Make sentence_transformers an optional dependency

docs/docs/getting_started/quickstart.md

@@ -21,7 +21,7 @@ We'll need to install a bunch of dependencies for this project.
 1.  Install CocoIndex:
 
     ```bash
-    pip install -U cocoindex
+    pip install -U 'cocoindex[embeddings]'
     ```
 
 2.  You can skip this step if you already have a Postgres database with pgvector extension installed.

docs/docs/ops/functions.md

@@ -77,6 +77,15 @@ Return: [*KTable*](/docs/core/data_types#ktable), each row represents a chunk, w
 
 `SentenceTransformerEmbed` embeds a text into a vector space using the [SentenceTransformer](https://huggingface.co/sentence-transformers) library.
 
+:::note Optional Dependency Required
+
+This function requires the 'sentence-transformers' library, which is an optional dependency. Install CocoIndex with:
+
+```bash
+pip install 'cocoindex[embeddings]'
+```
+:::
+
 The spec takes the following fields:
 
 *   `model` (`str`): The name of the SentenceTransformer model to use.

examples/amazon_s3_embedding/pyproject.toml

@@ -3,7 +3,7 @@ name = "amazon-s3-text-embedding"
 version = "0.1.0"
 description = "Simple example for cocoindex: build embedding index based on Amazon S3 files."
 requires-python = ">=3.11"
-dependencies = ["cocoindex>=0.1.52", "python-dotenv>=1.0.1"]
+dependencies = ["cocoindex[embeddings]>=0.1.52", "python-dotenv>=1.0.1"]
 
 [tool.setuptools]
 packages = []

examples/code_embedding/pyproject.toml

@@ -3,7 +3,7 @@ name = "code-embedding"
 version = "0.1.0"
 description = "Simple example for cocoindex: build embedding index based on source code."
 requires-python = ">=3.11"
-dependencies = ["cocoindex>=0.1.56", "python-dotenv>=1.0.1"]
+dependencies = ["cocoindex[embeddings]>=0.1.56", "python-dotenv>=1.0.1"]
 
 [tool.setuptools]
 packages = []

examples/fastapi_server_docker/requirements.txt

@@ -1,4 +1,4 @@
-cocoindex>=0.1.52
+cocoindex[embeddings]>=0.1.52
 python-dotenv>=1.0.1
 fastapi==0.115.12
 fastapi-cli==0.0.7

examples/gdrive_text_embedding/pyproject.toml

@@ -3,7 +3,7 @@ name = "gdrive-text-embedding"
 version = "0.1.0"
 description = "Simple example for cocoindex: build embedding index based on Google Drive files."
 requires-python = ">=3.11"
-dependencies = ["cocoindex>=0.1.52", "python-dotenv>=1.0.1"]
+dependencies = ["cocoindex[embeddings]>=0.1.52", "python-dotenv>=1.0.1"]
 
 [tool.setuptools]
 packages = []

examples/pdf_embedding/pyproject.toml

@@ -4,7 +4,7 @@ version = "0.1.0"
 description = "Simple example for cocoindex: build embedding index based on local PDF files."
 requires-python = ">=3.11"
 dependencies = [
-    "cocoindex>=0.1.52",
+    "cocoindex[embeddings]>=0.1.52",
     "python-dotenv>=1.0.1",
     "marker-pdf>=1.5.2",
 ]

examples/text_embedding/pyproject.toml

@@ -4,7 +4,7 @@ version = "0.1.0"
 description = "Simple example for cocoindex: build embedding index based on local text files."
 requires-python = ">=3.11"
 dependencies = [
-    "cocoindex>=0.1.52",
+    "cocoindex[embeddings]>=0.1.52",
     "python-dotenv>=1.0.1",
     "pgvector>=0.4.1",
     "psycopg[binary,pool]",

examples/text_embedding_qdrant/README.md

@@ -37,6 +37,7 @@ We use Qdrant client to query the index, and reuse the embedding operation in th
    pip install -e .
    ```
 
+
 - Setup:
 
    ```bash

examples/text_embedding_qdrant/pyproject.toml

@@ -4,7 +4,7 @@ version = "0.1.0"
 description = "Simple example for cocoindex: build embedding index based on local text files."
 requires-python = ">=3.11"
 dependencies = [
-    "cocoindex>=0.1.52",
+    "cocoindex[embeddings]>=0.1.52",
     "python-dotenv>=1.0.1",
     "qdrant-client>=1.6.0",
 ]

pyproject.toml

@@ -10,7 +10,6 @@ authors = [{ name = "CocoIndex", email = "cocoindex.io@gmail.com" }]
 readme = "README.md"
 requires-python = ">=3.11"
 dependencies = [
-    "sentence-transformers>=3.3.1",
     "click>=8.1.8",
     "rich>=14.0.0",
     "python-dotenv>=1.1.0",
@@ -31,6 +30,7 @@ features = ["pyo3/extension-module"]
 [project.optional-dependencies]
 test = ["pytest"]
 dev = ["ruff", "pre-commit"]
+embeddings = ["sentence-transformers>=3.3.1"]
 
 [tool.mypy]
 python_version = "3.11"

python/cocoindex/functions.py

@@ -12,6 +12,14 @@
 if TYPE_CHECKING:
     import sentence_transformers
 
+# Check if sentence_transformers is available
+try:
+    import sentence_transformers
+
+    _SENTENCE_TRANSFORMERS_AVAILABLE = True
+except ImportError:
+    _SENTENCE_TRANSFORMERS_AVAILABLE = False
+
 
 class ParseJson(op.FunctionSpec):
     """Parse a text into a JSON object."""
@@ -58,6 +66,10 @@ class SentenceTransformerEmbed(op.FunctionSpec):
 
         model: The name of the SentenceTransformer model to use.
         args: Additional arguments to pass to the SentenceTransformer constructor. e.g. {"trust_remote_code": True}
+
+    Note:
+        This function requires the optional sentence-transformers dependency.
+        Install it with: pip install 'cocoindex[embeddings]'
     """
 
     model: str
@@ -72,6 +84,14 @@ class SentenceTransformerEmbedExecutor:
     _model: "sentence_transformers.SentenceTransformer"
 
     def analyze(self, text: Any) -> type:
+        if not _SENTENCE_TRANSFORMERS_AVAILABLE:
+            raise ImportError(
+                "sentence_transformers is required for SentenceTransformerEmbed function. "
+                "Install it with one of these commands:\n"
+                "  pip install 'cocoindex[embeddings]'\n"
+                "  pip install sentence-transformers"
+            )
+
         import sentence_transformers  # pylint: disable=import-outside-toplevel
 
         args = self.spec.args or {}