possibly relevant
-                snippet, and metadata
-        Returns:
-            a `Document` object containing the answer,
-            and metadata containing source citations
-
-        """
-
-        # Define an auxiliary function to transform the list of
-        # passages into a single string
-        def stringify_passages(passages: List[Document]) -> str:
-            return "\n".join(
-                [
-                    f"""
-                Extract: {p.content}
-                Source: {p.metadata.source}
-                """
-                    for p in passages
-                ]
-            )
-
-        passages_str = stringify_passages(passages)
-        # Substitute Q and P into the templatized prompt
-
-        final_prompt = SUMMARY_ANSWER_PROMPT_GPT4.format(
-            question=f"Question:{question}", extracts=passages_str
-        )
-        show_if_debug(final_prompt, "SUMMARIZE_PROMPT= ")
-        # Generate the final verbatim extract based on the final prompt
-        llm_response = self.generate(prompt=final_prompt, max_tokens=1024)
-        final_answer = llm_response.message.strip()
-        show_if_debug(final_answer, "SUMMARIZE_RESPONSE= ")
-        parts = final_answer.split("SOURCE:", maxsplit=1)
-        if len(parts) > 1:
-            content = parts[0].strip()
-            sources = parts[1].strip()
-        else:
-            content = final_answer
-            sources = ""
-        return Document(
-            content=content,
-            metadata={
-                "source": "SOURCE: " + sources,
-                "cached": llm_response.cached,
-            },
-        )
-
 
 class StreamingIfAllowed:
     """Context to temporarily enable or disable streaming, if allowed globally via

{langroid-0.1.262 → langroid-0.1.265}/langroid/language_models/config.py

@@ -1,4 +1,4 @@
-from pydantic import BaseSettings
+from langroid.pydantic_v1 import BaseSettings
 
 
 class PromptFormatterConfig(BaseSettings):

{langroid-0.1.262 → langroid-0.1.265}/langroid/language_models/openai_gpt.py

@@ -24,7 +24,6 @@ import openai
 from groq import AsyncGroq, Groq
 from httpx import Timeout
 from openai import AsyncOpenAI, OpenAI
-from pydantic import BaseModel
 from rich import print
 from rich.markup import escape
 
@@ -50,6 +49,7 @@ from langroid.language_models.utils import (
     async_retry_with_exponential_backoff,
     retry_with_exponential_backoff,
 )
+from langroid.pydantic_v1 import BaseModel
 from langroid.utils.configuration import settings
 from langroid.utils.constants import Colors
 from langroid.utils.system import friendly_error

{langroid-0.1.262 → langroid-0.1.265}/langroid/mytypes.py

@@ -4,7 +4,7 @@ from enum import Enum
 from textwrap import dedent
 from typing import Any, Callable, Dict, List, Union
 
-from pydantic import BaseModel, Extra
+from langroid.pydantic_v1 import BaseModel, Extra
 
 Number = Union[int, float]
 Embedding = List[Number]

{langroid-0.1.262 → langroid-0.1.265}/langroid/parsing/code_parser.py

@@ -2,12 +2,12 @@ from functools import reduce
 from typing import Callable, List
 
 import tiktoken
-from pydantic import BaseSettings
 from pygments import lex
 from pygments.lexers import get_lexer_by_name
 from pygments.token import Token
 
 from langroid.mytypes import Document
+from langroid.pydantic_v1 import BaseSettings
 
 
 def chunk_code(

{langroid-0.1.262 → langroid-0.1.265}/langroid/parsing/parser.py

@@ -3,10 +3,10 @@ from enum import Enum
 from typing import Dict, List, Literal
 
 import tiktoken
-from pydantic import BaseSettings
 
 from langroid.mytypes import Document
 from langroid.parsing.para_sentence_split import create_chunks, remove_extra_whitespace
+from langroid.pydantic_v1 import BaseSettings
 
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.WARNING)

{langroid-0.1.262 → langroid-0.1.265}/langroid/parsing/repo_loader.py

@@ -15,11 +15,11 @@ from github import Github
 from github.ContentFile import ContentFile
 from github.Label import Label
 from github.Repository import Repository
-from pydantic import BaseModel, BaseSettings, Field
 
 from langroid.mytypes import DocMetaData, Document
 from langroid.parsing.document_parser import DocumentParser, DocumentType
 from langroid.parsing.parser import Parser, ParsingConfig
+from langroid.pydantic_v1 import BaseModel, BaseSettings, Field
 
 logger = logging.getLogger(__name__)