embedkit 0.1.2__tar.gz → 0.1.4__tar.gz

embedkit-0.1.4/PKG-INFO

@@ -0,0 +1,137 @@
+Metadata-Version: 2.4
+Name: embedkit
+Version: 0.1.4
+Summary: A simple toolkit for generating vector embeddings across multiple providers and models
+Author-email: JP Hwang <me@jphwang.com>
+License: MIT
+License-File: LICENSE
+Keywords: ai,cohere,colpali,embeddings,machine-learning,vector
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Requires-Python: >=3.10
+Requires-Dist: accelerate>=1.7.0
+Requires-Dist: cohere>=5.15.0
+Requires-Dist: colpali-engine<0.4.0,>=0.3.0
+Requires-Dist: pdf2image>=1.17.0
+Requires-Dist: pillow>=11.2.1
+Requires-Dist: torch<=2.5
+Requires-Dist: transformers
+Description-Content-Type: text/markdown
+
+# EmbedKit
+
+A unified interface for text and image embeddings, supporting multiple providers.
+
+## Installation
+
+```bash
+pip install embedkit
+```
+
+## Usage
+
+### Text Embeddings
+
+```python
+from embedkit import EmbedKit
+from embedkit.classes import Model, CohereInputType
+
+# Initialize with ColPali
+kit = EmbedKit.colpali(
+    model=Model.ColPali.V1_3,
+    text_batch_size=16,  # Optional: process text in batches of 16
+    image_batch_size=8,  # Optional: process images in batches of 8
+)
+
+# Get embeddings
+result = kit.embed_text("Hello world")
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+
+# Initialize with Cohere
+kit = EmbedKit.cohere(
+    model=Model.Cohere.EMBED_V4_0,
+    api_key="your-api-key",
+    text_input_type=CohereInputType.SEARCH_QUERY,  # or SEARCH_DOCUMENT
+    text_batch_size=64,  # Optional: process text in batches of 64
+    image_batch_size=8,  # Optional: process images in batches of 8
+)
+
+# Get embeddings
+result = kit.embed_text("Hello world")
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+```
+
+### Image Embeddings
+
+```python
+from pathlib import Path
+
+# Get embeddings for an image
+image_path = Path("path/to/image.png")
+result = kit.embed_image(image_path)
+
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+```
+
+### PDF Embeddings
+
+```python
+from pathlib import Path
+
+# Get embeddings for a PDF
+pdf_path = Path("path/to/document.pdf")
+result = kit.embed_pdf(pdf_path)
+
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+```
+
+## Response Format
+
+The embedding methods return an `EmbeddingResponse` object with the following structure:
+
+```python
+class EmbeddingResponse:
+    model_name: str
+    model_provider: str
+    input_type: str
+    objects: List[EmbeddingObject]
+
+class EmbeddingObject:
+    embedding: np.ndarray
+    source_b64: Optional[str]
+```
+
+## Supported Models
+
+### ColPali
+- `Model.ColPali.V1_3`
+
+### Cohere
+- `Model.Cohere.EMBED_V4_0`
+
+## Requirements
+
+- Python 3.10+
+
+## License
+
+MIT

embedkit-0.1.4/README.md

@@ -0,0 +1,110 @@
+# EmbedKit
+
+A unified interface for text and image embeddings, supporting multiple providers.
+
+## Installation
+
+```bash
+pip install embedkit
+```
+
+## Usage
+
+### Text Embeddings
+
+```python
+from embedkit import EmbedKit
+from embedkit.classes import Model, CohereInputType
+
+# Initialize with ColPali
+kit = EmbedKit.colpali(
+    model=Model.ColPali.V1_3,
+    text_batch_size=16,  # Optional: process text in batches of 16
+    image_batch_size=8,  # Optional: process images in batches of 8
+)
+
+# Get embeddings
+result = kit.embed_text("Hello world")
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+
+# Initialize with Cohere
+kit = EmbedKit.cohere(
+    model=Model.Cohere.EMBED_V4_0,
+    api_key="your-api-key",
+    text_input_type=CohereInputType.SEARCH_QUERY,  # or SEARCH_DOCUMENT
+    text_batch_size=64,  # Optional: process text in batches of 64
+    image_batch_size=8,  # Optional: process images in batches of 8
+)
+
+# Get embeddings
+result = kit.embed_text("Hello world")
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+```
+
+### Image Embeddings
+
+```python
+from pathlib import Path
+
+# Get embeddings for an image
+image_path = Path("path/to/image.png")
+result = kit.embed_image(image_path)
+
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+```
+
+### PDF Embeddings
+
+```python
+from pathlib import Path
+
+# Get embeddings for a PDF
+pdf_path = Path("path/to/document.pdf")
+result = kit.embed_pdf(pdf_path)
+
+print(result.model_provider)
+print(result.input_type)
+print(result.objects[0].embedding.shape)
+print(result.objects[0].source_b64)
+```
+
+## Response Format
+
+The embedding methods return an `EmbeddingResponse` object with the following structure:
+
+```python
+class EmbeddingResponse:
+    model_name: str
+    model_provider: str
+    input_type: str
+    objects: List[EmbeddingObject]
+
+class EmbeddingObject:
+    embedding: np.ndarray
+    source_b64: Optional[str]
+```
+
+## Supported Models
+
+### ColPali
+- `Model.ColPali.V1_3`
+
+### Cohere
+- `Model.Cohere.EMBED_V4_0`
+
+## Requirements
+
+- Python 3.10+
+
+## License
+
+MIT

{embedkit-0.1.2 → embedkit-0.1.4}/main.py

@@ -1,7 +1,6 @@
 # ./main.py
 from embedkit import EmbedKit
-from embedkit.models import Model
-from embedkit.providers.cohere import CohereInputType
+from embedkit.classes import Model, CohereInputType
 from pathlib import Path
 import os
 
@@ -38,23 +37,24 @@ long_pdf = Path("tmp/2407.01449v6.pdf")
 kit = EmbedKit.colpali(model=Model.ColPali.V1_3, text_batch_size=16, image_batch_size=8)
 
 results = kit.embed_text("Hello world")
-assert results.shape[0] == 1
-assert len(results.shape) == 3
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 2
+assert results.objects[0].source_b64 == None
 
 results = kit.embed_image(sample_image)
-assert results.shape[0] == 1
-assert len(results.shape) == 3
-assert len(results.source_images_b64) > 0
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 2
+assert type(results.objects[0].source_b64) == str
 
 results = kit.embed_pdf(sample_pdf)
-assert results.shape[0] == 1
-assert len(results.shape) == 3
-assert len(results.source_images_b64) > 0
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 2
+assert type(results.objects[0].source_b64) == str
 
-results = kit.embed_pdf(long_pdf)
-assert results.shape[0] == 26
-assert len(results.shape) == 3
-assert len(results.source_images_b64) > 0
+# results = kit.embed_pdf(long_pdf)
+# assert len(results.objects) == 26
+# assert len(results.objects[0].embedding.shape) == 2
+# assert type(results.objects[0].source_b64) == str
 
 
 kit = EmbedKit.cohere(
@@ -66,8 +66,9 @@ kit = EmbedKit.cohere(
 )
 
 results = kit.embed_text("Hello world")
-assert results.shape[0] == 1
-assert len(results.shape) == 2
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 1
+assert results.objects[0].source_b64 == None
 
 kit = EmbedKit.cohere(
     model=Model.Cohere.EMBED_V4_0,
@@ -78,20 +79,21 @@ kit = EmbedKit.cohere(
 )
 
 results = kit.embed_text("Hello world")
-assert results.shape[0] == 1
-assert len(results.shape) == 2
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 1
+assert results.objects[0].source_b64 == None
 
 results = kit.embed_image(sample_image)
-assert results.shape[0] == 1
-assert len(results.shape) == 2
-assert len(results.source_images_b64) > 0
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 1
+assert type(results.objects[0].source_b64) == str
 
 results = kit.embed_pdf(sample_pdf)
-assert results.shape[0] == 1
-assert len(results.shape) == 2
-assert len(results.source_images_b64) > 0
-
-results = kit.embed_pdf(long_pdf)
-assert results.shape[0] == 26
-assert len(results.shape) == 2
-assert len(results.source_images_b64) > 0
+assert len(results.objects) == 1
+assert len(results.objects[0].embedding.shape) == 1
+assert type(results.objects[0].source_b64) == str
+
+# results = kit.embed_pdf(long_pdf)
+# assert len(results.objects) == 1
+# assert len(results.objects[0].embedding.shape) == 1
+# assert type(results.objects[0].source_b64) == str

{embedkit-0.1.2 → embedkit-0.1.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "embedkit"
-version = "0.1.2"
+version = "0.1.4"
 description = "A simple toolkit for generating vector embeddings across multiple providers and models"
 readme = "README.md"
 requires-python = ">=3.10"

{embedkit-0.1.2 → embedkit-0.1.4}/src/embedkit/__init__.py

@@ -8,7 +8,7 @@ from pathlib import Path
 import numpy as np
 
 from .models import Model
-from .base import EmbeddingError, EmbeddingResult
+from .base import EmbeddingError, EmbeddingResponse
 from .providers import ColPaliProvider, CohereProvider
 from .providers.cohere import CohereInputType
 
@@ -84,8 +84,8 @@ class EmbedKit:
 
         provider = CohereProvider(
             api_key=api_key, model_name=model_name,
-            text_batch_size=48,
-            image_batch_size=8,
+            text_batch_size=text_batch_size,
+            image_batch_size=image_batch_size,
             text_input_type=text_input_type
         )
         return cls(provider)
@@ -103,7 +103,7 @@ class EmbedKit:
     #     provider = HuggingFaceProvider(model_name=model_name, device=device)
     #     return cls(provider)
 
-    def embed_text(self, texts: Union[str, List[str]], **kwargs) -> EmbeddingResult:
+    def embed_text(self, texts: Union[str, List[str]], **kwargs) -> EmbeddingResponse:
         """Generate document text embeddings using the configured provider.
 
         Args:
@@ -117,11 +117,11 @@ class EmbedKit:
 
     def embed_image(
         self, images: Union[Path, str, List[Union[Path, str]]]
-    ) -> EmbeddingResult:
+    ) -> EmbeddingResponse:
         """Generate image embeddings using the configured provider."""
         return self._provider.embed_image(images)
 
-    def embed_pdf(self, pdf: Union[Path, str]) -> EmbeddingResult:
+    def embed_pdf(self, pdf: Union[Path, str]) -> EmbeddingResponse:
         """Generate image embeddings from PDFsusing the configured provider. Takes a single PDF file."""
         return self._provider.embed_pdf(pdf)
 

{embedkit-0.1.2 → embedkit-0.1.4}/src/embedkit/base.py

@@ -9,35 +9,40 @@ from dataclasses import dataclass
 
 
 @dataclass
-class EmbeddingResult:
-    embeddings: np.ndarray
+class EmbeddingObject:
+    embedding: np.ndarray
+    source_b64: str = None
+
+
+@dataclass
+class EmbeddingResponse:
     model_name: str
     model_provider: str
     input_type: str
-    source_images_b64: Optional[List[str]] = None
+    objects: List[EmbeddingObject]
 
     @property
     def shape(self) -> tuple:
-        return self.embeddings.shape
+        return self.objects[0].embedding.shape
 
 
 class EmbeddingProvider(ABC):
     """Abstract base class for embedding providers."""
 
     @abstractmethod
-    def embed_text(self, texts: Union[str, List[str]], **kwargs) -> EmbeddingResult:
+    def embed_text(self, texts: Union[str, List[str]], **kwargs) -> EmbeddingResponse:
         """Generate document text embeddings using the configured provider."""
         pass
 
     @abstractmethod
     def embed_image(
         self, images: Union[Path, str, List[Union[Path, str]]]
-    ) -> EmbeddingResult:
+    ) -> EmbeddingResponse:
         """Generate image embeddings using the configured provider."""
         pass
 
     @abstractmethod
-    def embed_pdf(self, pdf: Union[Path, str]) -> EmbeddingResult:
+    def embed_pdf(self, pdf: Union[Path, str]) -> EmbeddingResponse:
         """Generate image embeddings from PDFsusing the configured provider. Takes a single PDF file."""
         pass
 

embedkit-0.1.4/src/embedkit/classes.py

@@ -0,0 +1,21 @@
+# ./src/embedkit/classes.py
+
+"""Core types and enums for the EmbedKit library.
+
+This module provides the main types and enums that users should interact with:
+- EmbeddingResult: The result type returned by embedding operations
+- EmbeddingError: Exception type for embedding operations
+- Model: Enum of supported embedding models
+- CohereInputType: Enum for Cohere's input types
+"""
+
+from . import EmbeddingResponse, EmbeddingError
+from .models import Model
+from .providers.cohere import CohereInputType
+
+__all__ = [
+    "EmbeddingResponse",
+    "EmbeddingError",
+    "Model",
+    "CohereInputType"
+]

{embedkit-0.1.2 → embedkit-0.1.4}/src/embedkit/providers/cohere.py

@@ -7,7 +7,7 @@ import numpy as np
 from enum import Enum
 
 from ..utils import pdf_to_images, image_to_base64
-from ..base import EmbeddingProvider, EmbeddingError, EmbeddingResult
+from ..base import EmbeddingProvider, EmbeddingError, EmbeddingResponse, EmbeddingObject
 
 
 class CohereInputType(Enum):
@@ -51,7 +51,7 @@ class CohereProvider(EmbeddingProvider):
                 raise EmbeddingError(f"Failed to initialize Cohere client: {e}") from e
         return self._client
 
-    def embed_text(self, texts: Union[str, List[str]], **kwargs) -> EmbeddingResult:
+    def embed_text(self, texts: Union[str, List[str]], **kwargs) -> EmbeddingResponse:
         """Generate text embeddings using the Cohere API."""
         client = self._get_client()
 
@@ -70,13 +70,17 @@ class CohereProvider(EmbeddingProvider):
                     input_type=self.input_type.value,
                     embedding_types=["float"],
                 )
-                all_embeddings.extend(response.embeddings.float_)
+                all_embeddings.extend(np.array(response.embeddings.float_))
 
-            return EmbeddingResult(
-                embeddings=np.array(all_embeddings),
+            return EmbeddingResponse(
                 model_name=self.model_name,
                 model_provider=self.provider_name,
                 input_type=self.input_type.value,
+                objects=[
+                    EmbeddingObject(
+                        embedding=e,
+                    ) for e in all_embeddings
+                ]
             )
 
         except Exception as e:
@@ -85,7 +89,7 @@ class CohereProvider(EmbeddingProvider):
     def embed_image(
         self,
         images: Union[Path, str, List[Union[Path, str]]],
-    ) -> EmbeddingResult:
+    ) -> EmbeddingResponse:
         """Generate embeddings for images using Cohere API."""
         client = self._get_client()
         input_type = "image"
@@ -116,21 +120,25 @@ class CohereProvider(EmbeddingProvider):
                     embedding_types=["float"],
                 )
 
-                all_embeddings.extend(response.embeddings.float_)
+                all_embeddings.extend(np.array(response.embeddings.float_))
                 all_b64_images.extend(b64_images)
 
-            return EmbeddingResult(
-                embeddings=np.array(all_embeddings),
+            return EmbeddingResponse(
                 model_name=self.model_name,
                 model_provider=self.provider_name,
                 input_type=input_type,
-                source_images_b64=all_b64_images,
+                objects=[
+                    EmbeddingObject(
+                        embedding=all_embeddings[i],
+                        source_b64=all_b64_images[i]
+                    ) for i in range(len(all_embeddings))
+                ]
             )
 
         except Exception as e:
             raise EmbeddingError(f"Failed to embed image with Cohere: {e}") from e
 
-    def embed_pdf(self, pdf_path: Path) -> EmbeddingResult:
+    def embed_pdf(self, pdf_path: Path) -> EmbeddingResponse:
         """Generate embeddings for a PDF file using Cohere API."""
         image_paths = pdf_to_images(pdf_path)
         return self.embed_image(image_paths)

{embedkit-0.1.2 → embedkit-0.1.4}/src/embedkit/providers/colpali.py

@@ -9,7 +9,7 @@ import torch
 from PIL import Image
 
 from ..utils import pdf_to_images, image_to_base64
-from ..base import EmbeddingProvider, EmbeddingError, EmbeddingResult
+from ..base import EmbeddingProvider, EmbeddingError, EmbeddingResponse, EmbeddingObject
 
 logger = logging.getLogger(__name__)
 
@@ -64,7 +64,7 @@ class ColPaliProvider(EmbeddingProvider):
             except Exception as e:
                 raise EmbeddingError(f"Failed to load model: {e}") from e
 
-    def embed_text(self, texts: Union[str, List[str]]) -> np.ndarray:
+    def embed_text(self, texts: Union[str, List[str]]) -> EmbeddingResponse:
         """Generate embeddings for text inputs."""
         self._load_model()
 
@@ -86,11 +86,15 @@ class ColPaliProvider(EmbeddingProvider):
             # Concatenate all batch embeddings
             final_embeddings = np.concatenate(all_embeddings, axis=0)
 
-            return EmbeddingResult(
-                embeddings=final_embeddings,
+            return EmbeddingResponse(
                 model_name=self.model_name,
                 model_provider=self.provider_name,
                 input_type="text",
+                objects=[
+                    EmbeddingObject(
+                        embedding=e,
+                    ) for e in final_embeddings
+                ]
             )
 
         except Exception as e:
@@ -98,7 +102,7 @@ class ColPaliProvider(EmbeddingProvider):
 
     def embed_image(
         self, images: Union[Path, str, List[Union[Path, str]]]
-    ) -> np.ndarray:
+    ) -> EmbeddingResponse:
         """Generate embeddings for images."""
         self._load_model()
 
@@ -135,18 +139,22 @@ class ColPaliProvider(EmbeddingProvider):
             # Concatenate all batch embeddings
             final_embeddings = np.concatenate(all_embeddings, axis=0)
 
-            return EmbeddingResult(
-                embeddings=final_embeddings,
+            return EmbeddingResponse(
                 model_name=self.model_name,
                 model_provider=self.provider_name,
                 input_type="image",
-                source_images_b64=all_b64_images,
+                objects=[
+                    EmbeddingObject(
+                        embedding=final_embeddings[i],
+                        source_b64=all_b64_images[i]
+                    ) for i in range(len(final_embeddings))
+                ]
             )
 
         except Exception as e:
             raise EmbeddingError(f"Failed to embed images: {e}") from e
 
-    def embed_pdf(self, pdf_path: Path) -> EmbeddingResult:
+    def embed_pdf(self, pdf_path: Path) -> EmbeddingResponse:
         """Generate embeddings for a PDF file using ColPali API."""
         images = pdf_to_images(pdf_path)
         return self.embed_image(images)

{embedkit-0.1.2 → embedkit-0.1.4}/tests/test_embedkit.py

@@ -1,6 +1,7 @@
 # tests/test_embedkit.py
 import os
 import pytest
+import numpy as np
 from pathlib import Path
 from embedkit import EmbedKit
 from embedkit.models import Model
@@ -57,10 +58,13 @@ def cohere_kit_search_document():
 def test_cohere_text_embedding(request, cohere_kit_fixture):
     """Test text embedding with Cohere models."""
     kit = request.getfixturevalue(cohere_kit_fixture)
-    embeddings = kit.embed_text("Hello world")
+    result = kit.embed_text("Hello world")
 
-    assert embeddings.shape[0] == 1
-    assert len(embeddings.shape) == 2
+    assert len(result.objects) == 1
+    assert len(result.objects[0].embedding.shape) == 1
+    assert result.objects[0].source_b64 is None
+    assert result.model_provider == "Cohere"
+    assert result.input_type in ["search_query", "search_document"]
 
 
 @pytest.mark.parametrize(
@@ -76,10 +80,14 @@ def test_cohere_search_document_file_embedding(
     """Test file embedding with Cohere search document model."""
     file_path = request.getfixturevalue(file_fixture)
     embed_func = getattr(cohere_kit_search_document, embed_method)
-    embeddings = embed_func(file_path)
+    result = embed_func(file_path)
 
-    assert embeddings.shape[0] == 1
-    assert len(embeddings.shape) == 2
+    assert len(result.objects) == 1
+    assert len(result.objects[0].embedding.shape) == 1
+    assert result.model_provider == "Cohere"
+    assert result.input_type == "image"
+    if hasattr(result.objects[0], "source_b64"):
+        assert result.objects[0].source_b64 is not None
 
 
 def test_cohere_invalid_model():
@@ -107,28 +115,34 @@ def test_cohere_missing_api_key():
 def test_colpali_text_embedding():
     """Test text embedding with Colpali model."""
     kit = EmbedKit.colpali(model=Model.ColPali.V1_3)
-    embeddings = kit.embed_text("Hello world")
+    result = kit.embed_text("Hello world")
 
-    assert embeddings.shape[0] == 1
-    assert len(embeddings.shape) == 3
+    assert len(result.objects) == 1
+    assert len(result.objects[0].embedding.shape) == 2
+    assert result.objects[0].source_b64 is None
+    assert result.model_provider == "ColPali"
+    assert result.input_type == "text"
 
 
 @pytest.mark.parametrize(
-    "embed_method,file_fixture,expected_dims",
+    "embed_method,file_fixture",
     [
-        ("embed_image", "sample_image_path", 3),
-        ("embed_pdf", "sample_pdf_path", 3),
+        ("embed_image", "sample_image_path"),
+        ("embed_pdf", "sample_pdf_path"),
     ],
 )
-def test_colpali_file_embedding(request, embed_method, file_fixture, expected_dims):
+def test_colpali_file_embedding(request, embed_method, file_fixture):
     """Test file embedding with Colpali model."""
     kit = EmbedKit.colpali(model=Model.ColPali.V1_3)
     file_path = request.getfixturevalue(file_fixture)
     embed_func = getattr(kit, embed_method)
-    embeddings = embed_func(file_path)
+    result = embed_func(file_path)
 
-    assert embeddings.shape[0] == 1
-    assert len(embeddings.shape) == expected_dims
+    assert len(result.objects) == 1
+    assert len(result.objects[0].embedding.shape) == 2
+    assert isinstance(result.objects[0].source_b64, str)
+    assert result.model_provider == "ColPali"
+    assert result.input_type == "image"
 
 
 def test_colpali_invalid_model():

embedkit-0.1.2/PKG-INFO

@@ -1,59 +0,0 @@
-Metadata-Version: 2.4
-Name: embedkit
-Version: 0.1.2
-Summary: A simple toolkit for generating vector embeddings across multiple providers and models
-Author-email: JP Hwang <me@jphwang.com>
-License: MIT
-License-File: LICENSE
-Keywords: ai,cohere,colpali,embeddings,machine-learning,vector
-Classifier: Development Status :: 4 - Beta
-Classifier: Intended Audience :: Developers
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
-Classifier: Topic :: Software Development :: Libraries :: Python Modules
-Requires-Python: >=3.10
-Requires-Dist: accelerate>=1.7.0
-Requires-Dist: cohere>=5.15.0
-Requires-Dist: colpali-engine<0.4.0,>=0.3.0
-Requires-Dist: pdf2image>=1.17.0
-Requires-Dist: pillow>=11.2.1
-Requires-Dist: torch<=2.5
-Requires-Dist: transformers
-Description-Content-Type: text/markdown
-
-# EmbedKit
-
-A Python library for generating embeddings from text, images, and PDFs using various models (e.g. from Cohere, ColPali).
-
-## Usage
-
-See [main.py](main.py) for examples.
-
-```python
-from embedkit import EmbedKit
-from embedkit.models import Model
-
-# Instantiate a kit
-# Using ColPali
-kit = EmbedKit.colpali(model=Model.ColPali.V1_3)
-
-# Using Cohere
-kit = EmbedKit.cohere(
-    model=Model.Cohere.EMBED_V4_0,
-    api_key="your_api_key",
-    text_input_type=CohereInputType.SEARCH_DOCUMENT,
-)
-
-# Then - the embedding API is consistent
-embeddings = kit.embed_text("Hello world") or kit.embed_text(["Hello world", "Hello world"])
-embeddings = kit.embed_image("path/to/image.png") or kit.embed_image(["path/to/image1.png", "path/to/image2.png"])
-embeddings = kit.embed_pdf("path/to/pdf.pdf")  # Single PDF only
-```
-
-## License
-
-MIT

embedkit-0.1.2/README.md

@@ -1,32 +0,0 @@
-# EmbedKit
-
-A Python library for generating embeddings from text, images, and PDFs using various models (e.g. from Cohere, ColPali).
-
-## Usage
-
-See [main.py](main.py) for examples.
-
-```python
-from embedkit import EmbedKit
-from embedkit.models import Model
-
-# Instantiate a kit
-# Using ColPali
-kit = EmbedKit.colpali(model=Model.ColPali.V1_3)
-
-# Using Cohere
-kit = EmbedKit.cohere(
-    model=Model.Cohere.EMBED_V4_0,
-    api_key="your_api_key",
-    text_input_type=CohereInputType.SEARCH_DOCUMENT,
-)
-
-# Then - the embedding API is consistent
-embeddings = kit.embed_text("Hello world") or kit.embed_text(["Hello world", "Hello world"])
-embeddings = kit.embed_image("path/to/image.png") or kit.embed_image(["path/to/image1.png", "path/to/image2.png"])
-embeddings = kit.embed_pdf("path/to/pdf.pdf")  # Single PDF only
-```
-
-## License
-
-MIT