langroid 0.1.175__py3-none-any.whl → 0.1.177__py3-none-any.whl

langroid/agent/special/doc_chat_agent.py

@@ -14,7 +14,7 @@ pip install "langroid[hf-embeddings]"
 """
 import logging
 from contextlib import ExitStack
-from typing import Dict, List, Optional, Set, Tuple, no_type_check
+from typing import Any, Dict, List, Optional, Set, Tuple, no_type_check
 
 import numpy as np
 import pandas as pd
@@ -52,7 +52,7 @@ from langroid.utils.configuration import settings
 from langroid.utils.constants import NO_ANSWER
 from langroid.utils.output.printing import show_if_debug
 from langroid.utils.pydantic_utils import dataframe_to_documents, extract_fields
-from langroid.vector_store.base import VectorStoreConfig
+from langroid.vector_store.base import VectorStore, VectorStoreConfig
 from langroid.vector_store.lancedb import LanceDBConfig
 
 logger = logging.getLogger(__name__)
@@ -181,16 +181,40 @@ class DocChatAgent(ChatAgent):
     ):
         super().__init__(config)
         self.config: DocChatAgentConfig = config
-        self.original_docs: None | List[Document] = None
+        self.original_docs: List[Document] = []
         self.original_docs_length = 0
         self.from_dataframe = False
         self.df_description = ""
-        self.chunked_docs: None | List[Document] = None
-        self.chunked_docs_clean: None | List[Document] = None
+        self.chunked_docs: List[Document] = []
+        self.chunked_docs_clean: List[Document] = []
         self.response: None | Document = None
         if len(config.doc_paths) > 0:
             self.ingest()
 
+    def clear(self) -> None:
+        """Clear the document collection and the specific collection in vecdb"""
+        if self.vecdb is None:
+            raise ValueError("VecDB not set")
+        self.original_docs = []
+        self.original_docs_length = 0
+        self.chunked_docs = []
+        self.chunked_docs_clean = []
+        collection_name = self.vecdb.config.collection_name
+        if collection_name is None:
+            return
+        try:
+            # Note we may have used a vecdb with a config.collection_name
+            # different from the agent's config.vecdb.collection_name!!
+            self.vecdb.delete_collection(collection_name)
+            self.vecdb = VectorStore.create(self.vecdb.config)
+        except Exception as e:
+            logger.warning(
+                f"""
+                Error while deleting collection {collection_name}:
+                {e}
+                """
+            )
+
     def ingest(self) -> None:
         """
         Chunk + embed + store docs specified by self.config.doc_paths
@@ -211,19 +235,46 @@ class DocChatAgent(ChatAgent):
             return
         self.ingest_doc_paths(self.config.doc_paths)
 
-    def ingest_doc_paths(self, paths: List[str]) -> None:
+    def ingest_doc_paths(
+        self,
+        paths: List[str],
+        metadata: List[Dict[str, Any]] | Dict[str, Any] = [],
+    ) -> None:
         """Split, ingest docs from specified paths,
-        do not add these to config.doc_paths
+        do not add these to config.doc_paths.
+
+        Args:
+            paths: List of file/folder paths or URLs
+            metadata: List of metadata dicts, one for each path.
+                If a single dict is passed in, it is used for all paths.
         """
+        paths_meta: Dict[str, Any] = {}
+        urls_meta: Dict[str, Any] = {}
         urls, paths = get_urls_and_paths(paths)
+        if len(metadata) > 0:
+            if isinstance(metadata, list):
+                path2meta = {p: m for p, m in zip(paths, metadata)}
+            else:
+                path2meta = {p: metadata for p in paths}
+            urls_meta = {u: path2meta[u] for u in urls}
+            paths_meta = {p: path2meta[p] for p in paths}
         docs: List[Document] = []
         parser = Parser(self.config.parsing)
         if len(urls) > 0:
-            loader = URLLoader(urls=urls, parser=parser)
-            docs = loader.load()
+            for u in urls:
+                meta = urls_meta.get(u, {})
+                loader = URLLoader(urls=[u], parser=parser)
+                docs = loader.load()
+                # update metadata of each doc with meta
+                for d in docs:
+                    d.metadata = d.metadata.copy(update=meta)
         if len(paths) > 0:
             for p in paths:
+                meta = paths_meta.get(p, {})
                 path_docs = RepoLoader.get_documents(p, parser=parser)
+                # update metadata of each doc with meta
+                for d in path_docs:
+                    d.metadata = d.metadata.copy(update=meta)
                 docs.extend(path_docs)
         n_docs = len(docs)
         n_splits = self.ingest_docs(docs)
@@ -240,7 +291,12 @@ class DocChatAgent(ChatAgent):
         print("\n".join(urls))
         print("\n".join(paths))
 
-    def ingest_docs(self, docs: List[Document], split: bool = True) -> int:
+    def ingest_docs(
+        self,
+        docs: List[Document],
+        split: bool = True,
+        metadata: List[Dict[str, Any]] | Dict[str, Any] = [],
+    ) -> int:
         """
         Chunk docs into pieces, map each chunk to vec-embedding, store in vec-db
 
@@ -248,8 +304,19 @@ class DocChatAgent(ChatAgent):
             docs: List of Document objects
             split: Whether to split docs into chunks. Default is True.
                 If False, docs are treated as "chunks" and are not split.
+            metadata: List of metadata dicts, one for each doc, to augment
+                whatever metadata is already in the doc.
+                [ASSUME no conflicting keys between the two metadata dicts.]
+                If a single dict is passed in, it is used for all docs.
         """
-        self.original_docs = docs
+        if isinstance(metadata, list) and len(metadata) > 0:
+            for d, m in zip(docs, metadata):
+                d.metadata = d.metadata.copy(update=m)
+        elif isinstance(metadata, dict):
+            for d in docs:
+                d.metadata = d.metadata.copy(update=metadata)
+
+        self.original_docs.extend(docs)
         if self.parser is None:
             raise ValueError("Parser not set")
         for d in docs:
@@ -374,7 +441,7 @@ class DocChatAgent(ChatAgent):
         """
         if filter is None and len(docs) > 0:
             # no filter, so just use the docs passed in
-            self.chunked_docs = docs
+            self.chunked_docs.extend(docs)
         else:
             if self.vecdb is None:
                 raise ValueError("VecDB not set")
@@ -1098,7 +1165,7 @@ class DocChatAgent(ChatAgent):
         """Summarize all docs"""
         if self.llm is None:
             raise ValueError("LLM not set")
-        if self.original_docs is None:
+        if len(self.original_docs) == 0:
             logger.warning(
                 """
                 No docs to summarize! Perhaps you are re-using a previously

langroid/agent/special/lance_doc_chat_agent.py

@@ -11,7 +11,7 @@ For usage see:
 """
 import json
 import logging
-from typing import List, Tuple
+from typing import Any, Dict, List, Tuple
 
 import pandas as pd
 
@@ -112,8 +112,13 @@ class LanceDocChatAgent(DocChatAgent):
             # pass on the query so LLM can handle it
             return plan.query
 
-    def ingest_docs(self, docs: List[Document], split: bool = True) -> int:
-        n = super().ingest_docs(docs, split)
+    def ingest_docs(
+        self,
+        docs: List[Document],
+        split: bool = True,
+        metadata: List[Dict[str, Any]] | Dict[str, Any] = [],
+    ) -> int:
+        n = super().ingest_docs(docs, split, metadata)
         tbl = self.vecdb.client.open_table(self.vecdb.config.collection_name)
         # We assume "content" is available as top-level field
         if "content" in tbl.schema.names:
@@ -126,6 +131,8 @@ class LanceDocChatAgent(DocChatAgent):
         content: str = "content",
         metadata: List[str] = [],
     ) -> int:
+        """Ingest from a dataframe. Assume we are doing this once, not incrementally"""
+
         self.from_dataframe = True
         if df.shape[0] == 0:
             raise ValueError(

langroid/agent/special/neo4j/neo4j_chat_agent.py

@@ -5,6 +5,8 @@ from pydantic import BaseSettings
 from rich import print
 from rich.console import Console
 
+from langroid.agent import ToolMessage
+
 if TYPE_CHECKING:
     import neo4j
 
@@ -16,10 +18,6 @@ from langroid.agent.special.neo4j.utils.system_message import (
     DEFAULT_SYS_MSG,
     SCHEMA_TOOLS_SYS_MSG,
 )
-from langroid.agent.special.neo4j.utils.tools import (
-    CypherQueryTool,
-    GraphSchemaTool,
-)
 from langroid.mytypes import Entity
 
 logger = logging.getLogger(__name__)
@@ -37,6 +35,21 @@ not_valid_query_response = [
 ]
 
 
+# TOOLS to be used by the agent
+
+
+class CypherQueryTool(ToolMessage):
+    request: str = "make_query"
+    purpose: str = """Use this tool to send the Generated Cypher query based on 
+    provided text description and schema."""
+    cypher_query: str
+
+
+class GraphSchemaTool(ToolMessage):
+    request: str = "get_schema"
+    purpose: str = """To get the schema of the graph database."""
+
+
 class Neo4jSettings(BaseSettings):
     uri: str = ""
     username: str = ""
@@ -133,10 +146,11 @@ class Neo4jChatAgent(ChatAgent):
         logger.error(f"Cypher Query failed: {query}\nException: {e}")
 
         # Construct the error message
-        error_message_template = f"""\
+        error_message_template = f"""
+        There were some errors running your Cypher query:
         {NEO4J_ERROR_MSG}: '{query}'
         {str(e)}
-        Run a new query, correcting the errors.
+        Send a new query, correcting the errors.
         """
 
         return error_message_template
@@ -155,7 +169,6 @@ class Neo4jChatAgent(ChatAgent):
         Returns:
             str: The result of executing the Cypher query.
         """
-        response_message = ""
         if not self.driver:
             raise ValueError("No database connection is established.")
 
@@ -271,7 +284,7 @@ class Neo4jChatAgent(ChatAgent):
             return "The database schema does not have any nodes or relationships."
 
     def _init_tool_messages(self) -> None:
-        """Initialize message tools used for chatting."""
+        """Attach ToolMessages to the Agent."""
         message = self._format_message()
         self.config.system_message = self.config.system_message.format(mode=message)
         super().__init__(self.config)

langroid/agent/special/table_chat_agent.py

@@ -10,7 +10,7 @@ the code and returns the result as a string.
 import io
 import logging
 import sys
-from typing import List, no_type_check
+from typing import List, Optional, no_type_check
 
 import numpy as np
 import pandas as pd
@@ -23,7 +23,7 @@ from langroid.agent.tool_message import ToolMessage
 from langroid.language_models.openai_gpt import OpenAIChatModel, OpenAIGPTConfig
 from langroid.parsing.table_loader import read_tabular_data
 from langroid.prompts.prompts_config import PromptsConfig
-from langroid.utils.constants import DONE, PASS
+from langroid.utils.constants import DONE
 from langroid.vector_store.base import VectorStoreConfig
 
 logger = logging.getLogger(__name__)
@@ -34,7 +34,7 @@ DEFAULT_TABLE_CHAT_SYSTEM_MESSAGE = f"""
 You are a savvy data scientist, with expertise in analyzing tabular datasets,
 using Python and the Pandas library for dataframe manipulation.
 Since you do not have access to the dataframe 'df', you
-will need to use the `run_code` tool/function-call to answer the question.
+will need to use the `run_code` tool/function-call to answer my questions.
 Here is a summary of the dataframe:
 {{summary}}
 Do not assume any columns other than those shown.
@@ -48,9 +48,9 @@ If you receive a null or other unexpected result, see if you have made an assump
 in your code, and try another way, or use `run_code` to explore the dataframe 
 before submitting your final code. 
 
-Once you have the answer to the question, say {DONE} and show me the answer.
-If you receive an error message, try using the `run_code` tool/function 
-again with the corrected code. 
+Once you have the answer to the question, possibly after a few steps,
+say {DONE} and show me the answer. If you receive an error message, 
+try using the `run_code` tool/function again with the corrected code. 
 
 VERY IMPORTANT: When using the `run_code` tool/function, DO NOT EXPLAIN ANYTHING,
    SIMPLY USE THE TOOL, with the CODE.
@@ -129,6 +129,7 @@ class RunCodeTool(ToolMessage):
     purpose: str = """
             To run <code> on the dataframe 'df' and 
             return the results to answer a question.
+            IMPORTANT: ALL the code should be in the <code> field.
             """
     code: str
 
@@ -145,6 +146,8 @@ class TableChatAgent(ChatAgent):
     Agent for chatting with a collection of documents.
     """
 
+    sent_code: bool = False
+
     def __init__(self, config: TableChatAgentConfig):
         if isinstance(config.data, pd.DataFrame):
             df = config.data
@@ -169,6 +172,15 @@ class TableChatAgent(ChatAgent):
         # enable the agent to use and handle the RunCodeTool
         self.enable_message(RunCodeTool)
 
+    def user_response(
+        self,
+        msg: Optional[str | ChatDocument] = None,
+    ) -> Optional[ChatDocument]:
+        response = super().user_response(msg)
+        if response is not None and response.content != "":
+            self.sent_code = False
+        return response
+
     def run_code(self, msg: RunCodeTool) -> str:
         """
         Handle a RunCodeTool message by running the code and returning the result.
@@ -178,6 +190,7 @@ class TableChatAgent(ChatAgent):
         Returns:
             str: The result of running the code along with any print output.
         """
+        self.sent_code = True
         code = msg.code
         # Create a dictionary that maps 'df' to the actual DataFrame
         local_vars = {"df": self.df}
@@ -230,14 +243,13 @@ class TableChatAgent(ChatAgent):
     ) -> str | ChatDocument | None:
         """Handle scenario where LLM forgets to say DONE or forgets to use run_code"""
         if isinstance(msg, ChatDocument) and msg.metadata.sender == lr.Entity.LLM:
-            return f"""
-                You either:
-                
-                (A) forgot to use the `run_code` tool/function to find the answer,
-                        ==> In this case re-try using the `run_code` tool/function. 
-                OR 
-                
-                (B) forgot to say {DONE} after you found the answer.
-                        ==> In this case say "{DONE} {PASS}"
-                """
+            if self.sent_code:
+                return DONE
+            else:
+                return """
+                    You forgot to use the `run_code` tool/function to find the answer.
+                    Try again using the `run_code` tool/function.
+                    Remember that ALL your code, including imports, 
+                    should be in the `code` field.
+                    """
         return None

langroid/agent/task.py

@@ -160,6 +160,8 @@ class Task:
         self.name = name or agent.config.name
         self.value: str = self.name
         self.default_human_response = default_human_response
+        if default_human_response is not None and default_human_response == "":
+            interactive = False
         self.interactive = interactive
         self.message_history_idx = -1
         if interactive:
@@ -957,8 +959,6 @@ class Task:
             bool: True if task is done, False otherwise
         """
         result = result or self.pending_message
-        if self.is_done:
-            return True
         user_quit = (
             result is not None
             and result.content in USER_QUIT
@@ -968,6 +968,9 @@ class Task:
             # for top-level task, only user can quit out
             return user_quit
 
+        if self.is_done:
+            return True
+
         if self.n_stalled_steps >= self.max_stalled_steps:
             # we are stuck, so bail to avoid infinite loop
             logger.warning(

langroid/agent/tools/metaphor_search_tool.py

@@ -0,0 +1,44 @@
+"""
+A tool to trigger a metaphor search for a given query, and return the top results with
+their titles, links, summaries. Since the tool is stateless (i.e. does not need
+access to agent state), it can be enabled for any agent, without having to define a
+special method inside the agent: `agent.enable_message(MetaphorSearchTool)`
+
+NOTE: Using this tool requires setting the METAPHOR_API_KEY environment variables in 
+your `.env` file, as explained in the
+[README](https://github.com/langroid/langroid#gear-installation-and-setup).
+
+This tool requires installing langroid with the `metaphor` extra, e.g.
+`pip install langroid[metaphor]` or `poetry add langroid[metaphor]`
+(it installs the `metaphor-python` package from pypi).
+
+For more information, please refer to the official docs:
+https://metaphor.systems/
+"""
+
+from langroid.agent.tool_message import ToolMessage
+from langroid.parsing.web_search import metaphor_search
+
+
+class MetaphorSearchTool(ToolMessage):
+    request: str = "metaphor_search"
+    purpose: str = """
+            To search the web by metaphor api and return up to <num_results> 
+            links relevant to the given <query>. 
+            """
+    query: str
+    num_results: int
+
+    def handle(self) -> str:
+        """
+        Conducts a search using the metaphor API based on the provided query
+        and number of results by triggering a metaphor_search.
+
+        Returns:
+            str: A formatted string containing the titles, links, and
+            summaries of each search result, separated by two newlines.
+        """
+
+        search_results = metaphor_search(self.query, self.num_results)
+        # return Title, Link, Summary of each result, separated by two newlines
+        return "\n\n".join(str(result) for result in search_results)

langroid/language_models/openai_gpt.py

@@ -60,7 +60,8 @@ class OpenAIChatModel(str, Enum):
 
     GPT3_5_TURBO = "gpt-3.5-turbo-1106"
     GPT4 = "gpt-4"
-    GPT4_TURBO = "gpt-4-1106-preview"
+    GPT4_32K = "gpt-4-32k"
+    GPT4_TURBO = "gpt-4-turbo-preview"
 
 
 class OpenAICompletionModel(str, Enum):
@@ -72,8 +73,9 @@ class OpenAICompletionModel(str, Enum):
 
 _context_length: Dict[str, int] = {
     # can add other non-openAI models here
-    OpenAIChatModel.GPT3_5_TURBO: 4096,
+    OpenAIChatModel.GPT3_5_TURBO: 16_385,
     OpenAIChatModel.GPT4: 8192,
+    OpenAIChatModel.GPT4_32K: 32_768,
     OpenAIChatModel.GPT4_TURBO: 128_000,
     OpenAICompletionModel.TEXT_DA_VINCI_003: 4096,
 }
@@ -81,7 +83,7 @@ _context_length: Dict[str, int] = {
 _cost_per_1k_tokens: Dict[str, Tuple[float, float]] = {
     # can add other non-openAI models here.
     # model => (prompt cost, generation cost) in USD
-    OpenAIChatModel.GPT3_5_TURBO: (0.0015, 0.002),
+    OpenAIChatModel.GPT3_5_TURBO: (0.001, 0.002),
     OpenAIChatModel.GPT4: (0.03, 0.06),  # 8K context
     OpenAIChatModel.GPT4_TURBO: (0.01, 0.03),  # 128K context
 }
@@ -197,7 +199,7 @@ class OpenAIGPTConfig(LLMConfig):
     api_base: str | None = None  # used for local or other non-OpenAI models
     litellm: bool = False  # use litellm api?
     max_output_tokens: int = 1024
-    min_output_tokens: int = 64
+    min_output_tokens: int = 1
     use_chat_for_completion = True  # do not change this, for OpenAI models!
     timeout: int = 20
     temperature: float = 0.2

langroid/parsing/web_search.py

@@ -77,3 +77,49 @@ def google_search(query: str, num_results: int = 5) -> List[WebSearchResult]:
         WebSearchResult(result["title"], result["link"], 3500, 300)
         for result in raw_results
     ]
+
+
+def metaphor_search(query: str, num_results: int = 5) -> List[WebSearchResult]:
+    """
+    Method that makes an API call by Metaphor client that queries
+    the top num_results links that matches the query. Returns a list
+    of WebSearchResult objects.
+
+    Args:
+        query (str): The query body that users wants to make.
+        num_results (int): Number of top matching results that we want
+            to grab
+    """
+
+    load_dotenv()
+
+    api_key = os.getenv("METAPHOR_API_KEY")
+    if not api_key:
+        raise ValueError(
+            """
+            METAPHOR_API_KEY is not set. 
+            Please set the METAPHOR_API_KEY environment variable.
+            """
+        )
+
+    try:
+        from metaphor_python import Metaphor
+    except ImportError:
+        raise ImportError(
+            "You are attempting to use the `metaphor_python` library;"
+            "To use it, please install langroid with the `metaphor` extra, e.g. "
+            "`pip install langroid[metaphor]` or `poetry add langroid[metaphor]` "
+            "(it installs the `metaphor_python` package from pypi)."
+        )
+
+    client = Metaphor(api_key=api_key)
+
+    response = client.search(
+        query=query,
+        num_results=num_results,
+    )
+    raw_results = response.results
+
+    return [
+        WebSearchResult(result.title, result.url, 3500, 300) for result in raw_results
+    ]

langroid/utils/logging.py

@@ -4,7 +4,6 @@ from typing import no_type_check
 
 import colorlog
 from rich.console import Console
-from rich.markup import escape
 
 
 # Define a function to set up the colored logger
@@ -126,6 +125,6 @@ class RichFileLogger:
         with open(self.log_file, "a") as f:
             if self.color:
                 console = Console(file=f, force_terminal=True, width=200)
-                console.print(escape(message))
+                console.print(message)
             else:
                 print(message, file=f)

langroid/vector_store/lancedb.py

@@ -6,7 +6,7 @@ import pandas as pd
 from dotenv import load_dotenv
 from lancedb.pydantic import LanceModel, Vector
 from lancedb.query import LanceVectorQueryBuilder
-from pydantic import BaseModel, create_model
+from pydantic import BaseModel, ValidationError, create_model
 
 from langroid.embedding_models.base import (
     EmbeddingModel,
@@ -354,7 +354,19 @@ class LanceDB(VectorStore):
                 self.unflattened_schema(**nested_dict_from_flat(rec)) for rec in records
             ]
         else:
-            docs = [self.schema(**rec) for rec in records]
+            try:
+                docs = [self.schema(**rec) for rec in records]
+            except ValidationError as e:
+                raise ValueError(
+                    f"""
+                Error validating LanceDB result: {e}
+                HINT: This could happen when you're re-using an 
+                existing LanceDB store with a different schema.
+                Try deleting your local lancedb storage at `{self.config.storage_path}`
+                re-ingesting your documents and/or replacing the collections.
+                """
+                )
+
         doc_cls = self.config.document_class
         doc_cls_field_names = doc_cls.__fields__.keys()
         return [

{langroid-0.1.175.dist-info → langroid-0.1.177.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: langroid
-Version: 0.1.175
+Version: 0.1.177
 Summary: Harness LLMs with Multi-Agent Programming
 License: MIT
 Author: Prasad Chalasani
@@ -12,6 +12,7 @@ Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Provides-Extra: hf-embeddings
 Provides-Extra: litellm
+Provides-Extra: metaphor
 Provides-Extra: mysql
 Provides-Extra: neo4j
 Provides-Extra: postgres
@@ -38,6 +39,7 @@ Requires-Dist: litellm (>=1.0.0,<2.0.0) ; extra == "litellm"
 Requires-Dist: lxml (>=4.9.3,<5.0.0)
 Requires-Dist: meilisearch (>=0.28.3,<0.29.0)
 Requires-Dist: meilisearch-python-sdk (>=2.2.3,<3.0.0)
+Requires-Dist: metaphor-python (>=0.1.23,<0.2.0) ; extra == "metaphor"
 Requires-Dist: mkdocs (>=1.4.2,<2.0.0)
 Requires-Dist: mkdocs-awesome-pages-plugin (>=2.8.0,<3.0.0)
 Requires-Dist: mkdocs-gen-files (>=0.4.0,<0.5.0)
@@ -160,11 +162,14 @@ Sponsorship is also accepted via [GitHub Sponsors](https://github.com/sponsors/l
 This is just a teaser; there's much more, like function-calling/tools, 
 Multi-Agent Collaboration, Structured Information Extraction, DocChatAgent 
 (RAG), SQLChatAgent, non-OpenAI local/remote LLMs, etc. Scroll down or see docs for more.
+See the Langroid Quick-Start [Colab](https://colab.research.google.com/github/langroid/langroid/blob/main/examples/Langroid_quick_start.ipynb)
+that builds up to a 2-agent information-extraction example using the OpenAI ChatCompletion API. 
+See also this [version](https://colab.research.google.com/drive/190Tk7t4AdY1P9F_NlZ33-YEoGnHweQQ0) that uses the OpenAI Assistants API instead.
 
-:fire: Just released! Updated Langroid Quick-Start [Colab](https://colab.research.google.com/github/langroid/langroid/blob/main/examples/Langroid_quick_start.ipynb)
-that builds up to a 2-agent chat example using the OpenAI ChatCompletion API. 
-See also this [version](https://colab.research.google.com/drive/190Tk7t4AdY1P9F_NlZ33-YEoGnHweQQ0) 
-that uses the OpenAI Assistants API instead.
+:fire: just released! [Example](https://github.com/langroid/langroid-examples/blob/main/examples/docqa/chat-multi-extract-local.py) 
+script showing how you can use Langroid multi-agents and tools
+to extract structured information from a document using **only a local LLM**
+(Mistral-7b-instruct-v0.2).
 
 ```python
 import langroid as lr
@@ -215,6 +220,15 @@ teacher_task.run()
 <summary> <b>Click to expand</b></summary>
 
 - **Jan 2024:**
+  - **0.1.175** 
+    - [Neo4jChatAgent](https://github.com/langroid/langroid/tree/main/langroid/agent/special/neo4j) to chat with a neo4j knowledge-graph.
+      (Thanks to [Mohannad](https://github.com/Mohannadcse)!). The agent uses tools to query the Neo4j schema and translate user queries to Cypher queries,
+      and the tool handler executes these queries, returning them to the LLM to compose
+      a natural language response (analogous to how `SQLChatAgent` works).
+      See example [script](https://github.com/langroid/langroid/tree/main/examples/kg-chat) using this Agent to answer questions about Python pkg dependencies.
+    - Support for `.doc` file parsing (in addition to `.docx`)
+    - Specify optional [`formatter` param](https://github.com/langroid/langroid/releases/tag/0.1.171) 
+      in `OpenAIGPTConfig` to ensure accurate chat formatting for local LLMs. 
   - **[0.1.157](https://github.com/langroid/langroid/releases/tag/0.1.157):** `DocChatAgentConfig` 
      has a new param: `add_fields_to_content`, to specify additional document fields to insert into 
      the main `content` field, to help improve retrieval.
@@ -316,6 +330,8 @@ See [this test](tests/main/test_recipient_tool.py) for example usage.
 Suppose you want to extract structured information about the key terms 
 of a commercial lease document. You can easily do this with Langroid using a two-agent system,
 as we show in the [langroid-examples](https://github.com/langroid/langroid-examples/blob/main/examples/docqa/chat_multi_extract.py) repo.
+(See [this script](https://github.com/langroid/langroid-examples/blob/main/examples/docqa/chat-multi-extract-local.py)
+for a version with the same functionality using a local Mistral-7b model.)
 The demo showcases just a few of the many features of Langroid, such as:
 - Multi-agent collaboration: `LeaseExtractor` is in charge of the task, and its LLM (GPT4) generates questions 
 to be answered by the `DocAgent`.
@@ -332,7 +348,9 @@ Here is what it looks like in action
 
 
 # :zap: Highlights
-
+(For a more up-to-date list see the 
+[release](https://github.com/langroid/langroid?tab=readme-ov-file#fire-updatesreleases) 
+section above)
 - **Agents as first-class citizens:** The [Agent](https://langroid.github.io/langroid/reference/agent/base/#langroid.agent.base.Agent) class encapsulates LLM conversation state,
   and optionally a vector-store and tools. Agents are a core abstraction in Langroid;
   Agents act as _message transformers_, and by default provide 3 _responder_ methods, one corresponding to each entity: LLM, Agent, User.

{langroid-0.1.175.dist-info → langroid-0.1.177.dist-info}/RECORD

@@ -8,18 +8,17 @@ langroid/agent/helpers.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/agent/junk,sha256=LxfuuW7Cijsg0szAzT81OjWWv1PMNI-6w_-DspVIO2s,339
 langroid/agent/openai_assistant.py,sha256=yBtxis64XOnxtJzlkwUoTm-wCyvKr4DGo9-laXYMok0,32654
 langroid/agent/special/__init__.py,sha256=xj4TvQ_oQX_xYPySbhmQAi2CPhuy_3yQPqqwzb4wsc0,943
-langroid/agent/special/doc_chat_agent.py,sha256=S5BNRvELBZLFY-3TwrraGeW2XCYCU5NaOEO36zbSeRQ,45363
-langroid/agent/special/lance_doc_chat_agent.py,sha256=5wwmoPtHfc_AfQNwZ7L2GnJvtOONGnDsiKzwjgbmHt8,8168
+langroid/agent/special/doc_chat_agent.py,sha256=zW5qG0N8TpW1RfkK9vPBWcWtFqD0aM9KzcpaUeASMJE,48083
+langroid/agent/special/lance_doc_chat_agent.py,sha256=tUpn6vnbHzlkkpKEHJym7dtxS-s39AwPfH2PqyNJuxY,8371
 langroid/agent/special/lance_rag/__init__.py,sha256=-pq--upe-8vycYoTwxoomBnuUqrcRFUukmW3uBL1cFM,219
 langroid/agent/special/lance_rag/critic_agent.py,sha256=9izW4keCxVZEqrFOgyVUHD7N1vTXLkRynXYYd1Vpwzw,5785
 langroid/agent/special/lance_rag/lance_rag_task.py,sha256=l_HQgrYY-CX2FwIsS961aEF3bYog3GDYo98fj0C0mSk,2889
 langroid/agent/special/lance_rag/lance_tools.py,sha256=WypIS-3ZMDqY_PZEGB2K80-o4RfS43_OnER0dyFlsDY,1339
 langroid/agent/special/lance_rag/query_planner_agent.py,sha256=dZXVano2NbRZy91nBcEW6LrvedsHfxL1oNCgMQEHZ-U,8016
 langroid/agent/special/neo4j/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-langroid/agent/special/neo4j/neo4j_chat_agent.py,sha256=ytx-FBDCGsn5m2sUzaYj6zHQhwsOGgdX1anuIXtvebI,11391
+langroid/agent/special/neo4j/neo4j_chat_agent.py,sha256=ZLTTGwoKSW-Lw7G8o9Fi7kMiSZGNxsfK3GymanfcJvc,11738
 langroid/agent/special/neo4j/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/agent/special/neo4j/utils/system_message.py,sha256=7rwq4CYpb1-1AvGLJBI6-X8UIu5RP5LiWU9re-umXxs,1959
-langroid/agent/special/neo4j/utils/tools.py,sha256=dSbL9cNzu6ZUofyK2eaSrHPzFzS_-YEruGodsVvMrQQ,441
 langroid/agent/special/relevance_extractor_agent.py,sha256=uJ1jS_7S_gnJeuOUxEren-IHi1kfVx475B8L4H8RcEM,4795
 langroid/agent/special/retriever_agent.py,sha256=uu6vqFg85uCVM-_DrXesYe2gH_-WcoHhlsKRlLuZPXk,1867
 langroid/agent/special/sql/__init__.py,sha256=qUM-b4FfvIt0gYWP7_niyqR3OwVMMkuK2SyqUYWjyxs,207
@@ -29,13 +28,14 @@ langroid/agent/special/sql/utils/description_extractors.py,sha256=GcQ82IhKPInS_3
 langroid/agent/special/sql/utils/populate_metadata.py,sha256=zRjw31a1ZXvpx9bcmbtC2mngdHl-bp1ZNHStcPG8_Qk,2712
 langroid/agent/special/sql/utils/system_message.py,sha256=qKLHkvQWRQodTtPLPxr1GSLUYUFASZU8x-ybV67cB68,1885
 langroid/agent/special/sql/utils/tools.py,sha256=6uB2424SLtmapui9ggcEr0ZTiB6_dL1-JRGgN8RK9Js,1332
-langroid/agent/special/table_chat_agent.py,sha256=Gb4wx27A-L96s0vFbTQoiVPLMKQpif-DDwnnojxnaJU,8609
-langroid/agent/task.py,sha256=hW-vDwQkhFUtKvXUHr0Q9gylhtNfh6-bqoxM1c5SLHE,46425
+langroid/agent/special/table_chat_agent.py,sha256=GEUTP-VdtMXq4CcPV80gDQrCEn-ZFb9IhuRMtLN5I1o,9030
+langroid/agent/task.py,sha256=oa6SDHusxwvut2u8CGQ_0AQ-aONexFx5Bk2Yol3LeN8,46538
 langroid/agent/tool_message.py,sha256=ngmWdiqMYbjF4Am0hsLyA9zK0Q9QF2ziec6FW0lPD90,7399
 langroid/agent/tools/__init__.py,sha256=q-maq3k2BXhPAU99G0H6-j_ozoRvx15I1RFpPVicQIU,304
 langroid/agent/tools/extract_tool.py,sha256=u5lL9rKBzaLBOrRyLnTAZ97pQ1uxyLP39XsWMnpaZpw,3789
 langroid/agent/tools/generator_tool.py,sha256=y0fB0ZObjA0b3L0uSTtrqRCKHDUR95arBftqiUeKD2o,663
 langroid/agent/tools/google_search_tool.py,sha256=64F9oMNdS237BBOitrvYXN4Il_ES_fNrHkh35tBEDfA,1160
+langroid/agent/tools/metaphor_search_tool.py,sha256=3leL7aqGUu31VZLsQJA8-zUgdMGDhLRpV8MdgX-_oPE,1780
 langroid/agent/tools/recipient_tool.py,sha256=o4qWuz-Lp2QkOQNy6PkRMT9AEc7btuk8EJXpTg0uPV4,9170
 langroid/agent/tools/run_python_code.py,sha256=V3mHdHQYn0M0PAtyoHxjNvk6KvWWcQ4ugo0TOKc8HyI,1752
 langroid/agent/tools/sciphi_search_rag_tool.py,sha256=_BpQJtBui7A92iY0qOHh3iNn7HTszQG9Ddm7YQmHSgk,2533
@@ -54,7 +54,7 @@ langroid/language_models/azure_openai.py,sha256=ncRCbKooqLVOY-PWQUIo9C3yTuKEFbAw
 langroid/language_models/base.py,sha256=N9Jn-veKBCd_ky9mhHVSmf1iRQ1RBdaxm45yrY9Njl0,20616
 langroid/language_models/config.py,sha256=5UF3DzO1a-Dfsc3vghE0XGq7g9t_xDsRCsuRiU4dgBg,366
 langroid/language_models/openai_assistants.py,sha256=9K-DEAL2aSWHeXj2hwCo2RAlK9_1oCPtqX2u1wISCj8,36
-langroid/language_models/openai_gpt.py,sha256=yp1OzxqbLAET-kC09dkQGcoldfCV-1XRRVj17lAMjlI,47798
+langroid/language_models/openai_gpt.py,sha256=TnIvA0aNk6NlwgPpjqadf7nF7A-Rvr4kQLc41xDUyno,47864
 langroid/language_models/prompt_formatter/__init__.py,sha256=9JXFF22QNMmbQV1q4nrIeQVTtA3Tx8tEZABLtLBdFyc,352
 langroid/language_models/prompt_formatter/base.py,sha256=eDS1sgRNZVnoajwV_ZIha6cba5Dt8xjgzdRbPITwx3Q,1221
 langroid/language_models/prompt_formatter/hf_formatter.py,sha256=_XT5oHM6TmT1x4hU38lGR_b7ej6rAMHEBQUJ-w2nSDY,3587
@@ -78,7 +78,7 @@ langroid/parsing/url_loader.py,sha256=54c6yt9grfUyImauSdM5UM4_ulU4JEz0ehdugAkxKI
 langroid/parsing/url_loader_cookies.py,sha256=Lg4sNpRz9MByWq2mde6T0hKv68VZSV3mtMjNEHuFeSU,2327
 langroid/parsing/urls.py,sha256=Nv4yCWQLLBEjaiRdaZZVQNBEl_cfK_V6cVuPm91wGtU,7686
 langroid/parsing/utils.py,sha256=NVX4D43taqjnQJ0P4tRKB5tX6iXfVXWKxWgGNhQsc5c,10030
-langroid/parsing/web_search.py,sha256=hGUVoSJNdpoT5rsm-ikAteMiUropHrzKaxN8EVVqO2U,2496
+langroid/parsing/web_search.py,sha256=rQSDMrJTmhB8z5bXKxaAdt1Y5h13oaGxooRhsOHB2T0,3879
 langroid/prompts/__init__.py,sha256=B0vpJzIJlMR3mFRtoQwyALsFzBHvLp9f92acD8xJA_0,185
 langroid/prompts/dialog.py,sha256=SpfiSyofSgy2pwD1YboHR_yHO3LEEMbv6j2sm874jKo,331
 langroid/prompts/prompts_config.py,sha256=XRQHzod7KBnoKn3B_V878jZiqBA7rcn-CtGPkuAe_yM,131
@@ -93,7 +93,7 @@ langroid/utils/docker.py,sha256=kJQOLTgM0x9j9pgIIqp0dZNZCTvoUDhp6i8tYBq1Jr0,1105
 langroid/utils/globals.py,sha256=VkTHhlqSz86oOPq65sjul0XU8I52UNaFC5vwybMQ74w,1343
 langroid/utils/llms/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/utils/llms/strings.py,sha256=CSAX9Z6FQOLXOzbLMe_Opqtc3ruDAKTTk7cPqc6Blh0,263
-langroid/utils/logging.py,sha256=CXe8w2gjFhhvgyGr1iwU--Je916H2rOrDNJosv5Tl3Y,3956
+langroid/utils/logging.py,sha256=R8TN-FqVpwZ4Ajgls9TDMthLvPpQd0QVNXK-PJDj1Z8,3917
 langroid/utils/output/__init__.py,sha256=Z58-2ZKnGpGNaKw_nEjHV_CHTzjMz-WRSRQnazTLrWU,289
 langroid/utils/output/printing.py,sha256=5EsYB1O4qKhocW19aebOUzK82RD9U5nygbY21yo8gfg,2872
 langroid/utils/pandas_utils.py,sha256=nSA1tIgOUTkRDn-IKq7HP8XGJcL6bA110LcPfRF7h8I,707
@@ -105,12 +105,12 @@ langroid/utils/web/selenium_login.py,sha256=mYI6EvVmne34N9RajlsxxRqJQJvV-WG4LGp6
 langroid/vector_store/__init__.py,sha256=qOa3_BLvf8tjdUBT4Zq7pSLTY9TD2Fgw62UHHJWNu8w,557
 langroid/vector_store/base.py,sha256=JNk-2f6t_WCavizU332tOoZcXHP73RpobRk88Aus52w,13706
 langroid/vector_store/chromadb.py,sha256=Y80k6an5sN0cRWtcl78Xr-Ht87nd_hBjvkSU5OdCyY8,7312
-langroid/vector_store/lancedb.py,sha256=sKjscEpfDR560_pjkH6UyGiN3kSG_rYu1uxOJOo1zN0,16146
+langroid/vector_store/lancedb.py,sha256=dsdZVHfnvOcs1BhVK99UknbIjbND0j6I6d2_QozuL7A,16671
 langroid/vector_store/meilisearch.py,sha256=d2huA9P-NoYRuAQ9ZeXJmMKr7ry8u90RUSR28k2ecQg,11340
 langroid/vector_store/momento.py,sha256=j6Eo6oIDN2fe7lsBOlCXJn3uvvERHHTFL5QJfeREeOM,10044
 langroid/vector_store/qdrant_cloud.py,sha256=3im4Mip0QXLkR6wiqVsjV1QvhSElfxdFSuDKddBDQ-4,188
 langroid/vector_store/qdrantdb.py,sha256=_egbsP9SWBwmI827EDYSSOqfIQSmwNsmJfFTxrLpWYE,13457
-langroid-0.1.175.dist-info/LICENSE,sha256=EgVbvA6VSYgUlvC3RvPKehSg7MFaxWDsFuzLOsPPfJg,1065
-langroid-0.1.175.dist-info/METADATA,sha256=2eqvV9ZkRME_CyHkBqFz8Th76R2Im7qZahnI0qcyIVw,43613
-langroid-0.1.175.dist-info/WHEEL,sha256=FMvqSimYX_P7y0a7UY-_Mc83r5zkBZsCYPm7Lr0Bsq4,88
-langroid-0.1.175.dist-info/RECORD,,
+langroid-0.1.177.dist-info/LICENSE,sha256=EgVbvA6VSYgUlvC3RvPKehSg7MFaxWDsFuzLOsPPfJg,1065
+langroid-0.1.177.dist-info/METADATA,sha256=0ID1A5oyoMecA7dcEUQNlw_VQ2uB64VrV0IwjFtJhoQ,45218
+langroid-0.1.177.dist-info/WHEEL,sha256=FMvqSimYX_P7y0a7UY-_Mc83r5zkBZsCYPm7Lr0Bsq4,88
+langroid-0.1.177.dist-info/RECORD,,

langroid/agent/special/neo4j/utils/tools.py

@@ -1,13 +0,0 @@
-from langroid.agent.tool_message import ToolMessage
-
-
-class CypherQueryTool(ToolMessage):
-    request: str = "make_query"
-    purpose: str = """Use this tool to send me the Generated Cypher query based on 
-    text description and schema that I will provide you."""
-    cypher_query: str
-
-
-class GraphSchemaTool(ToolMessage):
-    request: str = "get_schema"
-    purpose: str = """Use this tool to get me the schema of the graph database."""