migrate memory banks to Resource and new registration (#411)

* migrate memory banks to Resource and new registration

* address feedback

* address feedback

* fix tests

* pgvector fix

* pgvector fix v2

* remove auto discovery

* change register signature to make params required

* update client

* client fix

* use annotated union to parse

* remove base MemoryBank inheritence

---------

Co-authored-by: Dinesh Yeduguru <dineshyv@fb.com>
This commit is contained in:
Dinesh Yeduguru 2024-11-11 17:10:44 -08:00 committed by GitHub
parent 6b9850e11b
commit 38cce97597
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
19 changed files with 240 additions and 129 deletions

View file

@ -5,11 +5,21 @@
# the root directory of this source tree.
from enum import Enum
from typing import List, Literal, Optional, Protocol, runtime_checkable, Union
from typing import (
Annotated,
List,
Literal,
Optional,
Protocol,
runtime_checkable,
Union,
)
from llama_models.schema_utils import json_schema_type, webmethod
from pydantic import BaseModel, Field
from typing_extensions import Annotated
from llama_stack.apis.resource import Resource, ResourceType
@json_schema_type
@ -20,59 +30,98 @@ class MemoryBankType(Enum):
graph = "graph"
class CommonDef(BaseModel):
identifier: str
# Hack: move this out later
provider_id: str = ""
@json_schema_type
class VectorMemoryBankDef(CommonDef):
type: Literal[MemoryBankType.vector.value] = MemoryBankType.vector.value
class VectorMemoryBank(Resource):
type: Literal[ResourceType.memory_bank.value] = ResourceType.memory_bank.value
memory_bank_type: Literal[MemoryBankType.vector.value] = MemoryBankType.vector.value
embedding_model: str
chunk_size_in_tokens: int
overlap_size_in_tokens: Optional[int] = None
@json_schema_type
class KeyValueMemoryBankDef(CommonDef):
type: Literal[MemoryBankType.keyvalue.value] = MemoryBankType.keyvalue.value
class KeyValueMemoryBank(Resource):
type: Literal[ResourceType.memory_bank.value] = ResourceType.memory_bank.value
memory_bank_type: Literal[MemoryBankType.keyvalue.value] = (
MemoryBankType.keyvalue.value
)
@json_schema_type
class KeywordMemoryBankDef(CommonDef):
type: Literal[MemoryBankType.keyword.value] = MemoryBankType.keyword.value
class KeywordMemoryBank(Resource):
type: Literal[ResourceType.memory_bank.value] = ResourceType.memory_bank.value
memory_bank_type: Literal[MemoryBankType.keyword.value] = (
MemoryBankType.keyword.value
)
@json_schema_type
class GraphMemoryBankDef(CommonDef):
type: Literal[MemoryBankType.graph.value] = MemoryBankType.graph.value
class GraphMemoryBank(Resource):
type: Literal[ResourceType.memory_bank.value] = ResourceType.memory_bank.value
memory_bank_type: Literal[MemoryBankType.graph.value] = MemoryBankType.graph.value
MemoryBankDef = Annotated[
@json_schema_type
class VectorMemoryBankParams(BaseModel):
memory_bank_type: Literal[MemoryBankType.vector.value] = MemoryBankType.vector.value
embedding_model: str
chunk_size_in_tokens: int
overlap_size_in_tokens: Optional[int] = None
@json_schema_type
class KeyValueMemoryBankParams(BaseModel):
memory_bank_type: Literal[MemoryBankType.keyvalue.value] = (
MemoryBankType.keyvalue.value
)
@json_schema_type
class KeywordMemoryBankParams(BaseModel):
memory_bank_type: Literal[MemoryBankType.keyword.value] = (
MemoryBankType.keyword.value
)
@json_schema_type
class GraphMemoryBankParams(BaseModel):
memory_bank_type: Literal[MemoryBankType.graph.value] = MemoryBankType.graph.value
MemoryBank = Annotated[
Union[
VectorMemoryBankDef,
KeyValueMemoryBankDef,
KeywordMemoryBankDef,
GraphMemoryBankDef,
VectorMemoryBank,
KeyValueMemoryBank,
KeywordMemoryBank,
GraphMemoryBank,
],
Field(discriminator="type"),
Field(discriminator="memory_bank_type"),
]
MemoryBankDefWithProvider = MemoryBankDef
BankParams = Annotated[
Union[
VectorMemoryBankParams,
KeyValueMemoryBankParams,
KeywordMemoryBankParams,
GraphMemoryBankParams,
],
Field(discriminator="memory_bank_type"),
]
@runtime_checkable
class MemoryBanks(Protocol):
@webmethod(route="/memory_banks/list", method="GET")
async def list_memory_banks(self) -> List[MemoryBankDefWithProvider]: ...
async def list_memory_banks(self) -> List[MemoryBank]: ...
@webmethod(route="/memory_banks/get", method="GET")
async def get_memory_bank(
self, identifier: str
) -> Optional[MemoryBankDefWithProvider]: ...
async def get_memory_bank(self, memory_bank_id: str) -> Optional[MemoryBank]: ...
@webmethod(route="/memory_banks/register", method="POST")
async def register_memory_bank(
self, memory_bank: MemoryBankDefWithProvider
) -> None: ...
self,
memory_bank_id: str,
params: BankParams,
provider_id: Optional[str] = None,
provider_memorybank_id: Optional[str] = None,
) -> MemoryBank: ...