Merge branch 'main' into fix/groupchat-model-registration

Author: Matteo-Frattaroli

.gitattributes

@@ -1,3 +1,91 @@
+# Source code
+*.bash            text eol=lf
+*.bat             text eol=crlf
+*.cmd             text eol=crlf
+*.coffee          text
+*.css             text diff=css eol=lf
+*.htm             text diff=html eol=lf
+*.html            text diff=html eol=lf
+*.inc             text
+*.ini             text
+*.js              text
+*.json            text eol=lf
+*.jsx             text
+*.less            text
+*.ls              text
+*.map             text -diff
+*.od              text
+*.onlydata        text
+*.php             text diff=php
+*.pl              text
+*.ps1             text eol=crlf
+*.py              text diff=python eol=lf
+*.rb              text diff=ruby eol=lf
+*.sass            text
+*.scm             text
+*.scss            text diff=css
+*.sh              text eol=lf
+.husky/*          text eol=lf
+*.sql             text
+*.styl            text
+*.tag             text
+*.ts              text
+*.tsx             text
+*.xml             text
+*.xhtml           text diff=html
+
+# Docker
+Dockerfile        text eol=lf
+
+# Documentation
+*.ipynb           text
+*.markdown        text diff=markdown eol=lf
+*.md              text diff=markdown eol=lf
+*.mdwn            text diff=markdown eol=lf
+*.mdown           text diff=markdown eol=lf
+*.mkd             text diff=markdown eol=lf
+*.mkdn            text diff=markdown eol=lf
+*.mdtxt           text eol=lf
+*.mdtext          text eol=lf
+*.txt             text eol=lf
+AUTHORS           text eol=lf
+CHANGELOG         text eol=lf
+CHANGES           text eol=lf
+CONTRIBUTING      text eol=lf
+COPYING           text eol=lf
+copyright         text eol=lf
+*COPYRIGHT*       text eol=lf
+INSTALL           text eol=lf
+license           text eol=lf
+LICENSE           text eol=lf
+NEWS              text eol=lf
+readme            text eol=lf
+*README*          text eol=lf
+TODO              text
+
+# Configs
+*.cnf             text eol=lf
+*.conf            text eol=lf
+*.config          text eol=lf
+.editorconfig     text
+.env              text eol=lf
+.gitattributes    text eol=lf
+.gitconfig        text eol=lf
+.htaccess         text
+*.lock            text -diff
+package.json      text eol=lf
+package-lock.json text eol=lf -diff
+pnpm-lock.yaml    text eol=lf -diff
+.prettierrc       text
+yarn.lock         text -diff
+*.toml            text eol=lf
+*.yaml            text eol=lf
+*.yml             text eol=lf
+browserslist      text
+Makefile          text eol=lf
+makefile          text eol=lf
+
+# Images
 *.png filter=lfs diff=lfs merge=lfs -text
 *.jpg filter=lfs diff=lfs merge=lfs -text
 *.jpeg filter=lfs diff=lfs merge=lfs -text

.github/workflows/contrib-tests.yml

@@ -107,7 +107,7 @@ jobs:
         run: |
           sudo apt-get update
           sudo apt-get install -y tesseract-ocr poppler-utils
-          pip install unstructured[all-docs]==0.13.0
+          pip install --no-cache-dir unstructured[all-docs]==0.13.0
       - name: Install packages and dependencies for RetrieveChat
         run: |
           pip install -e .[retrievechat]
@@ -418,9 +418,9 @@ jobs:
         os: [ubuntu-latest, macos-latest, windows-2019]
         python-version: ["3.11"]
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - name: Set up Python ${{ matrix.python-version }}
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
       - name: Install packages and dependencies for all tests

.pre-commit-config.yaml

@@ -32,6 +32,7 @@ repos:
       - id: ruff
         types_or: [ python, pyi, jupyter ]
         args: ["--fix", "--ignore=E402"]
+        exclude: notebook/agentchat_databricks_dbrx.ipynb
   - repo: https://github.com/codespell-project/codespell
     rev: v2.2.6
     hooks:

README.md

@@ -15,9 +15,12 @@
     <img src="https://github.com/microsoft/autogen/blob/main/website/static/img/flaml.svg"  width=200>
     <br>
 </p> -->
+
+:fire: May 24, 2024: Foundation Capital published an article on [Forbes: The Promise of Multi-Agent AI](https://www.forbes.com/sites/joannechen/2024/05/24/the-promise-of-multi-agent-ai/?sh=2c1e4f454d97) and a video [AI in the Real World Episode 2: Exploring Multi-Agent AI and AutoGen with Chi Wang](https://www.youtube.com/watch?v=RLwyXRVvlNk).
+
 :fire: May 13, 2024: [The Economist](https://www.economist.com/science-and-technology/2024/05/13/todays-ai-models-are-impressive-teams-of-them-will-be-formidable) published an article about multi-agent systems (MAS) following a January 2024 interview with [Chi Wang](https://github.com/sonichi).
 
-:fire: May 11, 2024: [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://openreview.net/pdf?id=uAjxFFing2) received the best paper award in [ICLR 2024 LLM Agents Workshop](https://llmagents.github.io/).
+:fire: May 11, 2024: [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://openreview.net/pdf?id=uAjxFFing2) received the best paper award at the [ICLR 2024 LLM Agents Workshop](https://llmagents.github.io/).
 
 :fire: Apr 26, 2024: [AutoGen.NET](https://microsoft.github.io/autogen-for-net/) is available for .NET developers!
 

autogen/agentchat/contrib/capabilities/context_handling.py

@@ -8,8 +8,8 @@
 from autogen import ConversableAgent, token_count_utils
 
 warn(
-    "Context handling with TransformChatHistory is deprecated. "
-    "Please use TransformMessages from autogen/agentchat/contrib/capabilities/transform_messages.py instead.",
+    "Context handling with TransformChatHistory is deprecated and will be removed in `0.2.30`. "
+    "Please use `TransformMessages`, documentation can be found at https://microsoft.github.io/autogen/docs/topics/handling_long_contexts/intro_to_transform_messages",
     DeprecationWarning,
     stacklevel=2,
 )

autogen/agentchat/contrib/capabilities/transforms.py

@@ -8,6 +8,7 @@
 
 from autogen import token_count_utils
 from autogen.cache import AbstractCache, Cache
+from autogen.oai.openai_utils import filter_config
 
 from .text_compressors import LLMLingua, TextCompressor
 
@@ -130,6 +131,8 @@ def __init__(
         max_tokens: Optional[int] = None,
         min_tokens: Optional[int] = None,
         model: str = "gpt-3.5-turbo-0613",
+        filter_dict: Optional[Dict] = None,
+        exclude_filter: bool = True,
     ):
         """
         Args:
@@ -140,11 +143,17 @@ def __init__(
             min_tokens (Optional[int]): Minimum number of tokens in messages to apply the transformation.
                 Must be greater than or equal to 0 if not None.
             model (str): The target OpenAI model for tokenization alignment.
+            filter_dict (None or dict): A dictionary to filter out messages that you want/don't want to compress.
+                If None, no filters will be applied.
+            exclude_filter (bool): If exclude filter is True (the default value), messages that match the filter will be
+                excluded from token truncation. If False, messages that match the filter will be truncated.
         """
         self._model = model
         self._max_tokens_per_message = self._validate_max_tokens(max_tokens_per_message)
         self._max_tokens = self._validate_max_tokens(max_tokens)
         self._min_tokens = self._validate_min_tokens(min_tokens, max_tokens)
+        self._filter_dict = filter_dict
+        self._exclude_filter = exclude_filter
 
     def apply_transform(self, messages: List[Dict]) -> List[Dict]:
         """Applies token truncation to the conversation history.
@@ -169,10 +178,15 @@ def apply_transform(self, messages: List[Dict]) -> List[Dict]:
 
         for msg in reversed(temp_messages):
             # Some messages may not have content.
-            if not isinstance(msg.get("content"), (str, list)):
+            if not _is_content_right_type(msg.get("content")):
                 processed_messages.insert(0, msg)
                 continue
 
+            if not _should_transform_message(msg, self._filter_dict, self._exclude_filter):
+                processed_messages.insert(0, msg)
+                processed_messages_tokens += _count_tokens(msg["content"])
+                continue
+
             expected_tokens_remained = self._max_tokens - processed_messages_tokens - self._max_tokens_per_message
 
             # If adding this message would exceed the token limit, truncate the last message to meet the total token
@@ -282,6 +296,8 @@ def __init__(
         min_tokens: Optional[int] = None,
         compression_params: Dict = dict(),
         cache: Optional[AbstractCache] = Cache.disk(),
+        filter_dict: Optional[Dict] = None,
+        exclude_filter: bool = True,
     ):
         """
         Args:
@@ -293,6 +309,10 @@ def __init__(
                 dictionary.
             cache (None or AbstractCache): The cache client to use to store and retrieve previously compressed messages.
                 If None, no caching will be used.
+            filter_dict (None or dict): A dictionary to filter out messages that you want/don't want to compress.
+                If None, no filters will be applied.
+            exclude_filter (bool): If exclude filter is True (the default value), messages that match the filter will be
+                excluded from compression. If False, messages that match the filter will be compressed.
         """
 
         if text_compressor is None:
@@ -303,6 +323,8 @@ def __init__(
         self._text_compressor = text_compressor
         self._min_tokens = min_tokens
         self._compression_args = compression_params
+        self._filter_dict = filter_dict
+        self._exclude_filter = exclude_filter
         self._cache = cache
 
         # Optimizing savings calculations to optimize log generation
@@ -334,7 +356,10 @@ def apply_transform(self, messages: List[Dict]) -> List[Dict]:
         processed_messages = messages.copy()
         for message in processed_messages:
             # Some messages may not have content.
-            if not isinstance(message.get("content"), (str, list)):
+            if not _is_content_right_type(message.get("content")):
+                continue
+
+            if not _should_transform_message(message, self._filter_dict, self._exclude_filter):
                 continue
 
             if _is_content_text_empty(message["content"]):
@@ -397,7 +422,7 @@ def _cache_set(
         self, content: Union[str, List[Dict]], compressed_content: Union[str, List[Dict]], tokens_saved: int
     ):
         if self._cache:
-            value = (tokens_saved, json.dumps(compressed_content))
+            value = (tokens_saved, compressed_content)
             self._cache.set(self._cache_key(content), value)
 
     def _cache_key(self, content: Union[str, List[Dict]]) -> str:
@@ -427,10 +452,21 @@ def _count_tokens(content: Union[str, List[Dict[str, Any]]]) -> int:
     return token_count
 
 
+def _is_content_right_type(content: Any) -> bool:
+    return isinstance(content, (str, list))
+
+
 def _is_content_text_empty(content: Union[str, List[Dict[str, Any]]]) -> bool:
     if isinstance(content, str):
         return content == ""
     elif isinstance(content, list):
         return all(_is_content_text_empty(item.get("text", "")) for item in content)
     else:
         return False
+
+
+def _should_transform_message(message: Dict[str, Any], filter_dict: Optional[Dict[str, Any]], exclude: bool) -> bool:
+    if not filter_dict:
+        return True
+
+    return len(filter_config([message], filter_dict, exclude)) > 0

autogen/agentchat/contrib/compressible_agent.py

@@ -13,8 +13,8 @@
 logger = logging.getLogger(__name__)
 
 warn(
-    "Context handling with CompressibleAgent is deprecated. "
-    "Please use `TransformMessages`, documentation can be found at https://microsoft.github.io/autogen/docs/reference/agentchat/contrib/capabilities/transform_messages",
+    "Context handling with CompressibleAgent is deprecated and will be removed in `0.2.30`. "
+    "Please use `TransformMessages`, documentation can be found at https://microsoft.github.io/autogen/docs/topics/handling_long_contexts/intro_to_transform_messages",
     DeprecationWarning,
     stacklevel=2,
 )

autogen/agentchat/contrib/gpt_assistant_agent.py

@@ -11,6 +11,7 @@
 from autogen.agentchat.agent import Agent
 from autogen.agentchat.assistant_agent import AssistantAgent, ConversableAgent
 from autogen.oai.openai_utils import create_gpt_assistant, retrieve_assistants_by_name, update_gpt_assistant
+from autogen.runtime_logging import log_new_agent, logging_enabled
 
 logger = logging.getLogger(__name__)
 
@@ -65,6 +66,8 @@ def __init__(
         super().__init__(
             name=name, system_message=instructions, human_input_mode="NEVER", llm_config=openai_client_cfg, **kwargs
         )
+        if logging_enabled():
+            log_new_agent(self, locals())
 
         # GPTAssistantAgent's azure_deployment param may cause NotFoundError (404) in client.beta.assistants.list()
         # See: https://github.com/microsoft/autogen/pull/1721
@@ -169,6 +172,7 @@ def __init__(
                 # Tools are specified but overwrite_tools is False; do not update the assistant's tools
                 logger.warning("overwrite_tools is False. Using existing tools from assistant API.")
 
+        self.update_system_message(self._openai_assistant.instructions)
         # lazily create threads
         self._openai_threads = {}
         self._unread_index = defaultdict(int)

autogen/agentchat/contrib/retrieve_user_proxy_agent.py

@@ -7,8 +7,8 @@
 
 try:
     import chromadb
-except ImportError:
-    raise ImportError("Please install dependencies first. `pip install pyautogen[retrievechat]`")
+except ImportError as e:
+    raise ImportError(f"{e}. You can try `pip install pyautogen[retrievechat]`, or install `chromadb` manually.")
 from autogen.agentchat import UserProxyAgent
 from autogen.agentchat.agent import Agent
 from autogen.agentchat.contrib.vectordb.base import Document, QueryResults, VectorDB, VectorDBFactory