camel-ai 0.2.78__tar.gz → 0.2.79a0__tar.gz

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: camel-ai
-Version: 0.2.78
+Version: 0.2.79a0
 Summary: Communicative Agents for AI Society Study
 Project-URL: Homepage, https://www.camel-ai.org/
 Project-URL: Repository, https://github.com/camel-ai/camel
@@ -9,7 +9,7 @@ Author: CAMEL-AI.org
 License-Expression: Apache-2.0
 License-File: LICENSE
 Keywords: ai-societies,artificial-intelligence,communicative-ai,cooperative-ai,deep-learning,large-language-models,multi-agent-systems,natural-language-processing
-Requires-Python: <3.13,>=3.10
+Requires-Python: <3.15,>=3.10
 Requires-Dist: astor>=0.8.1
 Requires-Dist: colorama<0.5,>=0.4.6
 Requires-Dist: docstring-parser<0.18,>=0.17.0
@@ -72,7 +72,7 @@ Requires-Dist: jupyter-client<9,>=8.6.2; extra == 'all'
 Requires-Dist: langfuse>=2.60.5; extra == 'all'
 Requires-Dist: linkup-sdk<0.3,>=0.2.1; extra == 'all'
 Requires-Dist: litellm<2,>=1.38.1; extra == 'all'
-Requires-Dist: markitdown>=0.1.1; extra == 'all'
+Requires-Dist: markitdown>=0.1.1; (python_version >= '3.13') and extra == 'all'
 Requires-Dist: math-verify<0.8,>=0.7.0; extra == 'all'
 Requires-Dist: mcp>=1.3.0; extra == 'all'
 Requires-Dist: mem0ai>=0.1.67; extra == 'all'
@@ -102,7 +102,7 @@ Requires-Dist: pygithub<3,>=2.6.0; extra == 'all'
 Requires-Dist: pylatex>=1.4.2; extra == 'all'
 Requires-Dist: pymilvus<3,>=2.4.0; extra == 'all'
 Requires-Dist: pymupdf<2,>=1.22.5; extra == 'all'
-Requires-Dist: pyobvector>=0.1.18; extra == 'all'
+Requires-Dist: pyobvector>=0.1.18; (python_version < '3.13') and extra == 'all'
 Requires-Dist: pyowm<4,>=3.3.0; extra == 'all'
 Requires-Dist: pytelegrambotapi<5,>=4.18.0; extra == 'all'
 Requires-Dist: pytesseract>=0.3.13; extra == 'all'
@@ -143,7 +143,7 @@ Requires-Dist: types-pyyaml<7,>=6.0.12; extra == 'all'
 Requires-Dist: types-requests<3,>=2.31.0; extra == 'all'
 Requires-Dist: types-setuptools<70,>=69.2.0; extra == 'all'
 Requires-Dist: types-tqdm<5,>=4.66.0; extra == 'all'
-Requires-Dist: unstructured==0.16.20; extra == 'all'
+Requires-Dist: unstructured==0.16.20; (python_version < '3.13') and extra == 'all'
 Requires-Dist: weaviate-client>=4.15.0; extra == 'all'
 Requires-Dist: websockets<15.1,>=13.0; extra == 'all'
 Requires-Dist: wikipedia<2,>=1; extra == 'all'
@@ -215,7 +215,7 @@ Requires-Dist: chunkr-ai<0.1.0,>=0.0.50; extra == 'document-tools'
 Requires-Dist: crawl4ai>=0.3.745; extra == 'document-tools'
 Requires-Dist: docx2txt<0.9,>=0.8; extra == 'document-tools'
 Requires-Dist: docx>=0.2.4; extra == 'document-tools'
-Requires-Dist: markitdown>=0.1.1; extra == 'document-tools'
+Requires-Dist: markitdown>=0.1.1; (python_version >= '3.13') and extra == 'document-tools'
 Requires-Dist: numpy<=2.2,>=1.2; extra == 'document-tools'
 Requires-Dist: onnxruntime<=1.19.2; extra == 'document-tools'
 Requires-Dist: openapi-spec-validator<0.8,>=0.7.1; extra == 'document-tools'
@@ -226,7 +226,7 @@ Requires-Dist: pymupdf<2,>=1.22.5; extra == 'document-tools'
 Requires-Dist: python-pptx>=1.0.2; extra == 'document-tools'
 Requires-Dist: reportlab>=4.4.2; extra == 'document-tools'
 Requires-Dist: tabulate>=0.9.0; extra == 'document-tools'
-Requires-Dist: unstructured==0.16.20; extra == 'document-tools'
+Requires-Dist: unstructured==0.16.20; (python_version < '3.13') and extra == 'document-tools'
 Requires-Dist: xls2xlsx>=0.2.0; extra == 'document-tools'
 Provides-Extra: eigent
 Requires-Dist: anthropic<0.50.0,>=0.47.0; extra == 'eigent'
@@ -238,7 +238,8 @@ Requires-Dist: google-api-python-client==2.166.0; extra == 'eigent'
 Requires-Dist: google-auth-httplib2==0.2.0; extra == 'eigent'
 Requires-Dist: google-auth-oauthlib==1.2.1; extra == 'eigent'
 Requires-Dist: imageio[pyav]<3,>=2.34.2; extra == 'eigent'
-Requires-Dist: markitdown[all]>=0.1.1; extra == 'eigent'
+Requires-Dist: markitdown>=0.1.1; (python_version >= '3.13') and extra == 'eigent'
+Requires-Dist: markitdown[all]>=0.1.1; (python_version < '3.13') and extra == 'eigent'
 Requires-Dist: mcp-server-fetch==2025.1.17; extra == 'eigent'
 Requires-Dist: mcp-simple-arxiv==0.2.2; extra == 'eigent'
 Requires-Dist: numpy<=2.2,>=1.2; extra == 'eigent'
@@ -297,7 +298,7 @@ Requires-Dist: exa-py<2,>=1.10.0; extra == 'owl'
 Requires-Dist: ffmpeg-python<0.3,>=0.2.0; extra == 'owl'
 Requires-Dist: html2text>=2024.2.26; extra == 'owl'
 Requires-Dist: imageio[pyav]<3,>=2.34.2; extra == 'owl'
-Requires-Dist: markitdown>=0.1.1; extra == 'owl'
+Requires-Dist: markitdown>=0.1.1; (python_version >= '3.13') and extra == 'owl'
 Requires-Dist: mcp-server-fetch==2025.1.17; extra == 'owl'
 Requires-Dist: mcp-simple-arxiv==0.2.2; extra == 'owl'
 Requires-Dist: numpy<=2.2,>=1.2; extra == 'owl'
@@ -326,7 +327,7 @@ Requires-Dist: transformers<5,>=4; extra == 'owl'
 Requires-Dist: tree-sitter-python<0.24,>=0.23.6; extra == 'owl'
 Requires-Dist: tree-sitter<0.24,>=0.23.2; extra == 'owl'
 Requires-Dist: typer>=0.15.2; extra == 'owl'
-Requires-Dist: unstructured==0.16.20; extra == 'owl'
+Requires-Dist: unstructured==0.16.20; (python_version < '3.13') and extra == 'owl'
 Requires-Dist: websockets<15.1,>=13.0; extra == 'owl'
 Requires-Dist: wikipedia<2,>=1; extra == 'owl'
 Requires-Dist: xls2xlsx>=0.2.0; extra == 'owl'
@@ -343,11 +344,11 @@ Requires-Dist: neo4j<6,>=5.18.0; extra == 'rag'
 Requires-Dist: numpy<=2.2,>=1.2; extra == 'rag'
 Requires-Dist: protobuf>=6.0.0; extra == 'rag'
 Requires-Dist: pymilvus<3,>=2.4.0; extra == 'rag'
-Requires-Dist: pyobvector>=0.1.18; extra == 'rag'
+Requires-Dist: pyobvector>=0.1.18; (python_version < '3.13') and extra == 'rag'
 Requires-Dist: pytidb>=0.0.13; extra == 'rag'
 Requires-Dist: qdrant-client<2,>=1.9.0; extra == 'rag'
 Requires-Dist: rank-bm25<0.3,>=0.2.2; extra == 'rag'
-Requires-Dist: unstructured==0.16.20; extra == 'rag'
+Requires-Dist: unstructured==0.16.20; (python_version < '3.13') and extra == 'rag'
 Requires-Dist: weaviate-client>=4.15.0; extra == 'rag'
 Provides-Extra: research-tools
 Requires-Dist: arxiv2text<0.2,>=0.1.14; extra == 'research-tools'
@@ -366,7 +367,7 @@ Requires-Dist: pgvector<0.3,>=0.2.4; extra == 'storage'
 Requires-Dist: protobuf>=6.0.0; extra == 'storage'
 Requires-Dist: psycopg[binary]<4,>=3.1.18; extra == 'storage'
 Requires-Dist: pymilvus<3,>=2.4.0; extra == 'storage'
-Requires-Dist: pyobvector>=0.1.18; extra == 'storage'
+Requires-Dist: pyobvector>=0.1.18; (python_version < '3.13') and extra == 'storage'
 Requires-Dist: pytidb>=0.0.13; extra == 'storage'
 Requires-Dist: qdrant-client<2,>=1.9.0; extra == 'storage'
 Requires-Dist: redis<6,>=5.0.6; extra == 'storage'

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/camel/__init__.py

@@ -14,7 +14,7 @@
 
 from camel.logger import disable_logging, enable_logging, set_log_level
 
-__version__ = '0.2.78'
+__version__ = '0.2.79a0'
 
 __all__ = [
     '__version__',

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/camel/agents/chat_agent.py

@@ -71,7 +71,6 @@ from camel.memories import (
     MemoryRecord,
     ScoreBasedContextCreator,
 )
-from camel.memories.blocks.chat_history_block import EmptyMemoryWarning
 from camel.messages import (
     BaseMessage,
     FunctionCallingMessage,
@@ -877,9 +876,7 @@ class ChatAgent(BaseAgent):
                 [message.to_openai_message(role)]
             )
 
-            with warnings.catch_warnings():
-                warnings.filterwarnings("ignore", category=EmptyMemoryWarning)
-                _, ctx_tokens = self.memory.get_context()
+            _, ctx_tokens = self.memory.get_context()
 
             remaining_budget = max(0, token_limit - ctx_tokens)
 
@@ -1077,6 +1074,10 @@ class ChatAgent(BaseAgent):
         r"""Summarize the agent's current conversation context and persist it
         to a markdown file.
 
+        .. deprecated:: 0.2.80
+            Use :meth:`asummarize` for async/await support and better
+            performance in parallel summarization workflows.
+
         Args:
             filename (Optional[str]): The base filename (without extension) to
                 use for the markdown file. Defaults to a timestamped name when
@@ -1096,8 +1097,18 @@ class ChatAgent(BaseAgent):
             Dict[str, Any]: A dictionary containing the summary text, file
                 path, status message, and optionally structured_summary if
                 response_format was provided.
+
+        See Also:
+            :meth:`asummarize`: Async version for non-blocking LLM calls.
         """
 
+        warnings.warn(
+            "summarize() is synchronous. Consider using asummarize() "
+            "for async/await support and better performance.",
+            DeprecationWarning,
+            stacklevel=2,
+        )
+
         result: Dict[str, Any] = {
             "summary": "",
             "file_path": None,
@@ -1247,11 +1258,29 @@ class ChatAgent(BaseAgent):
                 result["status"] = status_message
                 return result
 
-            base_filename = (
-                filename
-                if filename
-                else f"context_summary_{datetime.now().strftime('%Y%m%d_%H%M%S')}"  # noqa: E501
-            )
+            # handle structured output if response_format was provided
+            structured_output = None
+            if response_format and response.msgs[-1].parsed:
+                structured_output = response.msgs[-1].parsed
+
+            # determine filename: use provided filename, or extract from
+            # structured output, or generate timestamp
+            if filename:
+                base_filename = filename
+            elif structured_output and hasattr(
+                structured_output, 'task_title'
+            ):
+                # use task_title from structured output for filename
+                task_title = structured_output.task_title
+                clean_title = ContextUtility.sanitize_workflow_filename(
+                    task_title
+                )
+                base_filename = (
+                    f"{clean_title}_workflow" if clean_title else "workflow"
+                )
+            else:
+                base_filename = f"context_summary_{datetime.now().strftime('%Y%m%d_%H%M%S')}"  # noqa: E501
+
             base_filename = Path(base_filename).with_suffix("").name
 
             metadata = context_util.get_session_metadata()
@@ -1262,11 +1291,274 @@ class ChatAgent(BaseAgent):
                 }
             )
 
-            # Handle structured output if response_format was provided
+            # convert structured output to custom markdown if present
+            if structured_output:
+                # convert structured output to custom markdown
+                summary_content = context_util.structured_output_to_markdown(
+                    structured_data=structured_output, metadata=metadata
+                )
+
+            # Save the markdown (either custom structured or default)
+            save_status = context_util.save_markdown_file(
+                base_filename,
+                summary_content,
+                title="Conversation Summary"
+                if not structured_output
+                else None,
+                metadata=metadata if not structured_output else None,
+            )
+
+            file_path = (
+                context_util.get_working_directory() / f"{base_filename}.md"
+            )
+
+            # Prepare result dictionary
+            result_dict = {
+                "summary": summary_content,
+                "file_path": str(file_path),
+                "status": save_status,
+                "structured_summary": structured_output,
+            }
+
+            result.update(result_dict)
+            logger.info("Conversation summary saved to %s", file_path)
+            return result
+
+        except Exception as exc:
+            error_message = f"Failed to summarize conversation context: {exc}"
+            logger.error(error_message)
+            result["status"] = error_message
+            return result
+
+    async def asummarize(
+        self,
+        filename: Optional[str] = None,
+        summary_prompt: Optional[str] = None,
+        response_format: Optional[Type[BaseModel]] = None,
+        working_directory: Optional[Union[str, Path]] = None,
+    ) -> Dict[str, Any]:
+        r"""Asynchronously summarize the agent's current conversation context
+        and persist it to a markdown file.
+
+        This is the async version of summarize() that uses astep() for
+        non-blocking LLM calls, enabling parallel summarization of multiple
+        agents.
+
+        Args:
+            filename (Optional[str]): The base filename (without extension) to
+                use for the markdown file. Defaults to a timestamped name when
+                not provided.
+            summary_prompt (Optional[str]): Custom prompt for the summarizer.
+                When omitted, a default prompt highlighting key decisions,
+                action items, and open questions is used.
+            response_format (Optional[Type[BaseModel]]): A Pydantic model
+                defining the expected structure of the response. If provided,
+                the summary will be generated as structured output and included
+                in the result.
+            working_directory (Optional[str|Path]): Optional directory to save
+                the markdown summary file. If provided, overrides the default
+                directory used by ContextUtility.
+
+        Returns:
+            Dict[str, Any]: A dictionary containing the summary text, file
+                path, status message, and optionally structured_summary if
+                response_format was provided.
+        """
+
+        result: Dict[str, Any] = {
+            "summary": "",
+            "file_path": None,
+            "status": "",
+        }
+
+        try:
+            # Use external context if set, otherwise create local one
+            if self._context_utility is None:
+                if working_directory is not None:
+                    self._context_utility = ContextUtility(
+                        working_directory=str(working_directory)
+                    )
+                else:
+                    self._context_utility = ContextUtility()
+            context_util = self._context_utility
+
+            # Get conversation directly from agent's memory
+            messages, _ = self.memory.get_context()
+
+            if not messages:
+                status_message = (
+                    "No conversation context available to summarize."
+                )
+                result["status"] = status_message
+                return result
+
+            # Convert messages to conversation text
+            conversation_lines = []
+            for message in messages:
+                role = message.get('role', 'unknown')
+                content = message.get('content', '')
+
+                # Handle tool call messages (assistant calling tools)
+                tool_calls = message.get('tool_calls')
+                if tool_calls and isinstance(tool_calls, (list, tuple)):
+                    for tool_call in tool_calls:
+                        # Handle both dict and object formats
+                        if isinstance(tool_call, dict):
+                            func_name = tool_call.get('function', {}).get(
+                                'name', 'unknown_tool'
+                            )
+                            func_args_str = tool_call.get('function', {}).get(
+                                'arguments', '{}'
+                            )
+                        else:
+                            # Handle object format (Pydantic or similar)
+                            func_name = getattr(
+                                getattr(tool_call, 'function', None),
+                                'name',
+                                'unknown_tool',
+                            )
+                            func_args_str = getattr(
+                                getattr(tool_call, 'function', None),
+                                'arguments',
+                                '{}',
+                            )
+
+                        # Parse and format arguments for readability
+                        try:
+                            import json
+
+                            args_dict = json.loads(func_args_str)
+                            args_formatted = ', '.join(
+                                f"{k}={v}" for k, v in args_dict.items()
+                            )
+                        except (json.JSONDecodeError, ValueError, TypeError):
+                            args_formatted = func_args_str
+
+                        conversation_lines.append(
+                            f"[TOOL CALL] {func_name}({args_formatted})"
+                        )
+
+                # Handle tool response messages
+                elif role == 'tool':
+                    tool_name = message.get('name', 'unknown_tool')
+                    if not content:
+                        content = str(message.get('content', ''))
+                    conversation_lines.append(
+                        f"[TOOL RESULT] {tool_name} → {content}"
+                    )
+
+                # Handle regular content messages (user/assistant/system)
+                elif content:
+                    conversation_lines.append(f"{role}: {content}")
+
+            conversation_text = "\n".join(conversation_lines).strip()
+
+            if not conversation_text:
+                status_message = (
+                    "Conversation context is empty; skipping summary."
+                )
+                result["status"] = status_message
+                return result
+
+            if self._context_summary_agent is None:
+                self._context_summary_agent = ChatAgent(
+                    system_message=(
+                        "You are a helpful assistant that summarizes "
+                        "conversations"
+                    ),
+                    model=self.model_backend,
+                    agent_id=f"{self.agent_id}_context_summarizer",
+                )
+            else:
+                self._context_summary_agent.reset()
+
+            if summary_prompt:
+                prompt_text = (
+                    f"{summary_prompt.rstrip()}\n\n"
+                    f"AGENT CONVERSATION TO BE SUMMARIZED:\n"
+                    f"{conversation_text}"
+                )
+            else:
+                prompt_text = (
+                    "Summarize the context information in concise markdown "
+                    "bullet points highlighting key decisions, action items.\n"
+                    f"Context information:\n{conversation_text}"
+                )
+
+            try:
+                # Use structured output if response_format is provided
+                if response_format:
+                    response = await self._context_summary_agent.astep(
+                        prompt_text, response_format=response_format
+                    )
+                else:
+                    response = await self._context_summary_agent.astep(
+                        prompt_text
+                    )
+
+                # Handle streaming response
+                if isinstance(response, AsyncStreamingChatAgentResponse):
+                    # Collect final response
+                    final_response = await response
+                    response = final_response
+
+            except Exception as step_exc:
+                error_message = (
+                    f"Failed to generate summary using model: {step_exc}"
+                )
+                logger.error(error_message)
+                result["status"] = error_message
+                return result
+
+            if not response.msgs:
+                status_message = (
+                    "Failed to generate summary from model response."
+                )
+                result["status"] = status_message
+                return result
+
+            summary_content = response.msgs[-1].content.strip()
+            if not summary_content:
+                status_message = "Generated summary is empty."
+                result["status"] = status_message
+                return result
+
+            # handle structured output if response_format was provided
             structured_output = None
             if response_format and response.msgs[-1].parsed:
                 structured_output = response.msgs[-1].parsed
-                # Convert structured output to custom markdown
+
+            # determine filename: use provided filename, or extract from
+            # structured output, or generate timestamp
+            if filename:
+                base_filename = filename
+            elif structured_output and hasattr(
+                structured_output, 'task_title'
+            ):
+                # use task_title from structured output for filename
+                task_title = structured_output.task_title
+                clean_title = ContextUtility.sanitize_workflow_filename(
+                    task_title
+                )
+                base_filename = (
+                    f"{clean_title}_workflow" if clean_title else "workflow"
+                )
+            else:
+                base_filename = f"context_summary_{datetime.now().strftime('%Y%m%d_%H%M%S')}"  # noqa: E501
+
+            base_filename = Path(base_filename).with_suffix("").name
+
+            metadata = context_util.get_session_metadata()
+            metadata.update(
+                {
+                    "agent_id": self.agent_id,
+                    "message_count": len(messages),
+                }
+            )
+
+            # convert structured output to custom markdown if present
+            if structured_output:
+                # convert structured output to custom markdown
                 summary_content = context_util.structured_output_to_markdown(
                     structured_data=structured_output, metadata=metadata
                 )
@@ -1347,7 +1639,7 @@ class ChatAgent(BaseAgent):
         message.
         """
         self.memory.clear()
-        # avoid UserWarning: The `ChatHistoryMemory` is empty.
+        # Write system message to memory if provided
         if self.system_message is not None:
             self.memory.write_record(
                 MemoryRecord(

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/camel/datasets/base_generator.py

@@ -62,6 +62,7 @@ class BaseGenerator(abc.ABC, IterableDataset):
         self._buffer = buffer
         self._data: List[DataPoint] = []
         self._batch_to_save: List[DataPoint] = []
+        self._iter_position: int = 0
 
         if data_path:
             file_path = Path(data_path)
@@ -103,9 +104,9 @@ class BaseGenerator(abc.ABC, IterableDataset):
         r"""Async iterator that yields datapoints dynamically.
 
         If a `data_path` was provided during initialization, those datapoints
-        are yielded first. When self._data is empty, 20 new datapoints
-        are generated. Every 100 yields, the batch is appended to the
-        JSONL file or discarded if `cache` is None.
+        are yielded first. When self._iter_position reaches the end of _data,
+        new datapoints are generated. Every 100 yields, the batch is appended
+        to the JSONL file or discarded if `cache` is None.
 
         Yields:
             DataPoint: A single datapoint.
@@ -113,9 +114,10 @@ class BaseGenerator(abc.ABC, IterableDataset):
 
         async def generator():
             while True:
-                if not self._data:
+                if self._iter_position >= len(self._data):
                     await self.generate_new(self._buffer)
-                datapoint = self._data.pop(0)
+                datapoint = self._data[self._iter_position]
+                self._iter_position += 1
                 yield datapoint
                 self._batch_to_save.append(datapoint)
                 if len(self._batch_to_save) == 100:
@@ -132,9 +134,9 @@ class BaseGenerator(abc.ABC, IterableDataset):
         r"""Synchronous iterator for PyTorch IterableDataset compatibility.
 
         If a `data_path` was provided during initialization, those datapoints
-        are yielded first. When self._data is empty, 20 new datapoints
-        are generated. Every 100 yields, the batch is appended to the
-        JSONL file or discarded if `cache` is None.
+        are yielded first. When self._iter_position reaches the end of _data,
+        new datapoints are generated. Every 100 yields, the batch is appended
+        to the JSONL file or discarded if `cache` is None.
 
         Yields:
             DataPoint: A single datapoint.
@@ -150,9 +152,10 @@ class BaseGenerator(abc.ABC, IterableDataset):
                 raise
 
         while True:
-            if not self._data:
+            if self._iter_position >= len(self._data):
                 asyncio.run(self.generate_new(self._buffer))
-            datapoint = self._data.pop(0)
+            datapoint = self._data[self._iter_position]
+            self._iter_position += 1
             yield datapoint
             self._batch_to_save.append(datapoint)
             if len(self._batch_to_save) == 100:
@@ -248,6 +251,7 @@ class BaseGenerator(abc.ABC, IterableDataset):
 
         self.save_to_jsonl(file_path)
         self._data = []
+        self._iter_position = 0
         logger.info(f"Data flushed to {file_path} and cleared from the memory")
 
     def _init_from_jsonl(self, file_path: Path) -> List[Dict[str, Any]]:
@@ -290,3 +294,28 @@ class BaseGenerator(abc.ABC, IterableDataset):
             f"Successfully loaded {len(raw_data)} items from {file_path}"
         )
         return raw_data
+
+    def __getitem__(self, index: int) -> DataPoint:
+        r"""Get a datapoint by index without removing the datapoint from _data.
+
+        Args:
+            index (int): Index of the datapoint to retrieve.
+
+        Returns:
+            DataPoint: The datapoint at the specified index.
+
+        Raises:
+            IndexError: If the index is out of range.
+        """
+        if index < 0 or index >= len(self._data):
+            raise IndexError(f"Index {index} is out of range")
+
+        return self._data[index]
+
+    def __len__(self) -> int:
+        r"""Get the number of datapoints in the dataset.
+
+        Returns:
+            int: The number of datapoints.
+        """
+        return len(self._data)

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/camel/environments/single_step.py

@@ -218,9 +218,34 @@ class SingleStepEnv:
             return observations[0] if batch_size == 1 else observations
 
         elif isinstance(self.dataset, BaseGenerator):
-            self._states = [
-                await self.dataset.async_sample() for _ in range(batch_size)
-            ]
+            # Generate more data if needed
+            if batch_size > len(self.dataset):
+                new_datapoints_needed = batch_size - len(self.dataset)
+                await self.dataset.generate_new(n=new_datapoints_needed)
+
+                # Verify that enough data was generated
+                if len(self.dataset) < batch_size:
+                    raise RuntimeError(
+                        f"Failed to generate enough datapoints. "
+                        f"Requested {batch_size}, but only "
+                        f"{len(self.dataset)} available after generation."
+                    )
+
+            # Choose sampling strategy based on whether seed is provided
+            if seed is not None:
+                # Deterministic random sampling when seed is provided
+                random_indices = rng.sample(
+                    range(len(self.dataset)), batch_size
+                )
+                self._states = [self.dataset[ind] for ind in random_indices]
+            else:
+                # Sequential sampling when no seed (backward compatible)
+                # Use async_sample to maintain sequential behavior
+                self._states = [
+                    await self.dataset.async_sample()
+                    for _ in range(batch_size)
+                ]
+
             self.current_batch_size = batch_size
             self._states_done = [False] * batch_size
 

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/camel/memories/__init__.py

@@ -18,7 +18,7 @@ from .agent_memories import (
     VectorDBMemory,
 )
 from .base import AgentMemory, BaseContextCreator, MemoryBlock
-from .blocks.chat_history_block import ChatHistoryBlock, EmptyMemoryWarning
+from .blocks.chat_history_block import ChatHistoryBlock
 from .blocks.vectordb_block import VectorDBBlock
 from .context_creators.score_based import ScoreBasedContextCreator
 from .records import ContextRecord, MemoryRecord
@@ -35,5 +35,4 @@ __all__ = [
     'ChatHistoryBlock',
     'VectorDBBlock',
     'LongtermAgentMemory',
-    'EmptyMemoryWarning',
 ]

{camel_ai-0.2.78 → camel_ai-0.2.79a0}/camel/memories/blocks/chat_history_block.py

@@ -11,7 +11,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ========= Copyright 2023-2024 @ CAMEL-AI.org. All Rights Reserved. =========
-import warnings
 from typing import List, Optional
 
 from camel.memories.base import MemoryBlock
@@ -21,17 +20,6 @@ from camel.storages.key_value_storages.in_memory import InMemoryKeyValueStorage
 from camel.types import OpenAIBackendRole
 
 
-class EmptyMemoryWarning(UserWarning):
-    """Warning raised when attempting to access an empty memory.
-
-    This warning is raised when operations are performed on memory
-    that contains no records. It can be safely caught and suppressed
-    in contexts where empty memory is expected.
-    """
-
-    pass
-
-
 class ChatHistoryBlock(MemoryBlock):
     r"""An implementation of the :obj:`MemoryBlock` abstract base class for
     maintaining a record of chat histories.
@@ -81,11 +69,8 @@ class ChatHistoryBlock(MemoryBlock):
         """
         record_dicts = self.storage.load()
         if len(record_dicts) == 0:
-            warnings.warn(
-                "The `ChatHistoryMemory` is empty.",
-                EmptyMemoryWarning,
-                stacklevel=1,
-            )
+            # Empty memory is a valid state (e.g., during initialization).
+            # Users can check if memory is empty by checking the returned list.
             return list()
 
         if window_size is not None and window_size >= 0: