corvic-engine 0.3.0rc78__tar.gz → 0.3.0rc80__tar.gz

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/Cargo.lock

@@ -16,7 +16,7 @@ checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
 
 [[package]]
 name = "corvic-engine"
-version = "0.3.0-rc.78"
+version = "0.3.0-rc.80"
 dependencies = [
  "numpy",
  "pyo3",
@@ -57,7 +57,7 @@ checksum = "48c757948c5ede0e46177b7add2e67155f70e33c07fea8284df6576da70b3719"
 
 [[package]]
 name = "engine-walk"
-version = "0.3.0-rc.78"
+version = "0.3.0-rc.80"
 dependencies = [
  "version_check",
 ]

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/Cargo.toml

@@ -3,7 +3,7 @@ resolver = "2"
 members = ["python"]
 
 [workspace.package]
-version = "0.3.0-rc.78"
+version = "0.3.0-rc.80"
 edition = "2021"
 publish = false
 

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: corvic-engine
-Version: 0.3.0rc78
+Version: 0.3.0rc80
 Classifier: Environment :: Console
 Classifier: License :: Other/Proprietary License
 Classifier: Programming Language :: Python :: Implementation :: CPython

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/pyproject.toml

@@ -18,7 +18,7 @@ authors = [
 ]
 readme = "README.md"
 description = "Seamless embedding generation and retrieval."
-version = "0.3.0-rc.78" # sync this with tool.poetry.version below
+version = "0.3.0-rc.80" # 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.
@@ -72,7 +72,7 @@ packages = [
 ]
 description = ""
 authors = []
-version = "0.3.0-rc.78"
+version = "0.3.0-rc.80"
 
 [tool.poetry.scripts]
 check = "corvic_check.cli:main"
@@ -111,7 +111,8 @@ typing-extensions = "^4.9.0"
 
 [tool.poetry.group.dev.dependencies]
 anyio = "^4"
-click = "^8.1.7"
+# TODO(thunt): update corvic_check to work with 8.2
+click = ">=8.1,<8.2.0"
 click_default_group = "^1.2.4"
 codespell = "^2.2.6"
 coverage = "^7.3.1"
@@ -121,7 +122,7 @@ maturin = "^1.4.0"
 pdoc = "^14.5.1"
 pre-commit = "^4.0.0"
 psycopg = { extras = ["binary"], version = "^3.2" }
-pyright = "^1.1.387"
+pyright = "1.1.400"
 pytest = "^8.1.1"
 pytest-asyncio = "^0.26"
 pytest-benchmark = "^4.0.0"
@@ -334,7 +335,7 @@ known-third-party = ["proto", "rockset"]
 
 [tool.pyright]
 include = ["python", "tests"]
-ignore = ["python/corvic_generated", "python/buf", "typings_generated"]
+ignore = ["python/corvic_generated", "python/buf", "typings_generated", "_typeshed"]
 typeCheckingMode = "strict"
 # Enable additional checks beyond strict
 reportCallInDefaultInitializer = "error"

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/python/corvic/orm/__init__.py

@@ -139,7 +139,7 @@ class Base(sa_orm.MappedAsDataclass, sa_orm.DeclarativeBase):
         return cls._updated_at
 
 
-class Org(Base, kw_only=True):
+class Org(SoftDeleteMixin, Base, kw_only=True):
     """An organization is a top-level grouping of resources."""
 
     __tablename__ = "org"

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/python/corvic/orm/_soft_delete.py

@@ -81,6 +81,11 @@ class SoftDeleteMixin(sa_orm.MappedAsDataclass):
             return None
         return self._deleted_at.replace(tzinfo=UTC)
 
+    @deleted_at.inplace.expression
+    @classmethod
+    def _deleted_at_at_expression(cls):
+        return cls._deleted_at
+
     def reset_delete(self):
         self._deleted_at = None
 

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/python/corvic/system/__init__.py

@@ -29,7 +29,6 @@ from corvic.system.client import Client
 from corvic.system.in_memory_executor import (
     InMemoryExecutionResult,
     InMemoryExecutor,
-    batch_to_proto_struct,  # pyright: ignore[reportDeprecated]
     get_polars_embedding,
     get_polars_embedding_length,
     make_dict_bytes_human_readable,
@@ -96,7 +95,6 @@ __all__ = [
     "UmapDimensionReducer",
     "ValidateFirstExecutor",
     "VectorSimilarityMetric",
-    "batch_to_proto_struct",
     "get_polars_embedding",
     "get_polars_embedding_length",
     "make_dict_bytes_human_readable",

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/python/corvic/system/_image_embedder.py

@@ -180,17 +180,53 @@ class Clip(ImageEmbedder):
         )
 
 
+class SigLIP2(Clip):
+    """SigLIP2 image embedder."""
+
+    def _load_models(self):
+        from transformers.models.auto.modeling_auto import AutoModel
+        from transformers.models.auto.processing_auto import AutoProcessor
+        from transformers.models.clip import (
+            CLIPModel,
+            CLIPProcessor,
+        )
+
+        model = cast(
+            CLIPModel,
+            AutoModel.from_pretrained(  # pyright: ignore[reportUnknownMemberType]
+                pretrained_model_name_or_path="google/siglip2-base-patch16-512",
+                revision="a89f5c5093f902bf39d3cd4d81d2c09867f0724b",
+                device_map="auto",
+            ),
+        )
+        processor = cast(
+            CLIPProcessor,
+            AutoProcessor.from_pretrained(  # pyright: ignore[reportUnknownMemberType]
+                pretrained_model_name_or_path="google/siglip2-base-patch16-512",
+                revision="a89f5c5093f902bf39d3cd4d81d2c09867f0724b",
+            ),
+        )
+        return ClipModels(model=model, processor=processor)
+
+
 class CombinedImageEmbedder(ImageEmbedder):
     def __init__(self):
         self._clip_embedder = Clip()
+        self._siglip2_embedder = SigLIP2()
         self._random_embedder = RandomImageEmbedder()
 
     def embed(
         self, context: EmbedImageContext
     ) -> Ok[EmbedImageResult] | InvalidArgumentError | InternalError:
-        if context.model_name == "random":
-            return self._random_embedder.embed(context)
-        return self._clip_embedder.embed(context)
+        match context.model_name:
+            case "random":
+                return self._random_embedder.embed(context)
+            case "clip":
+                return self._clip_embedder.embed(context)
+            case "siglip2":
+                return self._siglip2_embedder.embed(context)
+            case _:
+                return InvalidArgumentError(f"Unknown model name {context.model_name}")
 
     async def aembed(
         self,

{corvic_engine-0.3.0rc78 → corvic_engine-0.3.0rc80}/python/corvic/system/in_memory_executor.py

@@ -18,9 +18,7 @@ import polars.selectors as cs
 import pyarrow as pa
 import pyarrow.parquet as pq
 import structlog
-from google.protobuf import json_format, struct_pb2
 from more_itertools import flatten
-from typing_extensions import deprecated
 
 import corvic.system._column_encoding as column_encoding
 from corvic import embed, embedding_metric, op_graph, sql
@@ -99,17 +97,6 @@ def get_polars_embedding(
     )
 
 
-@deprecated("use pa_scalar.batch_to_structs instead")
-def batch_to_proto_struct(batch: pa.RecordBatch) -> list[struct_pb2.Struct]:
-    """Converts a RecordBatch to protobuf Structs safely."""
-    data = batch.to_pylist()
-    structs = [struct_pb2.Struct() for _ in range(len(data))]
-    for idx, datum in enumerate(data):
-        make_dict_bytes_human_readable(datum)
-        json_format.ParseDict(datum, structs[idx])
-    return structs
-
-
 def make_list_bytes_human_readable(data: list[Any]) -> None:
     """Utility function to cleanup list data types.