Merge pull request #46 from pinecone-io/seperate-openai-extra

add openai extra

README.md

@@ -27,6 +27,11 @@ If you wish to use `SentenceTransformerEncoder` dense encoder, you will need to
 pip install pinecone-text[dense]
 ```
 
+If you wish to use `OpenAIEncoder` dense encoder, you will need to install the `openai` extra:
+```bash
+pip install pinecone-text[openai]
+```
+
 ## Sparse Encoding
 
 To convert your own text corpus to sparse vectors, you can either use [BM25](https://www.pinecone.io/learn/semantic-search/#bm25) or [SPLADE](https://www.pinecone.io/learn/splade/).
@@ -41,7 +46,7 @@ To encode your documents and queries using BM25 as vector for dot product search
 > When conducting a search, you may come across queries that contain terms not found in the training corpus but are present in the database. To address this scenario, BM25Encoder uses a default document frequency value of 1 when encoding such terms. 
 #### Usage
 
-For an end-to-end example, you can refer to our Quora dataset generation with BM25 [notebook](https://github.com/pinecone-io/examples/blob/master/pinecone/sparse/bm25/bm25-vector-generation.ipynb).
+For an end-to-end example, you can refer to our Quora dataset generation with BM25 [notebook](https://colab.research.google.com/github/pinecone-io/examples/blob/master/learn/search/semantic-search/sparse/bm25/bm25-vector-generation.ipynb).
 
 ```python
 from pinecone_text.sparse import BM25Encoder
@@ -91,7 +96,7 @@ Currently the `SpladeEncoder` class supprts only the [naver/splade-cocondenser-e
 
 #### Usage
 
-For an end-to-end example, you can refer to our Quora dataset generation with SPLADE [notebook](https://github.com/pinecone-io/examples/blob/master/pinecone/sparse/splade/splade-vector-generation.ipynb).
+For an end-to-end example, you can refer to our Quora dataset generation with SPLADE [notebook](https://colab.research.google.com/github/pinecone-io/examples/blob/master/learn/search/semantic-search/sparse/splade/splade-vector-generation.ipynb).
 
 ```python
 from pinecone_text.sparse import SpladeEncoder
@@ -138,8 +143,9 @@ encoder.encode_queries(["Who jumped over the lazy dog?"])
 
 ### OpenAI models
 
-When using the `OpenAIEncoder`, you need to provide an API key for the OpenAI API, and store it in the `OPENAI_API_KEY` environment variable.
+When using the `OpenAIEncoder`, you need to provide an API key for the OpenAI API, and store it in the `OPENAI_API_KEY` environment variable before you import the encoder.
 
+By default the encoder will use `text-embedding-ada-002` as recommended by OpenAI. You can also specify a different model name using the `model_name` parameter.
 #### Usage
 ```python
 from pinecone_text.dense.openai_encoder import OpenAIEncoder

pinecone_text/dense/__init__.py

@@ -1,6 +0,0 @@
-"""
-Sentance Transformers are a class of models that take a sentence as input and output a vector representation of the sentence.
-These models are useful for tasks such as semantic search, clustering, and classification. The sentence transformer models are
-the work of the research team led by Nils Reimers at the University of Stuttgart. For more information, see the [Sentence Transformers paper](https://arxiv.org/abs/1908.10084).
-
-"""

pinecone_text/dense/openai_encoder.py

@@ -1,7 +1,14 @@
-import openai
 from typing import Union, List
 from pinecone_text.dense.base_dense_ecoder import BaseDenseEncoder
 
+try:
+    import openai
+except (OSError, ImportError, ModuleNotFoundError) as e:
+    raise ImportError(
+        "Failed to import openai. Make sure you install openai extra dependencies by running: "
+        "`pip install pinecone-text[openai]"
+    ) from e
+
 
 class OpenAIEncoder(BaseDenseEncoder):
     """

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pinecone-text"
-version = "0.5.3"
+version = "0.5.4"
 description = "Text utilities library by Pinecone.io"
 authors = ["Pinecone.io"]
 readme = "README.md"
@@ -19,6 +19,7 @@ openai =  { version = "^0.27.3", optional = true }
 [tool.poetry.extras]
 splade = ["torch", "transformers", "sentence-transformers"]
 dense = ["torch", "transformers", "sentence-transformers", "openai"]
+openai = ["openai"]
 
 [tool.poetry.group.dev]
 optional = true