langroid 0.37.2__tar.gz → 0.37.4__tar.gz

{langroid-0.37.2 → langroid-0.37.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: langroid
-Version: 0.37.2
+Version: 0.37.4
 Summary: Harness LLMs with Multi-Agent Programming
 Author-email: Prasad Chalasani <pchalasani@gmail.com>
 License: MIT

{langroid-0.37.2 → langroid-0.37.4}/langroid/vector_store/base.py

@@ -8,7 +8,7 @@ import pandas as pd
 
 from langroid.embedding_models.base import EmbeddingModel, EmbeddingModelsConfig
 from langroid.embedding_models.models import OpenAIEmbeddingsConfig
-from langroid.mytypes import DocMetaData, Document
+from langroid.mytypes import DocMetaData, Document, EmbeddingFunction
 from langroid.pydantic_v1 import BaseSettings
 from langroid.utils.algorithms.graph import components, topological_sort
 from langroid.utils.configuration import settings
@@ -51,6 +51,7 @@ class VectorStore(ABC):
             self.embedding_model = EmbeddingModel.create(config.embedding)
         else:
             self.embedding_model = config.embedding_model
+        self.embedding_fn: EmbeddingFunction = self.embedding_model.embedding_fn()
 
     @staticmethod
     def create(config: VectorStoreConfig) -> Optional["VectorStore"]:
@@ -86,6 +87,10 @@ class VectorStore(ABC):
             )
             return None
 
+    @property
+    def embedding_dim(self) -> int:
+        return len(self.embedding_fn(["test"])[0])
+
     @abstractmethod
     def clear_empty_collections(self) -> int:
         """Clear all empty collections in the vector store.

{langroid-0.37.2 → langroid-0.37.4}/langroid/vector_store/chromadb.py

@@ -35,7 +35,6 @@ class ChromaDB(VectorStore):
         except ImportError:
             raise LangroidImportError("chromadb", "chromadb")
         self.config = config
-        self.embedding_fn = self.embedding_model.embedding_fn()
         self.client = chromadb.Client(
             chromadb.config.Settings(
                 # chroma_db_impl="duckdb+parquet",

{langroid-0.37.2 → langroid-0.37.4}/langroid/vector_store/lancedb.py

@@ -26,7 +26,7 @@ from langroid.embedding_models.base import (
 )
 from langroid.embedding_models.models import OpenAIEmbeddingsConfig
 from langroid.exceptions import LangroidImportError
-from langroid.mytypes import Document, EmbeddingFunction
+from langroid.mytypes import Document
 from langroid.utils.configuration import settings
 from langroid.utils.pydantic_utils import (
     dataframe_to_document_model,
@@ -60,8 +60,6 @@ class LanceDB(VectorStore):
             raise LangroidImportError("lancedb", "lancedb")
 
         self.config: LanceDBConfig = config
-        self.embedding_fn: EmbeddingFunction = self.embedding_model.embedding_fn()
-        self.embedding_dim = self.embedding_model.embedding_dims
         self.host = config.host
         self.port = config.port
         self.is_from_dataframe = False  # were docs ingested from a dataframe?

{langroid-0.37.2 → langroid-0.37.4}/langroid/vector_store/momento.py

@@ -37,7 +37,7 @@ from langroid.embedding_models.base import (
     EmbeddingModelsConfig,
 )
 from langroid.embedding_models.models import OpenAIEmbeddingsConfig
-from langroid.mytypes import Document, EmbeddingFunction
+from langroid.mytypes import Document
 from langroid.utils.configuration import settings
 from langroid.utils.pydantic_utils import (
     flatten_pydantic_instance,
@@ -61,8 +61,6 @@ class MomentoVI(VectorStore):
             raise LangroidImportError("momento", "momento")
         self.distance = SimilarityMetric.COSINE_SIMILARITY
         self.config: MomentoVIConfig = config
-        self.embedding_fn: EmbeddingFunction = self.embedding_model.embedding_fn()
-        self.embedding_dim = self.embedding_model.embedding_dims
         self.host = config.host
         self.port = config.port
         load_dotenv()

{langroid-0.37.2 → langroid-0.37.4}/langroid/vector_store/qdrantdb.py

@@ -27,7 +27,7 @@ from langroid.embedding_models.base import (
     EmbeddingModelsConfig,
 )
 from langroid.embedding_models.models import OpenAIEmbeddingsConfig
-from langroid.mytypes import Document, EmbeddingFunction, Embeddings
+from langroid.mytypes import Document, Embeddings
 from langroid.utils.configuration import settings
 from langroid.vector_store.base import VectorStore, VectorStoreConfig
 
@@ -77,8 +77,6 @@ class QdrantDB(VectorStore):
     def __init__(self, config: QdrantDBConfig = QdrantDBConfig()):
         super().__init__(config)
         self.config: QdrantDBConfig = config
-        self.embedding_fn: EmbeddingFunction = self.embedding_model.embedding_fn()
-        self.embedding_dim = len(self.embedding_fn(["test"])[0])
         if self.config.use_sparse_embeddings:
             try:
                 from transformers import AutoModelForMaskedLM, AutoTokenizer

{langroid-0.37.2 → langroid-0.37.4}/langroid/vector_store/weaviatedb.py

@@ -10,7 +10,7 @@ from langroid.embedding_models.base import (
 )
 from langroid.embedding_models.models import OpenAIEmbeddingsConfig
 from langroid.exceptions import LangroidImportError
-from langroid.mytypes import DocMetaData, Document, EmbeddingFunction
+from langroid.mytypes import DocMetaData, Document
 from langroid.utils.configuration import settings
 from langroid.vector_store.base import VectorStore, VectorStoreConfig
 
@@ -38,8 +38,6 @@ class WeaviateDB(VectorStore):
     def __init__(self, config: WeaviateDBConfig = WeaviateDBConfig()):
         super().__init__(config)
         self.config: WeaviateDBConfig = config
-        self.embedding_fn: EmbeddingFunction = self.embedding_model.embedding_fn()
-        self.embedding_dim = self.embedding_model.embedding_dims
         load_dotenv()
         key = os.getenv("WEAVIATE_API_KEY")
         url = os.getenv("WEAVIATE_API_URL")
@@ -211,10 +209,10 @@ class WeaviateDB(VectorStore):
             return_properties=True,
             return_metadata=MetadataQuery(distance=True),
         )
-        return [
-            (self.weaviate_obj_to_doc(item), 1 - (item.metadata.distance or 1))
-            for item in response.objects
-        ]
+        maybe_distances = [item.metadata.distance for item in response.objects]
+        similarities = [0 if d is None else 1 - d for d in maybe_distances]
+        docs = [self.weaviate_obj_to_doc(item) for item in response.objects]
+        return list(zip(docs, similarities))
 
     def _create_valid_uuid_id(self, id: str) -> Any:
         try:

{langroid-0.37.2 → langroid-0.37.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "langroid"
-version = "0.37.2"
+version = "0.37.4"
 authors = [
     {name = "Prasad Chalasani", email = "pchalasani@gmail.com"},
 ]