[python] Bump version: 0.3.1 → 0.3.2

Closes #594

The embedding functions are pydantic models so multiple instances with
the same parameters are considered ==, which means that if you have
multiple embedding columns it's possible for the embeddings to get
overwritten. Instead we use `is` instead of == to avoid this problem.

testing: modified unit test to include this case

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.2.6
+current_version = 0.3.3
 commit = True
 message = Bump version: {current_version} → {new_version}
 tag = True

Cargo.toml

@@ -5,21 +5,23 @@ exclude = ["python"]
 resolver = "2"
 
 [workspace.dependencies]
-lance = { "version" = "=0.7.5", "features" = ["dynamodb"] }
+lance = { "version" = "=0.8.7", "features" = ["dynamodb"] }
-lance-linalg = { "version" = "=0.7.5" }
+lance-linalg = { "version" = "=0.8.7" }
+lance-testing = { "version" = "=0.8.7" }
 # Note that this one does not include pyarrow
-arrow = { version = "43.0.0", optional = false }
+arrow = { version = "47.0.0", optional = false }
-arrow-array = "43.0"
+arrow-array = "47.0"
-arrow-data = "43.0"
+arrow-data = "47.0"
-arrow-ipc = "43.0"
+arrow-ipc = "47.0"
-arrow-ord = "43.0"
+arrow-ord = "47.0"
-arrow-schema = "43.0"
+arrow-schema = "47.0"
-arrow-arith = "43.0"
+arrow-arith = "47.0"
-arrow-cast = "43.0"
+arrow-cast = "47.0"
-half = { "version" = "=2.2.1", default-features = false, features = [
+chrono = "0.4.23"
+half = { "version" = "=2.3.1", default-features = false, features = [
     "num-traits"
 ] }
 log = "0.4"
-object_store = "0.6.1"
+object_store = "0.7.1"
 snafu = "0.7.4"
 url = "2"

README.md

@@ -33,6 +33,8 @@ The key features of LanceDB include:
 
 * Zero-copy, automatic versioning, manage versions of your data without needing extra infrastructure.
 
+* GPU support in building vector index(*).
+
 * Ecosystem integrations with [LangChain 🦜️🔗](https://python.langchain.com/en/latest/modules/indexes/vectorstores/examples/lanecdb.html), [LlamaIndex 🦙](https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/LanceDBIndexDemo.html), Apache-Arrow, Pandas, Polars, DuckDB and more on the way.
 
 LanceDB's core is written in Rust 🦀 and is built using <a href="https://github.com/lancedb/lance">Lance</a>, an open-source columnar format designed for performant ML workloads.
@@ -52,8 +54,7 @@ const table = await db.createTable('vectors',
       [{ id: 1, vector: [0.1, 0.2], item: "foo", price: 10 },
        { id: 2, vector: [1.1, 1.2], item: "bar", price: 50 }])
 
-const query = table.search([0.1, 0.3]);
+const query = table.search([0.1, 0.3]).limit(2);
-query.limit = 20;
 const results = await query.execute();
 ```
 
@@ -70,7 +71,7 @@ db = lancedb.connect(uri)
 table = db.create_table("my_table",
                          data=[{"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
                                {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}])
-result = table.search([100, 100]).limit(2).to_df()
+result = table.search([100, 100]).limit(2).to_pandas()
 ```
 
 ## Blogs, Tutorials & Videos

docs/README.md

@@ -0,0 +1,26 @@
+# LanceDB Documentation
+
+LanceDB docs are deployed to https://lancedb.github.io/lancedb/.
+
+Docs is built and deployed automatically by [Github Actions](.github/workflows/docs.yml)
+whenever a commit is pushed to the `main` branch. So it is possible for the docs to show
+unreleased features.
+
+## Building the docs
+
+### Setup
+1. Install LanceDB. From LanceDB repo root: `pip install -e python`
+2. Install dependencies. From LanceDB repo root: `pip install -r docs/requirements.txt`
+3. Make sure you have node and npm setup
+4. Make sure protobuf and libssl are installed
+
+### Building node module and create markdown files
+
+See [Javascript docs README](docs/src/javascript/README.md)
+
+### Build docs
+From LanceDB repo root:
+
+Run: `PYTHONPATH=. mkdocs build -f docs/mkdocs.yml`
+
+If successful, you should see a `docs/site` directory that you can verify locally.

docs/mkdocs.yml

@@ -21,6 +21,7 @@ theme:
     - navigation.tracking
     - navigation.instant
     - navigation.indexes
+    - navigation.expand
   icon:
     repo: fontawesome/brands/github
   custom_dir: overrides
@@ -36,7 +37,7 @@ plugins:
           docstring_style: numpy
         rendering:
           heading_level: 4
-          show_source: false
+          show_source: true
           show_symbol_type_in_heading: true
           show_signature_annotations: true
           show_root_heading: true
@@ -68,11 +69,18 @@ nav:
   - 🏢 Home: index.md
   - 💡 Basics: basic.md
   - 📚 Guides:
-    - Tables: guides/tables.md
+    - Create Ingest Update Delete: guides/tables.md
     - Vector Search: search.md
     - SQL filters: sql.md
     - Indexing: ann_indexes.md
-  - 🧬 Embeddings: embedding.md
+    - Versioning & Reproducibility: notebooks/reproducibility.ipynb
+  - 🧬 Embeddings:
+    - embeddings/index.md
+    - Ingest Embedding Functions: embeddings/embedding_functions.md
+    - Available Functions: embeddings/default_embedding_functions.md
+    - Create Custom Embedding Functions: embeddings/api.md
+    - Example - Multi-lingual semantic search: notebooks/multi_lingual_example.ipynb
+    - Example - MultiModal CLIP Embeddings: notebooks/DisappearingEmbeddingFunction.ipynb
   - 🔍 Python full-text search: fts.md
   - 🔌 Integrations:
     - integrations/index.md
@@ -96,13 +104,22 @@ nav:
     - Serverless Website Chatbot: examples/serverless_website_chatbot.md
     - YouTube Transcript Search: examples/youtube_transcript_bot_with_nodejs.md
     - TransformersJS Embedding Search: examples/transformerjs_embedding_search_nodejs.md
+  - ⚙️ CLI & Config: cli_config.md
+
 - Basics: basic.md
 - Guides:
-  - Tables: guides/tables.md
+  - Create Ingest Update Delete: guides/tables.md
   - Vector Search: search.md
   - SQL filters: sql.md
   - Indexing: ann_indexes.md
-- Embeddings: embedding.md
+  - Versioning & Reproducibility: notebooks/reproducibility.ipynb
+- Embeddings:
+  - embeddings/index.md
+  - Ingest Embedding Functions: embeddings/embedding_functions.md
+  - Available Functions: embeddings/default_embedding_functions.md
+  - Create Custom Embedding Functions: embeddings/api.md
+  - Example - Multi-lingual semantic search: notebooks/multi_lingual_example.ipynb
+  - Example - MultiModal CLIP Embeddings: notebooks/DisappearingEmbeddingFunction.ipynb
 - Python full-text search: fts.md
 - Integrations:
   - integrations/index.md
@@ -133,6 +150,8 @@ nav:
 
 extra_css:
   - styles/global.css
+extra_javascript:
+  - scripts/posthog.js
 
 extra:
   analytics:

docs/src/ann_indexes.md

@@ -6,7 +6,7 @@ LanceDB provides many parameters to fine-tune the index's size, the speed of que
 
 Currently, LanceDB does *not* automatically create the ANN index.
 LanceDB has optimized code for KNN as well. For many use-cases, datasets under 100K vectors won't require index creation at all.
-If you can live with <100ms latency, skipping index creation is a simpler workflow while guaranteeing 100% recall.
+If you can live with < 100ms latency, skipping index creation is a simpler workflow while guaranteeing 100% recall.
 
 In the future we will look to automatically create and configure the ANN index.
 
@@ -68,6 +68,12 @@ a single PQ code.
   <figcaption>IVF_PQ index with <code>num_partitions=2, num_sub_vectors=4</code></figcaption>
 </figure>
 
+### Use GPU to build vector index
+
+Lance Python SDK has experimental GPU support for creating IVF index.
+You can specify the GPU device to train IVF partitions via
+
+- **accelerator**: Specify to `"cuda"`` to enable GPU training.
 
 ## Querying an ANN Index
 
@@ -91,7 +97,7 @@ There are a couple of parameters that can be used to fine-tune the search:
          .limit(2) \
          .nprobes(20) \
          .refine_factor(10) \
-         .to_df()
+         .to_pandas()
      ```
      ```
                                               vector       item       _distance
@@ -118,7 +124,7 @@ You can further filter the elements returned by a search using a where clause.
 
 === "Python"
      ```python
-     tbl.search(np.random.random((1536))).where("item != 'item 1141'").to_df()
+     tbl.search(np.random.random((1536))).where("item != 'item 1141'").to_pandas()
      ```
 
 === "Javascript"
@@ -135,7 +141,7 @@ You can select the columns returned by the query using a select clause.
 
 === "Python"
      ```python
-     tbl.search(np.random.random((1536))).select(["vector"]).to_df()
+     tbl.search(np.random.random((1536))).select(["vector"]).to_pandas()
      ```
      ```
         vector                                             _distance
@@ -154,28 +160,28 @@ You can select the columns returned by the query using a select clause.
 
 ## FAQ
 
-### When is it necessary to create an ANN vector index.
+### When is it necessary to create an ANN vector index?
 
-`LanceDB` has manually tuned SIMD code for computing vector distances.
+`LanceDB` has manually-tuned SIMD code for computing vector distances.
-In our benchmarks, computing 100K pairs of 1K dimension vectors only take less than 20ms.
+In our benchmarks, computing 100K pairs of 1K dimension vectors takes **less than 20ms**.
-For small dataset (<100K rows) or the applications which can accept 100ms latency, vector indices are usually not necessary.
+For small datasets (< 100K rows) or applications that can accept 100ms latency, vector indices are usually not necessary.
 
 For large-scale or higher dimension vectors, it is beneficial to create vector index.
 
-### How big is my index, and how many memory will it take.
+### How big is my index, and how many memory will it take?
 
-In LanceDB, all vector indices are disk-based, meaning that when responding to a vector query, only the relevant pages from the index file are loaded from disk and cached in memory. Additionally, each sub-vector is usually encoded into 1 byte PQ code.
+In LanceDB, all vector indices are **disk-based**, meaning that when responding to a vector query, only the relevant pages from the index file are loaded from disk and cached in memory. Additionally, each sub-vector is usually encoded into 1 byte PQ code.
 
 For example, with a 1024-dimension dataset, if we choose `num_sub_vectors=64`, each sub-vector has `1024 / 64 = 16` float32 numbers.
 Product quantization can lead to approximately `16 * sizeof(float32) / 1 = 64` times of space reduction.
 
-### How to choose `num_partitions` and `num_sub_vectors` for `IVF_PQ` index.
+### How to choose `num_partitions` and `num_sub_vectors` for `IVF_PQ` index?
 
 `num_partitions` is used to decide how many partitions the first level `IVF` index uses.
 Higher number of partitions could lead to more efficient I/O during queries and better accuracy, but it takes much more time to train.
 On `SIFT-1M` dataset, our benchmark shows that keeping each partition 1K-4K rows lead to a good latency / recall.
 
-`num_sub_vectors` decides how many Product Quantization code to generate on each vector. Because
+`num_sub_vectors` specifies how many Product Quantization (PQ) short codes to generate on each vector. Because
-Product Quantization is a lossy compression of the original vector, the more `num_sub_vectors` usually results to
+PQ is a lossy compression of the original vector, a higher `num_sub_vectors` usually results in
-less space distortion, and thus yield better accuracy. However, similarly, more `num_sub_vectors` causes heavier I/O and
+less space distortion, and thus yields better accuracy. However, a higher `num_sub_vectors` also causes heavier I/O and
-more PQ computation, thus, higher latency. `dimension / num_sub_vectors` should be aligned with 8 for better SIMD efficiency.
+more PQ computation, and thus, higher latency. `dimension / num_sub_vectors` should be a multiple of 8 for optimum SIMD efficiency.

docs/src/basic.md

@@ -123,9 +123,15 @@ After a table has been created, you can always add more data to it using
 
 === "Python"
       ```python
-      df = pd.DataFrame([{"vector": [1.3, 1.4], "item": "fizz", "price": 100.0},
+
-                        {"vector": [9.5, 56.2], "item": "buzz", "price": 200.0}])
+      # Option 1: Add a list of dicts to a table
-      tbl.add(df)
+      data = [{"vector": [1.3, 1.4], "item": "fizz", "price": 100.0},
+            {"vector": [9.5, 56.2], "item": "buzz", "price": 200.0}]
+      tbl.add(data)
+
+      # Option 2: Add a pandas DataFrame to a table
+      df = pd.DataFrame(data)
+      tbl.add(data)
       ```
 
 === "Javascript"
@@ -140,7 +146,7 @@ Once you've embedded the query, you can find its nearest neighbors using the fol
 
 === "Python"
       ```python
-      tbl.search([100, 100]).limit(2).to_df()
+      tbl.search([100, 100]).limit(2).to_pandas()
       ```
 
       This returns a pandas DataFrame with the results.

docs/src/cli_config.md

@@ -0,0 +1,37 @@
+
+## LanceDB CLI
+Once lanceDB is installed, you can access the CLI using `lancedb` command on the console
+```
+lancedb
+```
+This lists out all the various command-line options available. You can get the usage or help for a particular command
+```
+lancedb {command} --help
+```
+
+## LanceDB config
+LanceDB uses a global config file to store certain settings. These settings are configurable using the lanceDB cli.
+To view your config settings, you can use:
+```
+lancedb config
+```
+These config parameters can be tuned using the cli.
+```
+lancedb {config_name} --{argument}
+```
+
+## LanceDB Opt-in Diagnostics
+When enabled, LanceDB will send anonymous events to help us improve LanceDB. These diagnostics are used only for error reporting and no data is collected. Error & stats allow us to automate certain aspects of bug reporting, prioritization of fixes and feature requests.
+These diagnostics are opt-in and can be enabled or disabled using the `lancedb diagnostics` command. These are enabled by default.
+Get usage help.
+```
+lancedb diagnostics --help
+```
+Disable diagnostics
+```
+lancedb diagnostics --disabled
+```
+Enable diagnostics
+```
+lancedb diagnostics --enabled
+```

docs/src/embeddings/api.md

@@ -0,0 +1,213 @@
+To use your own custom embedding function, you need to follow these 2 simple steps.
+1. Create your embedding function by implementing the `EmbeddingFunction` interface
+2. Register your embedding function in the global `EmbeddingFunctionRegistry`.
+
+Let us see how this looks like in action.
+
+![](../assets/embeddings_api.png)
+
+
+`EmbeddingFunction` & `EmbeddingFunctionRegistry` handle low-level details for serializing schema and model information as metadata. To build a custom embdding function, you don't need to worry about those details and simply focus on setting up the model.
+
+## `TextEmbeddingFunction` Interface
+
+There is another optional layer of abstraction provided in form of `TextEmbeddingFunction`. You can use this if your model isn't multi-modal in nature and only operates on text. In such case both source and vector fields will have the same pathway for vectorization, so you simply just need to setup the model and rest is handled by `TextEmbeddingFunction`. You can read more about the class and its attributes in the class reference.
+
+
+Let's implement `SentenceTransformerEmbeddings` class. All you need to do is implement the `generate_embeddings()` and `ndims` function to handle the input types you expect and register the class in the global `EmbeddingFunctionRegistry`
+
+```python
+from lancedb.embeddings import register
+
+@register("sentence-transformers")
+class SentenceTransformerEmbeddings(TextEmbeddingFunction):
+    name: str = "all-MiniLM-L6-v2"
+    # set more default instance vars like device, etc.
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self._ndims = None
+    
+    def generate_embeddings(self, texts):
+        return self._embedding_model().encode(list(texts), ...).tolist()
+
+    def ndims(self):
+        if self._ndims is None:
+            self._ndims = len(self.generate_embeddings("foo")[0])
+        return self._ndims
+
+    @cached(cache={}) 
+    def _embedding_model(self):
+        return sentence_transformers.SentenceTransformer(name)
+
+```
+
+This is a stripped down version of our implementation of `SentenceTransformerEmbeddings` that removes certain optimizations and defaul settings.
+
+Now you can use this embedding function to create your table schema and that's it! you can then ingest data and run queries without manually vectorizing the inputs.
+
+```python
+from lancedb.pydantic import LanceModel, Vector
+
+registry = EmbeddingFunctionRegistry.get_instance()
+stransformer = registry.get("sentence-transformers").create()
+
+class TextModelSchema(LanceModel):
+    vector: Vector(stransformer.ndims) = stransformer.VectorField()
+    text: str = stransformer.SourceField()
+
+tbl = db.create_table("table", schema=TextModelSchema)
+
+tbl.add(pd.DataFrame({"text": ["halo", "world"]}))
+result = tbl.search("world").limit(5)
+```
+
+NOTE:
+
+You can always implement the `EmbeddingFunction` interface directly if you want or need to, `TextEmbeddingFunction` just makes it much simpler and faster for you to do so, by setting up the boiler plat for text-specific use case
+
+## Multi-modal embedding function example
+You can also use the `EmbeddingFunction` interface to implement more complex workflows such as multi-modal embedding function support. LanceDB implements `OpenClipEmeddingFunction` class that suppports multi-modal seach. Here's the implementation that you can use as a reference to build your own multi-modal embedding functions.
+
+```python
+@register("open-clip")
+class OpenClipEmbeddings(EmbeddingFunction):
+    name: str = "ViT-B-32"
+    pretrained: str = "laion2b_s34b_b79k"
+    device: str = "cpu"
+    batch_size: int = 64
+    normalize: bool = True
+    _model = PrivateAttr()
+    _preprocess = PrivateAttr()
+    _tokenizer = PrivateAttr()
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        open_clip = self.safe_import("open_clip", "open-clip") # EmbeddingFunction util to import external libs and raise if not found
+        model, _, preprocess = open_clip.create_model_and_transforms(
+            self.name, pretrained=self.pretrained
+        )
+        model.to(self.device)
+        self._model, self._preprocess = model, preprocess
+        self._tokenizer = open_clip.get_tokenizer(self.name)
+        self._ndims = None
+
+    def ndims(self):
+        if self._ndims is None:
+            self._ndims = self.generate_text_embeddings("foo").shape[0]
+        return self._ndims
+
+    def compute_query_embeddings(
+        self, query: Union[str, "PIL.Image.Image"], *args, **kwargs
+    ) -> List[np.ndarray]:
+        """
+        Compute the embeddings for a given user query
+
+        Parameters
+        ----------
+        query : Union[str, PIL.Image.Image]
+            The query to embed. A query can be either text or an image.
+        """
+        if isinstance(query, str):
+            return [self.generate_text_embeddings(query)]
+        else:
+            PIL = self.safe_import("PIL", "pillow")
+            if isinstance(query, PIL.Image.Image):
+                return [self.generate_image_embedding(query)]
+            else:
+                raise TypeError("OpenClip supports str or PIL Image as query")
+
+    def generate_text_embeddings(self, text: str) -> np.ndarray:
+        torch = self.safe_import("torch")
+        text = self.sanitize_input(text)
+        text = self._tokenizer(text)
+        text.to(self.device)
+        with torch.no_grad():
+            text_features = self._model.encode_text(text.to(self.device))
+            if self.normalize:
+                text_features /= text_features.norm(dim=-1, keepdim=True)
+            return text_features.cpu().numpy().squeeze()
+
+    def sanitize_input(self, images: IMAGES) -> Union[List[bytes], np.ndarray]:
+        """
+        Sanitize the input to the embedding function.
+        """
+        if isinstance(images, (str, bytes)):
+            images = [images]
+        elif isinstance(images, pa.Array):
+            images = images.to_pylist()
+        elif isinstance(images, pa.ChunkedArray):
+            images = images.combine_chunks().to_pylist()
+        return images
+
+    def compute_source_embeddings(
+        self, images: IMAGES, *args, **kwargs
+    ) -> List[np.array]:
+        """
+        Get the embeddings for the given images
+        """
+        images = self.sanitize_input(images)
+        embeddings = []
+        for i in range(0, len(images), self.batch_size):
+            j = min(i + self.batch_size, len(images))
+            batch = images[i:j]
+            embeddings.extend(self._parallel_get(batch))
+        return embeddings
+
+    def _parallel_get(self, images: Union[List[str], List[bytes]]) -> List[np.ndarray]:
+        """
+        Issue concurrent requests to retrieve the image data
+        """
+        with concurrent.futures.ThreadPoolExecutor() as executor:
+            futures = [
+                executor.submit(self.generate_image_embedding, image)
+                for image in images
+            ]
+            return [future.result() for future in futures]
+
+    def generate_image_embedding(
+        self, image: Union[str, bytes, "PIL.Image.Image"]
+    ) -> np.ndarray:
+        """
+        Generate the embedding for a single image
+
+        Parameters
+        ----------
+        image : Union[str, bytes, PIL.Image.Image]
+            The image to embed. If the image is a str, it is treated as a uri.
+            If the image is bytes, it is treated as the raw image bytes.
+        """
+        torch = self.safe_import("torch")
+        # TODO handle retry and errors for https
+        image = self._to_pil(image)
+        image = self._preprocess(image).unsqueeze(0)
+        with torch.no_grad():
+            return self._encode_and_normalize_image(image)
+
+    def _to_pil(self, image: Union[str, bytes]):
+        PIL = self.safe_import("PIL", "pillow")
+        if isinstance(image, bytes):
+            return PIL.Image.open(io.BytesIO(image))
+        if isinstance(image, PIL.Image.Image):
+            return image
+        elif isinstance(image, str):
+            parsed = urlparse.urlparse(image)
+            # TODO handle drive letter on windows.
+            if parsed.scheme == "file":
+                return PIL.Image.open(parsed.path)
+            elif parsed.scheme == "":
+                return PIL.Image.open(image if os.name == "nt" else parsed.path)
+            elif parsed.scheme.startswith("http"):
+                return PIL.Image.open(io.BytesIO(url_retrieve(image)))
+            else:
+                raise NotImplementedError("Only local and http(s) urls are supported")
+
+    def _encode_and_normalize_image(self, image_tensor: "torch.Tensor"):
+        """
+        encode a single image tensor and optionally normalize the output
+        """
+        image_features = self._model.encode_image(image_tensor)
+        if self.normalize:
+            image_features /= image_features.norm(dim=-1, keepdim=True)
+        return image_features.cpu().numpy().squeeze()
+```

docs/src/embeddings/default_embedding_functions.md

@@ -0,0 +1,156 @@
+There are various Embedding functions available out of the box with lancedb. We're working on supporting other popular embedding APIs.
+
+## Text Embedding Functions
+Here are the text embedding functions registered by default
+
+### Sentence Transformers
+Here are the parameters that you can set when registering a `sentence-transformers` object, and their default values:
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"all-MiniLM-L6-v2"` | The name of the model. |
+| `device` | `str` | `"cpu"` | The device to run the model on. Can be `"cpu"` or `"gpu"`. |
+| `normalize` | `bool` | `True` | Whether to normalize the input text before feeding it to the model. |
+
+
+```python
+db = lancedb.connect("/tmp/db")
+registry = EmbeddingFunctionRegistry.get_instance()
+func = registry.get("sentence-transformers").create(device="cpu")
+
+class Words(LanceModel):
+    text: str = func.SourceField()
+    vector: Vector(func.ndims()) = func.VectorField()
+
+table = db.create_table("words", schema=Words)
+table.add(
+    [
+        {"text": "hello world"}
+        {"text": "goodbye world"}
+    ]
+)
+
+query = "greetings"
+actual = table.search(query).limit(1).to_pydantic(Words)[0]
+print(actual.text)
+```
+
+### OpenAIEmbeddings
+LanceDB has OpenAI embeddings function in the registry by default. It is registered as `openai` and here are the parameters that you can customize when creating the instances
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"text-embedding-ada-002"` | The name of the model. |
+
+
+
+```python
+db = lancedb.connect("/tmp/db")
+registry = EmbeddingFunctionRegistry.get_instance()
+func = registry.get("openai").create()
+
+class Words(LanceModel):
+    text: str = func.SourceField()
+    vector: Vector(func.ndims()) = func.VectorField()
+
+table = db.create_table("words", schema=Words)
+table.add(
+    [
+        {"text": "hello world"}
+        {"text": "goodbye world"}
+    ]
+    )
+
+query = "greetings"
+actual = table.search(query).limit(1).to_pydantic(Words)[0]
+print(actual.text)
+```
+
+## Multi-modal embedding functions
+Multi-modal embedding functions allow you query your table using both images and text.
+
+### OpenClipEmbeddings
+We support CLIP model embeddings using the open souce alternbative, open-clip which support various customizations. It is registered as `open-clip` and supports following customizations.
+
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"ViT-B-32"` | The name of the model. |
+| `pretrained` | `str` | `"laion2b_s34b_b79k"` | The name of the pretrained model to load. |
+| `device` | `str` | `"cpu"` | The device to run the model on. Can be `"cpu"` or `"gpu"`. |
+| `batch_size` | `int` | `64` | The number of images to process in a batch. |
+| `normalize` | `bool` | `True` | Whether to normalize the input images before feeding them to the model. |
+
+
+This embedding function supports ingesting images as both bytes and urls. You can query them using both test and other images.
+
+NOTE:
+LanceDB supports ingesting images directly from accessible links.
+
+
+```python
+
+db = lancedb.connect(tmp_path)
+registry = EmbeddingFunctionRegistry.get_instance()
+func = registry.get("open-clip").create()
+
+class Images(LanceModel):
+    label: str
+    image_uri: str = func.SourceField() # image uri as the source
+    image_bytes: bytes = func.SourceField() # image bytes as the source
+    vector: Vector(func.ndims()) = func.VectorField() # vector column 
+    vec_from_bytes: Vector(func.ndims()) = func.VectorField() # Another vector column 
+
+table = db.create_table("images", schema=Images)
+labels = ["cat", "cat", "dog", "dog", "horse", "horse"]
+uris = [
+    "http://farm1.staticflickr.com/53/167798175_7c7845bbbd_z.jpg",
+    "http://farm1.staticflickr.com/134/332220238_da527d8140_z.jpg",
+    "http://farm9.staticflickr.com/8387/8602747737_2e5c2a45d4_z.jpg",
+    "http://farm5.staticflickr.com/4092/5017326486_1f46057f5f_z.jpg",
+    "http://farm9.staticflickr.com/8216/8434969557_d37882c42d_z.jpg",
+    "http://farm6.staticflickr.com/5142/5835678453_4f3a4edb45_z.jpg",
+]
+# get each uri as bytes
+image_bytes = [requests.get(uri).content for uri in uris]
+table.add(
+    [{"label": labels, "image_uri": uris, "image_bytes": image_bytes}]
+)
+```
+Now we can search using text from both the default vector column and the custom vector column
+```python
+
+# text search
+actual = table.search("man's best friend").limit(1).to_pydantic(Images)[0]
+print(actual.label) # prints "dog"
+
+frombytes = (
+    table.search("man's best friend", vector_column_name="vec_from_bytes")
+    .limit(1)
+    .to_pydantic(Images)[0]
+)
+print(frombytes.label)
+
+```
+
+Because we're using a multi-modal embedding function, we can also search using images
+
+```python
+# image search
+query_image_uri = "http://farm1.staticflickr.com/200/467715466_ed4a31801f_z.jpg"
+image_bytes = requests.get(query_image_uri).content
+query_image = Image.open(io.BytesIO(image_bytes))
+actual = table.search(query_image).limit(1).to_pydantic(Images)[0]
+print(actual.label == "dog")
+
+# image search using a custom vector column
+other = (
+    table.search(query_image, vector_column_name="vec_from_bytes")
+    .limit(1)
+    .to_pydantic(Images)[0]
+)
+print(actual.label)
+
+```
+
+If you have any questions about the embeddings API, supported models, or see a relevant model missing, please raise an issue.

docs/src/embeddings/embedding_functions.md

@@ -0,0 +1,82 @@
+Representing multi-modal data as vector embeddings is becoming a standard practice. Embedding functions themselves be thought of as a part of the processing pipeline that each request(input) has to be passed through. After initial setup these components are not expected to change for a particular project. 
+
+This is main motivation behind our new embedding functions API, that allow you simply set it up once and the table remembers it, effectively making the **embedding functions disappear in the background** so you don't have to worry about modelling and simply focus on the DB aspects of VectorDB.
+
+
+You can simply follow these steps and forget about the details of your embedding functions as long as you don't intend to change it.
+
+### Step 1 - Define the embedding function
+We have some pre-defined embedding functions in the global registry with more coming soon. Here's let's an implementation of CLIP as example.
+```
+registry = EmbeddingFunctionRegistry.get_instance()
+clip = registry.get("open-clip").create()
+
+```
+You can also define your own embedding function by implementing the `EmbeddingFunction` abstract base interface. It subclasses PyDantic Model which can be utilized to write complex schemas simply as we'll see next!
+
+### Step 2 - Define the Data Model or Schema
+Our embedding function from the previous section abstracts away all the details about the models and dimensions required to define the schema. You can simply set a feild as **source** or **vector** column. Here's how
+
+```python
+class Pets(LanceModel):
+    vector: Vector(clip.ndims) = clip.VectorField()
+    image_uri: str = clip.SourceField()
+
+```
+`VectorField` tells LanceDB to use the clip embedding function to generate query embeddings for `vector` column & `SourceField` tells that when adding data, automatically use the embedding function to encode `image_uri`.
+
+
+### Step 3 - Create LanceDB Table
+Now that we have chosen/defined our embedding function and the schema, we can create the table
+
+```python
+db = lancedb.connect("~/lancedb")
+table = db.create_table("pets", schema=Pets)
+
+```
+That's it! We have ingested all the information needed to embed source and query inputs. We can now forget about the model and dimension details and start to build or VectorDB
+
+### Step 4 - Ingest lots of data and run vector search!
+Now you can just add the data and it'll be vectorized automatically
+
+```python
+table.add([{"image_uri": u} for u in uris])
+```
+
+Our OpenCLIP query embedding function support querying via both text and images. 
+
+```python
+result = table.search("dog")
+```
+
+Let's query an image
+
+```python
+p = Path("path/to/images/samoyed_100.jpg")
+query_image = Image.open(p)
+table.search(query_image)
+
+```
+
+### A little fun with PyDantic
+LanceDB is integrated with PyDantic. Infact we've used the integration in the above example to define the schema. It is also being used behing the scene by the embdding function API to ingest useful information as table metadata.
+You can also use it for adding utility operations in the schema. For example, in our multi-modal example, you can search images using text or another image. Let us define a utility function to plot the image.
+```python
+class Pets(LanceModel):
+    vector: Vector(clip.ndims) = clip.VectorField()
+    image_uri: str = clip.SourceField()
+
+    @property
+    def image(self):
+        return Image.open(self.image_uri)
+```
+Now, you can covert your search results to pydantic model and use this property.
+
+```python
+rs = table.search(query_image).limit(3).to_pydantic(Pets)
+rs[2].image
+```
+
+![](../assets/dog_clip_output.png)
+
+Now that you've the basic idea about LanceDB embedding function, let us now dive deeper into the API that you can use to implement your own embedding functions!

docs/src/embeddings/index.md

@@ -1,13 +1,20 @@
-# Embedding Functions
+# Embedding
 
-Embeddings are high dimensional floating-point vector representations of your data or query.
+Embeddings are high dimensional floating-point vector representations of your data or query. Anything can be embedded using some embedding model or function. Position of embedding in a high dimensional vector space has semantic significance to a degree that depends on the type of modal and training. These embeddings when projected in a 2-D space generally group similar entities close-by forming groups.
-Anything can be embedded using some embedding model or function.
-For a given embedding function, the output will always have the same number of dimensions.
 
-## Creating an embedding function
+![](../assets/embedding_intro.png)
 
-Any function that takes as input a batch (list) of data and outputs a batch (list) of embeddings
+# Creating an embedding function
-can be used by LanceDB as an embedding function. The input and output batch sizes should be the same.
+
+LanceDB supports 2 major ways of vectorizing your data, explicit and implicit.
+
+1. By manually embedding the data before ingesting in the table
+2. By automatically embedding the data and query as they come, by ingesting embedding function information in the table itself! Covered in [Next Section](embedding_functions.md)
+
+Whatever workflow you prefer, we have the tools to support you.
+## Explicit Vectorization
+
+In this workflow, you can create your embedding function and vectorize your data using lancedb's `with_embedding` function. Let's look at some examples.
 
 ### HuggingFace example
 
@@ -118,7 +125,7 @@ belong in the same latent space and your results will be nonsensical.
      ```python
      query = "What's the best pizza topping?"
      query_vector = embed_func([query])[0]
-     tbl.search(query_vector).limit(10).to_df()
+     tbl.search(query_vector).limit(10).to_pandas()
      ```
 
      The above snippet returns a pandas DataFrame with the 10 closest vectors to the query.
@@ -134,9 +141,9 @@ belong in the same latent space and your results will be nonsensical.
      The above snippet returns an array of records with the 10 closest vectors to the query.
 
 
-## Roadmap
+## Implicit vectorization / Ingesting embedding functions
+Representing multi-modal data as vector embeddings is becoming a standard practice. Embedding functions themselves be thought of as a part of the processing pipeline that each request(input) has to be passed through. After initial setup these components are not expected to change for a particular project. 
 
-In the near future, we'll be integrating the embedding functions deeper into LanceDB<br/>.
+This is main motivation behind our new embedding functions API, that allow you simply set it up once and the table remembers it, effectively making the **embedding functions disappear in the background** so you don't have to worry about modelling and simply focus on the DB aspects of VectorDB.
-The goal is that you just have to configure the function once when you create the table,
+
-and then you'll never have to deal with embeddings / vectors after that unless you want to.
+Learn more in the Next Section
-We'll also integrate more popular models and APIs.

docs/src/examples/serverless_lancedb_with_s3_and_lambda.md

@@ -80,14 +80,14 @@ def handler(event, context):
     # Shape of SIFT is (128,1M), d=float32
     query_vector = np.array(event['query_vector'], dtype=np.float32)
 
-    rs = table.search(query_vector).limit(2).to_df()
+    rs = table.search(query_vector).limit(2).to_list()
 
     return {
         "statusCode": status_code,
         "headers": {
             "Content-Type": "application/json"
         },
-        "body": rs.to_json()
+        "body": json.dumps(rs)
     }
 ``` 
 

docs/src/fts.md

@@ -6,17 +6,19 @@ to make this available for JS as well.
 
 ## Installation
 
-To use full text search, you must install optional dependency tantivy-py:
+To use full text search, you must install the dependency `tantivy-py`:
 
-# tantivy 0.19.2
+# tantivy 0.20.1
-pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
+```sh
+pip install tantivy==0.20.1
+```
 
 
 ## Quickstart
 
 Assume:
 1. `table` is a LanceDB Table
-2. `text` is the name of the Table column that we want to index
+2. `text` is the name of the `Table` column that we want to index
 
 For example,
 
@@ -41,7 +43,13 @@ table.create_fts_index("text")
 To search:
 
 ```python
-df = table.search("puppy").limit(10).select(["text"]).to_df()
+table.search("puppy").limit(10).select(["text"]).to_list()
+```
+
+Which returns a list of dictionaries:
+
+```python
+[{'text': 'Frodo was a happy puppy', 'score': 0.6931471824645996}]
 ```
 
 LanceDB automatically looks for an FTS index if the input is str.

docs/src/guides/tables.md

@@ -42,7 +42,7 @@ A Table is a collection of Records in a LanceDB Database. You can follow along o
     import pandas as pd
 
     data = pd.DataFrame({
-        "vector": [[1.1, 1.2], [0.2, 1.8]],
+        "vector": [[1.1, 1.2, 1.3, 1.4], [0.2, 1.8, 0.4, 3.6]],
         "lat": [45.5, 40.1],
         "long": [-122.7, -74.1]
     })
@@ -56,7 +56,7 @@ A Table is a collection of Records in a LanceDB Database. You can follow along o
 
     ```python
     custom_schema = pa.schema([
-    pa.field("vector", pa.list_(pa.float32(), 2)),
+    pa.field("vector", pa.list_(pa.float32(), 4)),
     pa.field("lat", pa.float32()),
     pa.field("long", pa.float32())
     ])
@@ -70,8 +70,8 @@ A Table is a collection of Records in a LanceDB Database. You can follow along o
     ```python
     table = pa.Table.from_arrays(
             [
-                pa.array([[3.1, 4.1], [5.9, 26.5]],
+                pa.array([[3.1, 4.1, 5.1, 6.1], [5.9, 26.5, 4.7, 32.8]],
-                        pa.list_(pa.float32(), 2)),
+                        pa.list_(pa.float32(), 4)),
                 pa.array(["foo", "bar"]),
                 pa.array([10.0, 20.0]),
             ],
@@ -131,8 +131,8 @@ A Table is a collection of Records in a LanceDB Database. You can follow along o
         for i in range(5):
             yield pa.RecordBatch.from_arrays(
                 [
-                    pa.array([[3.1, 4.1], [5.9, 26.5]],
+                    pa.array([[3.1, 4.1, 5.1, 6.1], [5.9, 26.5, 4.7, 32.8]],
-                            pa.list_(pa.float32(), 2)),
+                            pa.list_(pa.float32(), 4)),
                     pa.array(["foo", "bar"]),
                     pa.array([10.0, 20.0]),
                 ],
@@ -140,7 +140,7 @@ A Table is a collection of Records in a LanceDB Database. You can follow along o
             )
 
     schema = pa.schema([
-        pa.field("vector", pa.list_(pa.float32(), 2)),
+        pa.field("vector", pa.list_(pa.float32(), 4)),
         pa.field("item", pa.utf8()),
         pa.field("price", pa.float32()),
     ])
@@ -251,8 +251,9 @@ After a table has been created, you can always add more data to it using
     ### Adding Pandas DataFrame
 
     ```python
-    df = pd.DataFrame([{"vector": [1.3, 1.4], "item": "fizz", "price": 100.0},
+    df = pd.DataFrame({
-                  {"vector": [9.5, 56.2], "item": "buzz", "price": 200.0}])
+        "vector": [[1.3, 1.4], [9.5, 56.2]], "item": ["fizz", "buzz"], "price": [100.0, 200.0]
+    })
     tbl.add(df)
     ```
 
@@ -261,17 +262,12 @@ After a table has been created, you can always add more data to it using
     ### Adding to table using Iterator
 
     ```python
-    import pandas as pd
-
     def make_batches():
         for i in range(5):
-            yield pd.DataFrame(
+            yield [
-                {
+                    {"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
-                    "vector": [[3.1, 4.1], [1, 1]],
+                    {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}
-                    "item": ["foo", "bar"],
+                ]
-                    "price": [10.0, 20.0],
-                })
-
     tbl.add(make_batches())
     ```
 
@@ -306,9 +302,10 @@ Use the `delete()` method on tables to delete rows from a table. To choose which
 
     ```python
     import lancedb
-    import pandas as pd
 
-    data = pd.DataFrame({"x": [1, 2, 3], "vector": [[1, 2], [3, 4], [5, 6]]})
+    data = [{"x": 1, "vector": [1, 2]},
+            {"x": 2, "vector": [3, 4]},
+            {"x": 3, "vector": [5, 6]}]
     db = lancedb.connect("./.lancedb")
     table = db.create_table("my_table", data)
     table.to_pandas()
@@ -364,6 +361,48 @@ Use the `delete()` method on tables to delete rows from a table. To choose which
     await tbl.countRows() // Returns 1
     ```
 
+### Updating a Table [Experimental]
+EXPERIMENTAL: Update rows in the table (not threadsafe).
+
+This can be used to update zero to all rows depending on how many rows match the where clause.
+
+| Parameter | Type | Description |
+|---|---|---|
+| `where` | `str` | The SQL where clause to use when updating rows. For example, `'x = 2'` or `'x IN (1, 2, 3)'`. The filter must not be empty, or it will error. |
+| `values` | `dict` | The values to update. The keys are the column names and the values are the values to set. |
+
+
+=== "Python"
+
+    ```python
+    import lancedb
+    import pandas as pd
+
+    # Create a lancedb connection
+    db = lancedb.connect("./.lancedb")
+
+    # Create a table from a pandas DataFrame
+    data = pd.DataFrame({"x": [1, 2, 3], "vector": [[1, 2], [3, 4], [5, 6]]})
+    table = db.create_table("my_table", data)
+
+    # Update the table where x = 2
+    table.update(where="x = 2", values={"vector": [10, 10]})
+
+    # Get the updated table as a pandas DataFrame
+    df = table.to_pandas()
+
+    # Print the DataFrame
+    print(df)
+    ```
+
+    Output
+    ```shell
+        x  vector
+    0  1  [1.0, 2.0]
+    1  3  [5.0, 6.0]
+    2  2  [10.0, 10.0]
+    ```
+
 ## What's Next?
 
 Learn how to Query your tables and create indices

docs/src/index.md

@@ -36,7 +36,7 @@ LanceDB's core is written in Rust 🦀 and is built using <a href="https://githu
       table = db.create_table("my_table",
                               data=[{"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
                                     {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}])
-      result = table.search([100, 100]).limit(2).to_df()
+      result = table.search([100, 100]).limit(2).to_list()
       ```
 
 === "Javascript"
@@ -67,7 +67,7 @@ LanceDB's core is written in Rust 🦀 and is built using <a href="https://githu
 
 ## Documentation Quick Links
 * [`Basic Operations`](basic.md) - basic functionality of LanceDB.
-* [`Embedding Functions`](embedding.md) - functions for working with embeddings.
+* [`Embedding Functions`](embeddings/index.md) - functions for working with embeddings.
 * [`Indexing`](ann_indexes.md) - create vector indexes to speed up queries.
 * [`Full text search`](fts.md) - [EXPERIMENTAL] full-text search API
 * [`Ecosystem Integrations`](python/integration.md) - integrating LanceDB with python data tooling ecosystem.

docs/src/notebooks/code_qa_bot.ipynb

@@ -144,7 +144,7 @@
    "source": [
     "# Pre-processing and loading the documentation\n",
     "\n",
-    "Next, let's pre-process and load the documentation. To make sure we don't need to do this repeatedly if we were updating code, we're caching it using pickle so we can retrieve it again (this could take a few minutes to run the first time yyou do it). We'll also add some more metadata to the docs here such as the title and version of the code:"
+    "Next, let's pre-process and load the documentation. To make sure we don't need to do this repeatedly if we were updating code, we're caching it using pickle so we can retrieve it again (this could take a few minutes to run the first time you do it). We'll also add some more metadata to the docs here such as the title and version of the code:"
    ]
   },
   {
@@ -255,7 +255,7 @@
    "id": "28d93b85",
    "metadata": {},
    "source": [
-    "And thats it! We're all setup. The next step is to run some queries, let's try a few:"
+    "And that's it! We're all set up. The next step is to run some queries, let's try a few:"
    ]
   },
   {

docs/src/notebooks/multimodal_search.ipynb

@@ -19,11 +19,11 @@
      "output_type": "stream",
      "text": [
       "\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip available: \u001b[0m\u001b[31;49m22.3.1\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.1.2\u001b[0m\n",
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m A new release of pip available: \u001B[0m\u001B[31;49m22.3.1\u001B[0m\u001B[39;49m -> \u001B[0m\u001B[32;49m23.1.2\u001B[0m\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n",
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m To update, run: \u001B[0m\u001B[32;49mpip install --upgrade pip\u001B[0m\n",
       "\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip available: \u001b[0m\u001b[31;49m22.3.1\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.1.2\u001b[0m\n",
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m A new release of pip available: \u001B[0m\u001B[31;49m22.3.1\u001B[0m\u001B[39;49m -> \u001B[0m\u001B[32;49m23.1.2\u001B[0m\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n"
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m To update, run: \u001B[0m\u001B[32;49mpip install --upgrade pip\u001B[0m\n"
      ]
     }
    ],
@@ -39,6 +39,7 @@
    "outputs": [],
    "source": [
     "import io\n",
+    "\n",
     "import PIL\n",
     "import duckdb\n",
     "import lancedb"
@@ -158,18 +159,18 @@
     "        \"db = lancedb.connect('~/datasets/demo')\\n\"\n",
     "        \"tbl = db.open_table('diffusiondb')\\n\\n\"\n",
     "        f\"embedding = embed_func('{query}')\\n\"\n",
-    "        \"tbl.search(embedding).limit(9).to_df()\"\n",
+    "        \"tbl.search(embedding).limit(9).to_pandas()\"\n",
     "    )\n",
-    "    return (_extract(tbl.search(emb).limit(9).to_df()), code)\n",
+    "    return (_extract(tbl.search(emb).limit(9).to_pandas()), code)\n",
     "\n",
     "def find_image_keywords(query):\n",
     "    code = (\n",
     "        \"import lancedb\\n\"\n",
     "        \"db = lancedb.connect('~/datasets/demo')\\n\"\n",
     "        \"tbl = db.open_table('diffusiondb')\\n\\n\"\n",
-    "        f\"tbl.search('{query}').limit(9).to_df()\"\n",
+    "        f\"tbl.search('{query}').limit(9).to_pandas()\"\n",
     "    )\n",
-    "    return (_extract(tbl.search(query).limit(9).to_df()), code)\n",
+    "    return (_extract(tbl.search(query).limit(9).to_pandas()), code)\n",
     "\n",
     "def find_image_sql(query):\n",
     "    code = (\n",

docs/src/notebooks/tables_guide.ipynb

@@ -114,13 +114,10 @@
     }
    ],
    "source": [
-    "import pandas as pd\n",
+    "data = [\n",
-    "\n",
+    "    {\"vector\": [1.1, 1.2], \"lat\": 45.5, \"long\": -122.7},\n",
-    "data = pd.DataFrame({\n",
+    "    {\"vector\": [0.2, 1.8], \"lat\": 40.1, \"long\": -74.1},\n",
-    "    \"vector\": [[1.1, 1.2], [0.2, 1.8]],\n",
+    "]\n",
-    "    \"lat\": [45.5, 40.1],\n",
-    "    \"long\": [-122.7, -74.1]\n",
-    "})\n",
     "\n",
     "db.create_table(\"table2\", data)\n",
     "\n",
@@ -572,9 +569,11 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = pd.DataFrame([{\"vector\": [1.3, 1.4], \"item\": \"fizz\", \"price\": 100.0},\n",
+    "data = [\n",
-    "              {\"vector\": [9.5, 56.2], \"item\": \"buzz\", \"price\": 200.0}])\n",
+    "        {\"vector\": [1.3, 1.4], \"item\": \"fizz\", \"price\": 100.0},\n",
-    "tbl.add(df)"
+    "        {\"vector\": [9.5, 56.2], \"item\": \"buzz\", \"price\": 200.0}\n",
+    "]\n",
+    "tbl.add(data)"
    ]
   },
   {
@@ -596,17 +595,12 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "\n",
-    "import pandas as pd\n",
-    "\n",
     "def make_batches():\n",
     "    for i in range(5):\n",
-    "        yield pd.DataFrame(\n",
+    "        yield [\n",
-    "                {\n",
+    "                  {\"vector\": [3.1, 4.1], \"item\": \"foo\", \"price\": 10.0},\n",
-    "                    \"vector\": [[3.1, 4.1], [1, 1]],\n",
+    "                  {\"vector\": [1, 1], \"item\": \"bar\", \"price\": 20.0},\n",
-    "                    \"item\": [\"foo\", \"bar\"],\n",
+    "              ]\n",
-    "                    \"price\": [10.0, 20.0],\n",
-    "                })\n",
     "tbl.add(make_batches())"
    ]
   },

docs/src/notebooks/youtube_transcript_search.ipynb

@@ -27,11 +27,11 @@
      "output_type": "stream",
      "text": [
       "\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.0\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.1.1\u001b[0m\n",
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m A new release of pip is available: \u001B[0m\u001B[31;49m23.0\u001B[0m\u001B[39;49m -> \u001B[0m\u001B[32;49m23.1.1\u001B[0m\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n",
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m To update, run: \u001B[0m\u001B[32;49mpip install --upgrade pip\u001B[0m\n",
       "\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.0\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.1.1\u001b[0m\n",
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m A new release of pip is available: \u001B[0m\u001B[31;49m23.0\u001B[0m\u001B[39;49m -> \u001B[0m\u001B[32;49m23.1.1\u001B[0m\n",
-      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n"
+      "\u001B[1m[\u001B[0m\u001B[34;49mnotice\u001B[0m\u001B[1;39;49m]\u001B[0m\u001B[39;49m To update, run: \u001B[0m\u001B[32;49mpip install --upgrade pip\u001B[0m\n"
      ]
     }
    ],
@@ -184,7 +184,7 @@
     "df = (contextualize(data.to_pandas())\n",
     "      .groupby(\"title\").text_col(\"text\")\n",
     "      .window(20).stride(4)\n",
-    "      .to_df())\n",
+    "      .to_pandas())\n",
     "df.head(1)"
    ]
   },
@@ -603,7 +603,7 @@
    "outputs": [],
    "source": [
     "# Use LanceDB to get top 3 most relevant context\n",
-    "context = tbl.search(emb).limit(3).to_df()"
+    "context = tbl.search(emb).limit(3).to_pandas()"
    ]
   },
   {

docs/src/python/arrow.md

@@ -39,7 +39,6 @@ to lazily generate data:
 
 from typing import Iterable
 import pyarrow as pa
-import lancedb
 
 def make_batches() -> Iterable[pa.RecordBatch]:
     for i in range(5):
@@ -74,7 +73,7 @@ table = db.open_table("pd_table")
 
 query_vector = [100, 100]
 # Pandas DataFrame
-df = table.search(query_vector).limit(1).to_df()
+df = table.search(query_vector).limit(1).to_pandas()
 print(df)
 ```
 
@@ -89,12 +88,12 @@ If you have more complex criteria, you can always apply the filter to the result
 ```python
 
 # Apply the filter via LanceDB
-results = table.search([100, 100]).where("price < 15").to_df()
+results = table.search([100, 100]).where("price < 15").to_pandas()
 assert len(results) == 1
 assert results["item"].iloc[0] == "foo"
 
 # Apply the filter via Pandas
-df = results = table.search([100, 100]).to_df()
+df = results = table.search([100, 100]).to_pandas()
 results = df[df.price < 15]
 assert len(results) == 1
 assert results["item"].iloc[0] == "foo"

docs/src/python/duckdb.md

@@ -11,15 +11,13 @@ pip install duckdb lancedb
 We will re-use [the dataset created previously](./arrow.md):
 
 ```python
-import pandas as pd
 import lancedb
 
 db = lancedb.connect("data/sample-lancedb")
-data = pd.DataFrame({
+data = [
-    "vector": [[3.1, 4.1], [5.9, 26.5]],
+    {"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
-    "item": ["foo", "bar"],
+    {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}
-    "price": [10.0, 20.0]
+]
-})
 table = db.create_table("pd_table", data=data)
 arrow_table = table.to_arrow()
 ```

docs/src/python/python.md

@@ -26,17 +26,17 @@ pip install lancedb
 
 ## Embeddings
 
-::: lancedb.embeddings.functions.EmbeddingFunctionRegistry
+::: lancedb.embeddings.registry.EmbeddingFunctionRegistry
 
-::: lancedb.embeddings.functions.EmbeddingFunction
+::: lancedb.embeddings.base.EmbeddingFunction
 
-::: lancedb.embeddings.functions.TextEmbeddingFunction
+::: lancedb.embeddings.base.TextEmbeddingFunction
 
-::: lancedb.embeddings.functions.SentenceTransformerEmbeddings
+::: lancedb.embeddings.sentence_transformers.SentenceTransformerEmbeddings
 
-::: lancedb.embeddings.functions.OpenAIEmbeddings
+::: lancedb.embeddings.openai.OpenAIEmbeddings
 
-::: lancedb.embeddings.functions.OpenClipEmbeddings
+::: lancedb.embeddings.open_clip.OpenClipEmbeddings
 
 ::: lancedb.embeddings.with_embeddings
 

docs/src/scripts/posthog.js

@@ -0,0 +1,4 @@
+window.addEventListener("DOMContentLoaded", (event) => {
+    !function(t,e){var o,n,p,r;e.__SV||(window.posthog=e,e._i=[],e.init=function(i,s,a){function g(t,e){var o=e.split(".");2==o.length&&(t=t[o[0]],e=o[1]),t[e]=function(){t.push([e].concat(Array.prototype.slice.call(arguments,0)))}}(p=t.createElement("script")).type="text/javascript",p.async=!0,p.src=s.api_host+"/static/array.js",(r=t.getElementsByTagName("script")[0]).parentNode.insertBefore(p,r);var u=e;for(void 0!==a?u=e[a]=[]:a="posthog",u.people=u.people||[],u.toString=function(t){var e="posthog";return"posthog"!==a&&(e+="."+a),t||(e+=" (stub)"),e},u.people.toString=function(){return u.toString(1)+".people (stub)"},o="capture identify alias people.set people.set_once set_config register register_once unregister opt_out_capturing has_opted_out_capturing opt_in_capturing reset isFeatureEnabled onFeatureFlags getFeatureFlag getFeatureFlagPayload reloadFeatureFlags group updateEarlyAccessFeatureEnrollment getEarlyAccessFeatures getActiveMatchingSurveys getSurveys".split(" "),n=0;n<o.length;n++)g(u,o[n]);e._i.push([i,s,a])},e.__SV=1)}(document,window.posthog||[]);
+    posthog.init('phc_oENDjGgHtmIDrV6puUiFem2RB4JA8gGWulfdulmMdZP',{api_host:'https://app.posthog.com'})
+});

docs/src/search.md

@@ -4,7 +4,7 @@
 In a recommendation system or search engine, you can find similar products from
 the one you searched.
 In LLM and other AI applications,
-each data point can be [presented by the embeddings generated from some models](embedding.md),
+each data point can be [presented by the embeddings generated from some models](embeddings/index.md),
 it returns the most relevant features.
 
 A search in high-dimensional vector space, is to find `K-Nearest-Neighbors (KNN)` of the query vector.
@@ -25,8 +25,8 @@ Currently, we support the following metrics:
 
 ### Flat Search
 
-If LanceDB does not create a vector index, LanceDB would need to scan (`Flat Search`) the entire vector column
+If you do not create a vector index, LanceDB would need to exhaustively scan the entire vector column (via `Flat Search`)
-and compute the distance for each vector in order to find the closest matches.
+and compute the distance for *every* vector in order to find the closest matches. This is effectively a KNN search.
 
 
 <!-- Setup Code
@@ -67,7 +67,7 @@ await db_setup.createTable('my_vectors', data)
 
     df = tbl.search(np.random.random((1536))) \
         .limit(10) \
-        .to_df()
+        .to_list()
     ```
 
 === "JavaScript"
@@ -92,7 +92,7 @@ as well.
     df = tbl.search(np.random.random((1536))) \
         .metric("cosine") \
         .limit(10) \
-        .to_df()
+        .to_list()
     ```
 
 
@@ -110,7 +110,7 @@ as well.
 
 To accelerate vector retrievals, it is common to build vector indices.
 A vector index is a data structure specifically designed to efficiently organize and
-search vector data based on their similarity or distance metrics.
+search vector data based on their similarity via the chosen distance metric.
 By constructing a vector index, you can reduce the search space and avoid the need
 for brute-force scanning of the entire vector column.
 

docs/test/md_testing.js

@@ -8,6 +8,7 @@ const excludedGlobs = [
   "../src/embedding.md",
   "../src/examples/*.md",
   "../src/guides/tables.md",
+  "../src/embeddings/*.md",
 ];
 
 const nodePrefix = "javascript";

docs/test/md_testing.py

@@ -10,6 +10,7 @@ excluded_globs = [
     "../src/integrations/voxel51.md",
     "../src/guides/tables.md",
     "../src/python/duckdb.md",
+    "../src/embeddings/*.md",
 ]
 
 python_prefix = "py"

docs/test/requirements.txt

@@ -1,5 +1,8 @@
-lancedb @ git+https://github.com/lancedb/lancedb.git#egg=subdir&subdirectory=python
+-e ../../python
 numpy
 pandas
 pylance
 duckdb
+--extra-index-url https://download.pytorch.org/whl/cpu
+torch
+

node/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "vectordb",
-  "version": "0.2.6",
+  "version": "0.3.3",
   "lockfileVersion": 2,
   "requires": true,
   "packages": {
     "": {
       "name": "vectordb",
-      "version": "0.2.6",
+      "version": "0.3.3",
       "cpu": [
         "x64",
         "arm64"
@@ -53,11 +53,11 @@
         "uuid": "^9.0.0"
       },
       "optionalDependencies": {
-        "@lancedb/vectordb-darwin-arm64": "0.2.6",
+        "@lancedb/vectordb-darwin-arm64": "0.3.3",
-        "@lancedb/vectordb-darwin-x64": "0.2.6",
+        "@lancedb/vectordb-darwin-x64": "0.3.3",
-        "@lancedb/vectordb-linux-arm64-gnu": "0.2.6",
+        "@lancedb/vectordb-linux-arm64-gnu": "0.3.3",
-        "@lancedb/vectordb-linux-x64-gnu": "0.2.6",
+        "@lancedb/vectordb-linux-x64-gnu": "0.3.3",
-        "@lancedb/vectordb-win32-x64-msvc": "0.2.6"
+        "@lancedb/vectordb-win32-x64-msvc": "0.3.3"
       }
     },
     "node_modules/@apache-arrow/ts": {
@@ -317,9 +317,9 @@
       }
     },
     "node_modules/@lancedb/vectordb-darwin-arm64": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.3.3.tgz",
-      "integrity": "sha512-9KCUvDmhVMuGIhleib/Gq43QhrRXjy2QJz21S85HDwL3DTH4J9n00A0V6eyLTBUyctnvMTcp3XZijosYUy1A8Q==",
+      "integrity": "sha512-nvyj7xNX2/wb/PH5TjyhLR/NQ1jVuoBw2B5UaSg7qf8Tnm5SSXWQ7F25RVKcKwh72fz1qB+CWW24ftZnRzbT/Q==",
       "cpu": [
         "arm64"
       ],
@@ -329,9 +329,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-darwin-x64": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.3.3.tgz",
-      "integrity": "sha512-WCYRFV9w13STgVYn4WSYne39mp+g8ET6TgMLvSSQBYJKp3xEggpSCtACetaDfmNpkml9DK/b5R95Jc7PBbmYgA==",
+      "integrity": "sha512-7CW+nILyPHp6cua0Rl0xaTDWw/vajEn/jCsEjFYgDmE+rtf5Z5Fum41FxR9C2TtIAvUK+nWb5mkYeOLqU6vRvg==",
       "cpu": [
         "x64"
       ],
@@ -341,9 +341,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-linux-arm64-gnu": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.3.3.tgz",
-      "integrity": "sha512-SE9OUgsOT6dG1q9v3nFr9ew+kwPTA4ktvNiHiyQstNz9BniuLNldF/Wtxzk/Z7DhbkPci4MfkR6RdsPTHBatHg==",
+      "integrity": "sha512-MmhwbacKxZPkLwwOqysVY8mUb8lFoyFIPlYhSLV4xS1C8X4HWALljIul1qMl1RYudp9Uc3PsOzRexl+OvCGfUw==",
       "cpu": [
         "arm64"
       ],
@@ -353,9 +353,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-linux-x64-gnu": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.3.3.tgz",
-      "integrity": "sha512-hvUsRQbaJiQnSjjKHIRhJM/eObJOqDJUXcpzz1fWw/MMSoy/CFaQwf9Uen2IWTgcngGkJAkeEKG7N5GxQxVbBQ==",
+      "integrity": "sha512-OrNlsKi/QPw59Po040oRKn8IuqFEk4upc/4FaFKqVkcmQjjZrMg5Kgy9ZfWIhHdAnWXXggZZIPArpt0X1B0ceA==",
       "cpu": [
         "x64"
       ],
@@ -365,9 +365,9 @@
       ]
     },
     "node_modules/@lancedb/vectordb-win32-x64-msvc": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.3.3.tgz",
-      "integrity": "sha512-XPIzbBPt28nsAa7INuyvYMZyJ78bgLfxjSyazlydzO10orIBHvR+sjcrdnCK4l48YmvPXcSYnKxlKMa1oUeIWQ==",
+      "integrity": "sha512-lIT0A7a6eqX51IfGyhECtpXXgsr//kgbd+HZbcCdPy2GMmNezSch/7V22zExDSpF32hX8WfgcTLYCVWVilggDQ==",
       "cpu": [
         "x64"
       ],
@@ -4869,33 +4869,33 @@
       }
     },
     "@lancedb/vectordb-darwin-arm64": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-arm64/-/vectordb-darwin-arm64-0.3.3.tgz",
-      "integrity": "sha512-9KCUvDmhVMuGIhleib/Gq43QhrRXjy2QJz21S85HDwL3DTH4J9n00A0V6eyLTBUyctnvMTcp3XZijosYUy1A8Q==",
+      "integrity": "sha512-nvyj7xNX2/wb/PH5TjyhLR/NQ1jVuoBw2B5UaSg7qf8Tnm5SSXWQ7F25RVKcKwh72fz1qB+CWW24ftZnRzbT/Q==",
       "optional": true
     },
     "@lancedb/vectordb-darwin-x64": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-darwin-x64/-/vectordb-darwin-x64-0.3.3.tgz",
-      "integrity": "sha512-WCYRFV9w13STgVYn4WSYne39mp+g8ET6TgMLvSSQBYJKp3xEggpSCtACetaDfmNpkml9DK/b5R95Jc7PBbmYgA==",
+      "integrity": "sha512-7CW+nILyPHp6cua0Rl0xaTDWw/vajEn/jCsEjFYgDmE+rtf5Z5Fum41FxR9C2TtIAvUK+nWb5mkYeOLqU6vRvg==",
       "optional": true
     },
     "@lancedb/vectordb-linux-arm64-gnu": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-arm64-gnu/-/vectordb-linux-arm64-gnu-0.3.3.tgz",
-      "integrity": "sha512-SE9OUgsOT6dG1q9v3nFr9ew+kwPTA4ktvNiHiyQstNz9BniuLNldF/Wtxzk/Z7DhbkPci4MfkR6RdsPTHBatHg==",
+      "integrity": "sha512-MmhwbacKxZPkLwwOqysVY8mUb8lFoyFIPlYhSLV4xS1C8X4HWALljIul1qMl1RYudp9Uc3PsOzRexl+OvCGfUw==",
       "optional": true
     },
     "@lancedb/vectordb-linux-x64-gnu": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-linux-x64-gnu/-/vectordb-linux-x64-gnu-0.3.3.tgz",
-      "integrity": "sha512-hvUsRQbaJiQnSjjKHIRhJM/eObJOqDJUXcpzz1fWw/MMSoy/CFaQwf9Uen2IWTgcngGkJAkeEKG7N5GxQxVbBQ==",
+      "integrity": "sha512-OrNlsKi/QPw59Po040oRKn8IuqFEk4upc/4FaFKqVkcmQjjZrMg5Kgy9ZfWIhHdAnWXXggZZIPArpt0X1B0ceA==",
       "optional": true
     },
     "@lancedb/vectordb-win32-x64-msvc": {
-      "version": "0.2.6",
+      "version": "0.3.3",
-      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.2.6.tgz",
+      "resolved": "https://registry.npmjs.org/@lancedb/vectordb-win32-x64-msvc/-/vectordb-win32-x64-msvc-0.3.3.tgz",
-      "integrity": "sha512-XPIzbBPt28nsAa7INuyvYMZyJ78bgLfxjSyazlydzO10orIBHvR+sjcrdnCK4l48YmvPXcSYnKxlKMa1oUeIWQ==",
+      "integrity": "sha512-lIT0A7a6eqX51IfGyhECtpXXgsr//kgbd+HZbcCdPy2GMmNezSch/7V22zExDSpF32hX8WfgcTLYCVWVilggDQ==",
       "optional": true
     },
     "@neon-rs/cli": {

node/package.json

@@ -1,6 +1,6 @@
 {
   "name": "vectordb",
-  "version": "0.2.6",
+  "version": "0.3.3",
   "description": " Serverless, low-latency vector database for AI applications",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
@@ -81,10 +81,10 @@
     }
   },
   "optionalDependencies": {
-    "@lancedb/vectordb-darwin-arm64": "0.2.6",
+    "@lancedb/vectordb-darwin-arm64": "0.3.3",
-    "@lancedb/vectordb-darwin-x64": "0.2.6",
+    "@lancedb/vectordb-darwin-x64": "0.3.3",
-    "@lancedb/vectordb-linux-arm64-gnu": "0.2.6",
+    "@lancedb/vectordb-linux-arm64-gnu": "0.3.3",
-    "@lancedb/vectordb-linux-x64-gnu": "0.2.6",
+    "@lancedb/vectordb-linux-x64-gnu": "0.3.3",
-    "@lancedb/vectordb-win32-x64-msvc": "0.2.6"
+    "@lancedb/vectordb-win32-x64-msvc": "0.3.3"
   }
 }

node/src/arrow.ts

@@ -20,7 +20,7 @@ import {
   Utf8,
   type Vector,
   FixedSizeList,
-  vectorFromArray, type Schema, Table as ArrowTable
+  vectorFromArray, type Schema, Table as ArrowTable, RecordBatchStreamWriter
 } from 'apache-arrow'
 import { type EmbeddingFunction } from './index'
 
@@ -77,7 +77,9 @@ function newVectorBuilder (dim: number): FixedSizeListBuilder<Float32> {
 
 // Creates the Arrow Type for a Vector column with dimension `dim`
 function newVectorType (dim: number): FixedSizeList<Float32> {
-  const children = new Field<Float32>('item', new Float32())
+  // Somewhere we always default to have the elements nullable, so we need to set it to true
+  // otherwise we often get schema mismatches because the stored data always has schema with nullable elements
+  const children = new Field<Float32>('item', new Float32(), true)
   return new FixedSizeList(dim, children)
 }
 
@@ -88,6 +90,13 @@ export async function fromRecordsToBuffer<T> (data: Array<Record<string, unknown
   return Buffer.from(await writer.toUint8Array())
 }
 
+// Converts an Array of records into Arrow IPC stream format
+export async function fromRecordsToStreamBuffer<T> (data: Array<Record<string, unknown>>, embeddings?: EmbeddingFunction<T>): Promise<Buffer> {
+  const table = await convertToTable(data, embeddings)
+  const writer = RecordBatchStreamWriter.writeAll(table)
+  return Buffer.from(await writer.toUint8Array())
+}
+
 // Converts an Arrow Table into Arrow IPC format
 export async function fromTableToBuffer<T> (table: ArrowTable, embeddings?: EmbeddingFunction<T>): Promise<Buffer> {
   if (embeddings !== undefined) {
@@ -105,6 +114,23 @@ export async function fromTableToBuffer<T> (table: ArrowTable, embeddings?: Embe
   return Buffer.from(await writer.toUint8Array())
 }
 
+// Converts an Arrow Table into Arrow IPC stream format
+export async function fromTableToStreamBuffer<T> (table: ArrowTable, embeddings?: EmbeddingFunction<T>): Promise<Buffer> {
+  if (embeddings !== undefined) {
+    const source = table.getChild(embeddings.sourceColumn)
+
+    if (source === null) {
+      throw new Error(`The embedding source column ${embeddings.sourceColumn} was not found in the Arrow Table`)
+    }
+
+    const vectors = await embeddings.embed(source.toArray() as T[])
+    const column = vectorFromArray(vectors, newVectorType(vectors[0].length))
+    table = table.assign(new ArrowTable({ vector: column }))
+  }
+  const writer = RecordBatchStreamWriter.writeAll(table)
+  return Buffer.from(await writer.toUint8Array())
+}
+
 // Creates an empty Arrow Table
 export function createEmptyTable (schema: Schema): ArrowTable {
   return new ArrowTable(schema)

node/src/index.ts

@@ -23,7 +23,7 @@ import { Query } from './query'
 import { isEmbeddingFunction } from './embedding/embedding_function'
 
 // eslint-disable-next-line @typescript-eslint/no-var-requires
-const { databaseNew, databaseTableNames, databaseOpenTable, databaseDropTable, tableCreate, tableAdd, tableCreateVectorIndex, tableCountRows, tableDelete } = require('../native.js')
+const { databaseNew, databaseTableNames, databaseOpenTable, databaseDropTable, tableCreate, tableAdd, tableCreateVectorIndex, tableCountRows, tableDelete, tableCleanupOldVersions, tableCompactFiles } = require('../native.js')
 
 export { Query }
 export type { EmbeddingFunction }
@@ -459,6 +459,111 @@ export class LocalTable<T = number[]> implements Table<T> {
   async delete (filter: string): Promise<void> {
     return tableDelete.call(this._tbl, filter).then((newTable: any) => { this._tbl = newTable })
   }
+
+  /**
+   * Clean up old versions of the table, freeing disk space.
+   *
+   * @param olderThan The minimum age in minutes of the versions to delete. If not
+   *                  provided, defaults to two weeks.
+   * @param deleteUnverified Because they may be part of an in-progress
+   *                  transaction, uncommitted files newer than 7 days old are
+   *                  not deleted by default. This means that failed transactions
+   *                  can leave around data that takes up disk space for up to
+   *                  7 days. You can override this safety mechanism by setting
+   *                 this option to `true`, only if you promise there are no
+   *                 in progress writes while you run this operation. Failure to
+   *                 uphold this promise can lead to corrupted tables.
+   * @returns
+   */
+  async cleanupOldVersions (olderThan?: number, deleteUnverified?: boolean): Promise<CleanupStats> {
+    return tableCleanupOldVersions.call(this._tbl, olderThan, deleteUnverified)
+      .then((res: { newTable: any, metrics: CleanupStats }) => {
+        this._tbl = res.newTable
+        return res.metrics
+      })
+  }
+
+  /**
+   * Run the compaction process on the table.
+   *
+   * This can be run after making several small appends to optimize the table
+   * for faster reads.
+   *
+   * @param options Advanced options configuring compaction. In most cases, you
+   *               can omit this arguments, as the default options are sensible
+   *               for most tables.
+   * @returns Metrics about the compaction operation.
+   */
+  async compactFiles (options?: CompactionOptions): Promise<CompactionMetrics> {
+    const optionsArg = options ?? {}
+    return tableCompactFiles.call(this._tbl, optionsArg)
+      .then((res: { newTable: any, metrics: CompactionMetrics }) => {
+        this._tbl = res.newTable
+        return res.metrics
+      })
+  }
+}
+
+export interface CleanupStats {
+  /**
+   * The number of bytes removed from disk.
+   */
+  bytesRemoved: number
+  /**
+   * The number of old table versions removed.
+   */
+  oldVersions: number
+}
+
+export interface CompactionOptions {
+  /**
+   * The number of rows per fragment to target. Fragments that have fewer rows
+   * will be compacted into adjacent fragments to produce larger fragments.
+   * Defaults to 1024 * 1024.
+   */
+  targetRowsPerFragment?: number
+  /**
+   * The maximum number of rows per group. Defaults to 1024.
+   */
+  maxRowsPerGroup?: number
+  /**
+   * If true, fragments that have rows that are deleted may be compacted to
+   * remove the deleted rows. This can improve the performance of queries.
+   * Default is true.
+   */
+  materializeDeletions?: boolean
+  /**
+   * A number between 0 and 1, representing the proportion of rows that must be
+   * marked deleted before a fragment is a candidate for compaction to remove
+   * the deleted rows. Default is 10%.
+   */
+  materializeDeletionsThreshold?: number
+  /**
+   * The number of threads to use for compaction. If not provided, defaults to
+   * the number of cores on the machine.
+   */
+  numThreads?: number
+}
+
+export interface CompactionMetrics {
+  /**
+   * The number of fragments that were removed.
+   */
+  fragmentsRemoved: number
+  /**
+   * The number of new fragments that were created.
+   */
+  fragmentsAdded: number
+  /**
+   * The number of files that were removed. Each fragment may have more than one
+   * file.
+   */
+  filesRemoved: number
+  /**
+   * The number of files added. This is typically equal to the number of
+   * fragments added.
+   */
+  filesAdded: number
 }
 
 /// Config to build IVF_PQ index.

node/src/integration_test/test.ts

@@ -18,6 +18,9 @@ import * as chaiAsPromised from 'chai-as-promised'
 import { v4 as uuidv4 } from 'uuid'
 
 import * as lancedb from '../index'
+import { tmpdir } from 'os'
+import * as fs from 'fs'
+import * as path from 'path'
 
 const assert = chai.assert
 chai.use(chaiAsPromised)
@@ -41,3 +44,137 @@ describe('LanceDB AWS Integration test', function () {
     assert.equal(await table.countRows(), 6)
   })
 })
+
+describe('LanceDB Mirrored Store Integration test', function () {
+  it('s3://...?mirroredStore=... param is processed correctly', async function () {
+    this.timeout(600000)
+
+    const dir = tmpdir()
+    console.log(dir)
+    const conn = await lancedb.connect(`s3://lancedb-integtest?mirroredStore=${dir}`)
+    const data = Array(200).fill({ vector: Array(128).fill(1.0), id: 0 })
+    data.push(...Array(200).fill({ vector: Array(128).fill(1.0), id: 1 }))
+    data.push(...Array(200).fill({ vector: Array(128).fill(1.0), id: 2 }))
+    data.push(...Array(200).fill({ vector: Array(128).fill(1.0), id: 3 }))
+
+    const tableName = uuidv4()
+
+    // try create table and check if it's mirrored
+    const t = await conn.createTable(tableName, data, { writeMode: lancedb.WriteMode.Overwrite })
+
+    const mirroredPath = path.join(dir, `${tableName}.lance`)
+    fs.readdir(mirroredPath, { withFileTypes: true }, (err, files) => {
+      if (err != null) throw err
+      // there should be three dirs
+      assert.equal(files.length, 3)
+      assert.isTrue(files[0].isDirectory())
+      assert.isTrue(files[1].isDirectory())
+
+      fs.readdir(path.join(mirroredPath, '_transactions'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].name.endsWith('.txn'))
+      })
+
+      fs.readdir(path.join(mirroredPath, '_versions'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].name.endsWith('.manifest'))
+      })
+
+      fs.readdir(path.join(mirroredPath, 'data'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].name.endsWith('.lance'))
+      })
+    })
+
+    // try create index and check if it's mirrored
+    await t.createIndex({ column: 'vector', type: 'ivf_pq' })
+
+    fs.readdir(mirroredPath, { withFileTypes: true }, (err, files) => {
+      if (err != null) throw err
+      // there should be four dirs
+      assert.equal(files.length, 4)
+      assert.isTrue(files[0].isDirectory())
+      assert.isTrue(files[1].isDirectory())
+      assert.isTrue(files[2].isDirectory())
+
+      // Two TXs now
+      fs.readdir(path.join(mirroredPath, '_transactions'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 2)
+        assert.isTrue(files[0].name.endsWith('.txn'))
+        assert.isTrue(files[1].name.endsWith('.txn'))
+      })
+
+      fs.readdir(path.join(mirroredPath, 'data'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].name.endsWith('.lance'))
+      })
+
+      fs.readdir(path.join(mirroredPath, '_indices'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].isDirectory())
+
+        fs.readdir(path.join(mirroredPath, '_indices', files[0].name), { withFileTypes: true }, (err, files) => {
+          if (err != null) throw err
+
+          assert.equal(files.length, 1)
+          assert.isTrue(files[0].isFile())
+          assert.isTrue(files[0].name.endsWith('.idx'))
+        })
+      })
+    })
+
+    // try delete and check if it's mirrored
+    await t.delete('id = 0')
+
+    fs.readdir(mirroredPath, { withFileTypes: true }, (err, files) => {
+      if (err != null) throw err
+      // there should be five dirs
+      assert.equal(files.length, 5)
+      assert.isTrue(files[0].isDirectory())
+      assert.isTrue(files[1].isDirectory())
+      assert.isTrue(files[2].isDirectory())
+      assert.isTrue(files[3].isDirectory())
+      assert.isTrue(files[4].isDirectory())
+
+      // Three TXs now
+      fs.readdir(path.join(mirroredPath, '_transactions'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 3)
+        assert.isTrue(files[0].name.endsWith('.txn'))
+        assert.isTrue(files[1].name.endsWith('.txn'))
+      })
+
+      fs.readdir(path.join(mirroredPath, 'data'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].name.endsWith('.lance'))
+      })
+
+      fs.readdir(path.join(mirroredPath, '_indices'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].isDirectory())
+
+        fs.readdir(path.join(mirroredPath, '_indices', files[0].name), { withFileTypes: true }, (err, files) => {
+          if (err != null) throw err
+
+          assert.equal(files.length, 1)
+          assert.isTrue(files[0].isFile())
+          assert.isTrue(files[0].name.endsWith('.idx'))
+        })
+      })
+
+      fs.readdir(path.join(mirroredPath, '_deletions'), { withFileTypes: true }, (err, files) => {
+        if (err != null) throw err
+        assert.equal(files.length, 1)
+        assert.isTrue(files[0].name.endsWith('.arrow'))
+      })
+    })
+  })
+})

node/src/remote/client.ts

@@ -108,13 +108,18 @@ export class HttpLancedbClient {
   /**
    * Sent POST request.
    */
-  public async post (path: string, data?: any, params?: Record<string, string | number>): Promise<AxiosResponse> {
+  public async post (
+    path: string,
+    data?: any,
+    params?: Record<string, string | number>,
+    content?: string | undefined
+  ): Promise<AxiosResponse> {
     const response = await axios.post(
         `${this._url}${path}`,
         data,
         {
           headers: {
-            'Content-Type': 'application/json',
+            'Content-Type': content ?? 'application/json',
             'x-api-key': this._apiKey(),
             ...(this._dbName !== undefined ? { 'x-lancedb-database': this._dbName } : {})
           },

node/src/remote/index.ts

@@ -18,8 +18,10 @@ import {
 } from '../index'
 import { Query } from '../query'
 
-import { Vector } from 'apache-arrow'
+import { Vector, Table as ArrowTable } from 'apache-arrow'
 import { HttpLancedbClient } from './client'
+import { isEmbeddingFunction } from '../embedding/embedding_function'
+import { createEmptyTable, fromRecordsToStreamBuffer, fromTableToStreamBuffer } from '../arrow'
 
 /**
  * Remote connection.
@@ -66,8 +68,60 @@ export class RemoteConnection implements Connection {
     }
   }
 
-  async createTable<T> (name: string | CreateTableOptions<T>, data?: Array<Record<string, unknown>>, optsOrEmbedding?: WriteOptions | EmbeddingFunction<T>, opt?: WriteOptions): Promise<Table<T>> {
+  async createTable<T> (nameOrOpts: string | CreateTableOptions<T>, data?: Array<Record<string, unknown>>, optsOrEmbedding?: WriteOptions | EmbeddingFunction<T>, opt?: WriteOptions): Promise<Table<T>> {
-    throw new Error('Not implemented')
+    // Logic copied from LocatlConnection, refactor these to a base class + connectionImpl pattern
+    let schema
+    let embeddings: undefined | EmbeddingFunction<T>
+    let tableName: string
+    if (typeof nameOrOpts === 'string') {
+      if (optsOrEmbedding !== undefined && isEmbeddingFunction(optsOrEmbedding)) {
+        embeddings = optsOrEmbedding
+      }
+      tableName = nameOrOpts
+    } else {
+      schema = nameOrOpts.schema
+      embeddings = nameOrOpts.embeddingFunction
+      tableName = nameOrOpts.name
+    }
+
+    let buffer: Buffer
+
+    function isEmpty (data: Array<Record<string, unknown>> | ArrowTable<any>): boolean {
+      if (data instanceof ArrowTable) {
+        return data.data.length === 0
+      }
+      return data.length === 0
+    }
+
+    if ((data === undefined) || isEmpty(data)) {
+      if (schema === undefined) {
+        throw new Error('Either data or schema needs to defined')
+      }
+      buffer = await fromTableToStreamBuffer(createEmptyTable(schema))
+    } else if (data instanceof ArrowTable) {
+      buffer = await fromTableToStreamBuffer(data, embeddings)
+    } else {
+      // data is Array<Record<...>>
+      buffer = await fromRecordsToStreamBuffer(data, embeddings)
+    }
+
+    const res = await this._client.post(
+      `/v1/table/${tableName}/create/`,
+      buffer,
+      undefined,
+      'application/vnd.apache.arrow.stream'
+    )
+    if (res.status !== 200) {
+      throw new Error(`Server Error, status: ${res.status}, ` +
+      // eslint-disable-next-line @typescript-eslint/restrict-template-expressions
+        `message: ${res.statusText}: ${res.data}`)
+    }
+
+    if (embeddings === undefined) {
+      return new RemoteTable(this._client, tableName)
+    } else {
+      return new RemoteTable(this._client, tableName, embeddings)
+    }
   }
 
   async dropTable (name: string): Promise<void> {
@@ -141,11 +195,39 @@ export class RemoteTable<T = number[]> implements Table<T> {
   }
 
   async add (data: Array<Record<string, unknown>>): Promise<number> {
-    throw new Error('Not implemented')
+    const buffer = await fromRecordsToStreamBuffer(data, this._embeddings)
+    const res = await this._client.post(
+      `/v1/table/${this._name}/insert/`,
+      buffer,
+      {
+        mode: 'append'
+      },
+      'application/vnd.apache.arrow.stream'
+    )
+    if (res.status !== 200) {
+      throw new Error(`Server Error, status: ${res.status}, ` +
+      // eslint-disable-next-line @typescript-eslint/restrict-template-expressions
+        `message: ${res.statusText}: ${res.data}`)
+    }
+    return data.length
   }
 
   async overwrite (data: Array<Record<string, unknown>>): Promise<number> {
-    throw new Error('Not implemented')
+    const buffer = await fromRecordsToStreamBuffer(data, this._embeddings)
+    const res = await this._client.post(
+      `/v1/table/${this._name}/insert/`,
+      buffer,
+      {
+        mode: 'overwrite'
+      },
+      'application/vnd.apache.arrow.stream'
+    )
+    if (res.status !== 200) {
+      throw new Error(`Server Error, status: ${res.status}, ` +
+      // eslint-disable-next-line @typescript-eslint/restrict-template-expressions
+        `message: ${res.statusText}: ${res.data}`)
+    }
+    return data.length
   }
 
   async createIndex (indexParams: VectorIndexParams): Promise<any> {
@@ -157,6 +239,6 @@ export class RemoteTable<T = number[]> implements Table<T> {
   }
 
   async delete (filter: string): Promise<void> {
-    throw new Error('Not implemented')
+    await this._client.post(`/v1/table/${this._name}/delete/`, { predicate: filter })
   }
 }

node/src/test/test.ts

@@ -18,7 +18,7 @@ import * as chai from 'chai'
 import * as chaiAsPromised from 'chai-as-promised'
 
 import * as lancedb from '../index'
-import { type AwsCredentials, type EmbeddingFunction, MetricType, Query, WriteMode, DefaultWriteOptions, isWriteOptions } from '../index'
+import { type AwsCredentials, type EmbeddingFunction, MetricType, Query, WriteMode, DefaultWriteOptions, isWriteOptions, type LocalTable } from '../index'
 import { FixedSizeList, Field, Int32, makeVector, Schema, Utf8, Table as ArrowTable, vectorFromArray, Float32 } from 'apache-arrow'
 
 const expect = chai.expect
@@ -446,3 +446,45 @@ describe('WriteOptions', function () {
     })
   })
 })
+
+describe('Compact and cleanup', function () {
+  it('can cleanup after compaction', async function () {
+    const dir = await track().mkdir('lancejs')
+    const con = await lancedb.connect(dir)
+
+    const data = [
+      { price: 10, name: 'foo', vector: [1, 2, 3] },
+      { price: 50, name: 'bar', vector: [4, 5, 6] }
+    ]
+    const table = await con.createTable('t1', data) as LocalTable
+
+    const newData = [
+      { price: 30, name: 'baz', vector: [7, 8, 9] }
+    ]
+    await table.add(newData)
+
+    const compactionMetrics = await table.compactFiles({
+      numThreads: 2
+    })
+    assert.equal(compactionMetrics.fragmentsRemoved, 2)
+    assert.equal(compactionMetrics.fragmentsAdded, 1)
+    assert.equal(await table.countRows(), 3)
+
+    await table.cleanupOldVersions()
+    assert.equal(await table.countRows(), 3)
+
+    // should have no effect, but this validates the arguments are parsed.
+    await table.compactFiles({
+      targetRowsPerFragment: 1024 * 10,
+      maxRowsPerGroup: 1024,
+      materializeDeletions: true,
+      materializeDeletionsThreshold: 0.5,
+      numThreads: 2
+    })
+
+    const cleanupMetrics = await table.cleanupOldVersions(0, true)
+    assert.isAtLeast(cleanupMetrics.bytesRemoved, 1)
+    assert.isAtLeast(cleanupMetrics.oldVersions, 1)
+    assert.equal(await table.countRows(), 3)
+  })
+})

python/.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.2.6
+current_version = 0.3.2
 commit = True
 message = [python] Bump version: {current_version} → {new_version}
 tag = True

python/LICENSE

@@ -0,0 +1 @@
+../LICENSE

python/README.md

@@ -16,7 +16,7 @@ pip install lancedb
 import lancedb
 db = lancedb.connect('<PATH_TO_LANCEDB_DATASET>')
 table = db.open_table('my_table')
-results = table.search([0.1, 0.3]).limit(20).to_df()
+results = table.search([0.1, 0.3]).limit(20).to_list()
 print(results)
 ```
 

python/lancedb/__init__.py

@@ -14,11 +14,12 @@
 import importlib.metadata
 from typing import Optional
 
+__version__ = importlib.metadata.version("lancedb")
+
 from .db import URI, DBConnection, LanceDBConnection
 from .remote.db import RemoteDBConnection
 from .schema import vector
-
+from .utils import sentry_log
-__version__ = importlib.metadata.version("lancedb")
 
 
 def connect(

python/lancedb/cli/__init__.py

@@ -0,0 +1,12 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.

python/lancedb/cli/cli.py

@@ -0,0 +1,46 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import click
+
+from lancedb.utils import CONFIG
+
+
+@click.group()
+@click.version_option(help="LanceDB command line interface entry point")
+def cli():
+    "LanceDB command line interface"
+
+
+diagnostics_help = """
+Enable or disable LanceDB diagnostics. When enabled, LanceDB will send anonymous events to help us improve LanceDB.
+These diagnostics are used only for error reporting and no data is collected. You can find more about diagnosis on
+our docs: https://lancedb.github.io/lancedb/cli_config/
+"""
+
+
+@cli.command(help=diagnostics_help)
+@click.option("--enabled/--disabled", default=True)
+def diagnostics(enabled):
+    CONFIG.update({"diagnostics": True if enabled else False})
+    click.echo("LanceDB diagnostics is %s" % ("enabled" if enabled else "disabled"))
+
+
+@cli.command(help="Show current LanceDB configuration")
+def config():
+    # TODO: pretty print as table with colors and formatting
+    click.echo("Current LanceDB configuration:")
+    cfg = CONFIG.copy()
+    cfg.pop("uuid")  # Don't show uuid as it is not configurable
+    for item, amount in cfg.items():
+        click.echo("{} ({})".format(item, amount))

python/lancedb/context.py

@@ -12,6 +12,9 @@
 #  limitations under the License.
 from __future__ import annotations
 
+import deprecation
+
+from . import __version__
 from .exceptions import MissingColumnError, MissingValueError
 from .util import safe_import_pandas
 
@@ -43,7 +46,7 @@ def contextualize(raw_df: "pd.DataFrame") -> Contextualizer:
     this how many tokens, but depending on the input data, it could be sentences,
     paragraphs, messages, etc.
 
-    >>> contextualize(data).window(3).stride(1).text_col('token').to_df()
+    >>> contextualize(data).window(3).stride(1).text_col('token').to_pandas()
                     token  document_id
     0     The quick brown            1
     1     quick brown fox            1
@@ -56,7 +59,7 @@ def contextualize(raw_df: "pd.DataFrame") -> Contextualizer:
     8          dog I love            1
     9   I love sandwiches            2
     10    love sandwiches            2
-    >>> contextualize(data).window(7).stride(1).min_window_size(7).text_col('token').to_df()
+    >>> contextualize(data).window(7).stride(1).min_window_size(7).text_col('token').to_pandas()
                                       token  document_id
     0   The quick brown fox jumped over the            1
     1  quick brown fox jumped over the lazy            1
@@ -68,7 +71,7 @@ def contextualize(raw_df: "pd.DataFrame") -> Contextualizer:
     ``stride`` determines how many rows to skip between each window start. This can
     be used to reduce the total number of windows generated.
 
-    >>> contextualize(data).window(4).stride(2).text_col('token').to_df()
+    >>> contextualize(data).window(4).stride(2).text_col('token').to_pandas()
                         token  document_id
     0     The quick brown fox            1
     2   brown fox jumped over            1
@@ -81,7 +84,7 @@ def contextualize(raw_df: "pd.DataFrame") -> Contextualizer:
     context windows that don't cross document boundaries. In this case, we can
     pass ``document_id`` as the group by.
 
-    >>> contextualize(data).window(4).stride(2).text_col('token').groupby('document_id').to_df()
+    >>> contextualize(data).window(4).stride(2).text_col('token').groupby('document_id').to_pandas()
                        token  document_id
     0    The quick brown fox            1
     2  brown fox jumped over            1
@@ -93,14 +96,14 @@ def contextualize(raw_df: "pd.DataFrame") -> Contextualizer:
     This can be used to trim the last few context windows which have size less than
     ``min_window_size``. By default context windows of size 1 are skipped.
 
-    >>> contextualize(data).window(6).stride(3).text_col('token').groupby('document_id').to_df()
+    >>> contextualize(data).window(6).stride(3).text_col('token').groupby('document_id').to_pandas()
                                  token  document_id
     0  The quick brown fox jumped over            1
     3     fox jumped over the lazy dog            1
     6                     the lazy dog            1
     9                I love sandwiches            2
 
-    >>> contextualize(data).window(6).stride(3).min_window_size(4).text_col('token').groupby('document_id').to_df()
+    >>> contextualize(data).window(6).stride(3).min_window_size(4).text_col('token').groupby('document_id').to_pandas()
                                  token  document_id
     0  The quick brown fox jumped over            1
     3     fox jumped over the lazy dog            1
@@ -176,7 +179,16 @@ class Contextualizer:
         self._min_window_size = min_window_size
         return self
 
+    @deprecation.deprecated(
+        deprecated_in="0.3.1",
+        removed_in="0.4.0",
+        current_version=__version__,
+        details="Use the bar function instead",
+    )
     def to_df(self) -> "pd.DataFrame":
+        return self.to_pandas()
+
+    def to_pandas(self) -> "pd.DataFrame":
         """Create the context windows and return a DataFrame."""
         if pd is None:
             raise ImportError(

python/lancedb/embeddings/__init__.py

@@ -11,14 +11,10 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 
-
+from .base import EmbeddingFunction, EmbeddingFunctionConfig, TextEmbeddingFunction
-from .functions import (
+from .cohere import CohereEmbeddingFunction
-    EmbeddingFunction,
+from .open_clip import OpenClipEmbeddings
-    EmbeddingFunctionConfig,
+from .openai import OpenAIEmbeddings
-    EmbeddingFunctionRegistry,
+from .registry import EmbeddingFunctionRegistry, get_registry
-    OpenAIEmbeddings,
+from .sentence_transformers import SentenceTransformerEmbeddings
-    OpenClipEmbeddings,
-    SentenceTransformerEmbeddings,
-    TextEmbeddingFunction,
-)
 from .utils import with_embeddings

python/lancedb/embeddings/base.py

@@ -0,0 +1,138 @@
+import importlib
+from abc import ABC, abstractmethod
+from typing import List, Union
+
+import numpy as np
+import pyarrow as pa
+from pydantic import BaseModel, Field, PrivateAttr
+
+from .utils import TEXT
+
+
+class EmbeddingFunction(BaseModel, ABC):
+    """
+    An ABC for embedding functions.
+
+    All concrete embedding functions must implement the following:
+    1. compute_query_embeddings() which takes a query and returns a list of embeddings
+    2. get_source_embeddings() which returns a list of embeddings for the source column
+    For text data, the two will be the same. For multi-modal data, the source column
+    might be images and the vector column might be text.
+    3. ndims method which returns the number of dimensions of the vector column
+    """
+
+    _ndims: int = PrivateAttr()
+
+    @classmethod
+    def create(cls, **kwargs):
+        """
+        Create an instance of the embedding function
+        """
+        return cls(**kwargs)
+
+    @abstractmethod
+    def compute_query_embeddings(self, *args, **kwargs) -> List[np.array]:
+        """
+        Compute the embeddings for a given user query
+        """
+        pass
+
+    @abstractmethod
+    def compute_source_embeddings(self, *args, **kwargs) -> List[np.array]:
+        """
+        Compute the embeddings for the source column in the database
+        """
+        pass
+
+    def sanitize_input(self, texts: TEXT) -> Union[List[str], np.ndarray]:
+        """
+        Sanitize the input to the embedding function.
+        """
+        if isinstance(texts, str):
+            texts = [texts]
+        elif isinstance(texts, pa.Array):
+            texts = texts.to_pylist()
+        elif isinstance(texts, pa.ChunkedArray):
+            texts = texts.combine_chunks().to_pylist()
+        return texts
+
+    @classmethod
+    def safe_import(cls, module: str, mitigation=None):
+        """
+        Import the specified module. If the module is not installed,
+        raise an ImportError with a helpful message.
+
+        Parameters
+        ----------
+        module : str
+            The name of the module to import
+        mitigation : Optional[str]
+            The package(s) to install to mitigate the error.
+            If not provided then the module name will be used.
+        """
+        try:
+            return importlib.import_module(module)
+        except ImportError:
+            raise ImportError(f"Please install {mitigation or module}")
+
+    def safe_model_dump(self):
+        from ..pydantic import PYDANTIC_VERSION
+
+        if PYDANTIC_VERSION.major < 2:
+            return dict(self)
+        return self.model_dump()
+
+    @abstractmethod
+    def ndims(self):
+        """
+        Return the dimensions of the vector column
+        """
+        pass
+
+    def SourceField(self, **kwargs):
+        """
+        Creates a pydantic Field that can automatically annotate
+        the source column for this embedding function
+        """
+        return Field(json_schema_extra={"source_column_for": self}, **kwargs)
+
+    def VectorField(self, **kwargs):
+        """
+        Creates a pydantic Field that can automatically annotate
+        the target vector column for this embedding function
+        """
+        return Field(json_schema_extra={"vector_column_for": self}, **kwargs)
+
+
+class EmbeddingFunctionConfig(BaseModel):
+    """
+    This model encapsulates the configuration for a embedding function
+    in a lancedb table. It holds the embedding function, the source column,
+    and the vector column
+    """
+
+    vector_column: str
+    source_column: str
+    function: EmbeddingFunction
+
+
+class TextEmbeddingFunction(EmbeddingFunction):
+    """
+    A callable ABC for embedding functions that take text as input
+    """
+
+    def compute_query_embeddings(self, query: str, *args, **kwargs) -> List[np.array]:
+        return self.compute_source_embeddings(query, *args, **kwargs)
+
+    def compute_source_embeddings(self, texts: TEXT, *args, **kwargs) -> List[np.array]:
+        texts = self.sanitize_input(texts)
+        return self.generate_embeddings(texts)
+
+    @abstractmethod
+    def generate_embeddings(
+        self, texts: Union[List[str], np.ndarray]
+    ) -> List[np.array]:
+        """
+        Generate the embeddings for the given texts
+        """
+        pass

python/lancedb/embeddings/cohere.py

@@ -0,0 +1,87 @@
+#  Copyright (c) 2023. LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import os
+from typing import ClassVar, List, Union
+
+import numpy as np
+
+from .base import TextEmbeddingFunction
+from .registry import register
+from .utils import api_key_not_found_help
+
+
+@register("cohere")
+class CohereEmbeddingFunction(TextEmbeddingFunction):
+    """
+    An embedding function that uses the Cohere API
+
+    https://docs.cohere.com/docs/multilingual-language-models
+
+    Parameters
+    ----------
+    name: str, default "embed-multilingual-v2.0"
+        The name of the model to use. See the Cohere documentation for a list of available models.
+
+    Examples
+    --------
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import EmbeddingFunctionRegistry
+
+    cohere = EmbeddingFunctionRegistry.get_instance().get("cohere").create(name="embed-multilingual-v2.0")
+
+    class TextModel(LanceModel):
+        text: str = cohere.SourceField()
+        vector: Vector(cohere.ndims()) =  cohere.VectorField()
+
+    data = [ { "text": "hello world" },
+            { "text": "goodbye world" }]
+
+    db = lancedb.connect("~/.lancedb")
+    tbl = db.create_table("test", schema=TextModel, mode="overwrite")
+
+    tbl.add(data)
+
+    """
+
+    name: str = "embed-multilingual-v2.0"
+    client: ClassVar = None
+
+    def ndims(self):
+        # TODO: fix hardcoding
+        return 768
+
+    def generate_embeddings(
+        self, texts: Union[List[str], np.ndarray]
+    ) -> List[np.array]:
+        """
+        Get the embeddings for the given texts
+
+        Parameters
+        ----------
+        texts: list[str] or np.ndarray (of str)
+            The texts to embed
+        """
+        # TODO retry, rate limit, token limit
+        self._init_client()
+        rs = CohereEmbeddingFunction.client.embed(texts=texts, model=self.name)
+
+        return [emb for emb in rs.embeddings]
+
+    def _init_client(self):
+        cohere = self.safe_import("cohere")
+        if CohereEmbeddingFunction.client is None:
+            if os.environ.get("COHERE_API_KEY") is None:
+                api_key_not_found_help("cohere")
+            CohereEmbeddingFunction.client = cohere.Client(os.environ["COHERE_API_KEY"])

python/lancedb/embeddings/functions.py

@@ -1,578 +0,0 @@
-#  Copyright (c) 2023. LanceDB Developers
-#
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-import concurrent.futures
-import importlib
-import io
-import json
-import os
-import socket
-import urllib.error
-import urllib.parse as urlparse
-import urllib.request
-from abc import ABC, abstractmethod
-from typing import Dict, List, Optional, Union
-
-import numpy as np
-import pyarrow as pa
-from cachetools import cached
-from pydantic import BaseModel, Field, PrivateAttr
-from tqdm import tqdm
-
-
-class EmbeddingFunctionRegistry:
-    """
-    This is a singleton class used to register embedding functions
-    and fetch them by name. It also handles serializing and deserializing.
-    You can implement your own embedding function by subclassing EmbeddingFunction
-    or TextEmbeddingFunction and registering it with the registry.
-
-    Examples
-    --------
-    >>> registry = EmbeddingFunctionRegistry.get_instance()
-    >>> @registry.register("my-embedding-function")
-    ... class MyEmbeddingFunction(EmbeddingFunction):
-    ...     def ndims(self) -> int:
-    ...         return 128
-    ...
-    ...     def compute_query_embeddings(self, query: str, *args, **kwargs) -> List[np.array]:
-    ...         return self.compute_source_embeddings(query, *args, **kwargs)
-    ...
-    ...     def compute_source_embeddings(self, texts: TEXT, *args, **kwargs) -> List[np.array]:
-    ...         return [np.random.rand(self.ndims()) for _ in range(len(texts))]
-    ...
-    >>> registry.get("my-embedding-function")
-    <class 'lancedb.embeddings.functions.MyEmbeddingFunction'>
-    """
-
-    @classmethod
-    def get_instance(cls):
-        return __REGISTRY__
-
-    def __init__(self):
-        self._functions = {}
-
-    def register(self, alias: str = None):
-        """
-        This creates a decorator that can be used to register
-        an EmbeddingFunction.
-
-        Parameters
-        ----------
-        alias : Optional[str]
-            a human friendly name for the embedding function. If not
-            provided, the class name will be used.
-        """
-
-        # This is a decorator for a class that inherits from BaseModel
-        # It adds the class to the registry
-        def decorator(cls):
-            if not issubclass(cls, EmbeddingFunction):
-                raise TypeError("Must be a subclass of EmbeddingFunction")
-            if cls.__name__ in self._functions:
-                raise KeyError(f"{cls.__name__} was already registered")
-            key = alias or cls.__name__
-            self._functions[key] = cls
-            cls.__embedding_function_registry_alias__ = alias
-            return cls
-
-        return decorator
-
-    def reset(self):
-        """
-        Reset the registry to its initial state
-        """
-        self._functions = {}
-
-    def get(self, name: str):
-        """
-        Fetch an embedding function class by name
-
-        Parameters
-        ----------
-        name : str
-            The name of the embedding function to fetch
-            Either the alias or the class name if no alias was provided
-            during registration
-        """
-        return self._functions[name]
-
-    def parse_functions(
-        self, metadata: Optional[Dict[bytes, bytes]]
-    ) -> Dict[str, "EmbeddingFunctionConfig"]:
-        """
-        Parse the metadata from an arrow table and
-        return a mapping of the vector column to the
-        embedding function and source column
-
-        Parameters
-        ----------
-        metadata : Optional[Dict[bytes, bytes]]
-            The metadata from an arrow table. Note that
-            the keys and values are bytes (pyarrow api)
-
-        Returns
-        -------
-        functions : dict
-            A mapping of vector column name to embedding function.
-            An empty dict is returned if input is None or does not
-            contain b"embedding_functions".
-        """
-        if metadata is None or b"embedding_functions" not in metadata:
-            return {}
-        serialized = metadata[b"embedding_functions"]
-        raw_list = json.loads(serialized.decode("utf-8"))
-        return {
-            obj["vector_column"]: EmbeddingFunctionConfig(
-                vector_column=obj["vector_column"],
-                source_column=obj["source_column"],
-                function=self.get(obj["name"])(**obj["model"]),
-            )
-            for obj in raw_list
-        }
-
-    def function_to_metadata(self, conf: "EmbeddingFunctionConfig"):
-        """
-        Convert the given embedding function and source / vector column configs
-        into a config dictionary that can be serialized into arrow metadata
-        """
-        func = conf.function
-        name = getattr(
-            func, "__embedding_function_registry_alias__", func.__class__.__name__
-        )
-        json_data = func.safe_model_dump()
-        return {
-            "name": name,
-            "model": json_data,
-            "source_column": conf.source_column,
-            "vector_column": conf.vector_column,
-        }
-
-    def get_table_metadata(self, func_list):
-        """
-        Convert a list of embedding functions and source / vector configs
-        into a config dictionary that can be serialized into arrow metadata
-        """
-        if func_list is None or len(func_list) == 0:
-            return None
-        json_data = [self.function_to_metadata(func) for func in func_list]
-        # Note that metadata dictionary values must be bytes
-        # so we need to json dump then utf8 encode
-        metadata = json.dumps(json_data, indent=2).encode("utf-8")
-        return {"embedding_functions": metadata}
-
-
-# Global instance
-__REGISTRY__ = EmbeddingFunctionRegistry()
-
-
-TEXT = Union[str, List[str], pa.Array, pa.ChunkedArray, np.ndarray]
-IMAGES = Union[
-    str, bytes, List[str], List[bytes], pa.Array, pa.ChunkedArray, np.ndarray
-]
-
-
-class EmbeddingFunction(BaseModel, ABC):
-    """
-    An ABC for embedding functions.
-
-    All concrete embedding functions must implement the following:
-    1. compute_query_embeddings() which takes a query and returns a list of embeddings
-    2. get_source_embeddings() which returns a list of embeddings for the source column
-    For text data, the two will be the same. For multi-modal data, the source column
-    might be images and the vector column might be text.
-    3. ndims method which returns the number of dimensions of the vector column
-    """
-
-    _ndims: int = PrivateAttr()
-
-    @classmethod
-    def create(cls, **kwargs):
-        """
-        Create an instance of the embedding function
-        """
-        return cls(**kwargs)
-
-    @abstractmethod
-    def compute_query_embeddings(self, *args, **kwargs) -> List[np.array]:
-        """
-        Compute the embeddings for a given user query
-        """
-        pass
-
-    @abstractmethod
-    def compute_source_embeddings(self, *args, **kwargs) -> List[np.array]:
-        """
-        Compute the embeddings for the source column in the database
-        """
-        pass
-
-    def sanitize_input(self, texts: TEXT) -> Union[List[str], np.ndarray]:
-        """
-        Sanitize the input to the embedding function.
-        """
-        if isinstance(texts, str):
-            texts = [texts]
-        elif isinstance(texts, pa.Array):
-            texts = texts.to_pylist()
-        elif isinstance(texts, pa.ChunkedArray):
-            texts = texts.combine_chunks().to_pylist()
-        return texts
-
-    @classmethod
-    def safe_import(cls, module: str, mitigation=None):
-        """
-        Import the specified module. If the module is not installed,
-        raise an ImportError with a helpful message.
-
-        Parameters
-        ----------
-        module : str
-            The name of the module to import
-        mitigation : Optional[str]
-            The package(s) to install to mitigate the error.
-            If not provided then the module name will be used.
-        """
-        try:
-            return importlib.import_module(module)
-        except ImportError:
-            raise ImportError(f"Please install {mitigation or module}")
-
-    def safe_model_dump(self):
-        from ..pydantic import PYDANTIC_VERSION
-
-        if PYDANTIC_VERSION.major < 2:
-            return dict(self)
-        return self.model_dump()
-
-    @abstractmethod
-    def ndims(self):
-        """
-        Return the dimensions of the vector column
-        """
-        pass
-
-    def SourceField(self, **kwargs):
-        """
-        Creates a pydantic Field that can automatically annotate
-        the source column for this embedding function
-        """
-        return Field(json_schema_extra={"source_column_for": self}, **kwargs)
-
-    def VectorField(self, **kwargs):
-        """
-        Creates a pydantic Field that can automatically annotate
-        the target vector column for this embedding function
-        """
-        return Field(json_schema_extra={"vector_column_for": self}, **kwargs)
-
-
-class EmbeddingFunctionConfig(BaseModel):
-    """
-    This model encapsulates the configuration for a embedding function
-    in a lancedb table. It holds the embedding function, the source column,
-    and the vector column
-    """
-
-    vector_column: str
-    source_column: str
-    function: EmbeddingFunction
-
-
-class TextEmbeddingFunction(EmbeddingFunction):
-    """
-    A callable ABC for embedding functions that take text as input
-    """
-
-    def compute_query_embeddings(self, query: str, *args, **kwargs) -> List[np.array]:
-        return self.compute_source_embeddings(query, *args, **kwargs)
-
-    def compute_source_embeddings(self, texts: TEXT, *args, **kwargs) -> List[np.array]:
-        texts = self.sanitize_input(texts)
-        return self.generate_embeddings(texts)
-
-    @abstractmethod
-    def generate_embeddings(
-        self, texts: Union[List[str], np.ndarray]
-    ) -> List[np.array]:
-        """
-        Generate the embeddings for the given texts
-        """
-        pass
-
-
-# @EmbeddingFunctionRegistry.get_instance().register(name) doesn't work in 3.8
-register = lambda name: EmbeddingFunctionRegistry.get_instance().register(name)
-
-
-@register("sentence-transformers")
-class SentenceTransformerEmbeddings(TextEmbeddingFunction):
-    """
-    An embedding function that uses the sentence-transformers library
-
-    https://huggingface.co/sentence-transformers
-    """
-
-    name: str = "all-MiniLM-L6-v2"
-    device: str = "cpu"
-    normalize: bool = True
-
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-        self._ndims = None
-
-    @property
-    def embedding_model(self):
-        """
-        Get the sentence-transformers embedding model specified by the
-        name and device. This is cached so that the model is only loaded
-        once per process.
-        """
-        return self.__class__.get_embedding_model(self.name, self.device)
-
-    def ndims(self):
-        if self._ndims is None:
-            self._ndims = len(self.generate_embeddings("foo")[0])
-        return self._ndims
-
-    def generate_embeddings(
-        self, texts: Union[List[str], np.ndarray]
-    ) -> List[np.array]:
-        """
-        Get the embeddings for the given texts
-
-        Parameters
-        ----------
-        texts: list[str] or np.ndarray (of str)
-            The texts to embed
-        """
-        return self.embedding_model.encode(
-            list(texts),
-            convert_to_numpy=True,
-            normalize_embeddings=self.normalize,
-        ).tolist()
-
-    @classmethod
-    @cached(cache={})
-    def get_embedding_model(cls, name, device):
-        """
-        Get the sentence-transformers embedding model specified by the
-        name and device. This is cached so that the model is only loaded
-        once per process.
-
-        Parameters
-        ----------
-        name : str
-            The name of the model to load
-        device : str
-            The device to load the model on
-
-        TODO: use lru_cache instead with a reasonable/configurable maxsize
-        """
-        sentence_transformers = cls.safe_import(
-            "sentence_transformers", "sentence-transformers"
-        )
-        return sentence_transformers.SentenceTransformer(name, device=device)
-
-
-@register("openai")
-class OpenAIEmbeddings(TextEmbeddingFunction):
-    """
-    An embedding function that uses the OpenAI API
-
-    https://platform.openai.com/docs/guides/embeddings
-    """
-
-    name: str = "text-embedding-ada-002"
-
-    def ndims(self):
-        # TODO don't hardcode this
-        return 1536
-
-    def generate_embeddings(
-        self, texts: Union[List[str], np.ndarray]
-    ) -> List[np.array]:
-        """
-        Get the embeddings for the given texts
-
-        Parameters
-        ----------
-        texts: list[str] or np.ndarray (of str)
-            The texts to embed
-        """
-        # TODO retry, rate limit, token limit
-        openai = self.safe_import("openai")
-        rs = openai.Embedding.create(input=texts, model=self.name)["data"]
-        return [v["embedding"] for v in rs]
-
-
-@register("open-clip")
-class OpenClipEmbeddings(EmbeddingFunction):
-    """
-    An embedding function that uses the OpenClip API
-    For multi-modal text-to-image search
-
-    https://github.com/mlfoundations/open_clip
-    """
-
-    name: str = "ViT-B-32"
-    pretrained: str = "laion2b_s34b_b79k"
-    device: str = "cpu"
-    batch_size: int = 64
-    normalize: bool = True
-    _model = PrivateAttr()
-    _preprocess = PrivateAttr()
-    _tokenizer = PrivateAttr()
-
-    def __init__(self, *args, **kwargs):
-        super().__init__(*args, **kwargs)
-        open_clip = self.safe_import("open_clip", "open-clip")
-        model, _, preprocess = open_clip.create_model_and_transforms(
-            self.name, pretrained=self.pretrained
-        )
-        model.to(self.device)
-        self._model, self._preprocess = model, preprocess
-        self._tokenizer = open_clip.get_tokenizer(self.name)
-        self._ndims = None
-
-    def ndims(self):
-        if self._ndims is None:
-            self._ndims = self.generate_text_embeddings("foo").shape[0]
-        return self._ndims
-
-    def compute_query_embeddings(
-        self, query: Union[str, "PIL.Image.Image"], *args, **kwargs
-    ) -> List[np.ndarray]:
-        """
-        Compute the embeddings for a given user query
-
-        Parameters
-        ----------
-        query : Union[str, PIL.Image.Image]
-            The query to embed. A query can be either text or an image.
-        """
-        if isinstance(query, str):
-            return [self.generate_text_embeddings(query)]
-        else:
-            PIL = self.safe_import("PIL", "pillow")
-            if isinstance(query, PIL.Image.Image):
-                return [self.generate_image_embedding(query)]
-            else:
-                raise TypeError("OpenClip supports str or PIL Image as query")
-
-    def generate_text_embeddings(self, text: str) -> np.ndarray:
-        torch = self.safe_import("torch")
-        text = self.sanitize_input(text)
-        text = self._tokenizer(text)
-        text.to(self.device)
-        with torch.no_grad():
-            text_features = self._model.encode_text(text.to(self.device))
-            if self.normalize:
-                text_features /= text_features.norm(dim=-1, keepdim=True)
-            return text_features.cpu().numpy().squeeze()
-
-    def sanitize_input(self, images: IMAGES) -> Union[List[bytes], np.ndarray]:
-        """
-        Sanitize the input to the embedding function.
-        """
-        if isinstance(images, (str, bytes)):
-            images = [images]
-        elif isinstance(images, pa.Array):
-            images = images.to_pylist()
-        elif isinstance(images, pa.ChunkedArray):
-            images = images.combine_chunks().to_pylist()
-        return images
-
-    def compute_source_embeddings(
-        self, images: IMAGES, *args, **kwargs
-    ) -> List[np.array]:
-        """
-        Get the embeddings for the given images
-        """
-        images = self.sanitize_input(images)
-        embeddings = []
-        for i in range(0, len(images), self.batch_size):
-            j = min(i + self.batch_size, len(images))
-            batch = images[i:j]
-            embeddings.extend(self._parallel_get(batch))
-        return embeddings
-
-    def _parallel_get(self, images: Union[List[str], List[bytes]]) -> List[np.ndarray]:
-        """
-        Issue concurrent requests to retrieve the image data
-        """
-        with concurrent.futures.ThreadPoolExecutor() as executor:
-            futures = [
-                executor.submit(self.generate_image_embedding, image)
-                for image in images
-            ]
-            return [future.result() for future in tqdm(futures)]
-
-    def generate_image_embedding(
-        self, image: Union[str, bytes, "PIL.Image.Image"]
-    ) -> np.ndarray:
-        """
-        Generate the embedding for a single image
-
-        Parameters
-        ----------
-        image : Union[str, bytes, PIL.Image.Image]
-            The image to embed. If the image is a str, it is treated as a uri.
-            If the image is bytes, it is treated as the raw image bytes.
-        """
-        torch = self.safe_import("torch")
-        # TODO handle retry and errors for https
-        image = self._to_pil(image)
-        image = self._preprocess(image).unsqueeze(0)
-        with torch.no_grad():
-            return self._encode_and_normalize_image(image)
-
-    def _to_pil(self, image: Union[str, bytes]):
-        PIL = self.safe_import("PIL", "pillow")
-        if isinstance(image, bytes):
-            return PIL.Image.open(io.BytesIO(image))
-        if isinstance(image, PIL.Image.Image):
-            return image
-        elif isinstance(image, str):
-            parsed = urlparse.urlparse(image)
-            # TODO handle drive letter on windows.
-            if parsed.scheme == "file":
-                return PIL.Image.open(parsed.path)
-            elif parsed.scheme == "":
-                return PIL.Image.open(image if os.name == "nt" else parsed.path)
-            elif parsed.scheme.startswith("http"):
-                return PIL.Image.open(io.BytesIO(url_retrieve(image)))
-            else:
-                raise NotImplementedError("Only local and http(s) urls are supported")
-
-    def _encode_and_normalize_image(self, image_tensor: "torch.Tensor"):
-        """
-        encode a single image tensor and optionally normalize the output
-        """
-        image_features = self._model.encode_image(image_tensor.to(self.device))
-        if self.normalize:
-            image_features /= image_features.norm(dim=-1, keepdim=True)
-        return image_features.cpu().numpy().squeeze()
-
-
-def url_retrieve(url: str):
-    """
-    Parameters
-    ----------
-    url: str
-        URL to download from
-    """
-    try:
-        with urllib.request.urlopen(url) as conn:
-            return conn.read()
-    except (socket.gaierror, urllib.error.URLError) as err:
-        raise ConnectionError("could not download {} due to {}".format(url, err))

python/lancedb/embeddings/open_clip.py

@@ -0,0 +1,163 @@
+import concurrent.futures
+import io
+import os
+import urllib.parse as urlparse
+from typing import List, Union
+
+import numpy as np
+import pyarrow as pa
+from pydantic import PrivateAttr
+from tqdm import tqdm
+
+from .base import EmbeddingFunction
+from .registry import register
+from .utils import IMAGES, url_retrieve
+
+
+@register("open-clip")
+class OpenClipEmbeddings(EmbeddingFunction):
+    """
+    An embedding function that uses the OpenClip API
+    For multi-modal text-to-image search
+
+    https://github.com/mlfoundations/open_clip
+    """
+
+    name: str = "ViT-B-32"
+    pretrained: str = "laion2b_s34b_b79k"
+    device: str = "cpu"
+    batch_size: int = 64
+    normalize: bool = True
+    _model = PrivateAttr()
+    _preprocess = PrivateAttr()
+    _tokenizer = PrivateAttr()
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        open_clip = self.safe_import("open_clip", "open-clip")
+        model, _, preprocess = open_clip.create_model_and_transforms(
+            self.name, pretrained=self.pretrained
+        )
+        model.to(self.device)
+        self._model, self._preprocess = model, preprocess
+        self._tokenizer = open_clip.get_tokenizer(self.name)
+        self._ndims = None
+
+    def ndims(self):
+        if self._ndims is None:
+            self._ndims = self.generate_text_embeddings("foo").shape[0]
+        return self._ndims
+
+    def compute_query_embeddings(
+        self, query: Union[str, "PIL.Image.Image"], *args, **kwargs
+    ) -> List[np.ndarray]:
+        """
+        Compute the embeddings for a given user query
+
+        Parameters
+        ----------
+        query : Union[str, PIL.Image.Image]
+            The query to embed. A query can be either text or an image.
+        """
+        if isinstance(query, str):
+            return [self.generate_text_embeddings(query)]
+        else:
+            PIL = self.safe_import("PIL", "pillow")
+            if isinstance(query, PIL.Image.Image):
+                return [self.generate_image_embedding(query)]
+            else:
+                raise TypeError("OpenClip supports str or PIL Image as query")
+
+    def generate_text_embeddings(self, text: str) -> np.ndarray:
+        torch = self.safe_import("torch")
+        text = self.sanitize_input(text)
+        text = self._tokenizer(text)
+        text.to(self.device)
+        with torch.no_grad():
+            text_features = self._model.encode_text(text.to(self.device))
+            if self.normalize:
+                text_features /= text_features.norm(dim=-1, keepdim=True)
+            return text_features.cpu().numpy().squeeze()
+
+    def sanitize_input(self, images: IMAGES) -> Union[List[bytes], np.ndarray]:
+        """
+        Sanitize the input to the embedding function.
+        """
+        if isinstance(images, (str, bytes)):
+            images = [images]
+        elif isinstance(images, pa.Array):
+            images = images.to_pylist()
+        elif isinstance(images, pa.ChunkedArray):
+            images = images.combine_chunks().to_pylist()
+        return images
+
+    def compute_source_embeddings(
+        self, images: IMAGES, *args, **kwargs
+    ) -> List[np.array]:
+        """
+        Get the embeddings for the given images
+        """
+        images = self.sanitize_input(images)
+        embeddings = []
+        for i in range(0, len(images), self.batch_size):
+            j = min(i + self.batch_size, len(images))
+            batch = images[i:j]
+            embeddings.extend(self._parallel_get(batch))
+        return embeddings
+
+    def _parallel_get(self, images: Union[List[str], List[bytes]]) -> List[np.ndarray]:
+        """
+        Issue concurrent requests to retrieve the image data
+        """
+        with concurrent.futures.ThreadPoolExecutor() as executor:
+            futures = [
+                executor.submit(self.generate_image_embedding, image)
+                for image in images
+            ]
+            return [future.result() for future in tqdm(futures)]
+
+    def generate_image_embedding(
+        self, image: Union[str, bytes, "PIL.Image.Image"]
+    ) -> np.ndarray:
+        """
+        Generate the embedding for a single image
+
+        Parameters
+        ----------
+        image : Union[str, bytes, PIL.Image.Image]
+            The image to embed. If the image is a str, it is treated as a uri.
+            If the image is bytes, it is treated as the raw image bytes.
+        """
+        torch = self.safe_import("torch")
+        # TODO handle retry and errors for https
+        image = self._to_pil(image)
+        image = self._preprocess(image).unsqueeze(0)
+        with torch.no_grad():
+            return self._encode_and_normalize_image(image)
+
+    def _to_pil(self, image: Union[str, bytes]):
+        PIL = self.safe_import("PIL", "pillow")
+        if isinstance(image, bytes):
+            return PIL.Image.open(io.BytesIO(image))
+        if isinstance(image, PIL.Image.Image):
+            return image
+        elif isinstance(image, str):
+            parsed = urlparse.urlparse(image)
+            # TODO handle drive letter on windows.
+            if parsed.scheme == "file":
+                return PIL.Image.open(parsed.path)
+            elif parsed.scheme == "":
+                return PIL.Image.open(image if os.name == "nt" else parsed.path)
+            elif parsed.scheme.startswith("http"):
+                return PIL.Image.open(io.BytesIO(url_retrieve(image)))
+            else:
+                raise NotImplementedError("Only local and http(s) urls are supported")
+
+    def _encode_and_normalize_image(self, image_tensor: "torch.Tensor"):
+        """
+        encode a single image tensor and optionally normalize the output
+        """
+        image_features = self._model.encode_image(image_tensor.to(self.device))
+        if self.normalize:
+            image_features /= image_features.norm(dim=-1, keepdim=True)
+        return image_features.cpu().numpy().squeeze()

python/lancedb/embeddings/openai.py

@@ -0,0 +1,37 @@
+from typing import List, Union
+
+import numpy as np
+
+from .base import TextEmbeddingFunction
+from .registry import register
+
+
+@register("openai")
+class OpenAIEmbeddings(TextEmbeddingFunction):
+    """
+    An embedding function that uses the OpenAI API
+
+    https://platform.openai.com/docs/guides/embeddings
+    """
+
+    name: str = "text-embedding-ada-002"
+
+    def ndims(self):
+        # TODO don't hardcode this
+        return 1536
+
+    def generate_embeddings(
+        self, texts: Union[List[str], np.ndarray]
+    ) -> List[np.array]:
+        """
+        Get the embeddings for the given texts
+
+        Parameters
+        ----------
+        texts: list[str] or np.ndarray (of str)
+            The texts to embed
+        """
+        # TODO retry, rate limit, token limit
+        openai = self.safe_import("openai")
+        rs = openai.Embedding.create(input=texts, model=self.name)["data"]
+        return [v["embedding"] for v in rs]

python/lancedb/embeddings/registry.py

@@ -0,0 +1,186 @@
+#  Copyright (c) 2023. LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+import json
+from typing import Dict, Optional
+
+from .base import EmbeddingFunction, EmbeddingFunctionConfig
+
+
+class EmbeddingFunctionRegistry:
+    """
+    This is a singleton class used to register embedding functions
+    and fetch them by name. It also handles serializing and deserializing.
+    You can implement your own embedding function by subclassing EmbeddingFunction
+    or TextEmbeddingFunction and registering it with the registry.
+
+    NOTE: Here TEXT is a type alias for Union[str, List[str], pa.Array, pa.ChunkedArray, np.ndarray]
+    Examples
+    --------
+    >>> registry = EmbeddingFunctionRegistry.get_instance()
+    >>> @registry.register("my-embedding-function")
+    ... class MyEmbeddingFunction(EmbeddingFunction):
+    ...     def ndims(self) -> int:
+    ...         return 128
+    ...
+    ...     def compute_query_embeddings(self, query: str, *args, **kwargs):
+    ...         return self.compute_source_embeddings(query, *args, **kwargs)
+    ...
+    ...     def compute_source_embeddings(self, texts, *args, **kwargs):
+    ...         return [np.random.rand(self.ndims()) for _ in range(len(texts))]
+    ...
+    >>> registry.get("my-embedding-function")
+    <class 'lancedb.embeddings.registry.MyEmbeddingFunction'>
+    """
+
+    @classmethod
+    def get_instance(cls):
+        return __REGISTRY__
+
+    def __init__(self):
+        self._functions = {}
+
+    def register(self, alias: str = None):
+        """
+        This creates a decorator that can be used to register
+        an EmbeddingFunction.
+
+        Parameters
+        ----------
+        alias : Optional[str]
+            a human friendly name for the embedding function. If not
+            provided, the class name will be used.
+        """
+
+        # This is a decorator for a class that inherits from BaseModel
+        # It adds the class to the registry
+        def decorator(cls):
+            if not issubclass(cls, EmbeddingFunction):
+                raise TypeError("Must be a subclass of EmbeddingFunction")
+            if cls.__name__ in self._functions:
+                raise KeyError(f"{cls.__name__} was already registered")
+            key = alias or cls.__name__
+            self._functions[key] = cls
+            cls.__embedding_function_registry_alias__ = alias
+            return cls
+
+        return decorator
+
+    def reset(self):
+        """
+        Reset the registry to its initial state
+        """
+        self._functions = {}
+
+    def get(self, name: str):
+        """
+        Fetch an embedding function class by name
+
+        Parameters
+        ----------
+        name : str
+            The name of the embedding function to fetch
+            Either the alias or the class name if no alias was provided
+            during registration
+        """
+        return self._functions[name]
+
+    def parse_functions(
+        self, metadata: Optional[Dict[bytes, bytes]]
+    ) -> Dict[str, "EmbeddingFunctionConfig"]:
+        """
+        Parse the metadata from an arrow table and
+        return a mapping of the vector column to the
+        embedding function and source column
+
+        Parameters
+        ----------
+        metadata : Optional[Dict[bytes, bytes]]
+            The metadata from an arrow table. Note that
+            the keys and values are bytes (pyarrow api)
+
+        Returns
+        -------
+        functions : dict
+            A mapping of vector column name to embedding function.
+            An empty dict is returned if input is None or does not
+            contain b"embedding_functions".
+        """
+        if metadata is None or b"embedding_functions" not in metadata:
+            return {}
+        serialized = metadata[b"embedding_functions"]
+        raw_list = json.loads(serialized.decode("utf-8"))
+        return {
+            obj["vector_column"]: EmbeddingFunctionConfig(
+                vector_column=obj["vector_column"],
+                source_column=obj["source_column"],
+                function=self.get(obj["name"])(**obj["model"]),
+            )
+            for obj in raw_list
+        }
+
+    def function_to_metadata(self, conf: "EmbeddingFunctionConfig"):
+        """
+        Convert the given embedding function and source / vector column configs
+        into a config dictionary that can be serialized into arrow metadata
+        """
+        func = conf.function
+        name = getattr(
+            func, "__embedding_function_registry_alias__", func.__class__.__name__
+        )
+        json_data = func.safe_model_dump()
+        return {
+            "name": name,
+            "model": json_data,
+            "source_column": conf.source_column,
+            "vector_column": conf.vector_column,
+        }
+
+    def get_table_metadata(self, func_list):
+        """
+        Convert a list of embedding functions and source / vector configs
+        into a config dictionary that can be serialized into arrow metadata
+        """
+        if func_list is None or len(func_list) == 0:
+            return None
+        json_data = [self.function_to_metadata(func) for func in func_list]
+        # Note that metadata dictionary values must be bytes
+        # so we need to json dump then utf8 encode
+        metadata = json.dumps(json_data, indent=2).encode("utf-8")
+        return {"embedding_functions": metadata}
+
+
+# Global instance
+__REGISTRY__ = EmbeddingFunctionRegistry()
+
+
+# @EmbeddingFunctionRegistry.get_instance().register(name) doesn't work in 3.8
+register = lambda name: EmbeddingFunctionRegistry.get_instance().register(name)
+
+
+def get_registry():
+    """
+    Utility function to get the global instance of the registry
+
+    Returns
+    -------
+    EmbeddingFunctionRegistry
+        The global registry instance
+
+    Examples
+    --------
+    from lancedb.embeddings import get_registry
+
+    registry = get_registry()
+    openai = registry.get("openai").create()
+    """
+    return __REGISTRY__.get_instance()

python/lancedb/embeddings/sentence_transformers.py

@@ -0,0 +1,77 @@
+from typing import List, Union
+
+import numpy as np
+from cachetools import cached
+
+from .base import TextEmbeddingFunction
+from .registry import register
+
+
+@register("sentence-transformers")
+class SentenceTransformerEmbeddings(TextEmbeddingFunction):
+    """
+    An embedding function that uses the sentence-transformers library
+
+    https://huggingface.co/sentence-transformers
+    """
+
+    name: str = "all-MiniLM-L6-v2"
+    device: str = "cpu"
+    normalize: bool = True
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self._ndims = None
+
+    @property
+    def embedding_model(self):
+        """
+        Get the sentence-transformers embedding model specified by the
+        name and device. This is cached so that the model is only loaded
+        once per process.
+        """
+        return self.__class__.get_embedding_model(self.name, self.device)
+
+    def ndims(self):
+        if self._ndims is None:
+            self._ndims = len(self.generate_embeddings("foo")[0])
+        return self._ndims
+
+    def generate_embeddings(
+        self, texts: Union[List[str], np.ndarray]
+    ) -> List[np.array]:
+        """
+        Get the embeddings for the given texts
+
+        Parameters
+        ----------
+        texts: list[str] or np.ndarray (of str)
+            The texts to embed
+        """
+        return self.embedding_model.encode(
+            list(texts),
+            convert_to_numpy=True,
+            normalize_embeddings=self.normalize,
+        ).tolist()
+
+    @classmethod
+    @cached(cache={})
+    def get_embedding_model(cls, name, device):
+        """
+        Get the sentence-transformers embedding model specified by the
+        name and device. This is cached so that the model is only loaded
+        once per process.
+
+        Parameters
+        ----------
+        name : str
+            The name of the model to load
+        device : str
+            The device to load the model on
+
+        TODO: use lru_cache instead with a reasonable/configurable maxsize
+        """
+        sentence_transformers = cls.safe_import(
+            "sentence_transformers", "sentence-transformers"
+        )
+        return sentence_transformers.SentenceTransformer(name, device=device)

python/lancedb/embeddings/utils.py

@@ -12,8 +12,10 @@
 #  limitations under the License.
 
 import math
+import socket
 import sys
-from typing import Callable, Union
+import urllib.error
+from typing import Callable, List, Union
 
 import numpy as np
 import pyarrow as pa
@@ -21,9 +23,15 @@ from lance.vector import vec_to_table
 from retry import retry
 
 from ..util import safe_import_pandas
+from ..utils.general import LOGGER
 
 pd = safe_import_pandas()
+
 DATA = Union[pa.Table, "pd.DataFrame"]
+TEXT = Union[str, List[str], pa.Array, pa.ChunkedArray, np.ndarray]
+IMAGES = Union[
+    str, bytes, List[str], List[bytes], pa.Array, pa.ChunkedArray, np.ndarray
+]
 
 
 def with_embeddings(
@@ -152,3 +160,22 @@ class FunctionWrapper:
             yield from tqdm(_chunker(arr), total=math.ceil(length / self._batch_size))
         else:
             yield from _chunker(arr)
+
+
+def url_retrieve(url: str):
+    """
+    Parameters
+    ----------
+    url: str
+        URL to download from
+    """
+    try:
+        with urllib.request.urlopen(url) as conn:
+            return conn.read()
+    except (socket.gaierror, urllib.error.URLError) as err:
+        raise ConnectionError("could not download {} due to {}".format(url, err))
+
+
+def api_key_not_found_help(provider):
+    LOGGER.error(f"Could not find API key for {provider}.")
+    raise ValueError(f"Please set the {provider.upper()}_API_KEY environment variable.")

python/lancedb/pydantic.py

@@ -19,6 +19,7 @@ import inspect
 import sys
 import types
 from abc import ABC, abstractmethod
+from datetime import date, datetime
 from typing import Any, Callable, Dict, Generator, List, Type, Union, _GenericAlias
 
 import numpy as np
@@ -159,6 +160,10 @@ def _py_type_to_arrow_type(py_type: Type[Any]) -> pa.DataType:
         return pa.bool_()
     elif py_type == bytes:
         return pa.binary()
+    elif py_type == date:
+        return pa.date32()
+    elif py_type == datetime:
+        return pa.timestamp("us")
     raise TypeError(
         f"Converting Pydantic type to Arrow Type: unsupported type {py_type}"
     )
@@ -322,7 +327,12 @@ class LanceModel(pydantic.BaseModel):
         for vec, func in vec_and_function:
             for source, field_info in cls.safe_get_fields().items():
                 src_func = get_extras(field_info, "source_column_for")
-                if src_func == func:
+                if src_func is func:
+                    # note we can't use == here since the function is a pydantic
+                    # model so two instances of the same function are ==, so if you
+                    # have multiple vector columns from multiple sources, both will
+                    # be mapped to the same source column
+                    # GH594
                     configs.append(
                         EmbeddingFunctionConfig(
                             source_column=source, vector_column=vec, function=func

python/lancedb/query.py

@@ -16,10 +16,12 @@ from __future__ import annotations
 from abc import ABC, abstractmethod
 from typing import List, Literal, Optional, Type, Union
 
+import deprecation
 import numpy as np
 import pyarrow as pa
 import pydantic
 
+from . import __version__
 from .common import VECTOR_COLUMN_NAME
 from .pydantic import LanceModel
 from .util import safe_import_pandas
@@ -127,7 +129,24 @@ class LanceQueryBuilder(ABC):
         self._columns = None
         self._where = None
 
+    @deprecation.deprecated(
+        deprecated_in="0.3.1",
+        removed_in="0.4.0",
+        current_version=__version__,
+        details="Use the bar function instead",
+    )
     def to_df(self) -> "pd.DataFrame":
+        """
+        Deprecated alias for `to_pandas()`. Please use `to_pandas()` instead.
+
+        Execute the query and return the results as a pandas DataFrame.
+        In addition to the selected columns, LanceDB also returns a vector
+        and also the "_distance" column which is the distance between the query
+        vector and the returned vector.
+        """
+        return self.to_pandas()
+
+    def to_pandas(self) -> "pd.DataFrame":
         """
         Execute the query and return the results as a pandas DataFrame.
         In addition to the selected columns, LanceDB also returns a vector
@@ -148,6 +167,16 @@ class LanceQueryBuilder(ABC):
         """
         raise NotImplementedError
 
+    def to_list(self) -> List[dict]:
+        """
+        Execute the query and return the results as a list of dictionaries.
+
+        Each list entry is a dictionary with the selected column names as keys,
+        or all table columns if `select` is not called. The vector and the "_distance"
+        fields are returned whether or not they're explicitly selected.
+        """
+        return self.to_arrow().to_pylist()
+
     def to_pydantic(self, model: Type[LanceModel]) -> List[LanceModel]:
         """Return the table as a list of pydantic models.
 
@@ -232,7 +261,7 @@ class LanceVectorQueryBuilder(LanceQueryBuilder):
     ...       .where("b < 10")
     ...       .select(["b"])
     ...       .limit(2)
-    ...       .to_df())
+    ...       .to_pandas())
        b      vector  _distance
     0  6  [0.4, 0.4]        0.0
     """

python/lancedb/remote/client.py

@@ -151,10 +151,15 @@ class RestfulLanceDBClient:
             return await deserialize(resp)
 
     @_check_not_closed
-    async def list_tables(self):
+    async def list_tables(self, limit: int, page_token: str):
         """List all tables in the database."""
-        json = await self.get("/v1/table/", {})
+        try:
+            json = await self.get(
+                "/v1/table/", {"limit": limit, "page_token": page_token}
+            )
             return json["tables"]
+        except StopAsyncIteration:
+            return []
 
     @_check_not_closed
     async def query(self, table_name: str, query: VectorQuery) -> VectorQueryResult:

python/lancedb/remote/db.py

@@ -13,7 +13,7 @@
 
 import asyncio
 import uuid
-from typing import List, Optional
+from typing import Iterator, Optional
 from urllib.parse import urlparse
 
 import pyarrow as pa
@@ -52,10 +52,27 @@ class RemoteDBConnection(DBConnection):
     def __repr__(self) -> str:
         return f"RemoveConnect(name={self.db_name})"
 
-    def table_names(self) -> List[str]:
+    def table_names(self, last_token: str, limit=10) -> Iterator[str]:
-        """List the names of all tables in the database."""
+        """List the names of all tables in the database.
-        result = self._loop.run_until_complete(self._client.list_tables())
+        Parameters
-        return result
+        ----------
+        last_token: str
+            The last token to start the new page.
+
+        Returns
+        -------
+        An iterator of table names.
+        """
+        while True:
+            result = self._loop.run_until_complete(
+                self._client.list_tables(limit, last_token)
+            )
+            if len(result) > 0:
+                last_token = result[len(result) - 1]
+            else:
+                break
+            for item in result:
+                yield result
 
     def open_table(self, name: str) -> Table:
         """Open a Lance Table in the database.
@@ -122,3 +139,8 @@ class RemoteDBConnection(DBConnection):
                 f"/v1/table/{name}/drop/",
             )
         )
+
+    async def close(self):
+        """Close the connection to the database."""
+        self._loop.close()
+        await self._client.close()

python/lancedb/remote/table.py

@@ -13,7 +13,7 @@
 
 import uuid
 from functools import cached_property
-from typing import Union
+from typing import Optional, Union
 
 import pyarrow as pa
 from lance import json_to_schema
@@ -62,6 +62,7 @@ class RemoteTable(Table):
         num_sub_vectors=96,
         vector_column_name: str = VECTOR_COLUMN_NAME,
         replace: bool = True,
+        accelerator: Optional[str] = None,
     ):
         raise NotImplementedError
 
@@ -104,4 +105,8 @@ class RemoteTable(Table):
         return self._conn._loop.run_until_complete(result).to_arrow()
 
     def delete(self, predicate: str):
-        raise NotImplementedError
+        """Delete rows from the table."""
+        payload = {"predicate": predicate}
+        self._conn._loop.run_until_complete(
+            self._conn._client.post(f"/v1/table/{self._name}/delete/", data=payload)
+        )

python/lancedb/table.py

@@ -16,6 +16,7 @@ from __future__ import annotations
 import inspect
 import os
 from abc import ABC, abstractmethod
+from datetime import timedelta
 from functools import cached_property
 from typing import Any, Iterable, List, Optional, Union
 
@@ -24,15 +25,15 @@ import numpy as np
 import pyarrow as pa
 import pyarrow.compute as pc
 from lance import LanceDataset
-from lance.dataset import ReaderLike
+from lance.dataset import CleanupStats, ReaderLike
 from lance.vector import vec_to_table
 
 from .common import DATA, VEC, VECTOR_COLUMN_NAME
-from .embeddings import EmbeddingFunctionRegistry
+from .embeddings import EmbeddingFunctionConfig, EmbeddingFunctionRegistry
-from .embeddings.functions import EmbeddingFunctionConfig
 from .pydantic import LanceModel
 from .query import LanceQueryBuilder, Query
 from .util import fs_from_uri, safe_import_pandas
+from .utils.events import register_event
 
 pd = safe_import_pandas()
 
@@ -136,7 +137,7 @@ class Table(ABC):
 
     Can query the table with [Table.search][lancedb.table.Table.search].
 
-    >>> table.search([0.4, 0.4]).select(["b"]).to_df()
+    >>> table.search([0.4, 0.4]).select(["b"]).to_pandas()
        b      vector  _distance
     0  4  [0.5, 1.3]       0.82
     1  2  [1.1, 1.2]       1.13
@@ -149,7 +150,7 @@ class Table(ABC):
     @abstractmethod
     def schema(self) -> pa.Schema:
         """The [Arrow Schema](https://arrow.apache.org/docs/python/api/datatypes.html#) of
-        this [Table](Table)
+        this Table
 
         """
         raise NotImplementedError
@@ -180,6 +181,7 @@ class Table(ABC):
         num_sub_vectors=96,
         vector_column_name: str = VECTOR_COLUMN_NAME,
         replace: bool = True,
+        accelerator: Optional[str] = None,
     ):
         """Create an index on the table.
 
@@ -200,6 +202,9 @@ class Table(ABC):
         replace: bool, default True
             If True, replace the existing index if it exists.
             If False, raise an error if duplicate index exists.
+        accelerator: str, default None
+            If set, use the given accelerator to create the index.
+            Only support "cuda" for now.
         """
         raise NotImplementedError
 
@@ -286,8 +291,9 @@ class Table(ABC):
         Examples
         --------
         >>> import lancedb
-        >>> import pandas as pd
+        >>> data = [
-        >>> data = pd.DataFrame({"x": [1, 2, 3], "vector": [[1, 2], [3, 4], [5, 6]]})
+        ...   {"x": 1, "vector": [1, 2]}, {"x": 2, "vector": [3, 4]}, {"x": 3, "vector": [5, 6]}
+        ... ]
         >>> db = lancedb.connect("./.lancedb")
         >>> table = db.create_table("my_table", data)
         >>> table.to_pandas()
@@ -390,6 +396,17 @@ class LanceTable(Table):
             raise ValueError(f"Invalid version {version}")
         self._reset_dataset(version=version)
 
+        try:
+            # Accessing the property updates the cached value
+            _ = self._dataset
+        except Exception as e:
+            if "not found" in str(e):
+                raise ValueError(
+                    f"Version {version} no longer exists. Was it cleaned up?"
+                )
+            else:
+                raise e
+
     def restore(self, version: int = None):
         """Restore a version of the table. This is an in-place operation.
 
@@ -479,6 +496,7 @@ class LanceTable(Table):
         num_sub_vectors=96,
         vector_column_name=VECTOR_COLUMN_NAME,
         replace: bool = True,
+        accelerator: Optional[str] = None,
     ):
         """Create an index on the table."""
         self._dataset.create_index(
@@ -488,8 +506,10 @@ class LanceTable(Table):
             num_partitions=num_partitions,
             num_sub_vectors=num_sub_vectors,
             replace=replace,
+            accelerator=accelerator,
         )
         self._reset_dataset()
+        register_event("create_index")
 
     def create_fts_index(self, field_names: Union[str, List[str]]):
         """Create a full-text search index on the table.
@@ -508,6 +528,7 @@ class LanceTable(Table):
             field_names = [field_names]
         index = create_index(self._get_fts_index_path(), field_names)
         populate_index(index, self, field_names)
+        register_event("create_fts_index")
 
     def _get_fts_index_path(self):
         return os.path.join(self._dataset_uri, "_indices", "tantivy")
@@ -560,6 +581,7 @@ class LanceTable(Table):
         )
         lance.write_dataset(data, self._dataset_uri, schema=self.schema, mode=mode)
         self._reset_dataset()
+        register_event("add")
 
     def merge(
         self,
@@ -623,6 +645,7 @@ class LanceTable(Table):
             other_table, left_on=left_on, right_on=right_on, schema=schema
         )
         self._reset_dataset()
+        register_event("merge")
 
     @cached_property
     def embedding_functions(self) -> dict:
@@ -673,6 +696,7 @@ class LanceTable(Table):
             and also the "_distance" column which is the distance between the query
             vector and the returned vector.
         """
+        register_event("search")
         return LanceQueryBuilder.create(
             self, query, query_type, vector_column_name=vector_column_name
         )
@@ -695,8 +719,9 @@ class LanceTable(Table):
         Examples
         --------
         >>> import lancedb
-        >>> import pandas as pd
+        >>> data = [
-        >>> data = pd.DataFrame({"x": [1, 2, 3], "vector": [[1, 2], [3, 4], [5, 6]]})
+        ...   {"x": 1, "vector": [1, 2]}, {"x": 2, "vector": [3, 4]}, {"x": 3, "vector": [5, 6]}
+        ... ]
         >>> db = lancedb.connect("./.lancedb")
         >>> table = db.create_table("my_table", data)
         >>> table.to_pandas()
@@ -776,6 +801,7 @@ class LanceTable(Table):
         if data is not None:
             table.add(data)
 
+        register_event("create_table")
         return table
 
     @classmethod
@@ -811,8 +837,9 @@ class LanceTable(Table):
         Examples
         --------
         >>> import lancedb
-        >>> import pandas as pd
+        >>> data = [
-        >>> data = pd.DataFrame({"x": [1, 2, 3], "vector": [[1, 2], [3, 4], [5, 6]]})
+        ...   {"x": 1, "vector": [1, 2]}, {"x": 2, "vector": [3, 4]}, {"x": 3, "vector": [5, 6]}
+        ... ]
         >>> db = lancedb.connect("./.lancedb")
         >>> table = db.create_table("my_table", data)
         >>> table.to_pandas()
@@ -841,6 +868,7 @@ class LanceTable(Table):
         self.delete(where)
         self.add(orig_data, mode="append")
         self._reset_dataset()
+        register_event("update")
 
     def _execute_query(self, query: Query) -> pa.Table:
         ds = self.to_lance()
@@ -864,6 +892,48 @@ class LanceTable(Table):
             },
         )
 
+    def cleanup_old_versions(
+        self,
+        older_than: Optional[timedelta] = None,
+        *,
+        delete_unverified: bool = False,
+    ) -> CleanupStats:
+        """
+        Clean up old versions of the table, freeing disk space.
+
+        Parameters
+        ----------
+        older_than: timedelta, default None
+            The minimum age of the version to delete. If None, then this defaults
+            to two weeks.
+        delete_unverified: bool, default False
+            Because they may be part of an in-progress transaction, files newer
+            than 7 days old are not deleted by default. If you are sure that
+            there are no in-progress transactions, then you can set this to True
+            to delete all files older than `older_than`.
+
+        Returns
+        -------
+        CleanupStats
+            The stats of the cleanup operation, including how many bytes were
+            freed.
+        """
+        return self.to_lance().cleanup_old_versions(
+            older_than, delete_unverified=delete_unverified
+        )
+
+    def compact_files(self, *args, **kwargs):
+        """
+        Run the compaction process on the table.
+
+        This can be run after making several small appends to optimize the table
+        for faster reads.
+
+        Arguments are passed onto :meth:`lance.dataset.DatasetOptimizer.compact_files`.
+        For most cases, the default should be fine.
+        """
+        return self.to_lance().optimize.compact_files(*args, **kwargs)
+
 
 def _sanitize_schema(
     data: pa.Table,

python/lancedb/utils/__init__.py

@@ -0,0 +1,15 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+from .config import Config
+
+CONFIG = Config()

python/lancedb/utils/config.py

@@ -0,0 +1,116 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import copy
+import hashlib
+import os
+import platform
+import uuid
+from pathlib import Path
+
+from .general import LOGGER, is_dir_writeable, yaml_load, yaml_save
+
+
+def get_user_config_dir(sub_dir="lancedb"):
+    """
+    Get the user config directory.
+
+    Args:
+        sub_dir (str): The name of the subdirectory to create.
+
+    Returns:
+        (Path): The path to the user config directory.
+    """
+    # Return the appropriate config directory for each operating system
+    if platform.system() == "Windows":
+        path = Path.home() / "AppData" / "Roaming" / sub_dir
+    elif platform.system() == "Darwin":
+        path = Path.home() / "Library" / "Application Support" / sub_dir
+    elif platform.system() == "Linux":
+        path = Path.home() / ".config" / sub_dir
+    else:
+        raise ValueError(f"Unsupported operating system: {platform.system()}")
+
+    # GCP and AWS lambda fix, only /tmp is writeable
+    if not is_dir_writeable(path.parent):
+        LOGGER.warning(
+            f"WARNING ⚠️ user config directory '{path}' is not writeable, defaulting to '/tmp' or CWD."
+            "Alternatively you can define a LANCEDB_CONFIG_DIR environment variable for this path."
+        )
+        path = (
+            Path("/tmp") / sub_dir
+            if is_dir_writeable("/tmp")
+            else Path().cwd() / sub_dir
+        )
+
+    # Create the subdirectory if it does not exist
+    path.mkdir(parents=True, exist_ok=True)
+
+    return path
+
+
+USER_CONFIG_DIR = Path(os.getenv("LANCEDB_CONFIG_DIR") or get_user_config_dir())
+CONFIG_FILE = USER_CONFIG_DIR / "config.yaml"
+
+
+class Config(dict):
+    """
+    Manages lancedb config stored in a YAML file.
+
+    Args:
+        file (str | Path): Path to the lancedb config YAML file. Default is USER_CONFIG_DIR / 'config.yaml'.
+    """
+
+    def __init__(self, file=CONFIG_FILE):
+        self.file = Path(file)
+        self.defaults = {  # Default global config values
+            "diagnostics": True,
+            "uuid": hashlib.sha256(str(uuid.getnode()).encode()).hexdigest(),
+        }
+
+        super().__init__(copy.deepcopy(self.defaults))
+
+        if not self.file.exists():
+            self.save()
+
+        self.load()
+        correct_keys = self.keys() == self.defaults.keys()
+        correct_types = all(
+            type(a) is type(b) for a, b in zip(self.values(), self.defaults.values())
+        )
+        if not (correct_keys and correct_types):
+            LOGGER.warning(
+                "WARNING ⚠️ LanceDB settings reset to default values. This may be due to a possible problem "
+                "with your settings or a recent package update. "
+                f"\nView settings & usage with 'lancedb settings' or at '{self.file}'"
+            )
+            self.reset()
+
+    def load(self):
+        """Loads settings from the YAML file."""
+        super().update(yaml_load(self.file))
+
+    def save(self):
+        """Saves the current settings to the YAML file."""
+        yaml_save(self.file, dict(self))
+
+    def update(self, *args, **kwargs):
+        """Updates a setting value in the current settings."""
+        super().update(*args, **kwargs)
+        self.save()
+
+    def reset(self):
+        """Resets the settings to default and saves them."""
+        self.clear()
+        self.update(self.defaults)
+        self.save()

python/lancedb/utils/events.py

@@ -0,0 +1,161 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import datetime
+import importlib.metadata
+import platform
+import random
+import sys
+import time
+
+from lancedb.utils import CONFIG
+from lancedb.utils.general import TryExcept
+
+from .general import (
+    PLATFORMS,
+    get_git_origin_url,
+    is_git_dir,
+    is_github_actions_ci,
+    is_online,
+    is_pip_package,
+    is_pytest_running,
+    threaded_request,
+)
+
+
+class _Events:
+    """
+    A class for collecting anonymous event analytics. Event analytics are enabled when ``diagnostics=True`` in config and
+    disabled when ``diagnostics=False``.
+
+    You can enable or disable diagnostics by running ``lancedb diagnostics --enabled`` or ``lancedb diagnostics --disabled``.
+
+    Attributes
+    ----------
+    url : str
+        The URL to send anonymous events.
+    rate_limit : float
+        The rate limit in seconds for sending events.
+    metadata : dict
+        A dictionary containing metadata about the environment.
+    enabled : bool
+        A flag to enable or disable Events based on certain conditions.
+    """
+
+    _instance = None
+
+    url = "https://app.posthog.com/capture/"
+    headers = {"Content-Type": "application/json"}
+    api_key = "phc_oENDjGgHtmIDrV6puUiFem2RB4JA8gGWulfdulmMdZP"
+    # This api-key is write only and is safe to expose in the codebase.
+
+    def __init__(self):
+        """
+        Initializes the Events object with default values for events, rate_limit, and metadata.
+        """
+        self.events = []  # events list
+        self.max_events = 25  # max events to store in memory
+        self.rate_limit = 60.0  # rate limit (seconds)
+        self.time = 0.0
+
+        if is_git_dir():
+            install = "git"
+        elif is_pip_package():
+            install = "pip"
+        else:
+            install = "other"
+        self.metadata = {
+            "cli": sys.argv[0],
+            "install": install,
+            "python": ".".join(platform.python_version_tuple()[:2]),
+            "version": importlib.metadata.version("lancedb"),
+            "platforms": PLATFORMS,
+            "session_id": round(random.random() * 1e15),
+            # 'engagement_time_msec': 1000 # TODO: In future we might be interested in this metric
+        }
+
+        TESTS_RUNNING = is_pytest_running() or is_github_actions_ci()
+        ONLINE = is_online()
+        self.enabled = (
+            CONFIG["diagnostics"]
+            and not TESTS_RUNNING
+            and ONLINE
+            and (
+                is_pip_package()
+                or get_git_origin_url() == "https://github.com/lancedb/lancedb.git"
+            )
+        )
+
+    def __call__(self, event_name, params={}):
+        """
+        Attempts to add a new event to the events list and send events if the rate limit is reached.
+
+        Args
+        ----
+        event_name : str
+            The name of the event to be logged.
+        params : dict, optional
+            A dictionary of additional parameters to be logged with the event.
+        """
+        ### NOTE: We might need a way to tag a session with a label to check usage from a source. Setting label should be exposed to the user.
+        if not self.enabled:
+            return
+        if (
+            len(self.events) < self.max_events
+        ):  # Events list limited to 25 events (drop any events past this)
+            params.update(self.metadata)
+            self.events.append(
+                {
+                    "event": event_name,
+                    "properties": params,
+                    "timestamp": datetime.datetime.now(
+                        tz=datetime.timezone.utc
+                    ).isoformat(),
+                    "distinct_id": CONFIG["uuid"],
+                }
+            )
+
+        # Check rate limit
+        t = time.time()
+        if (t - self.time) < self.rate_limit:
+            return
+        # Time is over rate limiter, send now
+        data = {
+            "api_key": self.api_key,
+            "distinct_id": CONFIG["uuid"],  # posthog needs this to accepts the event
+            "batch": self.events,
+        }
+
+        # POST equivalent to requests.post(self.url, json=data).
+        # threaded request is used to avoid blocking, retries are disabled, and verbose is disabled
+        # to avoid any possible disruption in the console.
+        threaded_request(
+            method="post",
+            url=self.url,
+            headers=self.headers,
+            json=data,
+            retry=0,
+            verbose=False,
+        )
+
+        # Flush & Reset
+        self.events = []
+        self.time = t
+
+
+@TryExcept(verbose=False)
+def register_event(name: str, **kwargs):
+    if _Events._instance is None:
+        _Events._instance = _Events()
+
+    _Events._instance(name, **kwargs)

python/lancedb/utils/general.py

@@ -0,0 +1,445 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import contextlib
+import importlib
+import logging.config
+import os
+import platform
+import subprocess
+import sys
+import threading
+import time
+from pathlib import Path
+from typing import Union
+
+import requests
+import yaml
+
+LOGGING_NAME = "lancedb"
+VERBOSE = (
+    str(os.getenv("LANCEDB_VERBOSE", True)).lower() == "true"
+)  # global verbose mode
+
+
+def set_logging(name=LOGGING_NAME, verbose=True):
+    """Sets up logging for the given name.
+
+    Parameters
+    ----------
+    name : str, optional
+        The name of the logger. Default is 'lancedb'.
+    verbose : bool, optional
+        Whether to enable verbose logging. Default is True.
+    """
+
+    rank = int(os.getenv("RANK", -1))  # rank in world for Multi-GPU trainings
+    level = logging.INFO if verbose and rank in {-1, 0} else logging.ERROR
+    logging.config.dictConfig(
+        {
+            "version": 1,
+            "disable_existing_loggers": False,
+            "formatters": {name: {"format": "%(message)s"}},
+            "handlers": {
+                name: {
+                    "class": "logging.StreamHandler",
+                    "formatter": name,
+                    "level": level,
+                }
+            },
+            "loggers": {name: {"level": level, "handlers": [name], "propagate": False}},
+        }
+    )
+
+
+set_logging(LOGGING_NAME, verbose=VERBOSE)
+LOGGER = logging.getLogger(LOGGING_NAME)
+
+
+def is_pip_package(filepath: str = __name__) -> bool:
+    """Determines if the file at the given filepath is part of a pip package.
+
+    Parameters
+    ----------
+    filepath : str, optional
+        The filepath to check. Default is the current file.
+
+    Returns
+    -------
+    bool
+        True if the file is part of a pip package, False otherwise.
+    """
+    # Get the spec for the module
+    spec = importlib.util.find_spec(filepath)
+
+    # Return whether the spec is not None and the origin is not None (indicating it is a package)
+    return spec is not None and spec.origin is not None
+
+
+def is_pytest_running():
+    """Determines whether pytest is currently running or not.
+
+    Returns
+    -------
+    bool
+        True if pytest is running, False otherwise.
+    """
+    return (
+        ("PYTEST_CURRENT_TEST" in os.environ)
+        or ("pytest" in sys.modules)
+        or ("pytest" in Path(sys.argv[0]).stem)
+    )
+
+
+def is_github_actions_ci() -> bool:
+    """
+    Determine if the current environment is a GitHub Actions CI Python runner.
+
+    Returns
+    -------
+    bool
+        True if the current environment is a GitHub Actions CI Python runner, False otherwise.
+    """
+
+    return (
+        "GITHUB_ACTIONS" in os.environ
+        and "RUNNER_OS" in os.environ
+        and "RUNNER_TOOL_CACHE" in os.environ
+    )
+
+
+def is_git_dir():
+    """
+    Determines whether the current file is part of a git repository.
+    If the current file is not part of a git repository, returns None.
+
+    Returns
+    -------
+    bool
+        True if current file is part of a git repository.
+    """
+    return get_git_dir() is not None
+
+
+def is_online() -> bool:
+    """
+    Check internet connectivity by attempting to connect to a known online host.
+
+    Returns
+    -------
+    bool
+        True if connection is successful, False otherwise.
+    """
+    import socket
+
+    for host in "1.1.1.1", "8.8.8.8", "223.5.5.5":  # Cloudflare, Google, AliDNS:
+        try:
+            test_connection = socket.create_connection(address=(host, 53), timeout=2)
+        except (socket.timeout, socket.gaierror, OSError):
+            continue
+        else:
+            # If the connection was successful, close it to avoid a ResourceWarning
+            test_connection.close()
+            return True
+    return False
+
+
+def is_dir_writeable(dir_path: Union[str, Path]) -> bool:
+    """Check if a directory is writeable.
+
+    Parameters
+    ----------
+    dir_path : Union[str, Path]
+        The path to the directory.
+
+    Returns
+    -------
+    bool
+        True if the directory is writeable, False otherwise.
+    """
+    return os.access(str(dir_path), os.W_OK)
+
+
+def is_colab():
+    """Check if the current script is running inside a Google Colab notebook.
+
+    Returns
+    -------
+    bool
+        True if running inside a Colab notebook, False otherwise.
+    """
+    return "COLAB_RELEASE_TAG" in os.environ or "COLAB_BACKEND_VERSION" in os.environ
+
+
+def is_kaggle():
+    """Check if the current script is running inside a Kaggle kernel.
+
+    Returns
+    -------
+    bool
+        True if running inside a Kaggle kernel, False otherwise.
+    """
+    return (
+        os.environ.get("PWD") == "/kaggle/working"
+        and os.environ.get("KAGGLE_URL_BASE") == "https://www.kaggle.com"
+    )
+
+
+def is_jupyter():
+    """Check if the current script is running inside a Jupyter Notebook.
+
+    Returns
+    -------
+    bool
+        True if running inside a Jupyter Notebook, False otherwise.
+    """
+    with contextlib.suppress(Exception):
+        from IPython import get_ipython
+
+        return get_ipython() is not None
+    return False
+
+
+def is_docker() -> bool:
+    """Determine if the script is running inside a Docker container.
+
+    Returns
+    -------
+    bool
+        True if the script is running inside a Docker container, False otherwise.
+    """
+    file = Path("/proc/self/cgroup")
+    if file.exists():
+        with open(file) as f:
+            return "docker" in f.read()
+    else:
+        return False
+
+
+def get_git_dir():
+    """Determine whether the current file is part of a git repository and if so, returns the repository root directory.
+    If the current file is not part of a git repository, returns None.
+
+    Returns
+    -------
+    Path | None
+        Git root directory if found or None if not found.
+    """
+    for d in Path(__file__).parents:
+        if (d / ".git").is_dir():
+            return d
+
+
+def get_git_origin_url():
+    """Retrieve the origin URL of a git repository.
+
+    Returns
+    -------
+    str | None
+        The origin URL of the git repository or None if not git directory.
+    """
+    if is_git_dir():
+        with contextlib.suppress(subprocess.CalledProcessError):
+            origin = subprocess.check_output(
+                ["git", "config", "--get", "remote.origin.url"]
+            )
+            return origin.decode().strip()
+
+
+def yaml_save(file="data.yaml", data=None, header=""):
+    """Save YAML data to a file.
+
+    Parameters
+    ----------
+    file : str, optional
+        File name, by default 'data.yaml'.
+    data : dict, optional
+        Data to save in YAML format, by default None.
+    header : str, optional
+        YAML header to add, by default "".
+    """
+    if data is None:
+        data = {}
+    file = Path(file)
+    if not file.parent.exists():
+        # Create parent directories if they don't exist
+        file.parent.mkdir(parents=True, exist_ok=True)
+
+    # Convert Path objects to strings
+    for k, v in data.items():
+        if isinstance(v, Path):
+            data[k] = str(v)
+
+    # Dump data to file in YAML format
+    with open(file, "w", errors="ignore", encoding="utf-8") as f:
+        if header:
+            f.write(header)
+        yaml.safe_dump(data, f, sort_keys=False, allow_unicode=True)
+
+
+def yaml_load(file="data.yaml", append_filename=False):
+    """
+    Load YAML data from a file.
+
+    Parameters
+    ----------
+    file : str, optional
+        File name. Default is 'data.yaml'.
+    append_filename : bool, optional
+        Add the YAML filename to the YAML dictionary. Default is False.
+
+    Returns
+    -------
+    dict
+        YAML data and file name.
+    """
+    assert Path(file).suffix in (
+        ".yaml",
+        ".yml",
+    ), f"Attempting to load non-YAML file {file} with yaml_load()"
+    with open(file, errors="ignore", encoding="utf-8") as f:
+        s = f.read()  # string
+
+        # Add YAML filename to dict and return
+        data = (
+            yaml.safe_load(s) or {}
+        )  # always return a dict (yaml.safe_load() may return None for empty files)
+        if append_filename:
+            data["yaml_file"] = str(file)
+        return data
+
+
+def yaml_print(yaml_file: Union[str, Path, dict]) -> None:
+    """
+    Pretty prints a YAML file or a YAML-formatted dictionary.
+
+    Parameters
+    ----------
+    yaml_file : Union[str, Path, dict]
+        The file path of the YAML file or a YAML-formatted dictionary.
+
+    Returns
+    -------
+    None
+    """
+    yaml_dict = (
+        yaml_load(yaml_file) if isinstance(yaml_file, (str, Path)) else yaml_file
+    )
+    dump = yaml.dump(yaml_dict, sort_keys=False, allow_unicode=True)
+    LOGGER.info(f"Printing '{yaml_file}'\n\n{dump}")
+
+
+PLATFORMS = [platform.system()]
+if is_colab():
+    PLATFORMS.append("Colab")
+if is_kaggle():
+    PLATFORMS.append("Kaggle")
+if is_jupyter():
+    PLATFORMS.append("Jupyter")
+if is_docker():
+    PLATFORMS.append("Docker")
+
+PLATFORMS = "|".join(PLATFORMS)
+
+
+class TryExcept(contextlib.ContextDecorator):
+    """
+    TryExcept context manager.
+    Usage: @TryExcept() decorator or 'with TryExcept():' context manager.
+    """
+
+    def __init__(self, msg="", verbose=True):
+        """
+        Parameters
+        ----------
+        msg : str, optional
+            Custom message to display in case of exception, by default "".
+        verbose : bool, optional
+            Whether to display the message, by default True.
+        """
+        self.msg = msg
+        self.verbose = verbose
+
+    def __enter__(self):
+        pass
+
+    def __exit__(self, exc_type, value, traceback):
+        if self.verbose and value:
+            LOGGER.info(f"{self.msg}{': ' if self.msg else ''}{value}")
+        return True
+
+
+def threaded_request(
+    method, url, retry=3, timeout=30, thread=True, code=-1, verbose=True, **kwargs
+):
+    """
+    Makes an HTTP request using the 'requests' library, with exponential backoff retries up to a specified timeout.
+
+    Parameters
+    ----------
+    method : str
+        The HTTP method to use for the request. Choices are 'post' and 'get'.
+    url : str
+        The URL to make the request to.
+    retry : int, optional
+        Number of retries to attempt before giving up, by default 3.
+    timeout : int, optional
+        Timeout in seconds after which the function will give up retrying, by default 30.
+    thread : bool, optional
+        Whether to execute the request in a separate daemon thread, by default True.
+    code : int, optional
+        An identifier for the request, used for logging purposes, by default -1.
+    verbose : bool, optional
+        A flag to determine whether to print out to console or not, by default True.
+
+    Returns
+    -------
+    requests.Response
+        The HTTP response object. If the request is executed in a separate thread, returns the thread itself.
+    """
+    retry_codes = ()  # retry only these codes TODO: add codes if needed in future (500, 408)
+
+    @TryExcept(verbose=verbose)
+    def func(method, url, **kwargs):
+        """Make HTTP requests with retries and timeouts, with optional progress tracking."""
+        response = None
+        t0 = time.time()
+        for i in range(retry + 1):
+            if (time.time() - t0) > timeout:
+                break
+            response = requests.request(method, url, **kwargs)
+            if response.status_code < 300:  # good return codes in the 2xx range
+                break
+            try:
+                m = response.json().get("message", "No JSON message.")
+            except AttributeError:
+                m = "Unable to read JSON."
+            if i == 0:
+                if response.status_code in retry_codes:
+                    m += f" Retrying {retry}x for {timeout}s." if retry else ""
+                elif response.status_code == 429:  # rate limit
+                    m = f"Rate limit reached"
+                if verbose:
+                    LOGGER.warning(f"{response.status_code} #{code}")
+                if response.status_code not in retry_codes:
+                    return response
+            time.sleep(2**i)  # exponential standoff
+        return response
+
+    args = method, url
+    if thread:
+        return threading.Thread(
+            target=func, args=args, kwargs=kwargs, daemon=True
+        ).start()
+    else:
+        return func(*args, **kwargs)

python/lancedb/utils/sentry_log.py

@@ -0,0 +1,112 @@
+#  Copyright 2023 LanceDB Developers
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import bdb
+import importlib.metadata
+import logging
+import sys
+from pathlib import Path
+
+from lancedb.utils import CONFIG
+
+from .general import (
+    PLATFORMS,
+    TryExcept,
+    is_git_dir,
+    is_github_actions_ci,
+    is_online,
+    is_pip_package,
+    is_pytest_running,
+)
+
+
+@TryExcept(verbose=False)
+def set_sentry():
+    """
+    Initialize the Sentry SDK for error tracking and reporting. Only used if sentry_sdk package is installed and
+    sync=True in settings. Run 'lancedb settings' to see and update settings YAML file.
+
+    Conditions required to send errors (ALL conditions must be met or no errors will be reported):
+        - sentry_sdk package is installed
+        - sync=True in  settings
+        - pytest is not running
+        - running in a pip package installation
+        - running in a non-git directory
+        - online environment
+
+    The function also configures Sentry SDK to ignore KeyboardInterrupt and FileNotFoundError
+    exceptions for now.
+
+    Additionally, the function sets custom tags and user information for Sentry events.
+    """
+
+    def before_send(event, hint):
+        """
+        Modify the event before sending it to Sentry based on specific exception types and messages.
+
+        Args:
+            event (dict): The event dictionary containing information about the error.
+            hint (dict): A dictionary containing additional information about the error.
+
+        Returns:
+            dict: The modified event or None if the event should not be sent to Sentry.
+        """
+        if "exc_info" in hint:
+            exc_type, exc_value, tb = hint["exc_info"]
+            if "out of memory" in str(exc_value).lower():
+                return None
+
+        if is_git_dir():
+            install = "git"
+        elif is_pip_package():
+            install = "pip"
+        else:
+            install = "other"
+
+        event["tags"] = {
+            "sys_argv": sys.argv[0],
+            "sys_argv_name": Path(sys.argv[0]).name,
+            "install": install,
+            "platforms": PLATFORMS,
+            "version": importlib.metadata.version("lancedb"),
+        }
+        return event
+
+    TESTS_RUNNING = is_pytest_running() or is_github_actions_ci()
+    ONLINE = is_online()
+    if CONFIG["diagnostics"] and not TESTS_RUNNING and ONLINE and is_pip_package():
+        # and not is_git_dir(): # not running inside a git dir. Maybe too restrictive?
+
+        # If sentry_sdk package is not installed then return and do not use Sentry
+        try:
+            import sentry_sdk  # noqa
+        except ImportError:
+            return
+
+        sentry_sdk.init(
+            dsn="https://c63ef8c64e05d1aa1a96513361f3ca2f@o4505950840946688.ingest.sentry.io/4505950933614592",
+            debug=False,
+            include_local_variables=False,
+            traces_sample_rate=1.0,
+            environment="production",  # 'dev' or 'production'
+            before_send=before_send,
+            ignore_errors=[KeyboardInterrupt, FileNotFoundError, bdb.BdbQuit],
+        )
+        sentry_sdk.set_user({"id": CONFIG["uuid"]})  # SHA-256 anonymized UUID hash
+
+        # Disable all sentry logging
+        for logger in "sentry_sdk", "sentry_sdk.errors":
+            logging.getLogger(logger).setLevel(logging.CRITICAL)
+
+
+set_sentry()

python/pyproject.toml

@@ -1,8 +1,9 @@
 [project]
 name = "lancedb"
-version = "0.2.6"
+version = "0.3.2"
 dependencies = [
-    "pylance==0.8.0",
+    "deprecation",
+    "pylance==0.8.7",
     "ratelimiter~=1.0",
     "retry>=0.9.2",
     "tqdm>=4.1.0",
@@ -10,7 +11,10 @@ dependencies = [
     "pydantic>=1.10",
     "attrs>=21.3.0",
     "semver>=3.0",
-    "cachetools"
+    "cachetools",
+    "pyyaml>=6.0",
+    "click>=8.1.7",
+    "requests>=2.31.0"
 ]
 description = "lancedb"
 authors = [{ name = "LanceDB Devs", email = "dev@lancedb.com" }]
@@ -48,7 +52,10 @@ tests = ["pandas>=1.4", "pytest", "pytest-mock", "pytest-asyncio", "requests"]
 dev = ["ruff", "pre-commit", "black"]
 docs = ["mkdocs", "mkdocs-jupyter", "mkdocs-material", "mkdocstrings[python]"]
 clip = ["torch", "pillow", "open-clip"]
-embeddings = ["openai", "sentence-transformers", "torch", "pillow", "open-clip"]
+embeddings = ["openai", "sentence-transformers", "torch", "pillow", "open-clip-torch", "cohere"]
+
+[project.scripts]
+lancedb = "lancedb.cli.cli:cli"
 
 [build-system]
 requires = ["setuptools", "wheel"]

python/tests/test_cli.py

@@ -0,0 +1,35 @@
+from click.testing import CliRunner
+
+from lancedb.cli.cli import cli
+from lancedb.utils import CONFIG
+
+
+def test_entry():
+    runner = CliRunner()
+    result = runner.invoke(cli)
+    assert result.exit_code == 0  # Main check
+    assert "lancedb" in result.output.lower()  # lazy check
+
+
+def test_diagnostics():
+    runner = CliRunner()
+    result = runner.invoke(cli, ["diagnostics", "--disabled"])
+    assert result.exit_code == 0  # Main check
+    assert CONFIG["diagnostics"] == False
+
+    result = runner.invoke(cli, ["diagnostics", "--enabled"])
+    assert result.exit_code == 0  # Main check
+    assert CONFIG["diagnostics"] == True
+
+
+def test_config():
+    runner = CliRunner()
+    result = runner.invoke(cli, ["config"])
+    assert result.exit_code == 0  # Main check
+    cfg = CONFIG.copy()
+    cfg.pop("uuid")
+    for (
+        item,
+        _,
+    ) in cfg.items():  # check for keys only as formatting is subject to change
+        assert item in result.output

python/tests/test_context.py

@@ -47,7 +47,7 @@ def test_contextualizer(raw_df: pd.DataFrame):
         .stride(3)
         .text_col("token")
         .groupby("document_id")
-        .to_df()["token"]
+        .to_pandas()["token"]
         .to_list()
     )
 
@@ -67,7 +67,7 @@ def test_contextualizer_with_threshold(raw_df: pd.DataFrame):
         .text_col("token")
         .groupby("document_id")
         .min_window_size(4)
-        .to_df()["token"]
+        .to_pandas()["token"]
         .to_list()
     )
 

python/tests/test_db.py

@@ -33,11 +33,11 @@ def test_basic(tmp_path):
             {"vector": [5.9, 26.5], "item": "bar", "price": 20.0},
         ],
     )
-    rs = table.search([100, 100]).limit(1).to_df()
+    rs = table.search([100, 100]).limit(1).to_pandas()
     assert len(rs) == 1
     assert rs["item"].iloc[0] == "bar"
 
-    rs = table.search([100, 100]).where("price < 15").limit(2).to_df()
+    rs = table.search([100, 100]).where("price < 15").limit(2).to_pandas()
     assert len(rs) == 1
     assert rs["item"].iloc[0] == "foo"
 
@@ -62,11 +62,11 @@ def test_ingest_pd(tmp_path):
         }
     )
     table = db.create_table("test", data=data)
-    rs = table.search([100, 100]).limit(1).to_df()
+    rs = table.search([100, 100]).limit(1).to_pandas()
     assert len(rs) == 1
     assert rs["item"].iloc[0] == "bar"
 
-    rs = table.search([100, 100]).where("price < 15").limit(2).to_df()
+    rs = table.search([100, 100]).where("price < 15").limit(2).to_pandas()
     assert len(rs) == 1
     assert rs["item"].iloc[0] == "foo"
 
@@ -137,8 +137,8 @@ def test_ingest_iterator(tmp_path):
         db = lancedb.connect(tmp_path)
         tbl = db.create_table("table2", make_batches(), schema=schema, mode="overwrite")
         tbl.to_pandas()
-        assert tbl.search([3.1, 4.1]).limit(1).to_df()["_distance"][0] == 0.0
+        assert tbl.search([3.1, 4.1]).limit(1).to_pandas()["_distance"][0] == 0.0
-        assert tbl.search([5.9, 26.5]).limit(1).to_df()["_distance"][0] == 0.0
+        assert tbl.search([5.9, 26.5]).limit(1).to_pandas()["_distance"][0] == 0.0
         tbl_len = len(tbl)
         tbl.add(make_batches())
         assert tbl_len == 50

python/tests/test_e2e_remote_db.py

@@ -23,5 +23,5 @@ from lancedb import LanceDBConnection
 def test_against_local_server():
     conn = LanceDBConnection("lancedb+http://localhost:10024")
     table = conn.open_table("sift1m_ivf1024_pq16")
-    df = table.search(np.random.rand(128)).to_df()
+    df = table.search(np.random.rand(128)).to_pandas()
     assert len(df) == 10

python/tests/test_embeddings_slow.py

@@ -11,6 +11,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 import io
+import os
 
 import numpy as np
 import pandas as pd
@@ -18,7 +19,7 @@ import pytest
 import requests
 
 import lancedb
-from lancedb.embeddings import EmbeddingFunctionRegistry
+from lancedb.embeddings import get_registry
 from lancedb.pydantic import LanceModel, Vector
 
 # These are integration tests for embedding functions.
@@ -30,12 +31,15 @@ from lancedb.pydantic import LanceModel, Vector
 @pytest.mark.parametrize("alias", ["sentence-transformers", "openai"])
 def test_sentence_transformer(alias, tmp_path):
     db = lancedb.connect(tmp_path)
-    registry = EmbeddingFunctionRegistry.get_instance()
+    registry = get_registry()
     func = registry.get(alias).create()
+    func2 = registry.get(alias).create()
 
     class Words(LanceModel):
         text: str = func.SourceField()
+        text2: str = func2.SourceField()
         vector: Vector(func.ndims()) = func.VectorField()
+        vector2: Vector(func2.ndims()) = func2.VectorField()
 
     table = db.create_table("words", schema=Words)
     table.add(
@@ -49,7 +53,16 @@ def test_sentence_transformer(alias, tmp_path):
                     "foo",
                     "bar",
                     "baz",
-                ]
+                ],
+                "text2": [
+                    "to be or not to be",
+                    "that is the question",
+                    "for whether tis nobler",
+                    "in the mind to suffer",
+                    "the slings and arrows",
+                    "of outrageous fortune",
+                    "or to take arms",
+                ],
             }
         )
     )
@@ -61,6 +74,13 @@ def test_sentence_transformer(alias, tmp_path):
     expected = table.search(vec).limit(1).to_pydantic(Words)[0]
     assert actual.text == expected.text
     assert actual.text == "hello world"
+    assert not np.allclose(actual.vector, actual.vector2)
+
+    actual = (
+        table.search(query, vector_column_name="vector2").limit(1).to_pydantic(Words)[0]
+    )
+    assert actual.text != "hello world"
+    assert not np.allclose(actual.vector, actual.vector2)
 
 
 @pytest.mark.slow
@@ -68,7 +88,7 @@ def test_openclip(tmp_path):
     from PIL import Image
 
     db = lancedb.connect(tmp_path)
-    registry = EmbeddingFunctionRegistry.get_instance()
+    registry = get_registry()
     func = registry.get("open-clip").create()
 
     class Images(LanceModel):
@@ -123,3 +143,22 @@ def test_openclip(tmp_path):
         arrow_table["vector"].combine_chunks().values.to_numpy(),
         arrow_table["vec_from_bytes"].combine_chunks().values.to_numpy(),
     )
+
+
+@pytest.mark.slow
+@pytest.mark.skipif(
+    os.environ.get("COHERE_API_KEY") is None, reason="COHERE_API_KEY not set"
+)  # also skip if cohere not installed
+def test_cohere_embedding_function():
+    cohere = get_registry().get("cohere").create(name="embed-multilingual-v2.0")
+
+    class TextModel(LanceModel):
+        text: str = cohere.SourceField()
+        vector: Vector(cohere.ndims()) = cohere.VectorField()
+
+    df = pd.DataFrame({"text": ["hello world", "goodbye world"]})
+    db = lancedb.connect("~/lancedb")
+    tbl = db.create_table("test", schema=TextModel, mode="overwrite")
+
+    tbl.add(df)
+    assert len(tbl.to_pandas()["vector"][0]) == cohere.ndims()

python/tests/test_fts.py

@@ -71,14 +71,14 @@ def test_search_index(tmp_path, table):
 
 def test_create_index_from_table(tmp_path, table):
     table.create_fts_index("text")
-    df = table.search("puppy").limit(10).select(["text"]).to_df()
+    df = table.search("puppy").limit(10).select(["text"]).to_pandas()
     assert len(df) == 10
     assert "text" in df.columns
 
 
 def test_create_index_multiple_columns(tmp_path, table):
     table.create_fts_index(["text", "text2"])
-    df = table.search("puppy").limit(10).to_df()
+    df = table.search("puppy").limit(10).to_pandas()
     assert len(df) == 10
     assert "text" in df.columns
     assert "text2" in df.columns
@@ -87,5 +87,5 @@ def test_create_index_multiple_columns(tmp_path, table):
 def test_empty_rs(tmp_path, table, mocker):
     table.create_fts_index(["text", "text2"])
     mocker.patch("lancedb.fts.search_index", return_value=([], []))
-    df = table.search("puppy").limit(10).to_df()
+    df = table.search("puppy").limit(10).to_pandas()
     assert len(df) == 0

python/tests/test_io.py

@@ -36,11 +36,11 @@ def test_s3_io():
             {"vector": [5.9, 26.5], "item": "bar", "price": 20.0},
         ],
     )
-    rs = table.search([100, 100]).limit(1).to_df()
+    rs = table.search([100, 100]).limit(1).to_pandas()
     assert len(rs) == 1
     assert rs["item"].iloc[0] == "bar"
 
-    rs = table.search([100, 100]).where("price < 15").limit(2).to_df()
+    rs = table.search([100, 100]).where("price < 15").limit(2).to_pandas()
     assert len(rs) == 1
     assert rs["item"].iloc[0] == "foo"
 

python/tests/test_pydantic.py

@@ -14,6 +14,7 @@
 
 import json
 import sys
+from datetime import date, datetime
 from typing import List, Optional
 
 import pyarrow as pa
@@ -40,10 +41,18 @@ def test_pydantic_to_arrow():
         li: List[int]
         opt: Optional[str] = None
         st: StructModel
+        dt: date
+        dtt: datetime
         # d: dict
 
     m = TestModel(
-        id=1, s="hello", vec=[1.0, 2.0, 3.0], li=[2, 3, 4], st=StructModel(a="a", b=1.0)
+        id=1,
+        s="hello",
+        vec=[1.0, 2.0, 3.0],
+        li=[2, 3, 4],
+        st=StructModel(a="a", b=1.0),
+        dt=date.today(),
+        dtt=datetime.now(),
     )
 
     schema = pydantic_to_schema(TestModel)
@@ -62,6 +71,8 @@ def test_pydantic_to_arrow():
                 ),
                 False,
             ),
+            pa.field("dt", pa.date32(), False),
+            pa.field("dtt", pa.timestamp("us"), False),
         ]
     )
     assert schema == expect_schema
@@ -79,10 +90,18 @@ def test_pydantic_to_arrow_py38():
         li: List[int]
         opt: Optional[str] = None
         st: StructModel
+        dt: date
+        dtt: datetime
         # d: dict
 
     m = TestModel(
-        id=1, s="hello", vec=[1.0, 2.0, 3.0], li=[2, 3, 4], st=StructModel(a="a", b=1.0)
+        id=1,
+        s="hello",
+        vec=[1.0, 2.0, 3.0],
+        li=[2, 3, 4],
+        st=StructModel(a="a", b=1.0),
+        dt=date.today(),
+        dtt=datetime.now(),
     )
 
     schema = pydantic_to_schema(TestModel)
@@ -101,6 +120,8 @@ def test_pydantic_to_arrow_py38():
                 ),
                 False,
             ),
+            pa.field("dt", pa.date32(), False),
+            pa.field("dtt", pa.timestamp("us"), False),
         ]
     )
     assert schema == expect_schema

python/tests/test_query.py

@@ -85,17 +85,20 @@ def test_cast(table):
 
 
 def test_query_builder(table):
-    df = (
+    rs = (
-        LanceVectorQueryBuilder(table, [0, 0], "vector").limit(1).select(["id"]).to_df()
+        LanceVectorQueryBuilder(table, [0, 0], "vector")
+        .limit(1)
+        .select(["id"])
+        .to_list()
     )
-    assert df["id"].values[0] == 1
+    assert rs[0]["id"] == 1
-    assert all(df["vector"].values[0] == [1, 2])
+    assert all(np.array(rs[0]["vector"]) == [1, 2])
 
 
 def test_query_builder_with_filter(table):
-    df = LanceVectorQueryBuilder(table, [0, 0], "vector").where("id = 2").to_df()
+    rs = LanceVectorQueryBuilder(table, [0, 0], "vector").where("id = 2").to_list()
-    assert df["id"].values[0] == 2
+    assert rs[0]["id"] == 2
-    assert all(df["vector"].values[0] == [3, 4])
+    assert all(np.array(rs[0]["vector"]) == [3, 4])
 
 
 def test_query_builder_with_prefilter(table):
@@ -103,7 +106,7 @@ def test_query_builder_with_prefilter(table):
         LanceVectorQueryBuilder(table, [0, 0], "vector")
         .where("id = 2")
         .limit(1)
-        .to_df()
+        .to_pandas()
     )
     assert len(df) == 0
 
@@ -111,7 +114,7 @@ def test_query_builder_with_prefilter(table):
         LanceVectorQueryBuilder(table, [0, 0], "vector")
         .where("id = 2", prefilter=True)
         .limit(1)
-        .to_df()
+        .to_pandas()
     )
     assert df["id"].values[0] == 2
     assert all(df["vector"].values[0] == [3, 4])
@@ -120,9 +123,11 @@ def test_query_builder_with_prefilter(table):
 def test_query_builder_with_metric(table):
     query = [4, 8]
     vector_column_name = "vector"
-    df_default = LanceVectorQueryBuilder(table, query, vector_column_name).to_df()
+    df_default = LanceVectorQueryBuilder(table, query, vector_column_name).to_pandas()
     df_l2 = (
-        LanceVectorQueryBuilder(table, query, vector_column_name).metric("L2").to_df()
+        LanceVectorQueryBuilder(table, query, vector_column_name)
+        .metric("L2")
+        .to_pandas()
     )
     tm.assert_frame_equal(df_default, df_l2)
 
@@ -130,7 +135,7 @@ def test_query_builder_with_metric(table):
         LanceVectorQueryBuilder(table, query, vector_column_name)
         .metric("cosine")
         .limit(1)
-        .to_df()
+        .to_pandas()
     )
     assert df_cosine._distance[0] == pytest.approx(
         cosine_distance(query, df_cosine.vector[0]),

python/tests/test_remote_client.py

@@ -86,7 +86,7 @@ async def test_e2e_with_mock_server():
                     columns=["id", "vector"],
                 ),
             )
-        ).to_df()
+        ).to_pandas()
 
         assert "vector" in df.columns
         assert "id" in df.columns

python/tests/test_remote_db.py

@@ -32,4 +32,4 @@ def test_remote_db():
     setattr(conn, "_client", FakeLanceDBClient())
 
     table = conn["test"]
-    table.search([1.0, 2.0]).to_df()
+    table.search([1.0, 2.0]).to_pandas()

python/tests/test_table.py

@@ -12,6 +12,7 @@
 #  limitations under the License.
 
 import functools
+from datetime import timedelta
 from pathlib import Path
 from typing import List
 from unittest.mock import PropertyMock, patch
@@ -223,6 +224,7 @@ def test_create_index_method():
                 num_partitions=256,
                 num_sub_vectors=96,
                 replace=True,
+                accelerator=None,
             )
 
 
@@ -426,8 +428,8 @@ def test_multiple_vector_columns(db):
     table.add(df)
 
     q = np.random.randn(10)
-    result1 = table.search(q, vector_column_name="vector1").limit(1).to_df()
+    result1 = table.search(q, vector_column_name="vector1").limit(1).to_pandas()
-    result2 = table.search(q, vector_column_name="vector2").limit(1).to_df()
+    result2 = table.search(q, vector_column_name="vector2").limit(1).to_pandas()
 
     assert result1["text"].iloc[0] != result2["text"].iloc[0]
 
@@ -438,6 +440,35 @@ def test_empty_query(db):
         "my_table",
         data=[{"text": "foo", "id": 0}, {"text": "bar", "id": 1}],
     )
-    df = table.search().select(["id"]).where("text='bar'").limit(1).to_df()
+    df = table.search().select(["id"]).where("text='bar'").limit(1).to_pandas()
     val = df.id.iloc[0]
     assert val == 1
+
+
+def test_compact_cleanup(db):
+    table = LanceTable.create(
+        db,
+        "my_table",
+        data=[{"text": "foo", "id": 0}, {"text": "bar", "id": 1}],
+    )
+
+    table.add([{"text": "baz", "id": 2}])
+    assert len(table) == 3
+    assert table.version == 3
+
+    stats = table.compact_files()
+    assert len(table) == 3
+    # Compact_files bump 2 versions.
+    assert table.version == 5
+    assert stats.fragments_removed > 0
+    assert stats.fragments_added == 1
+
+    stats = table.cleanup_old_versions()
+    assert stats.bytes_removed == 0
+
+    stats = table.cleanup_old_versions(older_than=timedelta(0), delete_unverified=True)
+    assert stats.bytes_removed > 0
+    assert table.version == 5
+
+    with pytest.raises(Exception, match="Version 3 no longer exists"):
+        table.checkout(3)

python/tests/test_telemetry.py

@@ -0,0 +1,60 @@
+import json
+
+import pytest
+
+import lancedb
+from lancedb.utils.events import _Events
+
+
+@pytest.fixture(autouse=True)
+def request_log_path(tmp_path):
+    return tmp_path / "request.json"
+
+
+def mock_register_event(name: str, **kwargs):
+    if _Events._instance is None:
+        _Events._instance = _Events()
+
+    _Events._instance.enabled = True
+    _Events._instance.rate_limit = 0
+    _Events._instance(name, **kwargs)
+
+
+def test_event_reporting(monkeypatch, request_log_path, tmp_path) -> None:
+    def mock_request(**kwargs):
+        json_data = kwargs.get("json", {})
+        with open(request_log_path, "w") as f:
+            json.dump(json_data, f)
+
+    monkeypatch.setattr(
+        lancedb.table, "register_event", mock_register_event
+    )  # Force enable registering events and strip exception handling
+    monkeypatch.setattr(lancedb.utils.events, "threaded_request", mock_request)
+
+    db = lancedb.connect(tmp_path)
+    db.create_table(
+        "test",
+        data=[
+            {"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
+            {"vector": [5.9, 26.5], "item": "bar", "price": 20.0},
+        ],
+        mode="overwrite",
+    )
+
+    assert request_log_path.exists()  # test if event was registered
+
+    with open(request_log_path, "r") as f:
+        json_data = json.load(f)
+
+    # TODO: don't hardcode these here. Instead create a module level json scehma in lancedb.utils.events for better evolvability
+    batch_keys = ["api_key", "distinct_id", "batch"]
+    event_keys = ["event", "properties", "timestamp", "distinct_id"]
+    property_keys = ["cli", "install", "platforms", "version", "session_id"]
+
+    assert all([key in json_data for key in batch_keys])
+    assert all([key in json_data["batch"][0] for key in event_keys])
+    assert all([key in json_data["batch"][0]["properties"] for key in property_keys])
+
+    # cleanup & reset
+    monkeypatch.undo()
+    _Events._instance = None

rust/ffi/node/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "vectordb-node"
-version = "0.2.6"
+version = "0.3.3"
 description = "Serverless, low-latency vector database for AI applications"
 license = "Apache-2.0"
 edition = "2018"
@@ -13,6 +13,7 @@ crate-type = ["cdylib"]
 arrow-array = { workspace = true }
 arrow-ipc = { workspace = true }
 arrow-schema = { workspace = true }
+chrono = { workspace = true }
 conv = "0.3.3"
 once_cell = "1"
 futures = "0.3"

rust/ffi/node/src/index/vector.rs

@@ -12,8 +12,7 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use lance::index::vector::ivf::IvfBuildParams;
+use lance::index::vector::{ivf::IvfBuildParams, pq::PQBuildParams};
-use lance::index::vector::pq::PQBuildParams;
 use lance_linalg::distance::MetricType;
 use neon::context::FunctionContext;
 use neon::prelude::*;
@@ -82,7 +81,7 @@ fn get_index_params_builder(
                 let ivf_params = IvfBuildParams {
                     num_partitions: np,
                     max_iters,
-                    centroids: None,
+                    ..Default::default()
                 };
                 index_builder.ivf_params(ivf_params)
             });

rust/ffi/node/src/lib.rs

@@ -74,7 +74,7 @@ fn runtime<'a, C: Context<'a>>(cx: &mut C) -> NeonResult<&'static Runtime> {
     static RUNTIME: OnceCell<Runtime> = OnceCell::new();
     static LOG: OnceCell<()> = OnceCell::new();
 
-    LOG.get_or_init(|| env_logger::init());
+    LOG.get_or_init(env_logger::init);
 
     RUNTIME.get_or_try_init(|| Runtime::new().or_throw(cx))
 }
@@ -148,7 +148,7 @@ fn get_aws_creds(
     match (secret_key_id, secret_key, temp_token) {
         (Some(key_id), Some(key), optional_token) => Ok(Some(Arc::new(
             StaticCredentialProvider::new(AwsCredential {
-                key_id: key_id,
+                key_id,
                 secret_key: key,
                 token: optional_token,
             }),
@@ -195,7 +195,7 @@ fn database_open_table(mut cx: FunctionContext) -> JsResult<JsPromise> {
 
     let (deferred, promise) = cx.promise();
     rt.spawn(async move {
-        let table_rst = database.open_table_with_params(&table_name, &params).await;
+        let table_rst = database.open_table_with_params(&table_name, params).await;
 
         deferred.settle_with(&channel, move |mut cx| {
             let js_table = JsTable::from(table_rst.or_throw(&mut cx)?);
@@ -237,6 +237,8 @@ fn main(mut cx: ModuleContext) -> NeonResult<()> {
     cx.export_function("tableAdd", JsTable::js_add)?;
     cx.export_function("tableCountRows", JsTable::js_count_rows)?;
     cx.export_function("tableDelete", JsTable::js_delete)?;
+    cx.export_function("tableCleanupOldVersions", JsTable::js_cleanup)?;
+    cx.export_function("tableCompactFiles", JsTable::js_compact)?;
     cx.export_function(
         "tableCreateVectorIndex",
         index::vector::table_create_vector_index,

rust/ffi/node/src/table.rs

@@ -13,6 +13,7 @@
 // limitations under the License.
 
 use arrow_array::RecordBatchIterator;
+use lance::dataset::optimize::CompactionOptions;
 use lance::dataset::{WriteMode, WriteParams};
 use lance::io::object_store::ObjectStoreParams;
 
@@ -69,7 +70,7 @@ impl JsTable {
             store_params: Some(ObjectStoreParams::with_aws_credentials(
                 aws_creds, aws_region,
             )),
-            mode: mode,
+            mode,
             ..WriteParams::default()
         };
 
@@ -120,7 +121,7 @@ impl JsTable {
             let add_result = table.add(batch_reader, Some(params)).await;
 
             deferred.settle_with(&channel, move |mut cx| {
-                let _added = add_result.or_throw(&mut cx)?;
+                add_result.or_throw(&mut cx)?;
                 Ok(cx.boxed(JsTable::from(table)))
             });
         });
@@ -163,4 +164,116 @@ impl JsTable {
         });
         Ok(promise)
     }
+
+    pub(crate) fn js_cleanup(mut cx: FunctionContext) -> JsResult<JsPromise> {
+        let js_table = cx.this().downcast_or_throw::<JsBox<JsTable>, _>(&mut cx)?;
+        let rt = runtime(&mut cx)?;
+        let (deferred, promise) = cx.promise();
+        let table = js_table.table.clone();
+        let channel = cx.channel();
+
+        let older_than: i64 = cx
+            .argument_opt(0)
+            .and_then(|val| val.downcast::<JsNumber, _>(&mut cx).ok())
+            .map(|val| val.value(&mut cx) as i64)
+            .unwrap_or_else(|| 2 * 7 * 24 * 60); // 2 weeks
+        let older_than = chrono::Duration::minutes(older_than);
+        let delete_unverified: bool = cx
+            .argument_opt(1)
+            .and_then(|val| val.downcast::<JsBoolean, _>(&mut cx).ok())
+            .map(|val| val.value(&mut cx))
+            .unwrap_or_default();
+
+        rt.spawn(async move {
+            let stats = table
+                .cleanup_old_versions(older_than, Some(delete_unverified))
+                .await;
+
+            deferred.settle_with(&channel, move |mut cx| {
+                let stats = stats.or_throw(&mut cx)?;
+
+                let output_metrics = JsObject::new(&mut cx);
+                let bytes_removed = cx.number(stats.bytes_removed as f64);
+                output_metrics.set(&mut cx, "bytesRemoved", bytes_removed)?;
+
+                let old_versions = cx.number(stats.old_versions as f64);
+                output_metrics.set(&mut cx, "oldVersions", old_versions)?;
+
+                let output_table = cx.boxed(JsTable::from(table));
+
+                let output = JsObject::new(&mut cx);
+                output.set(&mut cx, "metrics", output_metrics)?;
+                output.set(&mut cx, "newTable", output_table)?;
+
+                Ok(output)
+            })
+        });
+        Ok(promise)
+    }
+
+    pub(crate) fn js_compact(mut cx: FunctionContext) -> JsResult<JsPromise> {
+        let js_table = cx.this().downcast_or_throw::<JsBox<JsTable>, _>(&mut cx)?;
+        let rt = runtime(&mut cx)?;
+        let (deferred, promise) = cx.promise();
+        let mut table = js_table.table.clone();
+        let channel = cx.channel();
+
+        let js_options = cx.argument::<JsObject>(0)?;
+        let mut options = CompactionOptions::default();
+
+        if let Some(target_rows) =
+            js_options.get_opt::<JsNumber, _, _>(&mut cx, "targetRowsPerFragment")?
+        {
+            options.target_rows_per_fragment = target_rows.value(&mut cx) as usize;
+        }
+        if let Some(max_per_group) =
+            js_options.get_opt::<JsNumber, _, _>(&mut cx, "maxRowsPerGroup")?
+        {
+            options.max_rows_per_group = max_per_group.value(&mut cx) as usize;
+        }
+        if let Some(materialize_deletions) =
+            js_options.get_opt::<JsBoolean, _, _>(&mut cx, "materializeDeletions")?
+        {
+            options.materialize_deletions = materialize_deletions.value(&mut cx);
+        }
+        if let Some(materialize_deletions_threshold) =
+            js_options.get_opt::<JsNumber, _, _>(&mut cx, "materializeDeletionsThreshold")?
+        {
+            options.materialize_deletions_threshold =
+                materialize_deletions_threshold.value(&mut cx) as f32;
+        }
+        if let Some(num_threads) = js_options.get_opt::<JsNumber, _, _>(&mut cx, "numThreads")? {
+            options.num_threads = num_threads.value(&mut cx) as usize;
+        }
+
+        rt.spawn(async move {
+            let stats = table.compact_files(options).await;
+
+            deferred.settle_with(&channel, move |mut cx| {
+                let stats = stats.or_throw(&mut cx)?;
+
+                let output_metrics = JsObject::new(&mut cx);
+                let fragments_removed = cx.number(stats.fragments_removed as f64);
+                output_metrics.set(&mut cx, "fragmentsRemoved", fragments_removed)?;
+
+                let fragments_added = cx.number(stats.fragments_added as f64);
+                output_metrics.set(&mut cx, "fragmentsAdded", fragments_added)?;
+
+                let files_removed = cx.number(stats.files_removed as f64);
+                output_metrics.set(&mut cx, "filesRemoved", files_removed)?;
+
+                let files_added = cx.number(stats.files_added as f64);
+                output_metrics.set(&mut cx, "filesAdded", files_added)?;
+
+                let output_table = cx.boxed(JsTable::from(table));
+
+                let output = JsObject::new(&mut cx);
+                output.set(&mut cx, "metrics", output_metrics)?;
+                output.set(&mut cx, "newTable", output_table)?;
+
+                Ok(output)
+            })
+        });
+        Ok(promise)
+    }
 }

rust/vectordb/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "vectordb"
-version = "0.2.6"
+version = "0.3.3"
 edition = "2021"
 description = "LanceDB: A serverless, low-latency vector database for AI applications"
 license = "Apache-2.0"
@@ -16,16 +16,22 @@ arrow-data = { workspace = true }
 arrow-schema = { workspace = true }
 arrow-ord = { workspace = true }
 arrow-cast = { workspace = true }
+chrono = { workspace = true }
 object_store = { workspace = true }
 snafu = { workspace = true }
 half = { workspace = true }
 lance = { workspace = true }
 lance-linalg = { workspace = true }
+lance-testing = { workspace = true }
 tokio = { version = "1.23", features = ["rt-multi-thread"] }
 log = { workspace = true }
+async-trait = "0"
+bytes = "1"
+futures = "0"
 num-traits = "0"
 url = { workspace = true }
 
 [dev-dependencies]
 tempfile = "3.5.0"
 rand = { version = "0.8.3", features = ["small_rng"] }
+walkdir = "2"

rust/vectordb/src/data/inspect.rs

@@ -18,9 +18,9 @@ use arrow::compute::kernels::{aggregate::bool_and, length::length};
 use arrow_array::{
     cast::AsArray,
     types::{ArrowPrimitiveType, Int32Type, Int64Type},
-    Array, GenericListArray, OffsetSizeTrait, RecordBatchReader,
+    Array, GenericListArray, OffsetSizeTrait, PrimitiveArray, RecordBatchReader,
 };
-use arrow_ord::comparison::eq_dyn_scalar;
+use arrow_ord::cmp::eq;
 use arrow_schema::DataType;
 use num_traits::{ToPrimitive, Zero};
 
@@ -38,7 +38,8 @@ where
     }
 
     let dim = len_arr.as_primitive::<T>().value(0);
-    if bool_and(&eq_dyn_scalar(len_arr.as_primitive::<T>(), dim)?) != Some(true) {
+    let datum = PrimitiveArray::<T>::new_scalar(dim);
+    if bool_and(&eq(len_arr.as_primitive::<T>(), &datum)?) != Some(true) {
         Ok(None)
     } else {
         Ok(Some(dim))

rust/vectordb/src/database.rs

@@ -14,13 +14,16 @@
 
 use std::fs::create_dir_all;
 use std::path::Path;
+use std::sync::Arc;
 
 use arrow_array::RecordBatchReader;
 use lance::dataset::WriteParams;
-use lance::io::object_store::ObjectStore;
+use lance::io::object_store::{ObjectStore, WrappingObjectStore};
+use object_store::local::LocalFileSystem;
 use snafu::prelude::*;
 
-use crate::error::{CreateDirSnafu, InvalidTableNameSnafu, Result};
+use crate::error::{CreateDirSnafu, Error, InvalidTableNameSnafu, Result};
+use crate::io::object_store::MirroringObjectStoreWrapper;
 use crate::table::{ReadParams, Table};
 
 pub const LANCE_FILE_EXTENSION: &str = "lance";
@@ -31,10 +34,14 @@ pub struct Database {
 
     pub(crate) uri: String,
     pub(crate) base_path: object_store::path::Path,
+
+    // the object store wrapper to use on write path
+    pub(crate) store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
 }
 
 const LANCE_EXTENSION: &str = "lance";
 const ENGINE: &str = "engine";
+const MIRRORED_STORE: &str = "mirroredStore";
 
 /// A connection to LanceDB
 impl Database {
@@ -55,6 +62,7 @@ impl Database {
             Ok(mut url) => {
                 // iter thru the query params and extract the commit store param
                 let mut engine = None;
+                let mut mirrored_store = None;
                 let mut filtered_querys = vec![];
 
                 // WARNING: specifying engine is NOT a publicly supported feature in lancedb yet
@@ -62,6 +70,13 @@ impl Database {
                 for (key, value) in url.query_pairs() {
                     if key == ENGINE {
                         engine = Some(value.to_string());
+                    } else if key == MIRRORED_STORE {
+                        if cfg!(windows) {
+                            return Err(Error::Lance {
+                                message: "mirrored store is not supported on windows".into(),
+                            });
+                        }
+                        mirrored_store = Some(value.to_string());
                     } else {
                         // to owned so we can modify the url
                         filtered_querys.push((key.to_string(), value.to_string()));
@@ -96,11 +111,21 @@ impl Database {
                     Self::try_create_dir(&plain_uri).context(CreateDirSnafu { path: plain_uri })?;
                 }
 
+                let write_store_wrapper = match mirrored_store {
+                    Some(path) => {
+                        let mirrored_store = Arc::new(LocalFileSystem::new_with_prefix(path)?);
+                        let wrapper = MirroringObjectStoreWrapper::new(mirrored_store);
+                        Some(Arc::new(wrapper) as Arc<dyn WrappingObjectStore>)
+                    }
+                    None => None,
+                };
+
                 Ok(Database {
                     uri: table_base_uri,
                     query_string,
                     base_path,
                     object_store,
+                    store_wrapper: write_store_wrapper,
                 })
             }
             Err(_) => Self::open_path(uri).await,
@@ -110,13 +135,14 @@ impl Database {
     async fn open_path(path: &str) -> Result<Database> {
         let (object_store, base_path) = ObjectStore::from_uri(path).await?;
         if object_store.is_local() {
-            Self::try_create_dir(path).context(CreateDirSnafu { path: path })?;
+            Self::try_create_dir(path).context(CreateDirSnafu { path })?;
         }
         Ok(Self {
             uri: path.to_string(),
             query_string: None,
             base_path,
             object_store,
+            store_wrapper: None,
         })
     }
 
@@ -166,7 +192,15 @@ impl Database {
         params: Option<WriteParams>,
     ) -> Result<Table> {
         let table_uri = self.table_uri(name)?;
-        Table::create(&table_uri, name, batches, params).await
+
+        Table::create(
+            &table_uri,
+            name,
+            batches,
+            self.store_wrapper.clone(),
+            params,
+        )
+        .await
     }
 
     /// Open a table in the database.
@@ -178,7 +212,7 @@ impl Database {
     ///
     /// * A [Table] object.
     pub async fn open_table(&self, name: &str) -> Result<Table> {
-        self.open_table_with_params(name, &ReadParams::default())
+        self.open_table_with_params(name, ReadParams::default())
             .await
     }
 
@@ -191,9 +225,9 @@ impl Database {
     /// # Returns
     ///
     /// * A [Table] object.
-    pub async fn open_table_with_params(&self, name: &str, params: &ReadParams) -> Result<Table> {
+    pub async fn open_table_with_params(&self, name: &str, params: ReadParams) -> Result<Table> {
         let table_uri = self.table_uri(name)?;
-        Table::open_with_params(&table_uri, name, params).await
+        Table::open_with_params(&table_uri, name, self.store_wrapper.clone(), params).await
     }
 
     /// Drop a table in the database.

rust/vectordb/src/index/vector.rs

@@ -95,8 +95,8 @@ impl VectorIndexBuilder for IvfPQIndexBuilder {
     }
 
     fn build(&self) -> VectorIndexParams {
-        let ivf_params = self.ivf_params.clone().unwrap_or(IvfBuildParams::default());
+        let ivf_params = self.ivf_params.clone().unwrap_or_default();
-        let pq_params = self.pq_params.clone().unwrap_or(PQBuildParams::default());
+        let pq_params = self.pq_params.clone().unwrap_or_default();
 
         VectorIndexParams::with_ivf_pq_params(pq_params.metric_type, ivf_params, pq_params)
     }

rust/vectordb/src/io.rs

@@ -0,0 +1 @@
+pub mod object_store;

rust/vectordb/src/io/object_store.rs

@@ -0,0 +1,401 @@
+// Copyright 2023 Lance Developers.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+//! A mirroring object store that mirror writes to a secondary object store
+
+use std::{
+    fmt::Formatter,
+    pin::Pin,
+    sync::Arc,
+    task::{Context, Poll},
+};
+
+use bytes::Bytes;
+use futures::{stream::BoxStream, FutureExt, StreamExt};
+use lance::io::object_store::WrappingObjectStore;
+use object_store::{
+    path::Path, GetOptions, GetResult, ListResult, MultipartId, ObjectMeta, ObjectStore, Result,
+};
+
+use async_trait::async_trait;
+use tokio::{
+    io::{AsyncWrite, AsyncWriteExt},
+    task::JoinHandle,
+};
+
+#[derive(Debug)]
+struct MirroringObjectStore {
+    primary: Arc<dyn ObjectStore>,
+    secondary: Arc<dyn ObjectStore>,
+}
+
+impl std::fmt::Display for MirroringObjectStore {
+    fn fmt(&self, f: &mut Formatter<'_>) -> std::fmt::Result {
+        writeln!(f, "MirrowingObjectStore")?;
+        writeln!(f, "primary:")?;
+        self.primary.fmt(f)?;
+        writeln!(f, "secondary:")?;
+        self.secondary.fmt(f)?;
+        Ok(())
+    }
+}
+
+trait PrimaryOnly {
+    fn primary_only(&self) -> bool;
+}
+
+impl PrimaryOnly for Path {
+    fn primary_only(&self) -> bool {
+        self.filename().unwrap_or("") == "_latest.manifest"
+    }
+}
+
+/// An object store that mirrors write to secondsry object store first
+/// and than commit to primary object store.
+///
+/// This is meant to mirrow writes to a less-durable but lower-latency
+/// store. We have primary store that is durable but slow, and a secondary
+/// store that is fast but not asdurable
+///
+/// Note: this object store does not mirror writes to *.manifest files
+#[async_trait]
+impl ObjectStore for MirroringObjectStore {
+    async fn put(&self, location: &Path, bytes: Bytes) -> Result<()> {
+        if location.primary_only() {
+            self.primary.put(location, bytes).await
+        } else {
+            self.secondary.put(location, bytes.clone()).await?;
+            self.primary.put(location, bytes).await?;
+            Ok(())
+        }
+    }
+
+    async fn put_multipart(
+        &self,
+        location: &Path,
+    ) -> Result<(MultipartId, Box<dyn AsyncWrite + Unpin + Send>)> {
+        if location.primary_only() {
+            return self.primary.put_multipart(location).await;
+        }
+
+        let (id, stream) = self.secondary.put_multipart(location).await?;
+
+        let mirroring_upload = MirroringUpload::new(
+            Pin::new(stream),
+            self.primary.clone(),
+            self.secondary.clone(),
+            location.clone(),
+        );
+
+        Ok((id, Box::new(mirroring_upload)))
+    }
+
+    async fn abort_multipart(&self, location: &Path, multipart_id: &MultipartId) -> Result<()> {
+        if location.primary_only() {
+            return self.primary.abort_multipart(location, multipart_id).await;
+        }
+
+        self.secondary.abort_multipart(location, multipart_id).await
+    }
+
+    // Reads are routed to primary only
+    async fn get_opts(&self, location: &Path, options: GetOptions) -> Result<GetResult> {
+        self.primary.get_opts(location, options).await
+    }
+
+    async fn head(&self, location: &Path) -> Result<ObjectMeta> {
+        self.primary.head(location).await
+    }
+
+    async fn delete(&self, location: &Path) -> Result<()> {
+        if !location.primary_only() {
+            self.secondary.delete(location).await?;
+        }
+        self.primary.delete(location).await
+    }
+
+    async fn list(&self, prefix: Option<&Path>) -> Result<BoxStream<'_, Result<ObjectMeta>>> {
+        self.primary.list(prefix).await
+    }
+
+    async fn list_with_delimiter(&self, prefix: Option<&Path>) -> Result<ListResult> {
+        self.primary.list_with_delimiter(prefix).await
+    }
+
+    async fn copy(&self, from: &Path, to: &Path) -> Result<()> {
+        if to.primary_only() {
+            self.primary.copy(from, to).await
+        } else {
+            self.secondary.copy(from, to).await?;
+            self.primary.copy(from, to).await?;
+            Ok(())
+        }
+    }
+
+    async fn copy_if_not_exists(&self, from: &Path, to: &Path) -> Result<()> {
+        if !to.primary_only() {
+            self.secondary.copy(from, to).await?;
+        }
+        self.primary.copy_if_not_exists(from, to).await
+    }
+}
+
+struct MirroringUpload {
+    secondary_stream: Pin<Box<dyn AsyncWrite + Unpin + Send>>,
+
+    primary_store: Arc<dyn ObjectStore>,
+    secondary_store: Arc<dyn ObjectStore>,
+    location: Path,
+
+    state: MirroringUploadShutdown,
+}
+
+// The state goes from
+// None
+// -> (secondary)ShutingDown
+// -> (secondary)ShutdownDone
+// -> Uploading(to primary)
+// -> Done
+#[derive(Debug)]
+enum MirroringUploadShutdown {
+    None,
+    ShutingDown,
+    ShutdownDone,
+    Uploading(Pin<Box<JoinHandle<()>>>),
+    Completed,
+}
+
+impl MirroringUpload {
+    pub fn new(
+        secondary_stream: Pin<Box<dyn AsyncWrite + Unpin + Send>>,
+        primary_store: Arc<dyn ObjectStore>,
+        secondary_store: Arc<dyn ObjectStore>,
+        location: Path,
+    ) -> Self {
+        Self {
+            secondary_stream,
+            primary_store,
+            secondary_store,
+            location,
+            state: MirroringUploadShutdown::None,
+        }
+    }
+}
+
+impl AsyncWrite for MirroringUpload {
+    fn poll_write(
+        self: Pin<&mut Self>,
+        cx: &mut Context<'_>,
+        buf: &[u8],
+    ) -> Poll<Result<usize, std::io::Error>> {
+        if !matches!(self.state, MirroringUploadShutdown::None) {
+            return Poll::Ready(Err(std::io::Error::new(
+                std::io::ErrorKind::Other,
+                "already shutdown",
+            )));
+        }
+        // Write to secondary first
+        let mut_self = self.get_mut();
+        mut_self.secondary_stream.as_mut().poll_write(cx, buf)
+    }
+
+    fn poll_flush(self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<Result<(), std::io::Error>> {
+        if !matches!(self.state, MirroringUploadShutdown::None) {
+            return Poll::Ready(Err(std::io::Error::new(
+                std::io::ErrorKind::Other,
+                "already shutdown",
+            )));
+        }
+
+        let mut_self = self.get_mut();
+        mut_self.secondary_stream.as_mut().poll_flush(cx)
+    }
+
+    fn poll_shutdown(
+        self: Pin<&mut Self>,
+        cx: &mut Context<'_>,
+    ) -> Poll<Result<(), std::io::Error>> {
+        let mut_self = self.get_mut();
+
+        loop {
+            // try to shutdown secondary first
+            match &mut mut_self.state {
+                MirroringUploadShutdown::None | MirroringUploadShutdown::ShutingDown => {
+                    match mut_self.secondary_stream.as_mut().poll_shutdown(cx) {
+                        Poll::Ready(Ok(())) => {
+                            mut_self.state = MirroringUploadShutdown::ShutdownDone;
+                            // don't return, no waker is setup
+                        }
+                        Poll::Ready(Err(e)) => return Poll::Ready(Err(e)),
+                        Poll::Pending => {
+                            mut_self.state = MirroringUploadShutdown::ShutingDown;
+                            return Poll::Pending;
+                        }
+                    }
+                }
+                MirroringUploadShutdown::ShutdownDone => {
+                    let primary_store = mut_self.primary_store.clone();
+                    let secondary_store = mut_self.secondary_store.clone();
+                    let location = mut_self.location.clone();
+
+                    let upload_future =
+                        Box::pin(tokio::runtime::Handle::current().spawn(async move {
+                            let mut source =
+                                secondary_store.get(&location).await.unwrap().into_stream();
+                            let upload_stream = primary_store.put_multipart(&location).await;
+                            let (_, mut stream) = upload_stream.unwrap();
+
+                            while let Some(buf) = source.next().await {
+                                let buf = buf.unwrap();
+                                stream.write_all(&buf).await.unwrap();
+                            }
+
+                            stream.shutdown().await.unwrap();
+                        }));
+                    mut_self.state = MirroringUploadShutdown::Uploading(upload_future);
+                    // don't return, no waker is setup
+                }
+                MirroringUploadShutdown::Uploading(ref mut join_handle) => {
+                    match join_handle.poll_unpin(cx) {
+                        Poll::Ready(Ok(())) => {
+                            mut_self.state = MirroringUploadShutdown::Completed;
+                            return Poll::Ready(Ok(()));
+                        }
+                        Poll::Ready(Err(e)) => {
+                            mut_self.state = MirroringUploadShutdown::Completed;
+                            return Poll::Ready(Err(e.into()));
+                        }
+                        Poll::Pending => {
+                            return Poll::Pending;
+                        }
+                    }
+                }
+                MirroringUploadShutdown::Completed => {
+                    return Poll::Ready(Err(std::io::Error::new(
+                        std::io::ErrorKind::Other,
+                        "shutdown already completed",
+                    )))
+                }
+            }
+        }
+    }
+}
+
+#[derive(Debug)]
+pub struct MirroringObjectStoreWrapper {
+    secondary: Arc<dyn ObjectStore>,
+}
+
+impl MirroringObjectStoreWrapper {
+    pub fn new(secondary: Arc<dyn ObjectStore>) -> Self {
+        Self { secondary }
+    }
+}
+
+impl WrappingObjectStore for MirroringObjectStoreWrapper {
+    fn wrap(&self, primary: Arc<dyn ObjectStore>) -> Arc<dyn ObjectStore> {
+        Arc::new(MirroringObjectStore {
+            primary,
+            secondary: self.secondary.clone(),
+        })
+    }
+}
+
+// windows pathing can't be simply concatenated
+#[cfg(all(test, not(windows)))]
+mod test {
+    use super::*;
+    use crate::Database;
+    use arrow_array::PrimitiveArray;
+    use futures::TryStreamExt;
+    use lance::{dataset::WriteParams, io::object_store::ObjectStoreParams};
+    use lance_testing::datagen::{BatchGenerator, IncrementingInt32, RandomVector};
+    use object_store::local::LocalFileSystem;
+    use tempfile;
+
+    #[tokio::test]
+    async fn test_e2e() {
+        let dir1 = tempfile::tempdir().unwrap().into_path();
+        let dir2 = tempfile::tempdir().unwrap().into_path();
+
+        let secondary_store = LocalFileSystem::new_with_prefix(dir2.to_str().unwrap()).unwrap();
+        let object_store_wrapper = Arc::new(MirroringObjectStoreWrapper {
+            secondary: Arc::new(secondary_store),
+        });
+
+        let db = Database::connect(dir1.to_str().unwrap()).await.unwrap();
+
+        let mut param = WriteParams::default();
+        let mut store_params = ObjectStoreParams::default();
+        store_params.object_store_wrapper = Some(object_store_wrapper);
+        param.store_params = Some(store_params);
+
+        let mut datagen = BatchGenerator::new();
+        datagen = datagen.col(Box::new(IncrementingInt32::default()));
+        datagen = datagen.col(Box::new(RandomVector::default().named("vector".into())));
+
+        let res = db
+            .create_table("test", datagen.batch(100), Some(param.clone()))
+            .await;
+
+        // leave this here for easy debugging
+        let t = res.unwrap();
+
+        assert_eq!(t.count_rows().await.unwrap(), 100);
+
+        let q = t
+            .search(PrimitiveArray::from_iter_values(vec![0.1, 0.1, 0.1, 0.1]))
+            .limit(10)
+            .execute()
+            .await
+            .unwrap();
+
+        let bateches = q.try_collect::<Vec<_>>().await.unwrap();
+        assert_eq!(bateches.len(), 1);
+        assert_eq!(bateches[0].num_rows(), 10);
+
+        use walkdir::WalkDir;
+
+        let primary_location = dir1.join("test.lance").canonicalize().unwrap();
+        let secondary_location = dir2.join(primary_location.strip_prefix("/").unwrap());
+
+        let mut primary_iter = WalkDir::new(&primary_location).into_iter();
+        let mut secondary_iter = WalkDir::new(&secondary_location).into_iter();
+
+        let mut primary_elem = primary_iter.next();
+        let mut secondary_elem = secondary_iter.next();
+
+        loop {
+            if primary_elem.is_none() && secondary_elem.is_none() {
+                break;
+            }
+            // primary has more data then secondary, should not run out before secondary
+            let primary_f = primary_elem.unwrap().unwrap();
+            // hit manifest, skip, _versions contains all the manifest and should not exist on secondary
+            let primary_raw_path = primary_f.file_name().to_str().unwrap();
+            if primary_raw_path.contains("_latest.manifest") {
+                primary_elem = primary_iter.next();
+                continue;
+            }
+            let secondary_f = secondary_elem.unwrap().unwrap();
+            assert_eq!(
+                primary_f.path().strip_prefix(&primary_location),
+                secondary_f.path().strip_prefix(&secondary_location)
+            );
+
+            primary_elem = primary_iter.next();
+            secondary_elem = secondary_iter.next();
+        }
+    }
+}

rust/vectordb/src/lib.rs

@@ -16,8 +16,10 @@ pub mod data;
 pub mod database;
 pub mod error;
 pub mod index;
+pub mod io;
 pub mod query;
 pub mod table;
+pub mod utils;
 
 pub use database::Database;
 pub use table::Table;

rust/vectordb/src/table.rs

@@ -12,17 +12,22 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+use chrono::Duration;
 use std::sync::Arc;
 
 use arrow_array::{Float32Array, RecordBatchReader};
 use arrow_schema::SchemaRef;
+use lance::dataset::cleanup::RemovalStats;
+use lance::dataset::optimize::{compact_files, CompactionMetrics, CompactionOptions};
 use lance::dataset::{Dataset, WriteParams};
 use lance::index::IndexType;
+use lance::io::object_store::WrappingObjectStore;
 use std::path::Path;
 
 use crate::error::{Error, Result};
 use crate::index::vector::VectorIndexBuilder;
 use crate::query::Query;
+use crate::utils::{PatchReadParam, PatchWriteParam};
 use crate::WriteMode;
 
 pub use lance::dataset::ReadParams;
@@ -35,6 +40,9 @@ pub struct Table {
     name: String,
     uri: String,
     dataset: Arc<Dataset>,
+
+    // the object store wrapper to use on write path
+    store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
 }
 
 impl std::fmt::Display for Table {
@@ -56,12 +64,12 @@ impl Table {
     /// * A [Table] object.
     pub async fn open(uri: &str) -> Result<Self> {
         let name = Self::get_table_name(uri)?;
-        Self::open_with_params(uri, &name, &ReadParams::default()).await
+        Self::open_with_params(uri, &name, None, ReadParams::default()).await
     }
 
     /// Open an Table with a given name.
     pub async fn open_with_name(uri: &str, name: &str) -> Result<Self> {
-        Self::open_with_params(uri, name, &ReadParams::default()).await
+        Self::open_with_params(uri, name, None, ReadParams::default()).await
     }
 
     /// Opens an existing Table
@@ -75,8 +83,18 @@ impl Table {
     /// # Returns
     ///
     /// * A [Table] object.
-    pub async fn open_with_params(uri: &str, name: &str, params: &ReadParams) -> Result<Self> {
+    pub async fn open_with_params(
-        let dataset = Dataset::open_with_params(uri, params)
+        uri: &str,
+        name: &str,
+        write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
+        params: ReadParams,
+    ) -> Result<Self> {
+        // patch the params if we have a write store wrapper
+        let params = match write_store_wrapper.clone() {
+            Some(wrapper) => params.patch_with_store_wrapper(wrapper)?,
+            None => params,
+        };
+        let dataset = Dataset::open_with_params(uri, &params)
             .await
             .map_err(|e| match e {
                 lance::Error::DatasetNotFound { .. } => Error::TableNotFound {
@@ -90,6 +108,7 @@ impl Table {
             name: name.to_string(),
             uri: uri.to_string(),
             dataset: Arc::new(dataset),
+            store_wrapper: write_store_wrapper,
         })
     }
 
@@ -97,20 +116,26 @@ impl Table {
     ///
     pub async fn checkout(uri: &str, version: u64) -> Result<Self> {
         let name = Self::get_table_name(uri)?;
-        Self::checkout_with_params(uri, &name, version, &ReadParams::default()).await
+        Self::checkout_with_params(uri, &name, version, None, ReadParams::default()).await
     }
 
     pub async fn checkout_with_name(uri: &str, name: &str, version: u64) -> Result<Self> {
-        Self::checkout_with_params(uri, name, version, &ReadParams::default()).await
+        Self::checkout_with_params(uri, name, version, None, ReadParams::default()).await
     }
 
     pub async fn checkout_with_params(
         uri: &str,
         name: &str,
         version: u64,
-        params: &ReadParams,
+        write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
+        params: ReadParams,
     ) -> Result<Self> {
-        let dataset = Dataset::checkout_with_params(uri, version, params)
+        // patch the params if we have a write store wrapper
+        let params = match write_store_wrapper.clone() {
+            Some(wrapper) => params.patch_with_store_wrapper(wrapper)?,
+            None => params,
+        };
+        let dataset = Dataset::checkout_with_params(uri, version, &params)
             .await
             .map_err(|e| match e {
                 lance::Error::DatasetNotFound { .. } => Error::TableNotFound {
@@ -124,6 +149,23 @@ impl Table {
             name: name.to_string(),
             uri: uri.to_string(),
             dataset: Arc::new(dataset),
+            store_wrapper: write_store_wrapper,
+        })
+    }
+
+    pub async fn checkout_latest(&self) -> Result<Self> {
+        let latest_version_id = self.dataset.latest_version_id().await?;
+        let dataset = if latest_version_id == self.dataset.version().version {
+            self.dataset.clone()
+        } else {
+            Arc::new(self.dataset.checkout_version(latest_version_id).await?)
+        };
+
+        Ok(Table {
+            name: self.name.clone(),
+            uri: self.uri.clone(),
+            dataset,
+            store_wrapper: self.store_wrapper.clone(),
         })
     }
 
@@ -157,8 +199,15 @@ impl Table {
         uri: &str,
         name: &str,
         batches: impl RecordBatchReader + Send + 'static,
+        write_store_wrapper: Option<Arc<dyn WrappingObjectStore>>,
         params: Option<WriteParams>,
     ) -> Result<Self> {
+        // patch the params if we have a write store wrapper
+        let params = match write_store_wrapper.clone() {
+            Some(wrapper) => params.patch_with_store_wrapper(wrapper)?,
+            None => params,
+        };
+
         let dataset = Dataset::write(batches, uri, params)
             .await
             .map_err(|e| match e {
@@ -173,6 +222,7 @@ impl Table {
             name: name.to_string(),
             uri: uri.to_string(),
             dataset: Arc::new(dataset),
+            store_wrapper: write_store_wrapper,
         })
     }
 
@@ -190,8 +240,8 @@ impl Table {
     pub async fn create_index(&mut self, index_builder: &impl VectorIndexBuilder) -> Result<()> {
         use lance::index::DatasetIndexExt;
 
-        let dataset = self
+        let mut dataset = self.dataset.as_ref().clone();
-            .dataset
+        dataset
             .create_index(
                 &[index_builder
                     .get_column()
@@ -221,12 +271,18 @@ impl Table {
         batches: impl RecordBatchReader + Send + 'static,
         params: Option<WriteParams>,
     ) -> Result<()> {
-        let params = params.unwrap_or(WriteParams {
+        let params = Some(params.unwrap_or(WriteParams {
             mode: WriteMode::Append,
             ..WriteParams::default()
-        });
+        }));
 
-        self.dataset = Arc::new(Dataset::write(batches, &self.uri, Some(params)).await?);
+        // patch the params if we have a write store wrapper
+        let params = match self.store_wrapper.clone() {
+            Some(wrapper) => params.patch_with_store_wrapper(wrapper)?,
+            None => params,
+        };
+
+        self.dataset = Arc::new(Dataset::write(batches, &self.uri, params).await?);
         Ok(())
     }
 
@@ -268,6 +324,41 @@ impl Table {
         self.dataset = Arc::new(dataset);
         Ok(())
     }
+
+    /// Remove old versions of the dataset from disk.
+    ///
+    /// # Arguments
+    /// * `older_than` - The duration of time to keep versions of the dataset.
+    /// * `delete_unverified` - Because they may be part of an in-progress
+    ///   transaction, files newer than 7 days old are not deleted by default.
+    ///   If you are sure that there are no in-progress transactions, then you
+    ///   can set this to True to delete all files older than `older_than`.
+    ///
+    /// This calls into [lance::dataset::Dataset::cleanup_old_versions] and
+    /// returns the result.
+    pub async fn cleanup_old_versions(
+        &self,
+        older_than: Duration,
+        delete_unverified: Option<bool>,
+    ) -> Result<RemovalStats> {
+        Ok(self
+            .dataset
+            .cleanup_old_versions(older_than, delete_unverified)
+            .await?)
+    }
+
+    /// Compact files in the dataset.
+    ///
+    /// This can be run after making several small appends to optimize the table
+    /// for faster reads.
+    ///
+    /// This calls into [lance::dataset::optimize::compact_files].
+    pub async fn compact_files(&mut self, options: CompactionOptions) -> Result<CompactionMetrics> {
+        let mut dataset = self.dataset.as_ref().clone();
+        let metrics = compact_files(&mut dataset, options, None).await?;
+        self.dataset = Arc::new(dataset);
+        Ok(metrics)
+    }
 }
 
 #[cfg(test)]
@@ -330,10 +421,12 @@ mod tests {
 
         let batches = make_test_batches();
         let _ = batches.schema().clone();
-        Table::create(&uri, "test", batches, None).await.unwrap();
+        Table::create(&uri, "test", batches, None, None)
+            .await
+            .unwrap();
 
         let batches = make_test_batches();
-        let result = Table::create(&uri, "test", batches, None).await;
+        let result = Table::create(&uri, "test", batches, None, None).await;
         assert!(matches!(
             result.unwrap_err(),
             Error::TableAlreadyExists { .. }
@@ -347,7 +440,9 @@ mod tests {
 
         let batches = make_test_batches();
         let schema = batches.schema().clone();
-        let mut table = Table::create(&uri, "test", batches, None).await.unwrap();
+        let mut table = Table::create(&uri, "test", batches, None, None)
+            .await
+            .unwrap();
         assert_eq!(table.count_rows().await.unwrap(), 10);
 
         let new_batches = RecordBatchIterator::new(
@@ -373,7 +468,9 @@ mod tests {
 
         let batches = make_test_batches();
         let schema = batches.schema().clone();
-        let mut table = Table::create(uri, "test", batches, None).await.unwrap();
+        let mut table = Table::create(uri, "test", batches, None, None)
+            .await
+            .unwrap();
         assert_eq!(table.count_rows().await.unwrap(), 10);
 
         let new_batches = RecordBatchIterator::new(
@@ -456,7 +553,9 @@ mod tests {
             ..Default::default()
         };
         assert!(!wrapper.called());
-        let _ = Table::open_with_params(uri, "test", &param).await.unwrap();
+        let _ = Table::open_with_params(uri, "test", None, param)
+            .await
+            .unwrap();
         assert!(wrapper.called());
     }
 
@@ -508,7 +607,9 @@ mod tests {
             schema,
         );
 
-        let mut table = Table::create(uri, "test", batches, None).await.unwrap();
+        let mut table = Table::create(uri, "test", batches, None, None)
+            .await
+            .unwrap();
         let mut i = IvfPQIndexBuilder::new();
 
         let index_builder = i

rust/vectordb/src/utils.rs

@@ -0,0 +1,67 @@
+use std::sync::Arc;
+
+use lance::{
+    dataset::{ReadParams, WriteParams},
+    io::object_store::{ObjectStoreParams, WrappingObjectStore},
+};
+
+use crate::error::{Error, Result};
+
+pub trait PatchStoreParam {
+    fn patch_with_store_wrapper(
+        self,
+        wrapper: Arc<dyn WrappingObjectStore>,
+    ) -> Result<Option<ObjectStoreParams>>;
+}
+
+impl PatchStoreParam for Option<ObjectStoreParams> {
+    fn patch_with_store_wrapper(
+        self,
+        wrapper: Arc<dyn WrappingObjectStore>,
+    ) -> Result<Option<ObjectStoreParams>> {
+        let mut params = self.unwrap_or_default();
+        if params.object_store_wrapper.is_some() {
+            return Err(Error::Lance {
+                message: "can not patch param because object store is already set".into(),
+            });
+        }
+        params.object_store_wrapper = Some(wrapper);
+
+        Ok(Some(params))
+    }
+}
+
+pub trait PatchWriteParam {
+    fn patch_with_store_wrapper(
+        self,
+        wrapper: Arc<dyn WrappingObjectStore>,
+    ) -> Result<Option<WriteParams>>;
+}
+
+impl PatchWriteParam for Option<WriteParams> {
+    fn patch_with_store_wrapper(
+        self,
+        wrapper: Arc<dyn WrappingObjectStore>,
+    ) -> Result<Option<WriteParams>> {
+        let mut params = self.unwrap_or_default();
+        params.store_params = params.store_params.patch_with_store_wrapper(wrapper)?;
+        Ok(Some(params))
+    }
+}
+
+// NOTE: we have some API inconsistency here.
+// WriteParam is found in the form of Option<WriteParam> and ReadParam is found in the form of ReadParam
+
+pub trait PatchReadParam {
+    fn patch_with_store_wrapper(self, wrapper: Arc<dyn WrappingObjectStore>) -> Result<ReadParams>;
+}
+
+impl PatchReadParam for ReadParams {
+    fn patch_with_store_wrapper(
+        mut self,
+        wrapper: Arc<dyn WrappingObjectStore>,
+    ) -> Result<ReadParams> {
+        self.store_options = self.store_options.patch_with_store_wrapper(wrapper)?;
+        Ok(self)
+    }
+}