Fix #2753: Handle large inputs in memory by chunking text before embedding

Co-Authored-By: Joe Moura <joao@crewai.com>

src/crewai/memory/storage/rag_storage.py

@@ -6,11 +6,12 @@ import shutil
 import uuid
 from typing import Any, Dict, List, Optional
 
+import numpy as np
 from chromadb.api import ClientAPI
 
 from crewai.memory.storage.base_rag_storage import BaseRAGStorage
 from crewai.utilities import EmbeddingConfigurator
-from crewai.utilities.constants import MAX_FILE_NAME_LENGTH
+from crewai.utilities.constants import MAX_FILE_NAME_LENGTH, MEMORY_CHUNK_SIZE, MEMORY_CHUNK_OVERLAP
 from crewai.utilities.paths import db_storage_path
 
 
@@ -138,15 +139,45 @@ class RAGStorage(BaseRAGStorage):
             logging.error(f"Error during {self.type} search: {str(e)}")
             return []
 
+    def _chunk_text(self, text: str) -> List[str]:
+        """
+        Split text into chunks to avoid token limits.
+        
+        Args:
+            text: Text to chunk
+            
+        Returns:
+            List of text chunks
+        """
+        if not text:
+            return []
+        
+        if len(text) <= MEMORY_CHUNK_SIZE:
+            return [text]
+            
+        chunks = []
+        for i in range(0, len(text), MEMORY_CHUNK_SIZE - MEMORY_CHUNK_OVERLAP):
+            chunk = text[i:i + MEMORY_CHUNK_SIZE]
+            if chunk:  # Only add non-empty chunks
+                chunks.append(chunk)
+                
+        return chunks
+
     def _generate_embedding(self, text: str, metadata: Dict[str, Any]) -> None:  # type: ignore
         if not hasattr(self, "app") or not hasattr(self, "collection"):
             self._initialize_app()
 
-        self.collection.add(
+        chunks = self._chunk_text(text)
-            documents=[text],
+        
-            metadatas=[metadata or {}],
+        if not chunks:
-            ids=[str(uuid.uuid4())],
+            return None
-        )
+            
+        for chunk in chunks:
+            self.collection.add(
+                documents=[chunk],
+                metadatas=[metadata or {}],
+                ids=[str(uuid.uuid4())],
+            )
 
     def reset(self) -> None:
         try:

src/crewai/utilities/constants.py

@@ -4,3 +4,5 @@ DEFAULT_SCORE_THRESHOLD = 0.35
 KNOWLEDGE_DIRECTORY = "knowledge"
 MAX_LLM_RETRY = 3
 MAX_FILE_NAME_LENGTH = 255
+MEMORY_CHUNK_SIZE = 4000
+MEMORY_CHUNK_OVERLAP = 200

tests/memory/large_input_memory_test.py

@@ -0,0 +1,52 @@
+import pytest
+import numpy as np
+from unittest.mock import patch, MagicMock
+
+from crewai.memory.short_term.short_term_memory import ShortTermMemory
+from crewai.agent import Agent
+from crewai.crew import Crew
+from crewai.task import Task
+from crewai.utilities.constants import MEMORY_CHUNK_SIZE
+
+
+@pytest.fixture
+def short_term_memory():
+    """Fixture to create a ShortTermMemory instance"""
+    agent = Agent(
+        role="Researcher",
+        goal="Search relevant data and provide results",
+        backstory="You are a researcher at a leading tech think tank.",
+        tools=[],
+        verbose=True,
+    )
+
+    task = Task(
+        description="Perform a search on specific topics.",
+        expected_output="A list of relevant URLs based on the search query.",
+        agent=agent,
+    )
+    return ShortTermMemory(crew=Crew(agents=[agent], tasks=[task]))
+
+
+def test_memory_with_large_input(short_term_memory):
+    """Test that memory can handle large inputs without token limit errors"""
+    large_input = "test value " * (MEMORY_CHUNK_SIZE + 1000)
+    
+    with patch.object(
+        short_term_memory.storage, '_chunk_text', 
+        return_value=["chunk1", "chunk2"]
+    ) as mock_chunk_text:
+        with patch.object(
+            short_term_memory.storage.collection, 'add'
+        ) as mock_add:
+            short_term_memory.save(value=large_input, agent="test_agent")
+            
+            assert mock_chunk_text.called
+        
+    with patch.object(
+        short_term_memory.storage, 'search',
+        return_value=[{"context": large_input, "metadata": {"agent": "test_agent"}, "score": 0.95}]
+    ):
+        result = short_term_memory.search(large_input[:100], score_threshold=0.01)
+        assert result[0]["context"] == large_input
+        assert result[0]["metadata"]["agent"] == "test_agent"