feat: Adding ChunkMetadata

Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>

llama_stack/apis/vector_io/vector_io.py

@@ -19,17 +19,52 @@ from llama_stack.schema_utils import json_schema_type, webmethod
 from llama_stack.strong_typing.schema import register_schema
 
 
+@json_schema_type
+class ChunkMetadata(BaseModel):
+    """
+    `ChunkMetadata` is backend metadata for a `Chunk` that is used to store additional information about the chunk that
+        will NOT be inserted into the context during inference, but is required for backend functionality.
+        Use `metadata` in `Chunk` for metadata that will be used during inference.
+    :param document_id: The ID of the document this chunk belongs to.
+    :param source: The source of the content, such as a URL or file path.
+    :param created_timestamp: An optional timestamp indicating when the chunk was created.
+    :param updated_timestamp: An optional timestamp indicating when the chunk was last updated.
+    :param chunk_window: The window of the chunk, which can be used to group related chunks together.
+    :param chunk_tokenizer: The tokenizer used to create the chunk. Default is Tiktoken.
+    :param chunk_embedding_model: The embedding model used to create the chunk's embedding.
+    :param chunk_embedding_dimension: The dimension of the embedding vector for the chunk.
+    :param content_token_count: The number of tokens in the content of the chunk.
+    :param metadata_token_count: The number of tokens in the metadata of the chunk.
+    """
+
+    document_id: str | None = None
+    chunk_id: str | None = None
+    source: str | None = None
+    created_timestamp: int | None = None
+    updated_timestamp: int | None = None
+    chunk_window: str | None = None
+    chunk_tokenizer: str | None = None
+    chunk_embedding_model: str | None = None
+    chunk_embedding_dimension: int | None = None
+    content_token_count: int | None = None
+    metadata_token_count: int | None = None
+
+
+@json_schema_type
 class Chunk(BaseModel):
     """
     A chunk of content that can be inserted into a vector database.
     :param content: The content of the chunk, which can be interleaved text, images, or other types.
     :param embedding: Optional embedding for the chunk. If not provided, it will be computed later.
-    :param metadata: Metadata associated with the chunk, such as document ID, source, or other relevant information.
+    :param metadata: Metadata associated with the chunk that will be used during inference.
+    :param chunk_metadata: Metadata for the chunk that will NOT be inserted into the context during inference
+        that is required backend functionality.
     """
 
     content: InterleavedContent
     metadata: dict[str, Any] = Field(default_factory=dict)
     embedding: list[float] | None = None
+    chunk_metadata: ChunkMetadata | None = None
 
 
 @json_schema_type

llama_stack/providers/inline/tool_runtime/rag/memory.py

@@ -148,6 +148,9 @@ class MemoryToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, RAGToolRunti
         ]
         for i, chunk in enumerate(chunks):
             metadata = chunk.metadata
+            # update chunk.metadata with the chunk.chunk_metadata if it exists
+            if chunk.chunk_metadata:
+                metadata = {**metadata, **chunk.chunk_metadata.dict()}
             tokens += metadata.get("token_count", 0)
             tokens += metadata.get("metadata_token_count", 0)
 
@@ -157,7 +160,19 @@ class MemoryToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, RAGToolRunti
                 )
                 break
 
-            metadata_subset = {k: v for k, v in metadata.items() if k not in ["token_count", "metadata_token_count"]}
+            metadata_fields_to_exclude_from_context = [
+                "chunk_tokenizer",
+                "chunk_window",
+                "token_count",
+                "metadata_token_count",
+                "chunk_tokenizer",
+                "chunk_embedding_model",
+                "created_timestamp",
+                "updated_timestamp",
+                "chunk_window",
+                "content_token_count",
+            ]
+            metadata_subset = {k: v for k, v in metadata.items() if k not in metadata_fields_to_exclude_from_context}
             text_content = query_config.chunk_template.format(index=i + 1, chunk=chunk, metadata=metadata_subset)
             picked.append(TextContentItem(text=text_content))
 

llama_stack/providers/inline/vector_io/sqlite_vec/sqlite_vec.py

@@ -5,12 +5,10 @@
 # the root directory of this source tree.
 
 import asyncio
-import hashlib
 import json
 import logging
 import sqlite3
 import struct
-import uuid
 from typing import Any
 
 import numpy as np
@@ -33,6 +31,7 @@ from llama_stack.providers.utils.memory.vector_store import (
     EmbeddingIndex,
     VectorDBWithIndex,
 )
+from llama_stack.providers.utils.vector_io.chunk_utils import extract_or_generate_chunk_id
 
 logger = logging.getLogger(__name__)
 
@@ -202,8 +201,7 @@ class SQLiteVecIndex(EmbeddingIndex):
 
                     # Insert metadata
                     metadata_data = [
-                        (generate_chunk_id(chunk.metadata["document_id"], chunk.content), chunk.model_dump_json())
-                        for chunk in batch_chunks
+                        (extract_or_generate_chunk_id(chunk), chunk.model_dump_json()) for chunk in batch_chunks
                     ]
                     cur.executemany(
                         f"""
@@ -218,7 +216,7 @@ class SQLiteVecIndex(EmbeddingIndex):
                     embedding_data = [
                         (
                             (
-                                generate_chunk_id(chunk.metadata["document_id"], chunk.content),
+                                extract_or_generate_chunk_id(chunk),
                                 serialize_vector(emb.tolist()),
                             )
                         )
@@ -230,10 +228,7 @@ class SQLiteVecIndex(EmbeddingIndex):
                     )
 
                     # Insert FTS content
-                    fts_data = [
-                        (generate_chunk_id(chunk.metadata["document_id"], chunk.content), chunk.content)
-                        for chunk in batch_chunks
-                    ]
+                    fts_data = [(extract_or_generate_chunk_id(chunk), chunk.content) for chunk in batch_chunks]
                     # DELETE existing entries with same IDs (FTS5 doesn't support ON CONFLICT)
                     cur.executemany(
                         f"DELETE FROM {self.fts_table} WHERE id = ?;",
@@ -383,11 +378,11 @@ class SQLiteVecIndex(EmbeddingIndex):
 
         # Convert responses to score dictionaries using generate_chunk_id
         vector_scores = {
-            generate_chunk_id(chunk.metadata["document_id"], str(chunk.content)): score
+            extract_or_generate_chunk_id(chunk): score
             for chunk, score in zip(vector_response.chunks, vector_response.scores, strict=False)
         }
         keyword_scores = {
-            generate_chunk_id(chunk.metadata["document_id"], str(chunk.content)): score
+            extract_or_generate_chunk_id(chunk): score
             for chunk, score in zip(keyword_response.chunks, keyword_response.scores, strict=False)
         }
 
@@ -410,10 +405,10 @@ class SQLiteVecIndex(EmbeddingIndex):
         # Create a map of chunk_id to chunk for both responses
         chunk_map = {}
         for c in vector_response.chunks:
-            chunk_id = generate_chunk_id(c.metadata["document_id"], str(c.content))
+            chunk_id = extract_or_generate_chunk_id(c)
             chunk_map[chunk_id] = c
         for c in keyword_response.chunks:
-            chunk_id = generate_chunk_id(c.metadata["document_id"], str(c.content))
+            chunk_id = extract_or_generate_chunk_id(c)
             chunk_map[chunk_id] = c
 
         # Use the map to look up chunks by their IDs
@@ -757,9 +752,3 @@ class SQLiteVecVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorDBsProtoc
         if vector_db_id not in self.cache:
             raise ValueError(f"Vector DB {vector_db_id} not found")
         return await self.cache[vector_db_id].query_chunks(query, params)
-
-
-def generate_chunk_id(document_id: str, chunk_text: str) -> str:
-    """Generate a unique chunk ID using a hash of document ID and chunk text."""
-    hash_input = f"{document_id}:{chunk_text}".encode()
-    return str(uuid.UUID(hashlib.md5(hash_input).hexdigest()))

llama_stack/providers/remote/vector_io/qdrant/qdrant.py

@@ -72,7 +72,11 @@ class QdrantIndex(EmbeddingIndex):
 
         points = []
         for i, (chunk, embedding) in enumerate(zip(chunks, embeddings, strict=False)):
-            chunk_id = f"{chunk.metadata['document_id']}:chunk-{i}"
+            chunk_id = (
+                f"{chunk.metadata.get('document_id')}:chunk-{i}"
+                if chunk.metadata
+                else f"{chunk.chunk_metadata.document_id}:chunk-{i}"
+            )
             points.append(
                 PointStruct(
                     id=convert_id(chunk_id),

llama_stack/providers/utils/memory/vector_store.py

@@ -7,6 +7,7 @@ import base64
 import io
 import logging
 import re
+import time
 from abc import ABC, abstractmethod
 from dataclasses import dataclass
 from typing import Any
@@ -23,12 +24,13 @@ from llama_stack.apis.common.content_types import (
 )
 from llama_stack.apis.tools import RAGDocument
 from llama_stack.apis.vector_dbs import VectorDB
-from llama_stack.apis.vector_io import Chunk, QueryChunksResponse
+from llama_stack.apis.vector_io import Chunk, ChunkMetadata, QueryChunksResponse
 from llama_stack.models.llama.llama3.tokenizer import Tokenizer
 from llama_stack.providers.datatypes import Api
 from llama_stack.providers.utils.inference.prompt_adapter import (
     interleaved_content_as_str,
 )
+from llama_stack.providers.utils.vector_io.chunk_utils import generate_chunk_id
 
 log = logging.getLogger(__name__)
 
@@ -148,6 +150,10 @@ async def content_from_doc(doc: RAGDocument) -> str:
 def make_overlapped_chunks(
     document_id: str, text: str, window_len: int, overlap_len: int, metadata: dict[str, Any]
 ) -> list[Chunk]:
+    default_tokenizer = "DEFAULT_TIKTOKEN_TOKENIZER"
+    default_embedding_model = (
+        "DEFAULT_EMBEDDING_MODEL"  # This will be correctly updated in `VectorDBWithIndex.insert_chunks`
+    )
     tokenizer = Tokenizer.get_instance()
     tokens = tokenizer.encode(text, bos=False, eos=False)
     try:
@@ -166,11 +172,25 @@ def make_overlapped_chunks(
         chunk_metadata["token_count"] = len(toks)
         chunk_metadata["metadata_token_count"] = len(metadata_tokens)
 
+        backend_chunk_metadata = ChunkMetadata(
+            document_id=document_id,
+            chunk_id=generate_chunk_id(chunk, text),
+            source=metadata.get("source", None),
+            created_timestamp=metadata.get("created_timestamp", int(time.time())),
+            updated_timestamp=int(time.time()),
+            chunk_window=f"{i}-{i + len(toks)}",
+            chunk_tokenizer=default_tokenizer,
+            chunk_embedding_model=default_embedding_model,
+            content_token_count=len(toks),
+            metadata_token_count=len(metadata_tokens),
+        )
+
         # chunk is a string
         chunks.append(
             Chunk(
                 content=chunk,
                 metadata=chunk_metadata,
+                chunk_metadata=backend_chunk_metadata,
             )
         )
 
@@ -235,9 +255,13 @@ class VectorDBWithIndex:
     ) -> None:
         chunks_to_embed = []
         for i, c in enumerate(chunks):
+            # this should be done in `make_overlapped_chunks` but we do it here for convenience
             if c.embedding is None:
                 chunks_to_embed.append(c)
             else:
+                if c.chunk_metadata:
+                    c.chunk_metadata.chunk_embedding_model = self.vector_db.embedding_model
+                    c.chunk_metadata.chunk_embedding_dimension = self.vector_db.embedding_dimension
                 _validate_embedding(c.embedding, i, self.vector_db.embedding_dimension)
 
         if chunks_to_embed:

llama_stack/providers/utils/vector_io/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.

llama_stack/providers/utils/vector_io/chunk_utils.py

@@ -0,0 +1,42 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+import hashlib
+import logging
+import uuid
+
+from llama_stack.apis.vector_io import Chunk
+
+
+def generate_chunk_id(document_id: str, chunk_text: str) -> str:
+    """Generate a unique chunk ID using a hash of document ID and chunk text."""
+    hash_input = f"{document_id}:{chunk_text}".encode()
+    return str(uuid.UUID(hashlib.md5(hash_input).hexdigest()))
+
+
+def extract_chunk_id_from_metadata(chunk: Chunk) -> str | None:
+    """Extract existing chunk ID from metadata. This is for compatibility with older Chunks
+    that stored the document_id in the metadata and not in the ChunkMetadata."""
+    if chunk.chunk_metadata is not None and hasattr(chunk.chunk_metadata, "chunk_id"):
+        return chunk.chunk_metadata.chunk_id
+
+    if "chunk_id" in chunk.metadata:
+        return str(chunk.metadata["chunk_id"])
+
+    return None
+
+
+def extract_or_generate_chunk_id(chunk: Chunk) -> str:
+    """Extract existing chunk ID or generate a new one if not present. This is for compatibility with older Chunks
+    that stored the document_id in the metadata."""
+    stored_chunk_id = extract_chunk_id_from_metadata(chunk)
+    if stored_chunk_id:
+        return stored_chunk_id
+    elif "document_id" in chunk.metadata:
+        return generate_chunk_id(chunk.metadata["document_id"], str(chunk.content))
+    else:
+        logging.warning("Chunk has no ID or document_id in metadata. Generating random ID.")
+        return str(uuid.uuid4())