corvic-engine 0.3.0rc48__tar.gz → 0.3.0rc51__tar.gz

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/Cargo.lock

@@ -22,7 +22,7 @@ checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
 
 [[package]]
 name = "corvic-engine"
-version = "0.3.0-rc.48"
+version = "0.3.0-rc.51"
 dependencies = [
  "numpy",
  "pyo3",
@@ -63,7 +63,7 @@ checksum = "60b1af1c220855b6ceac025d3f6ecdd2b7c4894bfe9cd9bda4fbb4bc7c0d4cf0"
 
 [[package]]
 name = "engine-walk"
-version = "0.3.0-rc.48"
+version = "0.3.0-rc.51"
 dependencies = [
  "version_check",
 ]

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/Cargo.toml

@@ -3,7 +3,7 @@ resolver = "2"
 members = ["python"]
 
 [workspace.package]
-version = "0.3.0-rc.48"
+version = "0.3.0-rc.51"
 edition = "2021"
 publish = false
 

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: corvic-engine
-Version: 0.3.0rc48
+Version: 0.3.0rc51
 Classifier: Environment :: Console
 Classifier: License :: Other/Proprietary License
 Classifier: Programming Language :: Python :: Implementation :: CPython

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/pyproject.toml

@@ -20,7 +20,7 @@ authors = [
 ]
 readme = "README.md"
 description = "Seamless embedding generation and retrieval."
-version = "0.3.0-rc.48" # sync this with tool.poetry.version below
+version = "0.3.0-rc.51" # sync this with tool.poetry.version below
 
 # TODO(ddn): Pull dependencies from tool.poetry.dependencies. We use maturin as
 # the build system but poetry for the development environment.
@@ -74,7 +74,7 @@ packages = [
 ]
 description = ""
 authors = []
-version = "0.3.0-rc.48"
+version = "0.3.0-rc.51"
 
 [tool.poetry.scripts]
 check = "corvic_check.cli:main"
@@ -167,7 +167,7 @@ torch = [
 ]
 torch-geometric = "^2.5.3"
 deepgnn-torch = "^0.1.60"
-transformers = "^4.45.0"
+transformers = "^4.48.0"
 umap-learn = "^0.5.5"
 pillow = "10.4.0"
 

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/python/corvic/system/storage.py

@@ -3,10 +3,9 @@
 from __future__ import annotations
 
 import contextlib
-import io
 import uuid
 from collections.abc import Iterator
-from typing import Any, Final, Literal
+from typing import Any, BinaryIO, Final, Literal
 
 from typing_extensions import Protocol
 
@@ -75,9 +74,7 @@ class Blob(Protocol):
     def url(self) -> str: ...
 
     @contextlib.contextmanager
-    def open(
-        self, mode: Literal["rb", "wb"], **kwargs: Any
-    ) -> Iterator[io.BytesIO]: ...
+    def open(self, mode: Literal["rb", "wb"], **kwargs: Any) -> Iterator[BinaryIO]: ...
 
     def delete(self) -> None: ...
 
@@ -176,3 +173,12 @@ class StorageManager:
 
     def get_vector_blob_from_blob_name(self, blob_name: str) -> Blob:
         return self.bucket.blob(f"{self.vector_prefix}/{blob_name}")
+
+    def tabular_prefix_for_room_id(self, room_id: orm.RoomID) -> str:
+        return f"{self.tabular_prefix}/{self._render_room_id(room_id)}"
+
+    def unstructured_prefix_for_room_id(self, room_id: orm.RoomID) -> str:
+        return f"{self.unstructured_prefix}/{self._render_room_id(room_id)}"
+
+    def vector_prefix_for_room_id(self, room_id: orm.RoomID) -> str:
+        return f"{self.vector_prefix}/{self._render_room_id(room_id)}"

corvic_engine-0.3.0rc51/python/corvic/system_sqlite/__init__.py

@@ -0,0 +1,19 @@
+"""Single-machine, sqlite-backed implementation of corvic.system."""
+
+from corvic.system_sqlite.client import (
+    TABULAR_PREFIX,
+    UNSTRUCTURED_PREFIX,
+    VECTOR_PREFIX,
+    Client,
+    FSBlobClient,
+    RDBMSBlobClient,
+)
+
+__all__ = [
+    "Client",
+    "FSBlobClient",
+    "RDBMSBlobClient",
+    "TABULAR_PREFIX",
+    "UNSTRUCTURED_PREFIX",
+    "VECTOR_PREFIX",
+]

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/python/corvic/system_sqlite/client.py

@@ -20,6 +20,10 @@ from corvic.system_sqlite.fs_blob_store import FSBlobClient
 from corvic.system_sqlite.rdbms_blob_store import RDBMSBlobClient
 from corvic.system_sqlite.staging import DuckDBStaging
 
+UNSTRUCTURED_PREFIX = "unstructured_data"
+TABULAR_PREFIX = "tabular_data"
+VECTOR_PREFIX = "vectors"
+
 
 @contextlib.contextmanager
 def _context_requester_org_id_is(value: str | corvic.orm.OrgID) -> Iterator[None]:
@@ -133,9 +137,9 @@ class Client(corvic.system.Client):
         self._storage_manager = corvic.system.StorageManager(
             self._blob_client,
             bucket_name=bucket.name,
-            unstructured_prefix="unstructured_data",
-            tabular_prefix="tabular_data",
-            vector_prefix="vectors",
+            unstructured_prefix=UNSTRUCTURED_PREFIX,
+            tabular_prefix=TABULAR_PREFIX,
+            vector_prefix=VECTOR_PREFIX,
         )
         corvic.orm.Base.metadata.create_all(self._sa_engine)
         duck_db_conn = duckdb.connect(":memory:")

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/python/corvic_generated/orm/v1/agent_pb2.py

@@ -14,7 +14,7 @@ _sym_db = _symbol_database.Default()
 
 
 
-DESCRIPTOR = _descriptor_pool.Default().AddSerializedFile(b'\n\x19\x63orvic/orm/v1/agent.proto\x12\rcorvic.orm.v1\"\xf3\x01\n\x12\x45xecutorParameters\x12L\n\x15\x63ompletion_model_type\x18\x01 \x01(\x0e\x32\x18.corvic.orm.v1.ModelTypeR\x13\x63ompletionModelType\x12\x42\n\x1d\x63ompletion_system_instruction\x18\x02 \x01(\tR\x1b\x63ompletionSystemInstruction\x12\x33\n\x13\x63ompletion_model_id\x18\x03 \x01(\tH\x00R\x11\x63ompletionModelId\x88\x01\x01\x42\x16\n\x14_completion_model_id\"V\n\x10SpaceInstruction\x12 \n\x0binstruction\x18\x01 \x01(\tR\x0binstruction\x12 \n\x0cspace_run_id\x18\x02 \x01(\tR\nspaceRunId\"\x89\x02\n\x16OrchestratorParameters\x12\x1b\n\tspace_ids\x18\x01 \x03(\tR\x08spaceIds\x12k\n\x12space_instructions\x18\x02 \x03(\x0b\x32<.corvic.orm.v1.OrchestratorParameters.SpaceInstructionsEntryR\x11spaceInstructions\x1a\x65\n\x16SpaceInstructionsEntry\x12\x10\n\x03key\x18\x01 \x01(\tR\x03key\x12\x35\n\x05value\x18\x02 \x01(\x0b\x32\x1f.corvic.orm.v1.SpaceInstructionR\x05value:\x02\x38\x01\"\xc5\x01\n\x0f\x41gentParameters\x12R\n\x13\x65xecutor_parameters\x18\x01 \x01(\x0b\x32!.corvic.orm.v1.ExecutorParametersR\x12\x65xecutorParameters\x12^\n\x17orchestrator_parameters\x18\x02 \x01(\x0b\x32%.corvic.orm.v1.OrchestratorParametersR\x16orchestratorParameters*\xe0\x04\n\tModelType\x12\x1a\n\x16MODEL_TYPE_UNSPECIFIED\x10\x00\x12\x1d\n\x19MODEL_TYPE_GEMINI_1_5_PRO\x10\x01\x12%\n!MODEL_TYPE_GEMINI_1_5_PRO_PREVIEW\x10\x02\x12\x1f\n\x1bMODEL_TYPE_GEMINI_1_5_FLASH\x10\x03\x12\'\n#MODEL_TYPE_GEMINI_1_5_FLASH_PREVIEW\x10\x04\x12\'\n#MODEL_TYPE_GEMINI_2_0_FLASH_PREVIEW\x10\r\x12\x30\n,MODEL_TYPE_GEMINI_2_0_FLASH_THINKING_PREVIEW\x10\x10\x12\x1b\n\x17MODEL_TYPE_GPT_4_O_MINI\x10\x05\x12\x16\n\x12MODEL_TYPE_GPT_4_O\x10\x06\x12\x1c\n\x18MODEL_TYPE_GPT_3_5_TURBO\x10\x07\x12\x12\n\x0eMODEL_TYPE_O_1\x10\x0e\x12\x17\n\x13MODEL_TYPE_O_1_MINI\x10\x0f\x12\x17\n\x13MODEL_TYPE_O_3_MINI\x10\x12\x12 \n\x1cMODEL_TYPE_CLAUDE_3_5_SONNET\x10\x08\x12\x1d\n\x19MODEL_TYPE_CLAUDE_3_HAIKU\x10\t\x12\x1c\n\x18MODEL_TYPE_MISTRAL_LARGE\x10\n\x12\x1b\n\x17MODEL_TYPE_MISTRAL_NEMO\x10\x0b\x12 \n\x1cMODEL_TYPE_MISTRAL_CODESTRAL\x10\x0c\x12\x15\n\x11MODEL_TYPE_CUSTOM\x10\x11\x62\x06proto3')
+DESCRIPTOR = _descriptor_pool.Default().AddSerializedFile(b'\n\x19\x63orvic/orm/v1/agent.proto\x12\rcorvic.orm.v1\"\xf3\x01\n\x12\x45xecutorParameters\x12L\n\x15\x63ompletion_model_type\x18\x01 \x01(\x0e\x32\x18.corvic.orm.v1.ModelTypeR\x13\x63ompletionModelType\x12\x42\n\x1d\x63ompletion_system_instruction\x18\x02 \x01(\tR\x1b\x63ompletionSystemInstruction\x12\x33\n\x13\x63ompletion_model_id\x18\x03 \x01(\tH\x00R\x11\x63ompletionModelId\x88\x01\x01\x42\x16\n\x14_completion_model_id\"V\n\x10SpaceInstruction\x12 \n\x0binstruction\x18\x01 \x01(\tR\x0binstruction\x12 \n\x0cspace_run_id\x18\x02 \x01(\tR\nspaceRunId\"\x89\x02\n\x16OrchestratorParameters\x12\x1b\n\tspace_ids\x18\x01 \x03(\tR\x08spaceIds\x12k\n\x12space_instructions\x18\x02 \x03(\x0b\x32<.corvic.orm.v1.OrchestratorParameters.SpaceInstructionsEntryR\x11spaceInstructions\x1a\x65\n\x16SpaceInstructionsEntry\x12\x10\n\x03key\x18\x01 \x01(\tR\x03key\x12\x35\n\x05value\x18\x02 \x01(\x0b\x32\x1f.corvic.orm.v1.SpaceInstructionR\x05value:\x02\x38\x01\"\xc5\x01\n\x0f\x41gentParameters\x12R\n\x13\x65xecutor_parameters\x18\x01 \x01(\x0b\x32!.corvic.orm.v1.ExecutorParametersR\x12\x65xecutorParameters\x12^\n\x17orchestrator_parameters\x18\x02 \x01(\x0b\x32%.corvic.orm.v1.OrchestratorParametersR\x16orchestratorParameters*\x81\x05\n\tModelType\x12\x1a\n\x16MODEL_TYPE_UNSPECIFIED\x10\x00\x12\x1d\n\x19MODEL_TYPE_GEMINI_1_5_PRO\x10\x01\x12%\n!MODEL_TYPE_GEMINI_1_5_PRO_PREVIEW\x10\x02\x12\x1f\n\x1bMODEL_TYPE_GEMINI_1_5_FLASH\x10\x03\x12\'\n#MODEL_TYPE_GEMINI_1_5_FLASH_PREVIEW\x10\x04\x12\'\n#MODEL_TYPE_GEMINI_2_0_FLASH_PREVIEW\x10\r\x12\x30\n,MODEL_TYPE_GEMINI_2_0_FLASH_THINKING_PREVIEW\x10\x10\x12\x1f\n\x1bMODEL_TYPE_GEMINI_2_0_FLASH\x10\x13\x12\x1b\n\x17MODEL_TYPE_GPT_4_O_MINI\x10\x05\x12\x16\n\x12MODEL_TYPE_GPT_4_O\x10\x06\x12\x1c\n\x18MODEL_TYPE_GPT_3_5_TURBO\x10\x07\x12\x12\n\x0eMODEL_TYPE_O_1\x10\x0e\x12\x17\n\x13MODEL_TYPE_O_1_MINI\x10\x0f\x12\x17\n\x13MODEL_TYPE_O_3_MINI\x10\x12\x12 \n\x1cMODEL_TYPE_CLAUDE_3_5_SONNET\x10\x08\x12\x1d\n\x19MODEL_TYPE_CLAUDE_3_HAIKU\x10\t\x12\x1c\n\x18MODEL_TYPE_MISTRAL_LARGE\x10\n\x12\x1b\n\x17MODEL_TYPE_MISTRAL_NEMO\x10\x0b\x12 \n\x1cMODEL_TYPE_MISTRAL_CODESTRAL\x10\x0c\x12\x15\n\x11MODEL_TYPE_CUSTOM\x10\x11\x62\x06proto3')
 
 _globals = globals()
 _builder.BuildMessageAndEnumDescriptors(DESCRIPTOR, _globals)
@@ -24,7 +24,7 @@ if _descriptor._USE_C_DESCRIPTORS == False:
   _globals['_ORCHESTRATORPARAMETERS_SPACEINSTRUCTIONSENTRY']._options = None
   _globals['_ORCHESTRATORPARAMETERS_SPACEINSTRUCTIONSENTRY']._serialized_options = b'8\001'
   _globals['_MODELTYPE']._serialized_start=847
-  _globals['_MODELTYPE']._serialized_end=1455
+  _globals['_MODELTYPE']._serialized_end=1488
   _globals['_EXECUTORPARAMETERS']._serialized_start=45
   _globals['_EXECUTORPARAMETERS']._serialized_end=288
   _globals['_SPACEINSTRUCTION']._serialized_start=290

{corvic_engine-0.3.0rc48 → corvic_engine-0.3.0rc51}/python/corvic_generated/orm/v1/agent_pb2.pyi

@@ -15,6 +15,7 @@ class ModelType(int, metaclass=_enum_type_wrapper.EnumTypeWrapper):
     MODEL_TYPE_GEMINI_1_5_FLASH_PREVIEW: _ClassVar[ModelType]
     MODEL_TYPE_GEMINI_2_0_FLASH_PREVIEW: _ClassVar[ModelType]
     MODEL_TYPE_GEMINI_2_0_FLASH_THINKING_PREVIEW: _ClassVar[ModelType]
+    MODEL_TYPE_GEMINI_2_0_FLASH: _ClassVar[ModelType]
     MODEL_TYPE_GPT_4_O_MINI: _ClassVar[ModelType]
     MODEL_TYPE_GPT_4_O: _ClassVar[ModelType]
     MODEL_TYPE_GPT_3_5_TURBO: _ClassVar[ModelType]
@@ -34,6 +35,7 @@ MODEL_TYPE_GEMINI_1_5_FLASH: ModelType
 MODEL_TYPE_GEMINI_1_5_FLASH_PREVIEW: ModelType
 MODEL_TYPE_GEMINI_2_0_FLASH_PREVIEW: ModelType
 MODEL_TYPE_GEMINI_2_0_FLASH_THINKING_PREVIEW: ModelType
+MODEL_TYPE_GEMINI_2_0_FLASH: ModelType
 MODEL_TYPE_GPT_4_O_MINI: ModelType
 MODEL_TYPE_GPT_4_O: ModelType
 MODEL_TYPE_GPT_3_5_TURBO: ModelType

corvic_engine-0.3.0rc48/python/corvic/system_sqlite/__init__.py

@@ -1,9 +0,0 @@
-"""Single-machine, sqlite-backed implementation of corvic.system."""
-
-from corvic.system_sqlite.client import Client, FSBlobClient, RDBMSBlobClient
-
-__all__ = [
-    "Client",
-    "FSBlobClient",
-    "RDBMSBlobClient",
-]