Bump version: 0.11.0-beta.0 → 0.11.0

Changelog: https://github.com/lancedb/lance/releases/tag/v0.15.0

* Fixes #1466
* Closes #1475
* Fixes #1446

.bumpversion.cfg

@@ -1,12 +0,0 @@
-[bumpversion]
-current_version = 0.4.7
-commit = True
-message = Bump version: {current_version} → {new_version}
-tag = True
-tag_name = v{new_version}
-
-[bumpversion:file:node/package.json]
-
-[bumpversion:file:rust/ffi/node/Cargo.toml]
-
-[bumpversion:file:rust/vectordb/Cargo.toml]

.bumpversion.toml

@@ -0,0 +1,57 @@
+[tool.bumpversion]
+current_version = "0.7.2"
+parse = """(?x)
+    (?P<major>0|[1-9]\\d*)\\.
+    (?P<minor>0|[1-9]\\d*)\\.
+    (?P<patch>0|[1-9]\\d*)
+    (?:-(?P<pre_l>[a-zA-Z-]+)\\.(?P<pre_n>0|[1-9]\\d*))?
+"""
+serialize = [
+    "{major}.{minor}.{patch}-{pre_l}.{pre_n}",
+    "{major}.{minor}.{patch}",
+]
+search = "{current_version}"
+replace = "{new_version}"
+regex = false
+ignore_missing_version = false
+ignore_missing_files = false
+tag = true
+sign_tags = false
+tag_name = "v{new_version}"
+tag_message = "Bump version: {current_version} → {new_version}"
+allow_dirty = true
+commit = true
+message = "Bump version: {current_version} → {new_version}"
+commit_args = ""
+
+[tool.bumpversion.parts.pre_l]
+values = ["beta", "final"]
+optional_value = "final"
+
+[[tool.bumpversion.files]]
+filename = "node/package.json"
+search = "\"version\": \"{current_version}\","
+replace = "\"version\": \"{new_version}\","
+
+[[tool.bumpversion.files]]
+filename = "nodejs/package.json"
+search = "\"version\": \"{current_version}\","
+replace = "\"version\": \"{new_version}\","
+
+# nodejs binary packages
+[[tool.bumpversion.files]]
+glob = "nodejs/npm/*/package.json"
+search = "\"version\": \"{current_version}\","
+replace = "\"version\": \"{new_version}\","
+
+# Cargo files
+# ------------
+[[tool.bumpversion.files]]
+filename = "rust/ffi/node/Cargo.toml"
+search = "\nversion = \"{current_version}\""
+replace = "\nversion = \"{new_version}\""
+
+[[tool.bumpversion.files]]
+filename = "rust/lancedb/Cargo.toml"
+search = "\nversion = \"{current_version}\""
+replace = "\nversion = \"{new_version}\""

.cargo/config.toml

@@ -0,0 +1,40 @@
+[profile.release]
+lto = "fat"
+codegen-units = 1
+
+[profile.release-with-debug]
+inherits = "release"
+debug = true
+# Prioritize compile time over runtime performance
+codegen-units = 16
+lto = "thin"
+
+[target.'cfg(all())']
+rustflags = [
+    "-Wclippy::all",
+    "-Wclippy::style",
+    "-Wclippy::fallible_impl_from",
+    "-Wclippy::manual_let_else",
+    "-Wclippy::redundant_pub_crate",
+    "-Wclippy::string_add_assign",
+    "-Wclippy::string_add",
+    "-Wclippy::string_lit_as_bytes",
+    "-Wclippy::string_to_string",
+    "-Wclippy::use_self",
+    "-Dclippy::cargo",
+    "-Dclippy::dbg_macro",
+    # not too much we can do to avoid multiple crate versions
+    "-Aclippy::multiple-crate-versions",
+    "-Aclippy::wildcard_dependencies",
+]
+
+[target.x86_64-unknown-linux-gnu]
+rustflags = ["-C", "target-cpu=haswell", "-C", "target-feature=+avx2,+fma,+f16c"]
+
+[target.aarch64-apple-darwin]
+rustflags = ["-C", "target-cpu=apple-m1", "-C", "target-feature=+neon,+fp16,+fhm,+dotprod"]
+
+# Not all Windows systems have the C runtime installed, so this avoids library
+# not found errors on systems that are missing it.
+[target.x86_64-pc-windows-msvc]
+rustflags = ["-Ctarget-feature=+crt-static"]

.github/labeler.yml

@@ -0,0 +1,33 @@
+version: 1
+appendOnly: true
+# Labels are applied based on conventional commits standard
+# https://www.conventionalcommits.org/en/v1.0.0/
+# These labels are later used in release notes. See .github/release.yml
+labels:
+# If the PR title has an ! before the : it will be considered a breaking change
+# For example, `feat!: add new feature` will be considered a breaking change
+- label: breaking-change
+  title: "^[^:]+!:.*"
+- label: breaking-change
+  body: "BREAKING CHANGE"
+- label: enhancement
+  title: "^feat(\\(.+\\))?!?:.*"
+- label: bug
+  title: "^fix(\\(.+\\))?!?:.*"
+- label: documentation
+  title: "^docs(\\(.+\\))?!?:.*"
+- label: performance
+  title: "^perf(\\(.+\\))?!?:.*"
+- label: ci
+  title: "^ci(\\(.+\\))?!?:.*"
+- label: chore
+  title: "^(chore|test|build|style)(\\(.+\\))?!?:.*"
+- label: Python
+  files:
+    - "^python\\/.*"
+- label: Rust
+  files:
+    - "^rust\\/.*"
+- label: typescript
+  files:
+    - "^node\\/.*"

.github/release_notes.json

@@ -0,0 +1,41 @@
+{
+    "ignore_labels": ["chore"],
+    "pr_template": "- ${{TITLE}} by @${{AUTHOR}} in ${{URL}}",
+    "categories": [
+        {
+            "title": "## 🏆 Highlights",
+            "labels": ["highlight"]
+        },
+        {
+            "title": "## 🛠 Breaking Changes",
+            "labels": ["breaking-change"]
+        },
+        {
+            "title": "## ⚠️ Deprecations ",
+            "labels": ["deprecation"]
+        },
+        {
+            "title": "## 🎉 New Features",
+            "labels": ["enhancement"]
+        },
+        {
+            "title": "## 🐛 Bug Fixes",
+            "labels": ["bug"]
+        },
+        {
+            "title": "## 📚 Documentation",
+            "labels": ["documentation"]
+        },
+        {
+            "title": "## 🚀 Performance Improvements",
+            "labels": ["performance"]
+        },
+        {
+            "title": "## Other Changes"
+        },
+        {
+            "title": "## 🔧 Build and CI",
+            "labels": ["ci"]
+        }
+    ]
+}

.github/workflows/build_linux_wheel/action.yml

@@ -0,0 +1,63 @@
+# We create a composite action to be re-used both for testing and for releasing
+name: build-linux-wheel
+description: "Build a manylinux wheel for lance"
+inputs:
+  python-minor-version:
+    description: "8, 9, 10, 11, 12"
+    required: true
+  args:
+    description: "--release"
+    required: false
+    default: ""
+  arm-build:
+    description: "Build for arm64 instead of x86_64"
+    # Note: this does *not* mean the host is arm64, since we might be cross-compiling.
+    required: false
+    default: "false"
+  manylinux:
+    description: "The manylinux version to build for"
+    required: false
+    default: "2_17"
+runs:
+  using: "composite"
+  steps:
+    - name: CONFIRM ARM BUILD
+      shell: bash
+      run: |
+        echo "ARM BUILD: ${{ inputs.arm-build }}"
+    - name: Build x86_64 Manylinux wheel
+      if: ${{ inputs.arm-build == 'false' }}
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        working-directory: python
+        target: x86_64-unknown-linux-gnu
+        manylinux: ${{ inputs.manylinux }}
+        args: ${{ inputs.args }}
+        before-script-linux: |
+          set -e
+          yum install -y openssl-devel \
+            && curl -L https://github.com/protocolbuffers/protobuf/releases/download/v24.4/protoc-24.4-linux-$(uname -m).zip > /tmp/protoc.zip \
+            && unzip /tmp/protoc.zip -d /usr/local \
+            && rm /tmp/protoc.zip
+    - name: Build Arm Manylinux Wheel
+      if: ${{ inputs.arm-build == 'true' }}
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        working-directory: python
+        docker-options: "-e PIP_EXTRA_INDEX_URL=https://pypi.fury.io/lancedb/"
+        target: aarch64-unknown-linux-gnu
+        manylinux: ${{ inputs.manylinux }}
+        args: ${{ inputs.args }}
+        before-script-linux: |
+          set -e
+          apt install -y unzip
+          if [ $(uname -m) = "x86_64" ]; then
+            PROTOC_ARCH="x86_64"
+          else
+            PROTOC_ARCH="aarch_64"
+          fi
+          curl -L https://github.com/protocolbuffers/protobuf/releases/download/v24.4/protoc-24.4-linux-$PROTOC_ARCH.zip > /tmp/protoc.zip \
+            && unzip /tmp/protoc.zip -d /usr/local \
+            && rm /tmp/protoc.zip

.github/workflows/build_mac_wheel/action.yml

@@ -0,0 +1,26 @@
+# We create a composite action to be re-used both for testing and for releasing
+name: build_wheel
+description: "Build a lance wheel"
+inputs:
+  python-minor-version:
+    description: "8, 9, 10, 11"
+    required: true
+  args:
+    description: "--release"
+    required: false
+    default: ""
+runs:
+  using: "composite"
+  steps:
+    - name: Install macos dependency
+      shell: bash
+      run: |
+        brew install protobuf
+    - name: Build wheel
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        args: ${{ inputs.args }}
+        docker-options: "-e PIP_EXTRA_INDEX_URL=https://pypi.fury.io/lancedb/"
+        working-directory: python
+        interpreter: 3.${{ inputs.python-minor-version }}

.github/workflows/build_windows_wheel/action.yml

@@ -0,0 +1,34 @@
+# We create a composite action to be re-used both for testing and for releasing
+name: build_wheel
+description: "Build a lance wheel"
+inputs:
+  python-minor-version:
+    description: "8, 9, 10, 11"
+    required: true
+  args:
+    description: "--release"
+    required: false
+    default: ""
+runs:
+  using: "composite"
+  steps:
+    - name: Install Protoc v21.12
+      working-directory: C:\
+      run: |
+        New-Item -Path 'C:\protoc' -ItemType Directory
+        Set-Location C:\protoc
+        Invoke-WebRequest https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-win64.zip -OutFile C:\protoc\protoc.zip
+        7z x protoc.zip
+        Add-Content $env:GITHUB_PATH "C:\protoc\bin"
+      shell: powershell
+    - name: Build wheel
+      uses: PyO3/maturin-action@v1
+      with:
+        command: build
+        args: ${{ inputs.args }}
+        docker-options: "-e PIP_EXTRA_INDEX_URL=https://pypi.fury.io/lancedb/"
+        working-directory: python
+    - uses: actions/upload-artifact@v3
+      with:
+        name: windows-wheels
+        path: python\target\wheels

.github/workflows/cargo-publish.yml

@@ -1,13 +1,20 @@
 name: Cargo Publish
 
 on:
-  release:
-    types: [ published ]
+  push:
+    tags-ignore:
+      # We don't publish pre-releases for Rust. Crates.io is just a source
+      # distribution, so we don't need to publish pre-releases.
+      - 'v*-beta*'
+      - '*-v*' # for example, python-vX.Y.Z
 
 env:
   # This env var is used by Swatinem/rust-cache@v2 for the cache
   # key, so we set it to make sure it is always consistent.
   CARGO_TERM_COLOR: always
+  # Up-to-date compilers needed for fp16kernels.
+  CC: gcc-12
+  CXX: g++-12
 
 jobs:
   build:
@@ -26,4 +33,4 @@ jobs:
           sudo apt install -y protobuf-compiler libssl-dev
       - name: Publish the package
         run: |
-          cargo publish -p vectordb --all-features --token ${{ secrets.CARGO_REGISTRY_TOKEN }}
+          cargo publish -p lancedb --all-features --token ${{ secrets.CARGO_REGISTRY_TOKEN }}

.github/workflows/dev.yml

@@ -0,0 +1,81 @@
+name: PR Checks
+
+on:
+  pull_request_target:
+    types: [opened, edited, synchronize, reopened]
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  labeler:
+    permissions:
+      pull-requests: write
+    name: Label PR
+    runs-on: ubuntu-latest
+    steps:
+      - uses: srvaroa/labeler@master
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+  commitlint:
+    permissions:
+      pull-requests: write
+    name: Verify PR title / description conforms to semantic-release
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/setup-node@v3
+        with:
+          node-version: "18"
+      # These rules are disabled because Github will always ensure there
+      # is a blank line between the title and the body and Github will
+      # word wrap the description field to ensure a reasonable max line
+      # length.
+      - run: npm install @commitlint/config-conventional
+      - run: >
+          echo 'module.exports = {
+            "rules": {
+              "body-max-line-length": [0, "always", Infinity],
+              "footer-max-line-length": [0, "always", Infinity],
+              "body-leading-blank": [0, "always"]
+            }
+          }' > .commitlintrc.js
+      - run: npx commitlint --extends @commitlint/config-conventional --verbose <<< $COMMIT_MSG
+        env:
+          COMMIT_MSG: >
+            ${{ github.event.pull_request.title }}
+
+            ${{ github.event.pull_request.body }}
+      - if: failure()
+        uses: actions/github-script@v6
+        with:
+          script: |
+            const message = `**ACTION NEEDED**
+              
+              Lance follows the [Conventional Commits specification](https://www.conventionalcommits.org/en/v1.0.0/) for release automation.
+
+              The PR title and description are used as the merge commit message.\
+              Please update your PR title and description to match the specification.
+
+              For details on the error please inspect the "PR Title Check" action.
+              `
+            // Get list of current comments
+            const comments = await github.paginate(github.rest.issues.listComments, {
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              issue_number: context.issue.number
+            });
+            // Check if this job already commented
+            for (const comment of comments) {
+              if (comment.body === message) {
+                return // Already commented
+              }
+            }
+            // Post the comment about Conventional Commits
+            github.rest.issues.createComment({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              issue_number: context.issue.number,
+              body: message
+            })
+            core.setFailed(message)

.github/workflows/docs.yml

@@ -24,10 +24,14 @@ jobs:
     environment:
       name: github-pages
       url: ${{ steps.deployment.outputs.page_url }}
-    runs-on: ubuntu-22.04
+    runs-on: buildjet-8vcpu-ubuntu-2204
     steps:
       - name: Checkout
         uses: actions/checkout@v4
+      - name: Install dependecies needed for ubuntu
+        run: |
+          sudo apt install -y protobuf-compiler libssl-dev
+          rustup update && rustup default        
       - name: Set up Python
         uses: actions/setup-python@v5
         with:

.github/workflows/docs_test.yml

@@ -18,22 +18,30 @@ on:
 env:
   # Disable full debug symbol generation to speed up CI build and keep memory down
   # "1" means line tables only, which is useful for panic tracebacks.
-  RUSTFLAGS: "-C debuginfo=1 -C target-cpu=native -C target-feature=+f16c,+avx2,+fma"
+  RUSTFLAGS: "-C debuginfo=1 -C target-cpu=haswell -C target-feature=+f16c,+avx2,+fma"
   RUST_BACKTRACE: "1"
 
 jobs:
   test-python:
     name: Test doc python code
-    runs-on: "ubuntu-latest"
+    runs-on: "warp-ubuntu-latest-x64-4x"
     steps:
     - name: Checkout
       uses: actions/checkout@v4
+    - name: Print CPU capabilities
+      run: cat /proc/cpuinfo
+    - name: Install dependecies needed for ubuntu
+      run: |
+        sudo apt install -y protobuf-compiler libssl-dev
+        rustup update && rustup default
     - name: Set up Python
       uses: actions/setup-python@v5
       with:
         python-version: 3.11
         cache: "pip"
         cache-dependency-path: "docs/test/requirements.txt"
+    - name: Rust cache
+      uses: swatinem/rust-cache@v2
     - name: Build Python
       working-directory: docs/test
       run:
@@ -48,13 +56,18 @@ jobs:
         for d in *; do cd "$d"; echo "$d".py; python "$d".py; cd ..; done
   test-node:
     name: Test doc nodejs code
-    runs-on: "ubuntu-latest"
+    runs-on: "warp-ubuntu-latest-x64-4x"
+    timeout-minutes: 60
+    strategy:
+      fail-fast: false
     steps:
     - name: Checkout
       uses: actions/checkout@v4
       with:
         fetch-depth: 0
         lfs: true
+    - name: Print CPU capabilities
+      run: cat /proc/cpuinfo
     - name: Set up Node
       uses: actions/setup-node@v4
       with:
@@ -62,10 +75,17 @@ jobs:
     - name: Install dependecies needed for ubuntu
       run: |
         sudo apt install -y protobuf-compiler libssl-dev
+        rustup update && rustup default
     - name: Rust cache
       uses: swatinem/rust-cache@v2
     - name: Install node dependencies
       run: |
+        sudo swapoff -a
+        sudo fallocate -l 8G /swapfile
+        sudo chmod 600 /swapfile
+        sudo mkswap /swapfile
+        sudo swapon /swapfile
+        sudo swapon --show
         cd node
         npm ci
         npm run build-release

.github/workflows/java.yml

@@ -0,0 +1,85 @@
+name: Build and Run Java JNI Tests
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    paths:
+      - java/**
+      - rust/**
+      - .github/workflows/java.yml
+env:
+  # This env var is used by Swatinem/rust-cache@v2 for the cache
+  # key, so we set it to make sure it is always consistent.
+  CARGO_TERM_COLOR: always
+  # Disable full debug symbol generation to speed up CI build and keep memory down
+  # "1" means line tables only, which is useful for panic tracebacks.
+  RUSTFLAGS: "-C debuginfo=1"
+  RUST_BACKTRACE: "1"
+  # according to: https://matklad.github.io/2021/09/04/fast-rust-builds.html
+  # CI builds are faster with incremental disabled.
+  CARGO_INCREMENTAL: "0"
+  CARGO_BUILD_JOBS: "1"
+jobs:
+  linux-build:
+    runs-on: ubuntu-22.04
+    name: ubuntu-22.04 + Java 11 & 17
+    defaults:
+      run:
+        working-directory: ./java
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: java/core/lancedb-jni
+      - name: Run cargo fmt
+        run: cargo fmt --check
+        working-directory: ./java/core/lancedb-jni
+      - name: Install dependencies
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler libssl-dev
+      - name: Install Java 17
+        uses: actions/setup-java@v4
+        with:
+          distribution: temurin
+          java-version: 17
+          cache: "maven"
+      - run: echo "JAVA_17=$JAVA_HOME" >> $GITHUB_ENV
+      - name: Install Java 11
+        uses: actions/setup-java@v4
+        with:
+          distribution: temurin
+          java-version: 11
+          cache: "maven"
+      - name: Java Style Check
+        run: mvn checkstyle:check
+      # Disable because of issues in lancedb rust core code
+      # - name: Rust Clippy
+      #   working-directory: java/core/lancedb-jni
+      #   run: cargo clippy --all-targets -- -D warnings
+      - name: Running tests with Java 11
+        run: mvn clean test
+      - name: Running tests with Java 17
+        run: |
+          export JAVA_TOOL_OPTIONS="$JAVA_TOOL_OPTIONS \
+          -XX:+IgnoreUnrecognizedVMOptions \
+          --add-opens=java.base/java.lang=ALL-UNNAMED \
+          --add-opens=java.base/java.lang.invoke=ALL-UNNAMED \
+          --add-opens=java.base/java.lang.reflect=ALL-UNNAMED \
+          --add-opens=java.base/java.io=ALL-UNNAMED \
+          --add-opens=java.base/java.net=ALL-UNNAMED \
+          --add-opens=java.base/java.nio=ALL-UNNAMED \
+          --add-opens=java.base/java.util=ALL-UNNAMED \
+          --add-opens=java.base/java.util.concurrent=ALL-UNNAMED \
+          --add-opens=java.base/java.util.concurrent.atomic=ALL-UNNAMED \
+          --add-opens=java.base/jdk.internal.ref=ALL-UNNAMED \
+          --add-opens=java.base/sun.nio.ch=ALL-UNNAMED \
+          --add-opens=java.base/sun.nio.cs=ALL-UNNAMED \
+          --add-opens=java.base/sun.security.action=ALL-UNNAMED \
+          --add-opens=java.base/sun.util.calendar=ALL-UNNAMED \
+          --add-opens=java.security.jgss/sun.security.krb5=ALL-UNNAMED \
+          -Djdk.reflect.useDirectMethodHandle=false \
+          -Dio.netty.tryReflectionSetAccessible=true"
+          JAVA_HOME=$JAVA_17 mvn clean test

.github/workflows/make-release-commit.yml

@@ -1,37 +1,62 @@
 name: Create release commit
 
+# This workflow increments versions, tags the version, and pushes it.
+# When a tag is pushed, another workflow is triggered that creates a GH release
+# and uploads the binaries. This workflow is only for creating the tag.
+
+# This script will enforce that a minor version is incremented if there are any
+# breaking changes since the last minor increment. However, it isn't able to
+# differentiate between breaking changes in Node versus Python. If you wish to
+# bypass this check, you can manually increment the version and push the tag.
 on:
   workflow_dispatch:
     inputs:
       dry_run:
         description: 'Dry run (create the local commit/tags but do not push it)'
         required: true
-        default: "false"
-        type: choice
-        options:
-          - "true"
-          - "false"
-      part:
+        default: false
+        type: boolean
+      type:
         description: 'What kind of release is this?'
         required: true
-        default: 'patch'
+        default: 'preview'
         type: choice
         options:
-          - patch
-          - minor
-          - major
+          - preview
+          - stable
+      python:
+        description: 'Make a Python release'
+        required: true
+        default: true
+        type: boolean
+      other:
+        description: 'Make a Node/Rust release'
+        required: true
+        default: true
+        type: boolean
+      bump-minor:
+        description: 'Bump minor version'
+        required: true
+        default: false
+        type: boolean
 
 jobs:
-  bump-version:
+  make-release:
+    # Creates tag and GH release. The GH release will trigger the build and release jobs.
     runs-on: ubuntu-latest
+    permissions:
+      contents: write
     steps:
-      - name: Check out main
-        uses: actions/checkout@v4
+      - name: Output Inputs
+        run: echo "${{ toJSON(github.event.inputs) }}"
+      - uses: actions/checkout@v4
         with:
-          ref: main
-          persist-credentials: false
           fetch-depth: 0
           lfs: true
+          # It's important we use our token here, as the default token will NOT
+          # trigger any workflows watching for new tags. See:
+          # https://docs.github.com/en/actions/using-workflows/triggering-a-workflow#triggering-a-workflow-from-a-workflow
+          token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
       - name: Set git configs for bumpversion
         shell: bash
         run: |
@@ -41,19 +66,34 @@ jobs:
         uses: actions/setup-python@v5
         with:
           python-version: "3.11"
-      - name: Bump version, create tag and commit
+      - name: Bump Python version
+        if: ${{ inputs.python }}
+        working-directory: python
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         run: |
-          pip install bump2version
-          bumpversion --verbose ${{ inputs.part }}
-      - name: Push new version and tag
-        if: ${{ inputs.dry_run }} == "false"
+          # Need to get the commit before bumping the version, so we can
+          # determine if there are breaking changes in the next step as well.
+          echo "COMMIT_BEFORE_BUMP=$(git rev-parse HEAD)" >> $GITHUB_ENV
+
+          pip install bump-my-version PyGithub packaging
+          bash ../ci/bump_version.sh ${{ inputs.type }} ${{ inputs.bump-minor }} python-v
+      - name: Bump Node/Rust version
+        if: ${{ inputs.other }}
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          pip install bump-my-version PyGithub packaging
+          bash ci/bump_version.sh ${{ inputs.type }} ${{ inputs.bump-minor }} v $COMMIT_BEFORE_BUMP
+      - name: Push new version tag
+        if: ${{ !inputs.dry_run }}
         uses: ad-m/github-push-action@master
         with:
+          # Need to use PAT here too to trigger next workflow. See comment above.
           github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
-          branch: main
+          branch: ${{ github.ref }}
           tags: true
       - uses: ./.github/workflows/update_package_lock
-        if: ${{ inputs.dry_run }} == "false"
+        if: ${{ !inputs.dry_run && inputs.other }}
         with:
-          github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
-
+          github_token: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/node.yml

@@ -20,31 +20,11 @@ env:
   # "1" means line tables only, which is useful for panic tracebacks.
   #
   # Use native CPU to accelerate tests if possible, especially for f16
-  RUSTFLAGS: "-C debuginfo=1 -C target-cpu=native -C target-feature=+f16c,+avx2,+fma"
+  # target-cpu=haswell fixes failing ci build
+  RUSTFLAGS: "-C debuginfo=1 -C target-cpu=haswell -C target-feature=+f16c,+avx2,+fma"
   RUST_BACKTRACE: "1"
 
 jobs:
-  lint:
-    name: Lint
-    runs-on: ubuntu-22.04
-    defaults:
-      run:
-        shell: bash
-        working-directory: node
-    steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-        lfs: true
-    - uses: actions/setup-node@v3
-      with:
-        node-version: 20
-        cache: 'npm'
-        cache-dependency-path: node/package-lock.json
-    - name: Lint
-      run: |
-        npm ci
-        npm run lint
   linux:
     name: Linux (Node ${{ matrix.node-version }})
     timeout-minutes: 30
@@ -127,6 +107,7 @@ jobs:
       AWS_ENDPOINT: http://localhost:4566
       # this one is for dynamodb
       DYNAMODB_ENDPOINT: http://localhost:4566
+      ALLOW_HTTP: true
     steps:
     - uses: actions/checkout@v4
       with:

.github/workflows/nodejs.yml

@@ -28,6 +28,10 @@ jobs:
       run:
         shell: bash
         working-directory: nodejs
+    env:
+      # Need up-to-date compilers for kernels
+      CC: gcc-12
+      CXX: g++-12
     steps:
     - uses: actions/checkout@v4
       with:
@@ -48,7 +52,7 @@ jobs:
         cargo fmt --all -- --check
         cargo clippy --all --all-features -- -D warnings
         npm ci
-        npm run lint
+        npm run lint-ci
   linux:
     name: Linux (NodeJS ${{ matrix.node-version }})
     timeout-minutes: 30
@@ -80,7 +84,12 @@ jobs:
       run: |
         npm ci
         npm run build
+    - name: Setup localstack
+      working-directory: .
+      run: docker compose up --detach --wait
     - name: Test
+      env:
+        S3_TEST: "1"
       run: npm run test
   macos:
     timeout-minutes: 30
@@ -111,4 +120,3 @@ jobs:
     - name: Test
       run: |
         npm run test
-

.github/workflows/npm-publish.yml

@@ -1,11 +1,13 @@
 name: NPM Publish
 
 on:
-  release:
-    types: [ published ]
+  push:
+    tags:
+      - "v*"
 
 jobs:
   node:
+    name: vectordb Typescript
     runs-on: ubuntu-latest
     # Only runs on tags that matches the make-release action
     if: startsWith(github.ref, 'refs/tags/v')
@@ -19,7 +21,7 @@ jobs:
       - uses: actions/setup-node@v3
         with:
           node-version: 20
-          cache: 'npm'
+          cache: "npm"
           cache-dependency-path: node/package-lock.json
       - name: Install dependencies
         run: |
@@ -31,13 +33,14 @@ jobs:
           npm run tsc
           npm pack
       - name: Upload Linux Artifacts
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
           name: node-package
           path: |
             node/vectordb-*.tgz
 
   node-macos:
+    name: vectordb ${{ matrix.config.arch }}
     strategy:
       matrix:
         config:
@@ -61,15 +64,45 @@ jobs:
       - name: Build MacOS native node modules
         run: bash ci/build_macos_artifacts.sh ${{ matrix.config.arch }}
       - name: Upload Darwin Artifacts
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
-          name: native-darwin
+          name: node-native-darwin-${{ matrix.config.arch }}
           path: |
             node/dist/lancedb-vectordb-darwin*.tgz
 
+  nodejs-macos:
+    name: lancedb ${{ matrix.config.arch }}
+    strategy:
+      matrix:
+        config:
+          - arch: x86_64-apple-darwin
+            runner: macos-13
+          - arch: aarch64-apple-darwin
+            # xlarge is implicitly arm64.
+            runner: macos-14
+    runs-on: ${{ matrix.config.runner }}
+    # Only runs on tags that matches the make-release action
+    if: startsWith(github.ref, 'refs/tags/v')
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - name: Install system dependencies
+        run: brew install protobuf
+      - name: Install npm dependencies
+        run: |
+          cd nodejs
+          npm ci
+      - name: Build MacOS native nodejs modules
+        run: bash ci/build_macos_artifacts_nodejs.sh ${{ matrix.config.arch }}
+      - name: Upload Darwin Artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: nodejs-native-darwin-${{ matrix.config.arch }}
+          path: |
+            nodejs/dist/*.node
 
   node-linux:
-    name: node-linux (${{ matrix.config.arch}}-unknown-linux-gnu
+    name: vectordb (${{ matrix.config.arch}}-unknown-linux-gnu)
     runs-on: ${{ matrix.config.runner }}
     # Only runs on tags that matches the make-release action
     if: startsWith(github.ref, 'refs/tags/v')
@@ -80,21 +113,87 @@ jobs:
           - arch: x86_64
             runner: ubuntu-latest
           - arch: aarch64
-            runner: buildjet-4vcpu-ubuntu-2204-arm
+            # For successful fat LTO builds, we need a large runner to avoid OOM errors.
+            runner: warp-ubuntu-latest-arm64-4x
     steps:
       - name: Checkout
         uses: actions/checkout@v4
+      # To avoid OOM errors on ARM, we create a swap file.
+      - name: Configure aarch64 build
+        if: ${{ matrix.config.arch == 'aarch64' }}
+        run: |
+          free -h
+          sudo fallocate -l 16G /swapfile
+          sudo chmod 600 /swapfile
+          sudo mkswap /swapfile
+          sudo swapon /swapfile
+          echo "/swapfile swap swap defaults 0 0" >> sudo /etc/fstab
+          # print info
+          swapon --show
+          free -h
       - name: Build Linux Artifacts
         run: |
           bash ci/build_linux_artifacts.sh ${{ matrix.config.arch }}
       - name: Upload Linux Artifacts
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
-          name: native-linux
+          name: node-native-linux-${{ matrix.config.arch }}
           path: |
             node/dist/lancedb-vectordb-linux*.tgz
 
+  nodejs-linux:
+    name: lancedb (${{ matrix.config.arch}}-unknown-linux-gnu
+    runs-on: ${{ matrix.config.runner }}
+    # Only runs on tags that matches the make-release action
+    if: startsWith(github.ref, 'refs/tags/v')
+    strategy:
+      fail-fast: false
+      matrix:
+        config:
+          - arch: x86_64
+            runner: ubuntu-latest
+          - arch: aarch64
+            # For successful fat LTO builds, we need a large runner to avoid OOM errors.
+            runner: buildjet-16vcpu-ubuntu-2204-arm
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      # Buildjet aarch64 runners have only 1.5 GB RAM per core, vs 3.5 GB per core for
+      # x86_64 runners. To avoid OOM errors on ARM, we create a swap file.
+      - name: Configure aarch64 build
+        if: ${{ matrix.config.arch == 'aarch64' }}
+        run: |
+          free -h
+          sudo fallocate -l 16G /swapfile
+          sudo chmod 600 /swapfile
+          sudo mkswap /swapfile
+          sudo swapon /swapfile
+          echo "/swapfile swap swap defaults 0 0" >> sudo /etc/fstab
+          # print info
+          swapon --show
+          free -h
+      - name: Build Linux Artifacts
+        run: |
+          bash ci/build_linux_artifacts_nodejs.sh ${{ matrix.config.arch }}
+      - name: Upload Linux Artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: nodejs-native-linux-${{ matrix.config.arch }}
+          path: |
+            nodejs/dist/*.node
+      # The generic files are the same in all distros so we just pick
+      # one to do the upload.
+      - name: Upload Generic Artifacts
+        if: ${{ matrix.config.arch == 'x86_64' }}
+        uses: actions/upload-artifact@v4
+        with:
+          name: nodejs-dist
+          path: |
+            nodejs/dist/*
+            !nodejs/dist/*.node
+
   node-windows:
+    name: vectordb ${{ matrix.target }}
     runs-on: windows-2022
     # Only runs on tags that matches the make-release action
     if: startsWith(github.ref, 'refs/tags/v')
@@ -121,37 +220,150 @@ jobs:
       - name: Build Windows native node modules
         run: .\ci\build_windows_artifacts.ps1 ${{ matrix.target }}
       - name: Upload Windows Artifacts
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
-          name: native-windows
+          name: node-native-windows
           path: |
             node/dist/lancedb-vectordb-win32*.tgz
 
+  nodejs-windows:
+    name: lancedb ${{ matrix.target }}
+    runs-on: windows-2022
+    # Only runs on tags that matches the make-release action
+    if: startsWith(github.ref, 'refs/tags/v')
+    strategy:
+      fail-fast: false
+      matrix:
+        target: [x86_64-pc-windows-msvc]
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - name: Install Protoc v21.12
+        working-directory: C:\
+        run: |
+          New-Item -Path 'C:\protoc' -ItemType Directory
+          Set-Location C:\protoc
+          Invoke-WebRequest https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-win64.zip -OutFile C:\protoc\protoc.zip
+          7z x protoc.zip
+          Add-Content $env:GITHUB_PATH "C:\protoc\bin"
+        shell: powershell
+      - name: Install npm dependencies
+        run: |
+          cd nodejs
+          npm ci
+      - name: Build Windows native node modules
+        run: .\ci\build_windows_artifacts_nodejs.ps1 ${{ matrix.target }}
+      - name: Upload Windows Artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: nodejs-native-windows
+          path: |
+            nodejs/dist/*.node
+
   release:
+    name: vectordb NPM Publish
     needs: [node, node-macos, node-linux, node-windows]
     runs-on: ubuntu-latest
     # Only runs on tags that matches the make-release action
     if: startsWith(github.ref, 'refs/tags/v')
     steps:
-      - uses: actions/download-artifact@v3
+      - uses: actions/download-artifact@v4
+        with:
+          pattern: node-*
       - name: Display structure of downloaded files
         run: ls -R
       - uses: actions/setup-node@v3
         with:
           node-version: 20
-          registry-url: 'https://registry.npmjs.org'
+          registry-url: "https://registry.npmjs.org"
       - name: Publish to NPM
         env:
           NODE_AUTH_TOKEN: ${{ secrets.LANCEDB_NPM_REGISTRY_TOKEN }}
         run: |
+          # Tag beta as "preview" instead of default "latest". See lancedb 
+          # npm publish step for more info.
+          if [[ $GITHUB_REF =~ refs/tags/v(.*)-beta.* ]]; then
+            PUBLISH_ARGS="--tag preview"
+          fi
+
           mv */*.tgz .
           for filename in *.tgz; do
-            npm publish $filename
+            npm publish $PUBLISH_ARGS $filename
           done
+      - name: Notify Slack Action
+        uses: ravsamhq/notify-slack-action@2.3.0
+        if: ${{ always() }}
+        with:
+          status: ${{ job.status }}
+          notify_when: "failure"
+          notification_title: "{workflow} is failing"
+        env:
+          SLACK_WEBHOOK_URL: ${{ secrets.ACTION_MONITORING_SLACK }}
+
+  release-nodejs:
+    name: lancedb NPM Publish
+    needs: [nodejs-macos, nodejs-linux, nodejs-windows]
+    runs-on: ubuntu-latest
+    # Only runs on tags that matches the make-release action
+    if: startsWith(github.ref, 'refs/tags/v')
+    defaults:
+      run:
+        shell: bash
+        working-directory: nodejs
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - uses: actions/download-artifact@v4
+        with:
+          name: nodejs-dist
+          path: nodejs/dist
+      - uses: actions/download-artifact@v4
+        name: Download arch-specific binaries
+        with:
+          pattern: nodejs-*
+          path: nodejs/nodejs-artifacts
+          merge-multiple: true
+      - name: Display structure of downloaded files
+        run: find .
+      - uses: actions/setup-node@v3
+        with:
+          node-version: 20
+          registry-url: "https://registry.npmjs.org"
+      - name: Install napi-rs
+        run: npm install -g @napi-rs/cli
+      - name: Prepare artifacts
+        run: npx napi artifacts -d nodejs-artifacts
+      - name: Display structure of staged files
+        run: find npm
+      - name: Publish to NPM
+        env:
+          NODE_AUTH_TOKEN: ${{ secrets.LANCEDB_NPM_REGISTRY_TOKEN }}
+        # By default, things are published to the latest tag. This is what is
+        # installed by default if the user does not specify a version. This is
+        # good for stable releases, but for pre-releases, we want to publish to
+        # the "preview" tag so they can install with `npm install lancedb@preview`.
+        # See: https://medium.com/@mbostock/prereleases-and-npm-e778fc5e2420
+        run: |
+          if [[ $GITHUB_REF =~ refs/tags/v(.*)-beta.* ]]; then
+            npm publish --access public --tag preview
+          else
+            npm publish --access public
+          fi
+      - name: Notify Slack Action
+        uses: ravsamhq/notify-slack-action@2.3.0
+        if: ${{ always() }}
+        with:
+          status: ${{ job.status }}
+          notify_when: "failure"
+          notification_title: "{workflow} is failing"
+        env:
+          SLACK_WEBHOOK_URL: ${{ secrets.ACTION_MONITORING_SLACK }}
 
   update-package-lock:
     needs: [release]
     runs-on: ubuntu-latest
+    permissions:
+      contents: write
     steps:
       - name: Checkout
         uses: actions/checkout@v4
@@ -162,4 +374,87 @@ jobs:
           lfs: true
       - uses: ./.github/workflows/update_package_lock
         with:
-          github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
+          github_token: ${{ secrets.GITHUB_TOKEN }}
+
+  update-package-lock-nodejs:
+    needs: [release-nodejs]
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          ref: main
+          persist-credentials: false
+          fetch-depth: 0
+          lfs: true
+      - uses: ./.github/workflows/update_package_lock_nodejs
+        with:
+          github_token: ${{ secrets.GITHUB_TOKEN }}
+
+  gh-release:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Extract version
+        id: extract_version
+        env:
+          GITHUB_REF: ${{ github.ref }}
+        run: |
+          set -e
+          echo "Extracting tag and version from $GITHUB_REF"
+          if [[ $GITHUB_REF =~ refs/tags/v(.*) ]]; then
+            VERSION=${BASH_REMATCH[1]}
+            TAG=v$VERSION
+            echo "tag=$TAG" >> $GITHUB_OUTPUT
+            echo "version=$VERSION" >> $GITHUB_OUTPUT
+          else
+            echo "Failed to extract version from $GITHUB_REF"
+            exit 1
+          fi
+          echo "Extracted version $VERSION from $GITHUB_REF"
+          if [[ $VERSION =~ beta ]]; then
+            echo "This is a beta release"
+
+            # Get last release (that is not this one)
+            FROM_TAG=$(git tag --sort='version:refname' \
+              | grep ^v \
+              | grep -vF "$TAG" \
+              | python ci/semver_sort.py v \
+              | tail -n 1)
+          else
+            echo "This is a stable release"
+            # Get last stable tag (ignore betas)
+            FROM_TAG=$(git tag --sort='version:refname' \
+              | grep ^v \
+              | grep -vF "$TAG" \
+              | grep -v beta \
+              | python ci/semver_sort.py v \
+              | tail -n 1)
+          fi
+          echo "Found from tag $FROM_TAG"
+          echo "from_tag=$FROM_TAG" >> $GITHUB_OUTPUT
+      - name: Create Release Notes
+        id: release_notes
+        uses: mikepenz/release-changelog-builder-action@v4
+        with:
+          configuration: .github/release_notes.json
+          toTag: ${{ steps.extract_version.outputs.tag }}
+          fromTag: ${{ steps.extract_version.outputs.from_tag }}
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      - name: Create GH release
+        uses: softprops/action-gh-release@v2
+        with:
+          prerelease: ${{ contains('beta', github.ref) }}
+          tag_name: ${{ steps.extract_version.outputs.tag }}
+          token: ${{ secrets.GITHUB_TOKEN }}
+          generate_release_notes: false
+          name: Node/Rust LanceDB v${{ steps.extract_version.outputs.version }}
+          body: ${{ steps.release_notes.outputs.changelog }}

.github/workflows/pypi-publish.yml

@@ -1,31 +1,160 @@
 name: PyPI Publish
 
 on:
-  release:
-    types: [ published ]
+  push:
+    tags:
+      - 'python-v*'
 
 jobs:
-  publish:
-    runs-on: ubuntu-latest
-    # Only runs on tags that matches the python-make-release action
-    if: startsWith(github.ref, 'refs/tags/python-v')
-    defaults:
-      run:
-        shell: bash
-        working-directory: python
+  linux:
+    name: Python ${{ matrix.config.platform }} manylinux${{ matrix.config.manylinux }}
+    timeout-minutes: 60
+    strategy:
+      matrix:
+        config:
+          - platform: x86_64
+            manylinux: "2_17"
+            extra_args: ""
+          - platform: x86_64
+            manylinux: "2_28"
+            extra_args: "--features fp16kernels"
+          - platform: aarch64
+            manylinux: "2_24"
+            extra_args: ""
+          # We don't build fp16 kernels for aarch64, because it uses 
+          # cross compilation image, which doesn't have a new enough compiler.
+    runs-on: "ubuntu-22.04"
     steps:
       - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v4
         with:
-          python-version: "3.8"
-      - name: Build distribution
+          python-version: 3.8
+      - uses: ./.github/workflows/build_linux_wheel
+        with:
+          python-minor-version: 8
+          args: "--release --strip ${{ matrix.config.extra_args }}"
+          arm-build: ${{ matrix.config.platform == 'aarch64' }}
+          manylinux: ${{ matrix.config.manylinux }}
+      - uses: ./.github/workflows/upload_wheel
+        with:
+          pypi_token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          fury_token: ${{ secrets.FURY_TOKEN }}
+  mac:
+    timeout-minutes: 60
+    runs-on: ${{ matrix.config.runner }}
+    strategy:
+      matrix:
+        config:
+          - target: x86_64-apple-darwin
+            runner: macos-13
+          - target: aarch64-apple-darwin
+            runner: macos-14
+    env:
+      MACOSX_DEPLOYMENT_TARGET: 10.15
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.12
+      - uses: ./.github/workflows/build_mac_wheel
+        with:
+          python-minor-version: 8
+          args: "--release --strip --target ${{ matrix.config.target }} --features fp16kernels"
+      - uses: ./.github/workflows/upload_wheel
+        with:
+          pypi_token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          fury_token: ${{ secrets.FURY_TOKEN }}
+  windows:
+    timeout-minutes: 60
+    runs-on: windows-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.8
+      - uses: ./.github/workflows/build_windows_wheel
+        with:
+          python-minor-version: 8
+          args: "--release --strip"
+          vcpkg_token: ${{ secrets.VCPKG_GITHUB_PACKAGES }}
+      - uses: ./.github/workflows/upload_wheel
+        with:
+          pypi_token: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
+          fury_token: ${{ secrets.FURY_TOKEN }}
+  gh-release:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Extract version
+        id: extract_version
+        env:
+          GITHUB_REF: ${{ github.ref }}
         run: |
-          ls -la
-          pip install wheel setuptools --upgrade
-          python setup.py sdist bdist_wheel
-      - name: Publish
-        uses: pypa/gh-action-pypi-publish@v1.8.5
+          set -e
+          echo "Extracting tag and version from $GITHUB_REF"
+          if [[ $GITHUB_REF =~ refs/tags/python-v(.*) ]]; then
+            VERSION=${BASH_REMATCH[1]}
+            TAG=python-v$VERSION
+            echo "tag=$TAG" >> $GITHUB_OUTPUT
+            echo "version=$VERSION" >> $GITHUB_OUTPUT
+          else
+            echo "Failed to extract version from $GITHUB_REF"
+            exit 1
+          fi
+          echo "Extracted version $VERSION from $GITHUB_REF"
+          if [[ $VERSION =~ beta ]]; then
+            echo "This is a beta release"
+
+            # Get last release (that is not this one)
+            FROM_TAG=$(git tag --sort='version:refname' \
+              | grep ^python-v \
+              | grep -vF "$TAG" \
+              | python ci/semver_sort.py python-v \
+              | tail -n 1)
+          else
+            echo "This is a stable release"
+            # Get last stable tag (ignore betas)
+            FROM_TAG=$(git tag --sort='version:refname' \
+              | grep ^python-v \
+              | grep -vF "$TAG" \
+              | grep -v beta \
+              | python ci/semver_sort.py python-v \
+              | tail -n 1)
+          fi
+          echo "Found from tag $FROM_TAG"
+          echo "from_tag=$FROM_TAG" >> $GITHUB_OUTPUT
+      - name: Create Python Release Notes
+        id: python_release_notes
+        uses: mikepenz/release-changelog-builder-action@v4
         with:
-          password: ${{ secrets.LANCEDB_PYPI_API_TOKEN }}
-          packages-dir: python/dist
+          configuration: .github/release_notes.json
+          toTag: ${{ steps.extract_version.outputs.tag }}
+          fromTag: ${{ steps.extract_version.outputs.from_tag }}
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      - name: Create Python GH release
+        uses: softprops/action-gh-release@v2
+        with:
+          prerelease: ${{ contains('beta', github.ref) }}
+          tag_name: ${{ steps.extract_version.outputs.tag }}
+          token: ${{ secrets.GITHUB_TOKEN }}
+          generate_release_notes: false
+          name: Python LanceDB v${{ steps.extract_version.outputs.version }}
+          body: ${{ steps.python_release_notes.outputs.changelog }}

.github/workflows/python-make-release-commit.yml

@@ -1,56 +0,0 @@
-name: Python - Create release commit
-
-on:
-  workflow_dispatch:
-    inputs:
-      dry_run:
-        description: 'Dry run (create the local commit/tags but do not push it)'
-        required: true
-        default: "false"
-        type: choice
-        options:
-          - "true"
-          - "false"
-      part:
-        description: 'What kind of release is this?'
-        required: true
-        default: 'patch'
-        type: choice
-        options:
-          - patch
-          - minor
-          - major
-
-jobs:
-  bump-version:
-    runs-on: ubuntu-latest
-    steps:
-    - name: Check out main
-      uses: actions/checkout@v4
-      with:
-        ref: main
-        persist-credentials: false
-        fetch-depth: 0
-        lfs: true
-    - name: Set git configs for bumpversion
-      shell: bash
-      run: |
-        git config user.name 'Lance Release'
-        git config user.email 'lance-dev@lancedb.com'
-    - name: Set up Python
-      uses: actions/setup-python@v5
-      with:
-        python-version: "3.11"
-    - name: Bump version, create tag and commit
-      working-directory: python
-      run: |
-        pip install bump2version
-        bumpversion --verbose ${{ inputs.part }}
-    - name: Push new version and tag
-      if: ${{ inputs.dry_run }} == "false"
-      uses: ad-m/github-push-action@master
-      with:
-        github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}
-        branch: main
-        tags: true
-

.github/workflows/python.yml

@@ -14,49 +14,135 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-  linux:
+  lint:
+    name: "Lint"
     timeout-minutes: 30
-    strategy:
-      matrix:
-        python-minor-version: [ "8", "11" ]
     runs-on: "ubuntu-22.04"
     defaults:
       run:
         shell: bash
         working-directory: python
     steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-        lfs: true
-    - name: Set up Python
-      uses: actions/setup-python@v5
-      with:
-        python-version: 3.${{ matrix.python-minor-version }}
-    - name: Install lancedb
-      run: |
-        pip install -e .[tests]
-        pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
-        pip install pytest pytest-mock ruff
-    - name: Format check
-      run: ruff format --check .
-    - name: Lint
-      run: ruff .
-    - name: Run tests
-      run: pytest -m "not slow" -x -v --durations=30 tests
-    - name: doctest
-      run: pytest --doctest-modules lancedb
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+      - name: Install ruff
+        run: |
+          pip install ruff==0.5.4
+      - name: Format check
+        run: ruff format --check .
+      - name: Lint
+        run: ruff check .
+  doctest:
+    name: "Doctest"
+    timeout-minutes: 30
+    runs-on: "ubuntu-22.04"
+    defaults:
+      run:
+        shell: bash
+        working-directory: python
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+          cache: "pip"
+      - name: Install protobuf
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - name: Install
+        run: |
+          pip install --extra-index-url https://pypi.fury.io/lancedb/ -e .[tests,dev,embeddings]
+          pip install tantivy
+          pip install mlx
+      - name: Doctest
+        run: pytest --doctest-modules python/lancedb
+  linux:
+    name: "Linux: python-3.${{ matrix.python-minor-version }}"
+    timeout-minutes: 30
+    strategy:
+      matrix:
+        python-minor-version: ["9", "11"]
+    runs-on: "ubuntu-22.04"
+    defaults:
+      run:
+        shell: bash
+        working-directory: python
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Install protobuf
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: 3.${{ matrix.python-minor-version }}
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - uses: ./.github/workflows/build_linux_wheel
+      - uses: ./.github/workflows/run_tests
+        with:
+          integration: true
+      # Make sure wheels are not included in the Rust cache
+      - name: Delete wheels
+        run: rm -rf target/wheels
   platform:
-    name: "Platform: ${{ matrix.config.name }}"
+    name: "Mac: ${{ matrix.config.name }}"
     timeout-minutes: 30
     strategy:
       matrix:
         config:
-          - name: x86 Mac
+          - name: x86
             runner: macos-13
-          - name: Arm Mac
+          - name: Arm
             runner: macos-14
-          - name: x86 Windows
+    runs-on: "${{ matrix.config.runner }}"
+    defaults:
+      run:
+        shell: bash
+        working-directory: python
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - uses: ./.github/workflows/build_mac_wheel
+      - uses: ./.github/workflows/run_tests
+      # Make sure wheels are not included in the Rust cache
+      - name: Delete wheels
+        run: rm -rf target/wheels
+  windows:
+    name: "Windows: ${{ matrix.config.name }}"
+    timeout-minutes: 30
+    strategy:
+      matrix:
+        config:
+          - name: x86
             runner: windows-latest
     runs-on: "${{ matrix.config.runner }}"
     defaults:
@@ -64,21 +150,22 @@ jobs:
         shell: bash
         working-directory: python
     steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-        lfs: true
-    - name: Set up Python
-      uses: actions/setup-python@v5
-      with:
-        python-version: "3.11"
-    - name: Install lancedb
-      run: |
-        pip install -e .[tests]
-        pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
-        pip install pytest pytest-mock
-    - name: Run tests
-      run: pytest -m "not slow" -x -v --durations=30 tests
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+      - uses: Swatinem/rust-cache@v2
+        with:
+          workspaces: python
+      - uses: ./.github/workflows/build_windows_wheel
+      - uses: ./.github/workflows/run_tests
+      # Make sure wheels are not included in the Rust cache
+      - name: Delete wheels
+        run: rm -rf target/wheels
   pydantic1x:
     timeout-minutes: 30
     runs-on: "ubuntu-22.04"
@@ -87,21 +174,22 @@ jobs:
         shell: bash
         working-directory: python
     steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-        lfs: true
-    - name: Set up Python
-      uses: actions/setup-python@v5
-      with:
-        python-version: 3.9
-    - name: Install lancedb
-      run: |
-        pip install "pydantic<2"
-        pip install -e .[tests]
-        pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
-        pip install pytest pytest-mock
-    - name: Run tests
-      run: pytest -m "not slow" -x -v --durations=30 tests
-    - name: doctest
-      run: pytest --doctest-modules lancedb
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Install dependencies
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: 3.9
+      - name: Install lancedb
+        run: |
+          pip install "pydantic<2"
+          pip install --extra-index-url https://pypi.fury.io/lancedb/ -e .[tests]
+          pip install tantivy
+      - name: Run tests
+        run: pytest -m "not slow and not s3_test" -x -v --durations=30 python/tests

.github/workflows/run_tests/action.yml

@@ -0,0 +1,31 @@
+name: run-tests
+
+description: "Install lance wheel and run unit tests"
+inputs:
+  python-minor-version:
+    required: true
+    description: "8 9 10 11 12"
+  integration:
+    required: false
+    description: "Run integration tests"
+    default: "false"
+runs:
+  using: "composite"
+  steps:
+    - name: Install lancedb
+      shell: bash
+      run: |
+        pip3 install --extra-index-url https://pypi.fury.io/lancedb/ $(ls target/wheels/lancedb-*.whl)[tests,dev]
+    - name: Setup localstack for integration tests
+      if: ${{ inputs.integration == 'true' }}
+      shell: bash
+      working-directory: .
+      run: docker compose up --detach --wait
+    - name: pytest (with integration)
+      shell: bash
+      if: ${{ inputs.integration == 'true' }}
+      run: pytest -m "not slow" -x -v --durations=30 python/python/tests
+    - name: pytest (no integration tests)
+      shell: bash
+      if: ${{ inputs.integration != 'true' }}
+      run: pytest -m "not slow and not s3_test" -x -v --durations=30 python/python/tests

.github/workflows/rust.yml

@@ -31,6 +31,10 @@ jobs:
       run:
         shell: bash
         working-directory: rust
+    env:
+      # Need up-to-date compilers for kernels
+      CC: gcc-12
+      CXX: g++-12
     steps:
     - uses: actions/checkout@v4
       with:
@@ -49,11 +53,18 @@ jobs:
       run: cargo clippy --all --all-features -- -D warnings
   linux:
     timeout-minutes: 30
-    runs-on: ubuntu-22.04
+    # To build all features, we need more disk space than is available
+    # on the GitHub-provided runner. This is mostly due to the the
+    # sentence-transformers feature.
+    runs-on: warp-ubuntu-latest-x64-4x
     defaults:
       run:
         shell: bash
         working-directory: rust
+    env:
+      # Need up-to-date compilers for kernels
+      CC: gcc-12
+      CXX: g++-12
     steps:
     - uses: actions/checkout@v4
       with:
@@ -66,6 +77,9 @@ jobs:
       run: |
           sudo apt update
           sudo apt install -y protobuf-compiler libssl-dev
+    - name: Start S3 integration test environment
+      working-directory: .
+      run: docker compose up --detach --wait
     - name: Build
       run: cargo build --all-features
     - name: Run tests
@@ -97,7 +111,8 @@ jobs:
       - name: Build
         run: cargo build --all-features
       - name: Run tests
-        run: cargo test --all-features
+        # Run with everything except the integration tests.
+        run: cargo test --features remote,fp16kernels
   windows:
     runs-on: windows-2022
     steps:

.github/workflows/update_package_lock_nodejs/action.yml

@@ -0,0 +1,33 @@
+name: update_package_lock_nodejs
+description: "Update nodejs's package.lock"
+
+inputs:
+  github_token:
+    required: true
+    description: "github token for the repo"
+
+runs:
+  using: "composite"
+  steps:
+    - uses: actions/setup-node@v3
+      with:
+        node-version: 20
+    - name: Set git configs
+      shell: bash
+      run: |
+        git config user.name 'Lance Release'
+        git config user.email 'lance-dev@lancedb.com'
+    - name: Update package-lock.json file
+      working-directory: ./nodejs
+      run: |
+        npm install
+        git add package-lock.json
+        git commit -m "Updating package-lock.json"
+      shell: bash
+    - name: Push changes
+      if: ${{ inputs.dry_run }} == "false"
+      uses: ad-m/github-push-action@master
+      with:
+        github_token: ${{ inputs.github_token }}
+        branch: main
+        tags: true

.github/workflows/update_package_lock_run_nodejs.yml

@@ -0,0 +1,19 @@
+name: Update NodeJs package-lock.json
+
+on:
+  workflow_dispatch:
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          ref: main
+          persist-credentials: false
+          fetch-depth: 0
+          lfs: true
+      - uses: ./.github/workflows/update_package_lock_nodejs
+        with:
+          github_token: ${{ secrets.LANCEDB_RELEASE_TOKEN }}

.github/workflows/upload_wheel/action.yml

@@ -0,0 +1,44 @@
+name: upload-wheel
+
+description: "Upload wheels to Pypi"
+inputs:
+  pypi_token:
+    required: true
+    description: "release token for the repo"
+  fury_token:
+    required: true
+    description: "release token for the fury repo"
+
+runs:
+  using: "composite"
+  steps:
+  - name: Install dependencies
+    shell: bash
+    run: |
+      python -m pip install --upgrade pip
+      pip install twine
+  - name: Choose repo
+    shell: bash
+    id: choose_repo
+    run: |
+      if [ ${{ github.ref }} == "*beta*" ]; then
+        echo "repo=fury" >> $GITHUB_OUTPUT
+      else
+        echo "repo=pypi" >> $GITHUB_OUTPUT
+      fi
+  - name: Publish to PyPI
+    shell: bash
+    env:
+      FURY_TOKEN: ${{ inputs.fury_token }}
+      PYPI_TOKEN: ${{ inputs.pypi_token }}
+    run: |
+      if [ ${{ steps.choose_repo.outputs.repo }} == "fury" ]; then
+        WHEEL=$(ls target/wheels/lancedb-*.whl 2> /dev/null | head -n 1)
+        echo "Uploading $WHEEL to Fury"
+        curl -f -F package=@$WHEEL https://$FURY_TOKEN@push.fury.io/lancedb/
+      else
+        twine upload --repository ${{ steps.choose_repo.outputs.repo }} \
+          --username __token__ \
+          --password $PYPI_TOKEN \
+          target/wheels/lancedb-*.whl
+      fi

.gitignore

@@ -4,9 +4,10 @@
 **/__pycache__
 .DS_Store
 venv
+.venv
 
 .vscode
-
+.zed
 rust/target
 rust/Cargo.lock
 
@@ -22,6 +23,11 @@ python/dist
 
 **/.hypothesis
 
+# Compiled Dynamic libraries
+*.so
+*.dylib
+*.dll
+
 ## Javascript
 *.node
 **/node_modules
@@ -29,9 +35,12 @@ python/dist
 node/dist
 node/examples/**/package-lock.json
 node/examples/**/dist
+nodejs/lancedb/native*
 dist
 
 ## Rust
 target
 
+**/sccache.log
+
 Cargo.lock

.pre-commit-config.yaml

@@ -5,17 +5,17 @@ repos:
     -   id: check-yaml
     -   id: end-of-file-fixer
     -   id: trailing-whitespace
--   repo: https://github.com/psf/black
-    rev: 22.12.0
-    hooks:
-    -   id: black
 -   repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: v0.0.277
+    rev: v0.2.2
     hooks:
     - id: ruff
--   repo: https://github.com/pycqa/isort
-    rev: 5.12.0
-    hooks:
-    - id: isort
-      name: isort (python)
+- repo: local
+  hooks:
+    - id: local-biome-check
+      name: biome check
+      entry: npx @biomejs/biome@1.8.3 check --config-path nodejs/biome.json nodejs/
+      language: system
+      types: [text]
+      files: "nodejs/.*"
+      exclude: nodejs/lancedb/native.d.ts|nodejs/dist/.*|nodejs/examples/.*

Cargo.toml

@@ -1,37 +1,51 @@
 [workspace]
-members = ["rust/ffi/node", "rust/vectordb", "nodejs"]
+members = [
+    "rust/ffi/node",
+    "rust/lancedb",
+    "nodejs",
+    "python",
+    "java/core/lancedb-jni",
+]
 # Python package needs to be built by maturin.
 exclude = ["python"]
 resolver = "2"
 
 [workspace.package]
 edition = "2021"
-authors = ["Lance Devs <dev@lancedb.com>"]
+authors = ["LanceDB Devs <dev@lancedb.com>"]
 license = "Apache-2.0"
 repository = "https://github.com/lancedb/lancedb"
+description = "Serverless, low-latency vector database for AI applications"
+keywords = ["lancedb", "lance", "database", "vector", "search"]
+categories = ["database-implementations"]
 
 [workspace.dependencies]
-lance = { "version" = "=0.9.12", "features" = ["dynamodb"] }
-lance-index = { "version" = "=0.9.12" }
-lance-linalg = { "version" = "=0.9.12" }
-lance-testing = { "version" = "=0.9.12" }
+lance = { "version" = "=0.15.0", "features" = ["dynamodb"] }
+lance-index = { "version" = "=0.15.0" }
+lance-linalg = { "version" = "=0.15.0" }
+lance-testing = { "version" = "=0.15.0" }
+lance-datafusion = { "version" = "=0.15.0" }
 # Note that this one does not include pyarrow
-arrow = { version = "50.0", optional = false }
-arrow-array = "50.0"
-arrow-data = "50.0"
-arrow-ipc = "50.0"
-arrow-ord = "50.0"
-arrow-schema = "50.0"
-arrow-arith = "50.0"
-arrow-cast = "50.0"
+arrow = { version = "52.1", optional = false }
+arrow-array = "52.1"
+arrow-data = "52.1"
+arrow-ipc = "52.1"
+arrow-ord = "52.1"
+arrow-schema = "52.1"
+arrow-arith = "52.1"
+arrow-cast = "52.1"
 async-trait = "0"
-chrono = "0.4.23"
-half = { "version" = "=2.3.1", default-features = false, features = [
+chrono = "0.4.35"
+datafusion-physical-plan = "40.0"
+half = { "version" = "=2.4.1", default-features = false, features = [
     "num-traits",
 ] }
 futures = "0"
 log = "0.4"
-object_store = "0.9.0"
+object_store = "0.10.1"
+pin-project = "1.0.7"
 snafu = "0.7.4"
 url = "2"
 num-traits = "0.2"
+regex = "1.10"
+lazy_static = "1"

README.md

@@ -1,14 +1,14 @@
 <div align="center">
 <p align="center">
 
-<img width="275" alt="LanceDB Logo" src="https://user-images.githubusercontent.com/917119/226205734-6063d87a-1ecc-45fe-85be-1dea6383a3d8.png">
+<img width="275" alt="LanceDB Logo" src="https://github.com/lancedb/lancedb/assets/5846846/37d7c7ad-c2fd-4f56-9f16-fffb0d17c73a">
 
-**Developer-friendly, serverless vector database for AI applications**
+**Developer-friendly, database for multimodal AI**
 
 <a href='https://github.com/lancedb/vectordb-recipes/tree/main' target="_blank"><img alt='LanceDB' src='https://img.shields.io/badge/VectorDB_Recipes-100000?style=for-the-badge&logo=LanceDB&logoColor=white&labelColor=645cfb&color=645cfb'/></a>
 <a href='https://lancedb.github.io/lancedb/' target="_blank"><img alt='lancdb' src='https://img.shields.io/badge/DOCS-100000?style=for-the-badge&logo=lancdb&logoColor=white&labelColor=645cfb&color=645cfb'/></a>
-[![Medium](https://img.shields.io/badge/Medium-12100E?style=for-the-badge&logo=medium&logoColor=white)](https://blog.lancedb.com/) 
-[![Discord](https://img.shields.io/badge/Discord-%235865F2.svg?style=for-the-badge&logo=discord&logoColor=white)](https://discord.gg/zMM32dvNtd) 
+[![Blog](https://img.shields.io/badge/Blog-12100E?style=for-the-badge&logoColor=white)](https://blog.lancedb.com/)
+[![Discord](https://img.shields.io/badge/Discord-%235865F2.svg?style=for-the-badge&logo=discord&logoColor=white)](https://discord.gg/zMM32dvNtd)
 [![Twitter](https://img.shields.io/badge/Twitter-%231DA1F2.svg?style=for-the-badge&logo=Twitter&logoColor=white)](https://twitter.com/lancedb)
 
 </p>
@@ -20,7 +20,7 @@
 
 <hr />
 
-LanceDB is an open-source database for vector-search built with persistent storage, which greatly simplifies retrevial, filtering and management of embeddings.
+LanceDB is an open-source database for vector-search built with persistent storage, which greatly simplifies retrieval, filtering and management of embeddings.
 
 The key features of LanceDB include:
 
@@ -36,7 +36,7 @@ The key features of LanceDB include:
 
 * 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.
+* Ecosystem integrations with [LangChain 🦜️🔗](https://python.langchain.com/docs/integrations/vectorstores/lancedb/), [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.
 
@@ -44,26 +44,24 @@ LanceDB's core is written in Rust 🦀 and is built using <a href="https://githu
 
 **Javascript**
 ```shell
-npm install vectordb
+npm install @lancedb/lancedb
 ```
 
 ```javascript
-const lancedb = require('vectordb');
-const db = await lancedb.connect('data/sample-lancedb');
+import * as lancedb from "@lancedb/lancedb";
 
-const table = await db.createTable({
-  name: 'vectors',
-  data:  [
-    { id: 1, vector: [0.1, 0.2], item: "foo", price: 10 },
-    { id: 2, vector: [1.1, 1.2], item: "bar", price: 50 }
-  ]
-})
+const db = await lancedb.connect("data/sample-lancedb");
+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 },
+], {mode: 'overwrite'});
 
-const query = table.search([0.1, 0.3]).limit(2);
-const results = await query.execute();
+
+const query = table.vectorSearch([0.1, 0.3]).limit(2);
+const results = await query.toArray();
 
 // You can also search for rows by specific criteria without involving a vector search.
-const rowsByCriteria = await table.search(undefined).where("price >= 10").execute();
+const rowsByCriteria = await table.query().where("price >= 10").toArray();
 ```
 
 **Python**
@@ -83,5 +81,5 @@ result = table.search([100, 100]).limit(2).to_pandas()
 ```
 
 ## Blogs, Tutorials & Videos
-* 📈 <a href="https://blog.eto.ai/benchmarking-random-access-in-lance-ed690757a826">2000x better performance with Lance over Parquet</a>
+* 📈 <a href="https://blog.lancedb.com/benchmarking-random-access-in-lance/">2000x better performance with Lance over Parquet</a>
 * 🤖 <a href="https://github.com/lancedb/lancedb/blob/main/docs/src/notebooks/youtube_transcript_search.ipynb">Build a question and answer bot with LanceDB</a>

ci/build_linux_artifacts.sh

@@ -13,7 +13,9 @@ docker build \
     .
 popd
 
+# We turn on memory swap to avoid OOM killer
 docker run \
     -v $(pwd):/io -w /io \
+    --memory-swap=-1 \
     lancedb-node-manylinux \
     bash ci/manylinux_node/build.sh $ARCH

ci/build_linux_artifacts_nodejs.sh

@@ -0,0 +1,21 @@
+#!/bin/bash
+set -e
+ARCH=${1:-x86_64}
+
+# We pass down the current user so that when we later mount the local files
+# into the container, the files are accessible by the current user.
+pushd ci/manylinux_nodejs
+docker build \
+    -t lancedb-nodejs-manylinux \
+    --build-arg="ARCH=$ARCH" \
+    --build-arg="DOCKER_USER=$(id -u)" \
+    --progress=plain \
+    .
+popd
+
+# We turn on memory swap to avoid OOM killer
+docker run \
+    -v $(pwd):/io -w /io \
+    --memory-swap=-1 \
+    lancedb-nodejs-manylinux \
+    bash ci/manylinux_nodejs/build.sh $ARCH

ci/build_macos_artifacts_nodejs.sh

@@ -0,0 +1,34 @@
+# Builds the macOS artifacts (nodejs binaries).
+# Usage: ./ci/build_macos_artifacts_nodejs.sh [target]
+# Targets supported: x86_64-apple-darwin aarch64-apple-darwin
+set -e
+
+prebuild_rust() {
+    # Building here for the sake of easier debugging.
+    pushd rust/lancedb
+    echo "Building rust library for $1"
+    export RUST_BACKTRACE=1
+    cargo build --release --target $1
+    popd
+}
+
+build_node_binaries() {
+    pushd nodejs
+    echo "Building nodejs library for $1"
+    export RUST_TARGET=$1
+    npm run build-release
+    popd
+}
+
+if [ -n "$1" ]; then
+    targets=$1
+else
+    targets="x86_64-apple-darwin aarch64-apple-darwin"
+fi
+
+echo "Building artifacts for targets: $targets"
+for target in $targets
+    do
+    prebuild_rust $target
+    build_node_binaries $target
+done

ci/build_windows_artifacts_nodejs.ps1

@@ -0,0 +1,41 @@
+# Builds the Windows artifacts (nodejs binaries).
+# Usage:  .\ci\build_windows_artifacts_nodejs.ps1 [target]
+# Targets supported:
+# - x86_64-pc-windows-msvc
+# - i686-pc-windows-msvc
+
+function Prebuild-Rust {
+    param (
+        [string]$target
+    )
+
+    # Building here for the sake of easier debugging.
+    Push-Location -Path "rust/lancedb"
+    Write-Host "Building rust library for $target"
+    $env:RUST_BACKTRACE=1
+    cargo build --release --target $target
+    Pop-Location
+}
+
+function Build-NodeBinaries {
+    param (
+        [string]$target
+    )
+
+    Push-Location -Path "nodejs"
+    Write-Host "Building nodejs library for $target"
+    $env:RUST_TARGET=$target
+    npm run build-release
+    Pop-Location
+}
+
+$targets = $args[0]
+if (-not $targets) {
+    $targets = "x86_64-pc-windows-msvc"
+}
+
+Write-Host "Building artifacts for targets: $targets"
+foreach ($target in $targets) {
+    Prebuild-Rust $target
+    Build-NodeBinaries $target
+}

ci/bump_version.sh

@@ -0,0 +1,51 @@
+set -e
+
+RELEASE_TYPE=${1:-"stable"}
+BUMP_MINOR=${2:-false}
+TAG_PREFIX=${3:-"v"} # Such as "python-v"
+HEAD_SHA=${4:-$(git rev-parse HEAD)}
+
+readonly SELF_DIR=$(cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )
+
+PREV_TAG=$(git tag --sort='version:refname' | grep ^$TAG_PREFIX | python $SELF_DIR/semver_sort.py $TAG_PREFIX | tail -n 1)
+echo "Found previous tag $PREV_TAG"
+
+# Initially, we don't want to tag if we are doing stable, because we will bump
+# again later. See comment at end for why.
+if [[ "$RELEASE_TYPE" == 'stable' ]]; then 
+  BUMP_ARGS="--no-tag"
+fi
+
+# If last is stable and not bumping minor
+if [[ $PREV_TAG != *beta* ]]; then
+    if [[ "$BUMP_MINOR" != "false" ]]; then
+      # X.Y.Z -> X.(Y+1).0-beta.0
+      bump-my-version bump -vv $BUMP_ARGS minor
+    else
+      # X.Y.Z -> X.Y.(Z+1)-beta.0
+      bump-my-version bump -vv $BUMP_ARGS patch
+    fi
+else
+  if [[ "$BUMP_MINOR" != "false" ]]; then
+    # X.Y.Z-beta.N -> X.(Y+1).0-beta.0
+    bump-my-version bump -vv $BUMP_ARGS minor
+  else
+    # X.Y.Z-beta.N -> X.Y.Z-beta.(N+1)
+    bump-my-version bump -vv $BUMP_ARGS pre_n
+  fi
+fi
+
+# The above bump will always bump to a pre-release version. If we are releasing
+# a stable version, bump the pre-release level ("pre_l") to make it stable.
+if [[ $RELEASE_TYPE == 'stable' ]]; then
+  # X.Y.Z-beta.N -> X.Y.Z
+  bump-my-version bump -vv pre_l
+fi
+
+# Validate that we have incremented version appropriately for breaking changes
+NEW_TAG=$(git describe --tags --exact-match HEAD)
+NEW_VERSION=$(echo $NEW_TAG | sed "s/^$TAG_PREFIX//")
+LAST_STABLE_RELEASE=$(git tag --sort='version:refname' | grep ^$TAG_PREFIX | grep -v beta | grep -vF "$NEW_TAG" | python $SELF_DIR/semver_sort.py $TAG_PREFIX | tail -n 1)
+LAST_STABLE_VERSION=$(echo $LAST_STABLE_RELEASE | sed "s/^$TAG_PREFIX//")
+
+python $SELF_DIR/check_breaking_changes.py $LAST_STABLE_RELEASE $HEAD_SHA $LAST_STABLE_VERSION $NEW_VERSION

ci/check_breaking_changes.py

@@ -0,0 +1,35 @@
+"""
+Check whether there are any breaking changes in the PRs between the base and head commits.
+If there are, assert that we have incremented the minor version.
+"""
+import argparse
+import os
+from packaging.version import parse
+
+from github import Github
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("base")
+    parser.add_argument("head")
+    parser.add_argument("last_stable_version")
+    parser.add_argument("current_version")
+    args = parser.parse_args()
+
+    repo = Github(os.environ["GITHUB_TOKEN"]).get_repo(os.environ["GITHUB_REPOSITORY"])
+    commits = repo.compare(args.base, args.head).commits
+    prs = (pr for commit in commits for pr in commit.get_pulls())
+
+    for pr in prs:
+        if any(label.name == "breaking-change" for label in pr.labels):
+            print(f"Breaking change in PR: {pr.html_url}")
+            break
+    else:
+        print("No breaking changes found.")
+        exit(0)
+    
+    last_stable_version = parse(args.last_stable_version)
+    current_version = parse(args.current_version)
+    if current_version.minor <= last_stable_version.minor:
+        print("Minor version is not greater than the last stable version.")
+        exit(1)

ci/manylinux_node/Dockerfile

@@ -18,8 +18,8 @@ COPY install_protobuf.sh install_protobuf.sh
 RUN ./install_protobuf.sh ${ARCH}
 
 ENV DOCKER_USER=${DOCKER_USER}
-# Create a group and user
-RUN echo ${ARCH} && adduser --user-group --create-home --uid ${DOCKER_USER} build_user
+# Create a group and user, but only if it doesn't exist
+RUN echo ${ARCH} && id -u ${DOCKER_USER} >/dev/null 2>&1 || adduser --user-group --create-home --uid ${DOCKER_USER} build_user
 
 # We switch to the user to install Rust and Node, since those like to be 
 # installed at the user level.

ci/manylinux_nodejs/Dockerfile

@@ -0,0 +1,31 @@
+# Many linux dockerfile with Rust, Node, and Lance dependencies installed.
+# This container allows building the node modules native libraries in an
+# environment with a very old glibc, so that we are compatible with a wide
+# range of linux distributions.
+ARG ARCH=x86_64
+
+FROM quay.io/pypa/manylinux2014_${ARCH}
+
+ARG ARCH=x86_64
+ARG DOCKER_USER=default_user
+
+# Install static openssl
+COPY install_openssl.sh install_openssl.sh
+RUN ./install_openssl.sh ${ARCH} > /dev/null
+
+# Protobuf is also installed as root.
+COPY install_protobuf.sh install_protobuf.sh
+RUN ./install_protobuf.sh ${ARCH}
+
+ENV DOCKER_USER=${DOCKER_USER}
+# Create a group and user
+RUN echo ${ARCH} && adduser --user-group --create-home --uid ${DOCKER_USER} build_user
+
+# We switch to the user to install Rust and Node, since those like to be
+# installed at the user level.
+USER ${DOCKER_USER}
+
+COPY prepare_manylinux_node.sh prepare_manylinux_node.sh
+RUN cp /prepare_manylinux_node.sh $HOME/ && \
+    cd $HOME && \
+    ./prepare_manylinux_node.sh ${ARCH}

ci/manylinux_nodejs/build.sh

@@ -0,0 +1,18 @@
+#!/bin/bash
+# Builds the nodejs module for manylinux. Invoked by ci/build_linux_artifacts_nodejs.sh.
+set -e
+ARCH=${1:-x86_64}
+
+if [ "$ARCH" = "x86_64" ]; then
+    export OPENSSL_LIB_DIR=/usr/local/lib64/
+else
+    export OPENSSL_LIB_DIR=/usr/local/lib/
+fi
+export OPENSSL_STATIC=1
+export OPENSSL_INCLUDE_DIR=/usr/local/include/openssl
+
+source $HOME/.bashrc
+
+cd nodejs
+npm ci
+npm run build-release

ci/manylinux_nodejs/install_openssl.sh

@@ -0,0 +1,26 @@
+#!/bin/bash
+# Builds openssl from source so we can statically link to it
+
+# this is to avoid the error we get with the system installation:
+# /usr/bin/ld: <library>: version node not found for symbol SSLeay@@OPENSSL_1.0.1
+# /usr/bin/ld: failed to set dynamic section sizes: Bad value
+set -e
+
+git clone -b OpenSSL_1_1_1u \
+    --single-branch \
+    https://github.com/openssl/openssl.git
+
+pushd openssl
+
+if [[ $1 == x86_64* ]]; then
+    ARCH=linux-x86_64
+else
+    # gnu target
+    ARCH=linux-aarch64
+fi
+
+./Configure no-shared $ARCH
+
+make
+
+make install

ci/manylinux_nodejs/install_protobuf.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+# Installs protobuf compiler. Should be run as root.
+set -e
+
+if [[ $1 == x86_64* ]]; then
+    ARCH=x86_64
+else
+    # gnu target
+    ARCH=aarch_64
+fi
+
+PB_REL=https://github.com/protocolbuffers/protobuf/releases
+PB_VERSION=23.1
+curl -LO $PB_REL/download/v$PB_VERSION/protoc-$PB_VERSION-linux-$ARCH.zip
+unzip protoc-$PB_VERSION-linux-$ARCH.zip -d /usr/local

ci/manylinux_nodejs/prepare_manylinux_node.sh

@@ -0,0 +1,21 @@
+#!/bin/bash
+set -e
+
+install_node() {
+    echo "Installing node..."
+
+    curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.34.0/install.sh | bash
+
+    source "$HOME"/.bashrc
+
+    nvm install --no-progress 16
+}
+
+install_rust() {
+    echo "Installing rust..."
+    curl https://sh.rustup.rs -sSf | bash -s -- -y
+    export PATH="$PATH:/root/.cargo/bin"
+}
+
+install_node
+install_rust

ci/semver_sort.py

@@ -0,0 +1,35 @@
+"""
+Takes a list of semver strings and sorts them in ascending order.
+"""
+
+import sys
+from packaging.version import parse, InvalidVersion
+
+if __name__ == "__main__":
+    import argparse
+    parser = argparse.ArgumentParser()
+    parser.add_argument("prefix", default="v")
+    args = parser.parse_args()
+
+    # Read the input from stdin
+    lines = sys.stdin.readlines()
+
+    # Parse the versions
+    versions = []
+    for line in lines:
+        line = line.strip()
+        try:
+            version_str = line.removeprefix(args.prefix)
+            version = parse(version_str)
+        except InvalidVersion:
+            # There are old tags that don't follow the semver format
+            print(f"Invalid version: {line}", file=sys.stderr)
+            continue
+        versions.append((line, version))
+
+    # Sort the versions
+    versions.sort(key=lambda x: x[1])
+
+    # Print the sorted versions as original strings
+    for line, _ in versions:
+        print(line)

docker-compose.yml

@@ -1,18 +1,18 @@
 version: "3.9"
 services:
   localstack:
-    image: localstack/localstack:0.14
+    image: localstack/localstack:3.3
     ports:
       - 4566:4566
     environment:
-      - SERVICES=s3,dynamodb
+      - SERVICES=s3,dynamodb,kms
       - DEBUG=1
       - LS_LOG=trace
       - DOCKER_HOST=unix:///var/run/docker.sock
       - AWS_ACCESS_KEY_ID=ACCESSKEY
       - AWS_SECRET_ACCESS_KEY=SECRETKEY
     healthcheck:
-      test: [ "CMD", "curl", "-f", "http://localhost:4566/health" ]
+      test: [ "CMD", "curl", "-s", "http://localhost:4566/_localstack/health" ]
       interval: 5s
       retries: 3
       start_period: 10s

dockerfiles/Dockerfile

@@ -0,0 +1,27 @@
+#Simple base dockerfile that supports basic dependencies required to run lance with FTS and Hybrid Search
+#Usage docker build -t lancedb:latest -f Dockerfile .
+FROM python:3.10-slim-buster
+
+# Install Rust
+RUN apt-get update && apt-get install -y curl build-essential && \
+  curl https://sh.rustup.rs -sSf | sh -s -- -y
+
+# Set the environment variable for Rust
+ENV PATH="/root/.cargo/bin:${PATH}"
+
+# Install protobuf compiler
+RUN apt-get install -y protobuf-compiler && \
+  apt-get clean && \
+  rm -rf /var/lib/apt/lists/*
+
+RUN apt-get -y update &&\
+  apt-get -y upgrade && \
+  apt-get -y install git
+
+
+# Verify installations
+RUN python --version && \
+  rustc --version && \
+  protoc --version
+
+RUN pip install tantivy lancedb

docs/mkdocs.yml

@@ -27,7 +27,6 @@ theme:
     - content.tabs.link
     - content.action.edit
     - toc.follow
-    # - toc.integrate
     - navigation.top
     - navigation.tabs
     - navigation.tabs.sticky
@@ -39,162 +38,209 @@ theme:
   custom_dir: overrides
 
 plugins:
-- search
-- autorefs
-- mkdocstrings:
-    handlers:
-      python:
-        paths: [../python]
-        options:
-          docstring_style: numpy
-          heading_level: 4
-          show_source: true
-          show_symbol_type_in_heading: true
-          show_signature_annotations: true
-          members_order: source
-        import:
-          # for cross references
-          - https://arrow.apache.org/docs/objects.inv
-          - https://pandas.pydata.org/docs/objects.inv
-- mkdocs-jupyter
+  - search
+  - autorefs
+  - mkdocstrings:
+      handlers:
+        python:
+          paths: [../python]
+          options:
+            docstring_style: numpy
+            heading_level: 3
+            show_source: true
+            show_symbol_type_in_heading: true
+            show_signature_annotations: true
+            show_root_heading: true
+            members_order: source
+          import:
+            # for cross references
+            - https://arrow.apache.org/docs/objects.inv
+            - https://pandas.pydata.org/docs/objects.inv
+  - mkdocs-jupyter
+  - render_swagger:
+      allow_arbitrary_locations : true
 
 markdown_extensions:
-- admonition
-- footnotes
-- pymdownx.details
-- pymdownx.highlight:
-    anchor_linenums: true
-    line_spans: __span
-    pygments_lang_class: true
-- pymdownx.inlinehilite
-- pymdownx.snippets:
-    base_path: ..
-    dedent_subsections: true
-- pymdownx.superfences
-- pymdownx.tabbed:
-    alternate_style: true
-- md_in_html
-- attr_list
+  - admonition
+  - footnotes
+  - pymdownx.details
+  - pymdownx.highlight:
+      anchor_linenums: true
+      line_spans: __span
+      pygments_lang_class: true
+  - pymdownx.inlinehilite
+  - pymdownx.snippets:
+      base_path: ..
+      dedent_subsections: true
+  - pymdownx.superfences
+  - pymdownx.tabbed:
+      alternate_style: true
+  - md_in_html
+  - attr_list
 
 nav:
-- Home:
-  - LanceDB: index.md
-  - 🏃🏼‍♂️ Quick start: basic.md
-  - 📚 Concepts:
-    - Vector search: concepts/vector_search.md
-    - Indexing: concepts/index_ivfpq.md
-    - Storage: concepts/storage.md
-    - Data management: concepts/data_management.md
-  - 🔨 Guides:
-    - Working with tables: guides/tables.md
-    - Building an ANN index: ann_indexes.md
-    - Vector Search: search.md
-    - Full-text search: fts.md
-    - Hybrid search: hybrid_search.md
-    - Filtering: sql.md
-    - Versioning & Reproducibility: notebooks/reproducibility.ipynb
-    - Configuring Storage: guides/storage.md
-  - 🧬 Managing embeddings:
-    - Overview: embeddings/index.md
-    - Explicit management: embeddings/embedding_explicit.md
-    - Implicit management: embeddings/embedding_functions.md
-    - Available Functions: embeddings/default_embedding_functions.md
-    - Custom Embedding Functions: embeddings/api.md
-    - "Example: Multi-lingual semantic search": notebooks/multi_lingual_example.ipynb
-    - "Example: MultiModal CLIP Embeddings": notebooks/DisappearingEmbeddingFunction.ipynb
-  - 🔌 Integrations:
-    - Tools and data formats: integrations/index.md
-    - Pandas and PyArrow: python/pandas_and_pyarrow.md
-    - Polars: python/polars_arrow.md
-    - DuckDB: python/duckdb.md
-    - LangChain 🔗: https://python.langchain.com/en/latest/modules/indexes/vectorstores/examples/lancedb.html
-    - LangChain JS/TS 🔗: https://js.langchain.com/docs/modules/data_connection/vectorstores/integrations/lancedb
-    - LlamaIndex 🦙: https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/LanceDBIndexDemo.html
-    - Pydantic: python/pydantic.md
-    - Voxel51: integrations/voxel51.md
-    - PromptTools: integrations/prompttools.md
-  - 🎯 Examples:
-    - Overview: examples/index.md
-    - 🐍 Python:
-      - Overview: examples/examples_python.md
+  - Home:
+      - LanceDB: index.md
+      - 🏃🏼‍♂️ Quick start: basic.md
+      - 📚 Concepts:
+          - Vector search: concepts/vector_search.md
+          - Indexing: concepts/index_ivfpq.md
+          - Storage: concepts/storage.md
+          - Data management: concepts/data_management.md
+      - 🔨 Guides:
+          - Working with tables: guides/tables.md
+          - Building an ANN index: ann_indexes.md
+          - Vector Search: search.md
+          - Full-text search: fts.md
+          - Hybrid search:
+              - Overview: hybrid_search/hybrid_search.md
+              - Comparing Rerankers: hybrid_search/eval.md
+              - Airbnb financial data example: notebooks/hybrid_search.ipynb
+          - Reranking:
+              - Quickstart: reranking/index.md
+              - Cohere Reranker: reranking/cohere.md
+              - Linear Combination Reranker: reranking/linear_combination.md
+              - Reciprocal Rank Fusion Reranker: reranking/rrf.md
+              - Cross Encoder Reranker: reranking/cross_encoder.md
+              - ColBERT Reranker: reranking/colbert.md
+              - Jina Reranker: reranking/jina.md
+              - OpenAI Reranker: reranking/openai.md
+              - Building Custom Rerankers: reranking/custom_reranker.md
+              - Example: notebooks/lancedb_reranking.ipynb
+          - Filtering: sql.md
+          - Versioning & Reproducibility: notebooks/reproducibility.ipynb
+          - Configuring Storage: guides/storage.md
+          - Migration Guide: migration.md
+          - Tuning retrieval performance:
+              - Choosing right query type: guides/tuning_retrievers/1_query_types.md
+              - Reranking: guides/tuning_retrievers/2_reranking.md
+              - Embedding fine-tuning: guides/tuning_retrievers/3_embed_tuning.md
+      - 🧬 Managing embeddings:
+          - Overview: embeddings/index.md
+          - Embedding functions: embeddings/embedding_functions.md
+          - Available models: embeddings/default_embedding_functions.md
+          - User-defined embedding functions: embeddings/custom_embedding_function.md
+          - "Example: Multi-lingual semantic search": notebooks/multi_lingual_example.ipynb
+          - "Example: MultiModal CLIP Embeddings": notebooks/DisappearingEmbeddingFunction.ipynb
+      - 🔌 Integrations:
+          - Tools and data formats: integrations/index.md
+          - Pandas and PyArrow: python/pandas_and_pyarrow.md
+          - Polars: python/polars_arrow.md
+          - DuckDB: python/duckdb.md
+          - LangChain:
+            - LangChain 🔗: integrations/langchain.md
+            - LangChain demo: notebooks/langchain_demo.ipynb
+            - LangChain JS/TS 🔗: https://js.langchain.com/docs/integrations/vectorstores/lancedb
+          - LlamaIndex 🦙:
+            - LlamaIndex docs: integrations/llamaIndex.md
+            - LlamaIndex demo: notebooks/llamaIndex_demo.ipynb
+          - Pydantic: python/pydantic.md
+          - Voxel51: integrations/voxel51.md
+          - PromptTools: integrations/prompttools.md
+      - 🎯 Examples:
+          - Overview: examples/index.md
+          - 🐍 Python:
+              - Overview: examples/examples_python.md
+              - YouTube Transcript Search: notebooks/youtube_transcript_search.ipynb
+              - Documentation QA Bot using LangChain: notebooks/code_qa_bot.ipynb
+              - Multimodal search using CLIP: notebooks/multimodal_search.ipynb
+              - Example - Calculate CLIP Embeddings with Roboflow Inference: examples/image_embeddings_roboflow.md
+              - Serverless QA Bot with S3 and Lambda: examples/serverless_lancedb_with_s3_and_lambda.md
+              - Serverless QA Bot with Modal: examples/serverless_qa_bot_with_modal_and_langchain.md
+          - 👾 JavaScript:
+              - Overview: examples/examples_js.md
+              - 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
+          - 🦀 Rust:
+              - Overview: examples/examples_rust.md
+      - 💭 FAQs: faq.md
+      - ⚙️ API reference:
+          - 🐍 Python: python/python.md
+          - 👾 JavaScript (vectordb): javascript/modules.md
+          - 👾 JavaScript (lancedb): js/globals.md
+          - 🦀 Rust: https://docs.rs/lancedb/latest/lancedb/
+      - ☁️ LanceDB Cloud:
+          - Overview: cloud/index.md
+          - API reference:
+              - 🐍 Python: python/saas-python.md
+              - 👾 JavaScript: javascript/modules.md
+              - REST API: cloud/rest.md
+
+  - Quick start: basic.md
+  - Concepts:
+      - Vector search: concepts/vector_search.md
+      - Indexing: concepts/index_ivfpq.md
+      - Storage: concepts/storage.md
+      - Data management: concepts/data_management.md
+  - Guides:
+      - Working with tables: guides/tables.md
+      - Building an ANN index: ann_indexes.md
+      - Vector Search: search.md
+      - Full-text search: fts.md
+      - Hybrid search:
+          - Overview: hybrid_search/hybrid_search.md
+          - Comparing Rerankers: hybrid_search/eval.md
+          - Airbnb financial data example: notebooks/hybrid_search.ipynb
+      - Reranking:
+          - Quickstart: reranking/index.md
+          - Cohere Reranker: reranking/cohere.md
+          - Linear Combination Reranker: reranking/linear_combination.md
+          - Reciprocal Rank Fusion Reranker: reranking/rrf.md
+          - Cross Encoder Reranker: reranking/cross_encoder.md
+          - ColBERT Reranker: reranking/colbert.md
+          - Jina Reranker: reranking/jina.md
+          - OpenAI Reranker: reranking/openai.md
+          - Building Custom Rerankers: reranking/custom_reranker.md
+          - Example: notebooks/lancedb_reranking.ipynb
+      - Filtering: sql.md
+      - Versioning & Reproducibility: notebooks/reproducibility.ipynb
+      - Configuring Storage: guides/storage.md
+      - Migration Guide: migration.md
+      - Tuning retrieval performance:
+          - Choosing right query type: guides/tuning_retrievers/1_query_types.md
+          - Reranking: guides/tuning_retrievers/2_reranking.md
+          - Embedding fine-tuning: guides/tuning_retrievers/3_embed_tuning.md
+  - Managing Embeddings:
+      - Overview: embeddings/index.md
+      - Embedding functions: embeddings/embedding_functions.md
+      - Available models: embeddings/default_embedding_functions.md
+      - User-defined embedding functions: embeddings/custom_embedding_function.md
+      - "Example: Multi-lingual semantic search": notebooks/multi_lingual_example.ipynb
+      - "Example: MultiModal CLIP Embeddings": notebooks/DisappearingEmbeddingFunction.ipynb
+  - Integrations:
+      - Overview: integrations/index.md
+      - Pandas and PyArrow: python/pandas_and_pyarrow.md
+      - Polars: python/polars_arrow.md
+      - DuckDB: python/duckdb.md
+      - LangChain 🦜️🔗↗: integrations/langchain.md
+      - LangChain.js 🦜️🔗↗: https://js.langchain.com/docs/integrations/vectorstores/lancedb
+      - LlamaIndex 🦙↗: integrations/llamaIndex.md
+      - Pydantic: python/pydantic.md
+      - Voxel51: integrations/voxel51.md
+      - PromptTools: integrations/prompttools.md
+  - Examples:
+      - examples/index.md
       - YouTube Transcript Search: notebooks/youtube_transcript_search.ipynb
       - Documentation QA Bot using LangChain: notebooks/code_qa_bot.ipynb
       - Multimodal search using CLIP: notebooks/multimodal_search.ipynb
-      - Example - Calculate CLIP Embeddings with Roboflow Inference: examples/image_embeddings_roboflow.md
       - Serverless QA Bot with S3 and Lambda: examples/serverless_lancedb_with_s3_and_lambda.md
       - Serverless QA Bot with Modal: examples/serverless_qa_bot_with_modal_and_langchain.md
-    - 👾 JavaScript:
-      - Overview: examples/examples_js.md
-      - Serverless Website Chatbot: examples/serverless_website_chatbot.md
-      - YouTube Transcript Search: examples/youtube_transcript_bot_with_nodejs.md
+      - YouTube Transcript Search (JS): examples/youtube_transcript_bot_with_nodejs.md
+      - Serverless Chatbot from any website: examples/serverless_website_chatbot.md
       - TransformersJS Embedding Search: examples/transformerjs_embedding_search_nodejs.md
-  - 🔧 CLI & Config: cli_config.md
-  - 💭 FAQs: faq.md
-  - ⚙️ API reference:
-    - 🐍 Python: python/python.md
-    - 👾 JavaScript: javascript/modules.md
-    - 🦀 Rust: https://docs.rs/vectordb/latest/vectordb/
-  - ☁️ LanceDB Cloud:
-    - Overview: cloud/index.md
-    - API reference:
-      - 🐍 Python: python/saas-python.md
-      - 👾 JavaScript: javascript/saas-modules.md
-
-
-- Quick start: basic.md
-- Concepts:
-    - Vector search: concepts/vector_search.md
-    - Indexing: concepts/index_ivfpq.md
-    - Storage: concepts/storage.md
-    - Data management: concepts/data_management.md
-- Guides:
-    - Working with tables: guides/tables.md
-    - Building an ANN index: ann_indexes.md
-    - Vector Search: search.md
-    - Full-text search: fts.md
-    - Hybrid search: hybrid_search.md
-    - Filtering: sql.md
-    - Versioning & Reproducibility: notebooks/reproducibility.ipynb
-    - Configuring Storage: guides/storage.md
-- Managing Embeddings:
-  - Overview: embeddings/index.md
-  - Explicit management: embeddings/embedding_explicit.md
-  - Implicit management: embeddings/embedding_functions.md
-  - Available Functions: embeddings/default_embedding_functions.md
-  - Custom Embedding Functions: embeddings/api.md
-  - "Example: Multi-lingual semantic search": notebooks/multi_lingual_example.ipynb
-  - "Example: MultiModal CLIP Embeddings": notebooks/DisappearingEmbeddingFunction.ipynb
-- Integrations:
-  - Overview: integrations/index.md
-  - Pandas and PyArrow: python/pandas_and_pyarrow.md
-  - Polars: python/polars_arrow.md
-  - DuckDB : python/duckdb.md
-  - LangChain 🦜️🔗↗: https://python.langchain.com/en/latest/modules/indexes/vectorstores/examples/lancedb.html
-  - LangChain.js 🦜️🔗↗: https://js.langchain.com/docs/modules/data_connection/vectorstores/integrations/lancedb
-  - LlamaIndex 🦙↗: https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/LanceDBIndexDemo.html
-  - Pydantic: python/pydantic.md
-  - Voxel51: integrations/voxel51.md
-  - PromptTools: integrations/prompttools.md
-- Python examples:
-  - examples/index.md
-  - YouTube Transcript Search: notebooks/youtube_transcript_search.ipynb
-  - Documentation QA Bot using LangChain: notebooks/code_qa_bot.ipynb
-  - Multimodal search using CLIP: notebooks/multimodal_search.ipynb
-  - Serverless QA Bot with S3 and Lambda: examples/serverless_lancedb_with_s3_and_lambda.md
-  - Serverless QA Bot with Modal: examples/serverless_qa_bot_with_modal_and_langchain.md
-- Javascript examples:
-  - Overview: examples/examples_js.md
-  - YouTube Transcript Search: examples/youtube_transcript_bot_with_nodejs.md
-  - Serverless Chatbot from any website: examples/serverless_website_chatbot.md
-  - TransformersJS Embedding Search: examples/transformerjs_embedding_search_nodejs.md
-- API reference:
-  - Python: python/python.md
-  - Javascript: javascript/modules.md
-- LanceDB Cloud:
-    - Overview: cloud/index.md
-    - API reference:
-      - 🐍 Python: python/saas-python.md
-      - 👾 JavaScript: javascript/saas-modules.md
+  - API reference:
+      - Overview: api_reference.md
+      - Python: python/python.md
+      - Javascript (vectordb): javascript/modules.md
+      - Javascript (lancedb): js/globals.md
+      - Rust: https://docs.rs/lancedb/latest/lancedb/index.html
+  - LanceDB Cloud:
+      - Overview: cloud/index.md
+      - API reference:
+          - 🐍 Python: python/saas-python.md
+          - 👾 JavaScript: javascript/modules.md
+          - REST API: cloud/rest.md
 
 extra_css:
   - styles/global.css
@@ -207,3 +253,10 @@ extra:
   analytics:
     provider: google
     property: G-B7NFM40W74
+  social:
+    - icon: fontawesome/brands/github
+      link: https://github.com/lancedb/lancedb
+    - icon: fontawesome/brands/x-twitter
+      link: https://twitter.com/lancedb
+    - icon: fontawesome/brands/linkedin
+      link: https://www.linkedin.com/company/lancedb

docs/openapi.yml

@@ -0,0 +1,487 @@
+openapi: 3.1.0
+info:
+  version: 1.0.0
+  title: LanceDB Cloud API
+  description: |
+    LanceDB Cloud API is a RESTful API that allows users to access and modify data stored in LanceDB Cloud.
+    Table actions are considered temporary resource creations and all use POST method.
+  contact:
+    name: LanceDB support
+    url: https://lancedb.com
+    email: contact@lancedb.com
+
+servers:
+  - url: https://{db}.{region}.api.lancedb.com
+    description: LanceDB Cloud REST endpoint.
+    variables:
+      db:
+        default: ""
+        description: the name of DB
+      region:
+        default: "us-east-1"
+        description: the service region of the DB
+
+security:
+  - key_auth: []
+
+components:
+  securitySchemes:
+    key_auth:
+      name: x-api-key
+      type: apiKey
+      in: header
+  parameters:
+    table_name:
+      name: name
+      in: path
+      description: name of the table
+      required: true
+      schema:
+        type: string
+  responses:
+    invalid_request:
+      description: Invalid request
+      content:
+        text/plain:
+          schema:
+            type: string
+    not_found:
+      description: Not found
+      content:
+        text/plain:
+          schema:
+            type: string
+    unauthorized:
+      description: Unauthorized
+      content:
+        text/plain:
+          schema:
+            type: string
+  requestBodies:
+    arrow_stream_buffer:
+      description: Arrow IPC stream buffer
+      required: true
+      content:
+        application/vnd.apache.arrow.stream:
+          schema:
+            type: string
+            format: binary
+
+paths:
+  /v1/table/:
+    get:
+      description: List tables, optionally, with pagination.
+      tags:
+        - Tables
+      summary: List Tables
+      operationId: listTables
+      parameters:
+        - name: limit
+          in: query
+          description: Limits the number of items to return.
+          schema:
+            type: integer
+        - name: page_token
+          in: query
+          description: Specifies the starting position of the next query
+          schema:
+            type: string
+      responses:
+        "200":
+          description: Successfully returned a list of tables in the DB
+          content:
+            application/json:
+              schema:
+                type: object
+                properties:
+                  tables:
+                    type: array
+                    items:
+                      type: string
+                  page_token:
+                    type: string
+
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+
+  /v1/table/{name}/create/:
+    post:
+      description: Create a new table
+      summary: Create a new table
+      operationId: createTable
+      tags:
+        - Tables
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        $ref: "#/components/requestBodies/arrow_stream_buffer"
+      responses:
+        "200":
+          description: Table successfully created
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+
+  /v1/table/{name}/query/:
+    post:
+      description: Vector Query
+      url: https://{db-uri}.{aws-region}.api.lancedb.com/v1/table/{name}/query/
+      tags:
+        - Data
+      summary: Vector Query
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                vector:
+                  type: FixedSizeList
+                  description: |
+                    The targetted vector to search for. Required.
+                vector_column:
+                  type: string
+                  description: |
+                    The column to query, it can be inferred from the schema if there is only one vector column.
+                prefilter:
+                  type: boolean
+                  description: |
+                    Whether to prefilter the data. Optional.
+                k:
+                  type: integer
+                  description: |
+                    The number of search results to return. Default is 10.
+                distance_type:
+                  type: string
+                  description: |
+                    The distance metric to use for search. L2, Cosine, Dot and Hamming are supported. Default is L2.
+                bypass_vector_index:
+                  type: boolean
+                  description: |
+                    Whether to bypass vector index. Optional.
+                filter:
+                  type: string
+                  description: |
+                    A filter expression that specifies the rows to query. Optional.
+                columns:
+                  type: array
+                  items:
+                    type: string
+                  description: |
+                    The columns to return. Optional.
+                nprobe:
+                  type: integer
+                  description: |
+                    The number of probes to use for search. Optional.
+                refine_factor:
+                  type: integer
+                  description: |
+                    The refine factor to use for search. Optional.
+                  default: null
+                fast_search:
+                  type: boolean
+                  description: |
+                    Whether to use fast search. Optional.
+                  default: false
+              required:
+                - vector
+
+      responses:
+        "200":
+          description: top k results if query is successfully executed
+          content:
+            application/json:
+              schema:
+                type: object
+                properties:
+                  results:
+                    type: array
+                    items:
+                      type: object
+                      properties:
+                        id:
+                          type: integer
+                        selected_col_1_to_return:
+                          type: col_1_type
+                        selected_col_n_to_return:
+                          type: col_n_type
+                        _distance:
+                          type: float
+
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+
+  /v1/table/{name}/insert/:
+    post:
+      description: Insert new data to the Table.
+      tags:
+        - Data
+      operationId: insertData
+      summary: Insert new data.
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        $ref: "#/components/requestBodies/arrow_stream_buffer"
+      responses:
+        "200":
+          description: Insert successful
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+  /v1/table/{name}/merge_insert/:
+    post:
+      description: Create a "merge insert" operation
+        This operation can add rows, update rows, and remove rows all in a single
+        transaction. See python method `lancedb.table.Table.merge_insert` for examples.
+      tags:
+        - Data
+      summary: Merge Insert
+      operationId: mergeInsert
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+        - name: on
+          in: query
+          description: |
+            The column to use as the primary key for the merge operation.
+          required: true
+          schema:
+            type: string
+        - name: when_matched_update_all
+          in: query
+          description: |
+            Rows that exist in both the source table (new data) and
+            the target table (old data) will be updated, replacing
+            the old row with the corresponding matching row.
+          required: false
+          schema:
+            type: boolean
+        - name: when_matched_update_all_filt
+          in: query
+          description: |
+            If present then only rows that satisfy the filter expression will
+            be updated
+          required: false
+          schema:
+            type: string
+        - name: when_not_matched_insert_all
+          in: query
+          description: |
+            Rows that exist only in the source table (new data) will be
+            inserted into the target table (old data).
+          required: false
+          schema:
+            type: boolean
+        - name: when_not_matched_by_source_delete
+          in: query
+          description: |
+            Rows that exist only in the target table (old data) will be
+            deleted. An optional condition (`when_not_matched_by_source_delete_filt`)
+            can be provided to limit what data is deleted.
+          required: false
+          schema:
+            type: boolean
+        - name: when_not_matched_by_source_delete_filt
+          in: query
+          description: |
+            The filter expression that specifies the rows to delete.
+          required: false
+          schema:
+            type: string
+      requestBody:
+        $ref: "#/components/requestBodies/arrow_stream_buffer"
+      responses:
+        "200":
+          description: Merge Insert successful
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+  /v1/table/{name}/delete/:
+    post:
+      description: Delete rows from a table.
+      tags:
+        - Data
+      summary: Delete rows from a table
+      operationId: deleteData
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                predicate:
+                  type: string
+                  description: |
+                    A filter expression that specifies the rows to delete.
+      responses:
+        "200":
+          description: Delete successful
+        "401":
+          $ref: "#/components/responses/unauthorized"
+  /v1/table/{name}/drop/:
+    post:
+      description: Drop a table
+      tags:
+        - Tables
+      summary: Drop a table
+      operationId: dropTable
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        $ref: "#/components/requestBodies/arrow_stream_buffer"
+      responses:
+        "200":
+          description: Drop successful
+        "401":
+          $ref: "#/components/responses/unauthorized"
+
+  /v1/table/{name}/describe/:
+    post:
+      description: Describe a table and return Table Information.
+      tags:
+        - Tables
+      summary: Describe a table
+      operationId: describeTable
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      responses:
+        "200":
+          description: Table information
+          content:
+            application/json:
+              schema:
+                type: object
+                properties:
+                  table:
+                    type: string
+                  version:
+                    type: integer
+                  schema:
+                    type: string
+                  stats:
+                    type: object
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+
+  /v1/table/{name}/index/list/:
+    post:
+      description: List indexes of a table
+      tags:
+        - Tables
+      summary: List indexes of a table
+      operationId: listIndexes
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      responses:
+        "200":
+          description: Available list of indexes on the table.
+          content:
+            application/json:
+              schema:
+                type: object
+                properties:
+                  indexes:
+                    type: array
+                    items:
+                      type: object
+                      properties:
+                        columns:
+                          type: array
+                          items:
+                            type: string
+                        index_name:
+                          type: string
+                        index_uuid:
+                          type: string
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+  /v1/table/{name}/create_index/:
+    post:
+      description: Create vector index on a Table
+      tags:
+        - Tables
+      summary: Create vector index on a Table
+      operationId: createIndex
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                column:
+                  type: string
+                metric_type:
+                  type: string
+                  nullable: false
+                  description: |
+                    The metric type to use for the index. L2, Cosine, Dot are supported.
+                index_type:
+                  type: string
+      responses:
+        "200":
+          description: Index successfully created
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"
+  /v1/table/{name}/create_scalar_index/:
+    post:
+      description: Create a scalar index on a table
+      tags:
+        - Tables
+      summary: Create a scalar index on a table
+      operationId: createScalarIndex
+      parameters:
+        - $ref: "#/components/parameters/table_name"
+      requestBody:
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                column:
+                  type: string
+                index_type:
+                  type: string
+                  required: false
+      responses:
+        "200":
+          description: Scalar Index successfully created
+        "400":
+          $ref: "#/components/responses/invalid_request"
+        "401":
+          $ref: "#/components/responses/unauthorized"
+        "404":
+          $ref: "#/components/responses/not_found"

docs/requirements.txt

@@ -2,4 +2,5 @@ mkdocs==1.5.3
 mkdocs-jupyter==0.24.1
 mkdocs-material==9.5.3
 mkdocstrings[python]==0.20.0
-pydantic
+mkdocs-render-swagger-plugin
+pydantic

docs/src/ann_indexes.md

@@ -7,20 +7,11 @@ for brute-force scanning of the entire vector space.
 A vector index is faster but less accurate than exhaustive search (kNN or flat search).
 LanceDB provides many parameters to fine-tune the index's size, the speed of queries, and the accuracy of results.
 
-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.
+## Disk-based Index
 
-In the future we will look to automatically create and configure the ANN index as data comes in.
-
-## Types of Index
-
-Lance can support multiple index types, the most widely used one is `IVF_PQ`.
-
-- `IVF_PQ`: use **Inverted File Index (IVF)** to first divide the dataset into `N` partitions,
-  and then use **Product Quantization** to compress vectors in each partition.
-- `DiskANN` (**Experimental**): organize the vector as a on-disk graph, where the vertices approximately
-  represent the nearest neighbors of each vector.
+Lance provides an `IVF_PQ` disk-based index. It uses **Inverted File Index (IVF)** to first divide
+the dataset into `N` partitions, and then applies **Product Quantization** to compress vectors in each partition.
+See the [indexing](concepts/index_ivfpq.md) concepts guide for more information on how this works.
 
 ## Creating an IVF_PQ Index
 
@@ -28,39 +19,75 @@ Lance supports `IVF_PQ` index type by default.
 
 === "Python"
 
-     Creating indexes is done via the [create_index](https://lancedb.github.io/lancedb/python/#lancedb.table.LanceTable.create_index) method.
+    Creating indexes is done via the [create_index](https://lancedb.github.io/lancedb/python/#lancedb.table.LanceTable.create_index) method.
 
-     ```python
-     import lancedb
-     import numpy as np
-     uri = "data/sample-lancedb"
-     db = lancedb.connect(uri)
+    ```python
+    import lancedb
+    import numpy as np
+    uri = "data/sample-lancedb"
+    db = lancedb.connect(uri)
 
-     # Create 10,000 sample vectors
-     data = [{"vector": row, "item": f"item {i}"}
+    # Create 10,000 sample vectors
+    data = [{"vector": row, "item": f"item {i}"}
         for i, row in enumerate(np.random.random((10_000, 1536)).astype('float32'))]
 
-     # Add the vectors to a table
-     tbl = db.create_table("my_vectors", data=data)
+    # Add the vectors to a table
+    tbl = db.create_table("my_vectors", data=data)
 
-     # Create and train the index - you need to have enough data in the table for an effective training step
-     tbl.create_index(num_partitions=256, num_sub_vectors=96)
-     ```
+    # Create and train the index - you need to have enough data in the table for an effective training step
+    tbl.create_index(num_partitions=256, num_sub_vectors=96)
+    ```
 
-=== "Typescript"
+=== "TypeScript"
 
-     ```typescript
-     --8<--- "docs/src/ann_indexes.ts:import"
+    === "@lancedb/lancedb"
 
-     --8<-- "docs/src/ann_indexes.ts:ingest"
-     ```
+        Creating indexes is done via the [lancedb.Table.createIndex](../js/classes/Table.md/#createIndex) method.
 
-- **metric** (default: "L2"): The distance metric to use. By default it uses euclidean distance "`L2`".
+        ```typescript
+        --8<--- "nodejs/examples/ann_indexes.ts:import"
+
+        --8<-- "nodejs/examples/ann_indexes.ts:ingest"
+        ```
+
+    === "vectordb (deprecated)"
+
+        Creating indexes is done via the [lancedb.Table.createIndex](../javascript/interfaces/Table.md/#createIndex) method.
+
+        ```typescript
+        --8<--- "docs/src/ann_indexes.ts:import"
+
+        --8<-- "docs/src/ann_indexes.ts:ingest"
+        ```
+
+=== "Rust"
+
+    ```rust
+    --8<-- "rust/lancedb/examples/ivf_pq.rs:create_index"
+    ```
+
+    IVF_PQ index parameters are more fully defined in the [crate docs](https://docs.rs/lancedb/latest/lancedb/index/vector/struct.IvfPqIndexBuilder.html).
+
+The following IVF_PQ paramters can be specified:
+
+- **distance_type**: The distance metric to use. By default it uses euclidean distance "`L2`".
   We also support "cosine" and "dot" distance as well.
-- **num_partitions** (default: 256): The number of partitions of the index.
-- **num_sub_vectors** (default: 96): The number of sub-vectors (M) that will be created during Product Quantization (PQ).
-  For D dimensional vector, it will be divided into `M` of `D/M` sub-vectors, each of which is presented by
-  a single PQ code.
+- **num_partitions**: The number of partitions in the index. The default is the square root
+  of the number of rows.
+
+!!! note
+
+    In the synchronous python SDK and node's `vectordb` the default is 256. This default has
+    changed in the asynchronous python SDK and node's `lancedb`.
+
+- **num_sub_vectors**: The number of sub-vectors (M) that will be created during Product Quantization (PQ).
+  For D dimensional vector, it will be divided into `M` subvectors with dimension `D/M`, each of which is replaced by
+  a single PQ code. The default is the dimension of the vector divided by 16.
+
+!!! note
+
+    In the synchronous python SDK and node's `vectordb` the default is currently 96. This default has
+    changed in the asynchronous python SDK and node's `lancedb`.
 
 <figure markdown>
   ![IVF PQ](./assets/ivf_pq.png)
@@ -78,29 +105,29 @@ You can specify the GPU device to train IVF partitions via
 
 === "Linux"
 
-     <!-- skip-test -->
-     ``` { .python .copy }
-     # Create index using CUDA on Nvidia GPUs.
-     tbl.create_index(
-          num_partitions=256,
-          num_sub_vectors=96,
-          accelerator="cuda"
-     )
-     ```
+    <!-- skip-test -->
+    ``` { .python .copy }
+    # Create index using CUDA on Nvidia GPUs.
+    tbl.create_index(
+        num_partitions=256,
+        num_sub_vectors=96,
+        accelerator="cuda"
+    )
+    ```
 
-=== "Macos"
+=== "MacOS"
 
-     <!-- skip-test -->
-     ```python
-     # Create index using MPS on Apple Silicon.
-     tbl.create_index(
-          num_partitions=256,
-          num_sub_vectors=96,
-          accelerator="mps"
-     )
-     ```
+    <!-- skip-test -->
+    ```python
+    # Create index using MPS on Apple Silicon.
+    tbl.create_index(
+        num_partitions=256,
+        num_sub_vectors=96,
+        accelerator="mps"
+    )
+    ```
 
-Trouble shootings:
+Troubleshooting:
 
 If you see `AssertionError: Torch not compiled with CUDA enabled`, you need to [install
 PyTorch with CUDA support](https://pytorch.org/get-started/locally/).
@@ -123,25 +150,41 @@ There are a couple of parameters that can be used to fine-tune the search:
 
 === "Python"
 
-     ```python
-     tbl.search(np.random.random((1536))) \
-         .limit(2) \
-         .nprobes(20) \
-         .refine_factor(10) \
-         .to_pandas()
-     ```
+    ```python
+    tbl.search(np.random.random((1536))) \
+        .limit(2) \
+        .nprobes(20) \
+        .refine_factor(10) \
+        .to_pandas()
+    ```
 
-     ```text
+    ```text
                                               vector       item       _distance
-     0  [0.44949695, 0.8444449, 0.06281311, 0.23338133...  item 1141  103.575333
-     1  [0.48587373, 0.269207, 0.15095535, 0.65531915,...  item 3953  108.393867
-     ```
+    0  [0.44949695, 0.8444449, 0.06281311, 0.23338133...  item 1141  103.575333
+    1  [0.48587373, 0.269207, 0.15095535, 0.65531915,...  item 3953  108.393867
+    ```
 
-=== "Typescript"
+=== "TypeScript"
 
-     ```typescript
-     --8<-- "docs/src/ann_indexes.ts:search1"
-     ```
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/ann_indexes.ts:search1"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/ann_indexes.ts:search1"
+        ```
+
+=== "Rust"
+
+    ```rust
+    --8<-- "rust/lancedb/examples/ivf_pq.rs:search1"
+    ```
+
+    Vector search options are more fully defined in the [crate docs](https://docs.rs/lancedb/latest/lancedb/query/struct.Query.html#method.nearest_to).
 
 The search will return the data requested in addition to the distance of each item.
 
@@ -151,15 +194,23 @@ 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_pandas()
-     ```
+    ```python
+    tbl.search(np.random.random((1536))).where("item != 'item 1141'").to_pandas()
+    ```
 
-=== "Typescript"
+=== "TypeScript"
 
-     ```javascript
-     --8<-- "docs/src/ann_indexes.ts:search2"
-     ```
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/ann_indexes.ts:search2"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```javascript
+        --8<-- "docs/src/ann_indexes.ts:search2"
+        ```
 
 ### Projections (select clause)
 
@@ -167,33 +218,49 @@ You can select the columns returned by the query using a select clause.
 
 === "Python"
 
-     ```python
-     tbl.search(np.random.random((1536))).select(["vector"]).to_pandas()
-     ```
+    ```python
+    tbl.search(np.random.random((1536))).select(["vector"]).to_pandas()
+    ```
 
 
-     ```text
-                                                   vector _distance
-     0  [0.30928212, 0.022668175, 0.1756372, 0.4911822...  93.971092
-     1  [0.2525465, 0.01723831, 0.261568, 0.002007689,...  95.173485
-     ...
-     ```
+    ```text
+                                                vector _distance
+    0  [0.30928212, 0.022668175, 0.1756372, 0.4911822...  93.971092
+    1  [0.2525465, 0.01723831, 0.261568, 0.002007689,...  95.173485
+    ...
+    ```
 
-=== "Typescript"
+=== "TypeScript"
 
-     ```typescript
-     --8<-- "docs/src/ann_indexes.ts:search3"
-     ```
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/ann_indexes.ts:search3"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/ann_indexes.ts:search3"
+        ```
 
 ## FAQ
 
+### Why do I need to manually create an index?
+
+Currently, LanceDB does _not_ automatically create the ANN index.
+LanceDB is well-optimized for kNN (exhaustive search) via a disk-based index. For many use-cases,
+datasets of the order of ~100K vectors don't require index creation. If you can live with up to
+100ms latency, skipping index creation is a simpler workflow while guaranteeing 100% recall.
+
 ### When is it necessary to create an ANN vector index?
 
-`LanceDB` has manually-tuned SIMD code for computing vector distances.
-In our benchmarks, computing 100K pairs of 1K dimension vectors takes **less than 20ms**.
-For small datasets (< 100K rows) or applications that can accept 100ms latency, vector indices are usually not necessary.
+`LanceDB` comes out-of-the-box with highly optimized SIMD code for computing vector similarity.
+In our benchmarks, computing distances for 100K pairs of 1K dimension vectors takes **less than 20ms**.
+We observe that for small datasets (~100K rows) or for 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.
+For large-scale or higher dimension vectors, it can beneficial to create vector index for performance.
 
 ### How big is my index, and how many memory will it take?
 

docs/src/api_reference.md

@@ -0,0 +1,8 @@
+# API Reference
+
+The API reference for the LanceDB client SDKs are available at the following locations:
+
+- [Python](python/python.md)
+- [JavaScript (legacy vectordb package)](javascript/modules.md)
+- [JavaScript (newer @lancedb/lancedb package)](js/globals.md)
+- [Rust](https://docs.rs/lancedb/latest/lancedb/index.html)

docs/src/basic.md

@@ -3,7 +3,7 @@
 !!! info "LanceDB can be run in a number of ways:"
 
     * Embedded within an existing backend (like your Django, Flask, Node.js or FastAPI application)
-    * Connected to directly from a client application like a Jupyter notebook for analytical workloads
+    * Directly from a client application like a Jupyter notebook for analytical workloads
     * Deployed as a remote serverless database
 
 ![](assets/lancedb_embedded_explanation.png)
@@ -16,21 +16,68 @@
       pip install lancedb
       ```
 
-=== "Typescript"
+=== "Typescript[^1]"
+    === "@lancedb/lancedb"
 
-      ```shell
-      npm install vectordb
-      ```
+        ```shell
+        npm install @lancedb/lancedb
+        ```
+        !!! note "Bundling `@lancedb/lancedb` apps with Webpack"
+
+            Since LanceDB contains a prebuilt Node binary, you must configure `next.config.js` to exclude it from webpack. This is required for both using Next.js and deploying a LanceDB app on Vercel.
+
+            ```javascript
+            /** @type {import('next').NextConfig} */
+            module.exports = ({
+            webpack(config) {
+                config.externals.push({ '@lancedb/lancedb': '@lancedb/lancedb' })
+                return config;
+            }
+            })
+            ```
+
+        !!! note "Yarn users"
+
+            Unlike other package managers, Yarn does not automatically resolve peer dependencies. If you are using Yarn, you will need to manually install 'apache-arrow':
+
+            ```shell
+            yarn add apache-arrow
+            ```
+
+    === "vectordb (deprecated)"
+
+        ```shell
+        npm install vectordb
+        ```
+        !!! note "Bundling `vectordb` apps with Webpack"
+
+            Since LanceDB contains a prebuilt Node binary, you must configure `next.config.js` to exclude it from webpack. This is required for both using Next.js and deploying a LanceDB app on Vercel.
+
+            ```javascript
+            /** @type {import('next').NextConfig} */
+            module.exports = ({
+            webpack(config) {
+                config.externals.push({ vectordb: 'vectordb' })
+                return config;
+            }
+            })
+            ```
+
+        !!! note "Yarn users"
+
+            Unlike other package managers, Yarn does not automatically resolve peer dependencies. If you are using Yarn, you will need to manually install 'apache-arrow':
+
+            ```shell
+            yarn add apache-arrow
+            ```
 
 === "Rust"
 
-    !!! warning "Rust SDK is experimental, might introduce breaking changes in the near future"
-
     ```shell
-    cargo add vectordb
+    cargo add lancedb
     ```
 
-    !!! info "To use the vectordb create, you first need to install protobuf."
+    !!! info "To use the lancedb create, you first need to install protobuf."
 
     === "macOS"
 
@@ -44,49 +91,109 @@
         sudo apt install -y protobuf-compiler libssl-dev
         ```
 
-    !!! info "Please also make sure you're using the same version of Arrow as in the [vectordb crate](https://github.com/lancedb/lancedb/blob/main/Cargo.toml)"
+    !!! info "Please also make sure you're using the same version of Arrow as in the [lancedb crate](https://github.com/lancedb/lancedb/blob/main/Cargo.toml)"
 
-## How to connect to a database
+### Preview releases
+
+Stable releases are created about every 2 weeks. For the latest features and bug
+fixes, you can install the preview release. These releases receive the same
+level of testing as stable releases, but are not guaranteed to be available for
+more than 6 months after they are released. Once your application is stable, we
+recommend switching to stable releases.
 
 === "Python"
 
-      ```python
-      import lancedb
-      uri = "data/sample-lancedb"
-      db = lancedb.connect(uri)
+      ```shell
+      pip install --pre --extra-index-url https://pypi.fury.io/lancedb/ lancedb
       ```
 
-=== "Typescript"
+=== "Typescript[^1]"
 
-    ```typescript
-    --8<-- "docs/src/basic_legacy.ts:import"
+    === "@lancedb/lancedb"
 
-    --8<-- "docs/src/basic_legacy.ts:open_db"
+        ```shell
+        npm install @lancedb/lancedb@preview
+        ```
+    === "vectordb (deprecated)"
+
+        ```shell
+        npm install vectordb@preview
+        ```
+
+=== "Rust"
+
+    We don't push preview releases to crates.io, but you can referent the tag
+    in GitHub within your Cargo dependencies:
+
+    ```toml
+    [dependencies]
+    lancedb = { git = "https://github.com/lancedb/lancedb.git", tag = "vX.Y.Z-beta.N" }
     ```
 
+## Connect to a database
+
+=== "Python"
+
+    ```python
+    --8<-- "python/python/tests/docs/test_basic.py:imports"
+    --8<-- "python/python/tests/docs/test_basic.py:connect"
+
+    --8<-- "python/python/tests/docs/test_basic.py:connect_async"
+    ```
+
+    !!! note "Asynchronous Python API"
+
+        The asynchronous Python API is new and has some slight differences compared
+        to the synchronous API.  Feel free to start using the asynchronous version.
+        Once all features have migrated we will start to move the synchronous API to
+        use the same syntax as the asynchronous API.  To help with this migration we
+        have created a [migration guide](migration.md) detailing the differences.
+
+=== "Typescript[^1]"
+
+    === "@lancedb/lancedb"
+
+        ```typescript
+        import * as lancedb from "@lancedb/lancedb";
+        import * as arrow from "apache-arrow";
+
+        --8<-- "nodejs/examples/basic.ts:connect"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:open_db"
+        ```
+
 === "Rust"
 
     ```rust
     #[tokio::main]
     async fn main() -> Result<()> {
-        --8<-- "rust/vectordb/examples/simple.rs:connect"
+        --8<-- "rust/lancedb/examples/simple.rs:connect"
     }
     ```
 
-    !!! info "See [examples/simple.rs](https://github.com/lancedb/lancedb/tree/main/rust/vectordb/examples/simple.rs) for a full working example."
+    !!! info "See [examples/simple.rs](https://github.com/lancedb/lancedb/tree/main/rust/lancedb/examples/simple.rs) for a full working example."
 
 LanceDB will create the directory if it doesn't exist (including parent directories).
 
 If you need a reminder of the uri, you can call `db.uri()`.
 
-## How to create a table
+## Create a table
+
+### Create a table from initial data
+
+If you have data to insert into the table at creation time, you can simultaneously create a
+table and insert the data into it. The schema of the data will be used as the schema of the
+table.
 
 === "Python"
 
     ```python
-    tbl = 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}])
+    --8<-- "python/python/tests/docs/test_basic.py:create_table"
+    --8<-- "python/python/tests/docs/test_basic.py:create_table_async"
     ```
 
     If the table already exists, LanceDB will raise an error by default.
@@ -96,80 +203,115 @@ If you need a reminder of the uri, you can call `db.uri()`.
     You can also pass in a pandas DataFrame directly:
 
     ```python
-    import pandas as pd
-    df = pd.DataFrame([{"vector": [3.1, 4.1], "item": "foo", "price": 10.0},
-                       {"vector": [5.9, 26.5], "item": "bar", "price": 20.0}])
-    tbl = db.create_table("table_from_df", data=df)
+    --8<-- "python/python/tests/docs/test_basic.py:create_table_pandas"
+    --8<-- "python/python/tests/docs/test_basic.py:create_table_async_pandas"
     ```
 
-=== "Typescript"
+=== "Typescript[^1]"
 
-    ```typescript
-    --8<-- "docs/src/basic_legacy.ts:create_table"
-    ```
+    === "@lancedb/lancedb"
 
-    If the table already exists, LanceDB will raise an error by default.
-    If you want to overwrite the table, you can pass in `mode="overwrite"`
-    to the `createTable` function.
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:create_table"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:create_table"
+        ```
+
+        If the table already exists, LanceDB will raise an error by default.
+        If you want to overwrite the table, you can pass in `mode:"overwrite"`
+        to the `createTable` function.
 
 === "Rust"
 
     ```rust
-    use arrow_schema::{DataType, Schema, Field};
-    use arrow_array::{RecordBatch, RecordBatchIterator};
-
-    --8<-- "rust/vectordb/examples/simple.rs:create_table"
+    --8<-- "rust/lancedb/examples/simple.rs:create_table"
     ```
 
-    If the table already exists, LanceDB will raise an error by default.
+    If the table already exists, LanceDB will raise an error by default.  See
+    [the mode option](https://docs.rs/lancedb/latest/lancedb/connection/struct.CreateTableBuilder.html#method.mode)
+    for details on how to overwrite (or open) existing tables instead.
 
-!!! info "Under the hood, LanceDB is converting the input data into an Apache Arrow table and persisting it to disk in [Lance format](https://www.github.com/lancedb/lance)."
+    !!! Providing table records in Rust
 
-### Creating an empty table
+        The Rust SDK currently expects data to be provided as an Arrow
+        [RecordBatchReader](https://docs.rs/arrow-array/latest/arrow_array/trait.RecordBatchReader.html)
+        Support for additional formats (such as serde or polars) is on the roadmap.
+
+!!! info "Under the hood, LanceDB reads in the Apache Arrow data and persists it to disk using the [Lance format](https://www.github.com/lancedb/lance)."
+
+!!! info "Automatic embedding generation with Embedding API"
+    When working with embedding models, it is recommended to use the LanceDB embedding API to automatically create vector representation of the data and queries in the background. See the [quickstart example](#using-the-embedding-api) or the embedding API [guide](./embeddings/)
+
+### Create an empty table
 
 Sometimes you may not have the data to insert into the table at creation time.
-In this case, you can create an empty table and specify the schema.
+In this case, you can create an empty table and specify the schema, so that you can add
+data to the table at a later time (as long as it conforms to the schema). This is
+similar to a `CREATE TABLE` statement in SQL.
 
 === "Python"
 
       ```python
-      import pyarrow as pa
-      schema = pa.schema([pa.field("vector", pa.list_(pa.float32(), list_size=2))])
-      tbl = db.create_table("empty_table", schema=schema)
+      --8<-- "python/python/tests/docs/test_basic.py:create_empty_table"
+      --8<-- "python/python/tests/docs/test_basic.py:create_empty_table_async"
       ```
 
-=== "Typescript"
+    !!! note "You can define schema in Pydantic"
+        LanceDB comes with Pydantic support, which allows you to define the schema of your data using Pydantic models. This makes it easy to work with LanceDB tables and data. Learn more about all supported types in [tables guide](./guides/tables.md).
 
-    ```typescript
-    --8<-- "docs/src/basic_legacy.ts:create_empty_table"
-    ```
+=== "Typescript[^1]"
+
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:create_empty_table"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:create_empty_table"
+        ```
 
 === "Rust"
 
     ```rust
-    --8<-- "rust/vectordb/examples/simple.rs:create_empty_table"
+    --8<-- "rust/lancedb/examples/simple.rs:create_empty_table"
     ```
 
-## How to open an existing table
+## Open an existing table
 
-Once created, you can open a table using the following code:
+Once created, you can open a table as follows:
 
 === "Python"
 
     ```python
-    tbl = db.open_table("my_table")
+    --8<-- "python/python/tests/docs/test_basic.py:open_table"
+    --8<-- "python/python/tests/docs/test_basic.py:open_table_async"
     ```
 
-=== "Typescript"
+=== "Typescript[^1]"
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:open_table"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        const tbl = await db.openTable("myTable");
+        ```
 
-    ```typescript
-    const tbl = await db.openTable("myTable");
-    ```
 
 === "Rust"
 
     ```rust
-    --8<-- "rust/vectordb/examples/simple.rs:open_with_existing_file"
+    --8<-- "rust/lancedb/examples/simple.rs:open_existing_tbl"
     ```
 
 If you forget the name of your table, you can always get a listing of all table names:
@@ -177,101 +319,139 @@ If you forget the name of your table, you can always get a listing of all table
 === "Python"
 
     ```python
-    print(db.table_names())
+    --8<-- "python/python/tests/docs/test_basic.py:table_names"
+    --8<-- "python/python/tests/docs/test_basic.py:table_names_async"
     ```
 
-=== "Javascript"
+=== "Typescript[^1]"
+    === "@lancedb/lancedb"
 
-    ```javascript
-    console.log(await db.tableNames());
-    ```
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:table_names"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        console.log(await db.tableNames());
+        ```
 
 === "Rust"
 
     ```rust
-    --8<-- "rust/vectordb/examples/simple.rs:list_names"
+    --8<-- "rust/lancedb/examples/simple.rs:list_names"
     ```
 
-## How to add data to a table
+## Add data to a table
 
-After a table has been created, you can always add more data to it using
+After a table has been created, you can always add more data to it as follows:
 
 === "Python"
 
     ```python
-
-    # Option 1: Add a list of dicts to a table
-    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)
+    --8<-- "python/python/tests/docs/test_basic.py:add_data"
+    --8<-- "python/python/tests/docs/test_basic.py:add_data_async"
     ```
 
-=== "Typescript"
+=== "Typescript[^1]"
+    === "@lancedb/lancedb"
 
-    ```typescript
-    --8<-- "docs/src/basic_legacy.ts:add"
-    ```
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:add_data"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:add"
+        ```
 
 === "Rust"
 
     ```rust
-    --8<-- "rust/vectordb/examples/simple.rs:add"
+    --8<-- "rust/lancedb/examples/simple.rs:add"
     ```
 
-## How to search for (approximate) nearest neighbors
+## Search for nearest neighbors
 
-Once you've embedded the query, you can find its nearest neighbors using the following code:
+Once you've embedded the query, you can find its nearest neighbors as follows:
 
 === "Python"
 
     ```python
-    tbl.search([100, 100]).limit(2).to_pandas()
+    --8<-- "python/python/tests/docs/test_basic.py:vector_search"
+    --8<-- "python/python/tests/docs/test_basic.py:vector_search_async"
     ```
 
     This returns a pandas DataFrame with the results.
 
-=== "Typescript"
+=== "Typescript[^1]"
+    === "@lancedb/lancedb"
 
-    ```typescript
-    --8<-- "docs/src/basic_legacy.ts:search"
-    ```
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:vector_search"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:search"
+        ```
 
 === "Rust"
 
     ```rust
     use futures::TryStreamExt;
 
-    --8<-- "rust/vectordb/examples/simple.rs:search"
+    --8<-- "rust/lancedb/examples/simple.rs:search"
     ```
 
+    !!! Query vectors in Rust
+        Rust does not yet support automatic execution of embedding functions.  You will need to
+        calculate embeddings yourself.  Support for this is on the roadmap and can be tracked at
+        https://github.com/lancedb/lancedb/issues/994
+
+        Query vectors can be provided as Arrow arrays or a Vec/slice of Rust floats.
+        Support for additional formats (e.g. `polars::series::Series`) is on the roadmap.
+
 By default, LanceDB runs a brute-force scan over dataset to find the K nearest neighbours (KNN).
 For tables with more than 50K vectors, creating an ANN index is recommended to speed up search performance.
+LanceDB allows you to create an ANN index on a table as follows:
 
 === "Python"
 
     ```py
-    tbl.create_index()
+    --8<-- "python/python/tests/docs/test_basic.py:create_index"
+    --8<-- "python/python/tests/docs/test_basic.py:create_index_async"
     ```
 
-=== "Typescript"
+=== "Typescript[^1]"
+    === "@lancedb/lancedb"
 
-    ```{.typescript .ignore}
-    --8<-- "docs/src/basic_legacy.ts:create_index"
-    ```
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:create_index"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```{.typescript .ignore}
+        --8<-- "docs/src/basic_legacy.ts:create_index"
+        ```
 
 === "Rust"
 
     ```rust
-     --8<-- "rust/vectordb/examples/simple.rs:create_index"
+    --8<-- "rust/lancedb/examples/simple.rs:create_index"
     ```
 
-Check [Approximate Nearest Neighbor (ANN) Indexes](/ann_indices.md) section for more details.
+!!! note "Why do I need to create an index manually?"
+LanceDB does not automatically create the ANN index for two reasons. The first is that it's optimized
+for really fast retrievals via a disk-based index, and the second is that data and query workloads can
+be very diverse, so there's no one-size-fits-all index configuration. LanceDB provides many parameters
+to fine-tune index size, query latency and accuracy. See the section on
+[ANN indexes](ann_indexes.md) for more details.
 
-## How to delete rows from a table
+## Delete rows from a table
 
 Use the `delete()` method on tables to delete rows from a table. To choose
 which rows to delete, provide a filter that matches on the metadata columns.
@@ -280,78 +460,125 @@ This can delete any number of rows that match the filter.
 === "Python"
 
     ```python
-    tbl.delete('item = "fizz"')
+    --8<-- "python/python/tests/docs/test_basic.py:delete_rows"
+    --8<-- "python/python/tests/docs/test_basic.py:delete_rows_async"
     ```
 
-=== "Typescript"
+=== "Typescript[^1]"
 
-    ```typescript
-    --8<-- "docs/src/basic_legacy.ts:delete"
-    ```
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:delete_rows"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:delete"
+        ```
 
 === "Rust"
 
     ```rust
-    --8<-- "rust/vectordb/examples/simple.rs:delete"
+    --8<-- "rust/lancedb/examples/simple.rs:delete"
     ```
 
 The deletion predicate is a SQL expression that supports the same expressions
-as the `where()` clause on a search. They can be as simple or complex as needed.
-To see what expressions are supported, see the [SQL filters](sql.md) section.
+as the `where()` clause (`only_if()` in Rust) on a search. They can be as
+simple or complex as needed. To see what expressions are supported, see the
+[SQL filters](sql.md) section.
 
 === "Python"
 
       Read more: [lancedb.table.Table.delete][]
 
-=== "Javascript"
+=== "Typescript[^1]"
 
-      Read more: [vectordb.Table.delete](javascript/interfaces/Table.md#delete)
+    === "@lancedb/lancedb"
 
-## How to remove a table
+        Read more: [lancedb.Table.delete](javascript/interfaces/Table.md#delete)
+
+    === "vectordb (deprecated)"
+
+        Read more: [vectordb.Table.delete](javascript/interfaces/Table.md#delete)
+
+=== "Rust"
+
+      Read more: [lancedb::Table::delete](https://docs.rs/lancedb/latest/lancedb/table/struct.Table.html#method.delete)
+
+## Drop a table
 
 Use the `drop_table()` method on the database to remove a table.
 
 === "Python"
 
-      ```python
-      db.drop_table("my_table")
-      ```
+    ```python
+    --8<-- "python/python/tests/docs/test_basic.py:drop_table"
+    --8<-- "python/python/tests/docs/test_basic.py:drop_table_async"
+    ```
 
-      This permanently removes the table and is not recoverable, unlike deleting rows.
-      By default, if the table does not exist an exception is raised. To suppress this,
-      you can pass in `ignore_missing=True`.
+    This permanently removes the table and is not recoverable, unlike deleting rows.
+    By default, if the table does not exist an exception is raised. To suppress this,
+    you can pass in `ignore_missing=True`.
 
-=== "Typescript"
+=== "Typescript[^1]"
 
-      ```typescript
-      --8<-- "docs/src/basic_legacy.ts:drop_table"
-      ```
+    === "@lancedb/lancedb"
 
-      This permanently removes the table and is not recoverable, unlike deleting rows.
-      If the table does not exist an exception is raised.
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:drop_table"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:drop_table"
+        ```
+
+        This permanently removes the table and is not recoverable, unlike deleting rows.
+        If the table does not exist an exception is raised.
 
 === "Rust"
 
     ```rust
-    --8<-- "rust/vectordb/examples/simple.rs:drop_table"
+    --8<-- "rust/lancedb/examples/simple.rs:drop_table"
     ```
 
-!!! note "Bundling `vectordb` apps with Webpack"
 
-    If you're using the `vectordb` module in JavaScript, since LanceDB contains a prebuilt Node binary, you must configure `next.config.js` to exclude it from webpack. This is required for both using Next.js and deploying a LanceDB app on Vercel.
+## Using the Embedding API
+You can use the embedding API when working with embedding models. It automatically vectorizes the data at ingestion and query time and comes with built-in integrations with popular embedding models like Openai, Hugging Face, Sentence Transformers, CLIP and more.
 
-    ```javascript
-    /** @type {import('next').NextConfig} */
-    module.exports = ({
-    webpack(config) {
-        config.externals.push({ vectordb: 'vectordb' })
-        return config;
-    }
-    })
+=== "Python"
+
+    ```python
+    --8<-- "python/python/tests/docs/test_embeddings_optional.py:imports"
+    --8<-- "python/python/tests/docs/test_embeddings_optional.py:openai_embeddings"
     ```
 
+=== "Typescript[^1]"
+
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/embedding.ts:imports"
+        --8<-- "nodejs/examples/embedding.ts:openai_embeddings"
+        ```
+
+=== "Rust"
+
+    ```rust
+    --8<-- "rust/lancedb/examples/openai.rs:imports"
+    --8<-- "rust/lancedb/examples/openai.rs:openai_embeddings"
+    ```
+
+Learn about using the existing integrations and creating custom embedding functions in the [embedding API guide](./embeddings/).
+
+
 ## What's next
 
 This section covered the very basics of using LanceDB. If you're learning about vector databases for the first time, you may want to read the page on [indexing](concepts/index_ivfpq.md) to get familiar with the concepts.
 
 If you've already worked with other vector databases, you may want to read the [guides](guides/tables.md) to learn how to work with LanceDB in more detail.
+
+[^1]: The `vectordb` package is a legacy package that is  deprecated in favor of `@lancedb/lancedb`.  The `vectordb` package will continue to receive bug fixes and security updates until September 2024.  We recommend all new projects use `@lancedb/lancedb`.  See the [migration guide](migration.md) for more information.

docs/src/basic_legacy.ts

@@ -1,6 +1,14 @@
 // --8<-- [start:import]
 import * as lancedb from "vectordb";
-import { Schema, Field, Float32, FixedSizeList, Int32, Float16 } from "apache-arrow";
+import {
+  Schema,
+  Field,
+  Float32,
+  FixedSizeList,
+  Int32,
+  Float16,
+} from "apache-arrow";
+import * as arrow from "apache-arrow";
 // --8<-- [end:import]
 import * as fs from "fs";
 import { Table as ArrowTable, Utf8 } from "apache-arrow";
@@ -20,9 +28,33 @@ const example = async () => {
       { vector: [3.1, 4.1], item: "foo", price: 10.0 },
       { vector: [5.9, 26.5], item: "bar", price: 20.0 },
     ],
-    { writeMode: lancedb.WriteMode.Overwrite }
+    { writeMode: lancedb.WriteMode.Overwrite },
   );
   // --8<-- [end:create_table]
+  {
+    // --8<-- [start:create_table_with_schema]
+    const schema = new arrow.Schema([
+      new arrow.Field(
+        "vector",
+        new arrow.FixedSizeList(
+          2,
+          new arrow.Field("item", new arrow.Float32(), true),
+        ),
+      ),
+      new arrow.Field("item", new arrow.Utf8(), true),
+      new arrow.Field("price", new arrow.Float32(), true),
+    ]);
+    const data = [
+      { vector: [3.1, 4.1], item: "foo", price: 10.0 },
+      { vector: [5.9, 26.5], item: "bar", price: 20.0 },
+    ];
+    const tbl = await db.createTable({
+      name: "myTableWithSchema",
+      data,
+      schema,
+    });
+    // --8<-- [end:create_table_with_schema]
+  }
 
   // --8<-- [start:add]
   const newData = Array.from({ length: 500 }, (_, i) => ({
@@ -42,33 +74,35 @@ const example = async () => {
   // --8<-- [end:create_index]
 
   // --8<-- [start:create_empty_table]
-  const schema = new Schema([
-    new Field("id", new Int32()),
-    new Field("name", new Utf8()),
+  const schema = new arrow.Schema([
+    new arrow.Field("id", new arrow.Int32()),
+    new arrow.Field("name", new arrow.Utf8()),
   ]);
+
   const empty_tbl = await db.createTable({ name: "empty_table", schema });
   // --8<-- [end:create_empty_table]
-
-  // --8<-- [start:create_f16_table]
-  const dim = 16
-  const total = 10
-  const f16_schema = new Schema([
-      new Field('id', new Int32()),
+  {
+    // --8<-- [start:create_f16_table]
+    const dim = 16;
+    const total = 10;
+    const schema = new Schema([
+      new Field("id", new Int32()),
       new Field(
-        'vector',
-        new FixedSizeList(dim, new Field('item', new Float16(), true)),
-        false
-      )
-    ])
-  const data = lancedb.makeArrowTable(
+        "vector",
+        new FixedSizeList(dim, new Field("item", new Float16(), true)),
+        false,
+      ),
+    ]);
+    const data = lancedb.makeArrowTable(
       Array.from(Array(total), (_, i) => ({
         id: i,
-        vector: Array.from(Array(dim), Math.random)
+        vector: Array.from(Array(dim), Math.random),
       })),
-      { f16_schema }
-    )
-  const table = await db.createTable('f16_tbl', data)
-  // --8<-- [end:create_f16_table]
+      { schema },
+    );
+    const table = await db.createTable("f16_tbl", data);
+    // --8<-- [end:create_f16_table]
+  }
 
   // --8<-- [start:search]
   const query = await tbl.search([100, 100]).limit(2).execute();

docs/src/cli_config.md

@@ -1,51 +0,0 @@
-
-# CLI & Config
-
-## 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/cloud/rest.md

@@ -0,0 +1 @@
+!!swagger ../../openapi.yml!!

docs/src/concepts/index_ivfpq.md

@@ -31,7 +31,7 @@ As an example, consider starting with 128-dimensional vector consisting of 32-bi
 
 While PQ helps with reducing the size of the index, IVF primarily addresses search performance. The primary purpose of an inverted file index is to facilitate rapid and effective nearest neighbor search by narrowing down the search space.
 
-In IVF, the PQ vector space is divided into *Voronoi cells*, which are essentially partitions that consist of all the points in the space that are within a threshold distance of the given region's seed point. These seed points are used to create an inverted index that correlates each centroid with a list of vectors in the space, allowing a search to be restricted to just a subset of vectors in the index.
+In IVF, the PQ vector space is divided into *Voronoi cells*, which are essentially partitions that consist of all the points in the space that are within a threshold distance of the given region's seed point. These seed points are initialized by running K-means over the stored vectors. The centroids of K-means turn into the seed points which then each define a region. These regions are then are used to create an inverted index that correlates each centroid with a list of vectors in the space, allowing a search to be restricted to just a subset of vectors in the index.
 
 ![](../assets/ivfpq_ivf_desc.webp)
 
@@ -81,24 +81,4 @@ The above query will perform a search on the table `tbl` using the given query v
 * `to_pandas()`: Convert the results to a pandas DataFrame
 
 And there you have it! You now understand what an IVF-PQ index is, and how to create and query it in LanceDB.
-
-
-## FAQ
-
-### When is it necessary to create a vector index?
-
-LanceDB has manually-tuned SIMD code for computing vector distances. In our benchmarks, computing 100K pairs of 1K dimension vectors takes **<20ms**. For small datasets (<100K rows) or applications that can accept up to 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 much 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.
-
-For example, with 1024-dimension vectors, 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?
-
-`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` specifies how many PQ short codes to generate on each vector. Because PQ is a lossy compression of the original vector, a higher `num_sub_vectors` usually results in less space distortion, and thus yields better accuracy. However, a higher `num_sub_vectors` also causes heavier I/O and more PQ computation, and thus, higher latency. `dimension / num_sub_vectors` should be a multiple of 8 for optimum SIMD efficiency.
+To see how to create an IVF-PQ index in LanceDB, take a look at the [ANN indexes](../ann_indexes.md) section.

docs/src/embeddings/custom_embedding_function.md

@@ -17,6 +17,7 @@ Let's implement `SentenceTransformerEmbeddings` class. All you need to do is imp
 
 ```python
 from lancedb.embeddings import register
+from lancedb.util import attempt_import_or_raise
 
 @register("sentence-transformers")
 class SentenceTransformerEmbeddings(TextEmbeddingFunction):
@@ -81,7 +82,7 @@ class OpenClipEmbeddings(EmbeddingFunction):
 
     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
+        open_clip = attempt_import_or_raise("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
         )
@@ -109,14 +110,14 @@ class OpenClipEmbeddings(EmbeddingFunction):
         if isinstance(query, str):
             return [self.generate_text_embeddings(query)]
         else:
-            PIL = self.safe_import("PIL", "pillow")
+            PIL = attempt_import_or_raise("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")
+        torch = attempt_import_or_raise("torch")
         text = self.sanitize_input(text)
         text = self._tokenizer(text)
         text.to(self.device)
@@ -175,7 +176,7 @@ class OpenClipEmbeddings(EmbeddingFunction):
             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")
+        torch = attempt_import_or_raise("torch")
         # TODO handle retry and errors for https
         image = self._to_pil(image)
         image = self._preprocess(image).unsqueeze(0)
@@ -183,7 +184,7 @@ class OpenClipEmbeddings(EmbeddingFunction):
             return self._encode_and_normalize_image(image)
 
     def _to_pil(self, image: Union[str, bytes]):
-        PIL = self.safe_import("PIL", "pillow")
+        PIL = attempt_import_or_raise("PIL", "pillow")
         if isinstance(image, bytes):
             return PIL.Image.open(io.BytesIO(image))
         if isinstance(image, PIL.Image.Image):

docs/src/embeddings/default_embedding_functions.md

@@ -9,56 +9,267 @@ Contains the text embedding functions registered by default.
 ### Sentence transformers
 Allows you to set parameters when registering a `sentence-transformers` object.
 
+!!! info
+    Sentence transformer embeddings are normalized by default. It is recommended to use normalized embeddings for similarity search.
+
 | 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 |
+| `trust_remote_code` | `bool` | `False` | Whether to trust and execute remote code from the model's Huggingface repository |
 
 
+??? "Check out available sentence-transformer models here!"
+    ```markdown
+    - sentence-transformers/all-MiniLM-L12-v2
+    - sentence-transformers/paraphrase-mpnet-base-v2
+    - sentence-transformers/gtr-t5-base
+    - sentence-transformers/LaBSE
+    - sentence-transformers/all-MiniLM-L6-v2
+    - sentence-transformers/bert-base-nli-max-tokens
+    - sentence-transformers/bert-base-nli-mean-tokens
+    - sentence-transformers/bert-base-nli-stsb-mean-tokens
+    - sentence-transformers/bert-base-wikipedia-sections-mean-tokens
+    - sentence-transformers/bert-large-nli-cls-token
+    - sentence-transformers/bert-large-nli-max-tokens
+    - sentence-transformers/bert-large-nli-mean-tokens
+    - sentence-transformers/bert-large-nli-stsb-mean-tokens
+    - sentence-transformers/distilbert-base-nli-max-tokens
+    - sentence-transformers/distilbert-base-nli-mean-tokens
+    - sentence-transformers/distilbert-base-nli-stsb-mean-tokens
+    - sentence-transformers/distilroberta-base-msmarco-v1
+    - sentence-transformers/distilroberta-base-msmarco-v2
+    - sentence-transformers/nli-bert-base-cls-pooling
+    - sentence-transformers/nli-bert-base-max-pooling
+    - sentence-transformers/nli-bert-base
+    - sentence-transformers/nli-bert-large-cls-pooling
+    - sentence-transformers/nli-bert-large-max-pooling
+    - sentence-transformers/nli-bert-large
+    - sentence-transformers/nli-distilbert-base-max-pooling
+    - sentence-transformers/nli-distilbert-base
+    - sentence-transformers/nli-roberta-base
+    - sentence-transformers/nli-roberta-large
+    - sentence-transformers/roberta-base-nli-mean-tokens
+    - sentence-transformers/roberta-base-nli-stsb-mean-tokens
+    - sentence-transformers/roberta-large-nli-mean-tokens
+    - sentence-transformers/roberta-large-nli-stsb-mean-tokens
+    - sentence-transformers/stsb-bert-base
+    - sentence-transformers/stsb-bert-large
+    - sentence-transformers/stsb-distilbert-base
+    - sentence-transformers/stsb-roberta-base
+    - sentence-transformers/stsb-roberta-large
+    - sentence-transformers/xlm-r-100langs-bert-base-nli-mean-tokens
+    - sentence-transformers/xlm-r-100langs-bert-base-nli-stsb-mean-tokens
+    - sentence-transformers/xlm-r-base-en-ko-nli-ststb
+    - sentence-transformers/xlm-r-bert-base-nli-mean-tokens
+    - sentence-transformers/xlm-r-bert-base-nli-stsb-mean-tokens
+    - sentence-transformers/xlm-r-large-en-ko-nli-ststb
+    - sentence-transformers/bert-base-nli-cls-token
+    - sentence-transformers/all-distilroberta-v1
+    - sentence-transformers/multi-qa-MiniLM-L6-dot-v1
+    - sentence-transformers/multi-qa-distilbert-cos-v1
+    - sentence-transformers/multi-qa-distilbert-dot-v1
+    - sentence-transformers/multi-qa-mpnet-base-cos-v1
+    - sentence-transformers/multi-qa-mpnet-base-dot-v1
+    - sentence-transformers/nli-distilroberta-base-v2
+    - sentence-transformers/all-MiniLM-L6-v1
+    - sentence-transformers/all-mpnet-base-v1
+    - sentence-transformers/all-mpnet-base-v2
+    - sentence-transformers/all-roberta-large-v1
+    - sentence-transformers/allenai-specter
+    - sentence-transformers/average_word_embeddings_glove.6B.300d
+    - sentence-transformers/average_word_embeddings_glove.840B.300d
+    - sentence-transformers/average_word_embeddings_komninos
+    - sentence-transformers/average_word_embeddings_levy_dependency
+    - sentence-transformers/clip-ViT-B-32-multilingual-v1
+    - sentence-transformers/clip-ViT-B-32
+    - sentence-transformers/distilbert-base-nli-stsb-quora-ranking
+    - sentence-transformers/distilbert-multilingual-nli-stsb-quora-ranking
+    - sentence-transformers/distilroberta-base-paraphrase-v1
+    - sentence-transformers/distiluse-base-multilingual-cased-v1
+    - sentence-transformers/distiluse-base-multilingual-cased-v2
+    - sentence-transformers/distiluse-base-multilingual-cased
+    - sentence-transformers/facebook-dpr-ctx_encoder-multiset-base
+    - sentence-transformers/facebook-dpr-ctx_encoder-single-nq-base
+    - sentence-transformers/facebook-dpr-question_encoder-multiset-base
+    - sentence-transformers/facebook-dpr-question_encoder-single-nq-base
+    - sentence-transformers/gtr-t5-large
+    - sentence-transformers/gtr-t5-xl
+    - sentence-transformers/gtr-t5-xxl
+    - sentence-transformers/msmarco-MiniLM-L-12-v3
+    - sentence-transformers/msmarco-MiniLM-L-6-v3
+    - sentence-transformers/msmarco-MiniLM-L12-cos-v5
+    - sentence-transformers/msmarco-MiniLM-L6-cos-v5
+    - sentence-transformers/msmarco-bert-base-dot-v5
+    - sentence-transformers/msmarco-bert-co-condensor
+    - sentence-transformers/msmarco-distilbert-base-dot-prod-v3
+    - sentence-transformers/msmarco-distilbert-base-tas-b
+    - sentence-transformers/msmarco-distilbert-base-v2
+    - sentence-transformers/msmarco-distilbert-base-v3
+    - sentence-transformers/msmarco-distilbert-base-v4
+    - sentence-transformers/msmarco-distilbert-cos-v5
+    - sentence-transformers/msmarco-distilbert-dot-v5
+    - sentence-transformers/msmarco-distilbert-multilingual-en-de-v2-tmp-lng-aligned
+    - sentence-transformers/msmarco-distilbert-multilingual-en-de-v2-tmp-trained-scratch
+    - sentence-transformers/msmarco-distilroberta-base-v2
+    - sentence-transformers/msmarco-roberta-base-ance-firstp
+    - sentence-transformers/msmarco-roberta-base-v2
+    - sentence-transformers/msmarco-roberta-base-v3
+    - sentence-transformers/multi-qa-MiniLM-L6-cos-v1
+    - sentence-transformers/nli-mpnet-base-v2
+    - sentence-transformers/nli-roberta-base-v2
+    - sentence-transformers/nq-distilbert-base-v1
+    - sentence-transformers/paraphrase-MiniLM-L12-v2
+    - sentence-transformers/paraphrase-MiniLM-L3-v2
+    - sentence-transformers/paraphrase-MiniLM-L6-v2
+    - sentence-transformers/paraphrase-TinyBERT-L6-v2
+    - sentence-transformers/paraphrase-albert-base-v2
+    - sentence-transformers/paraphrase-albert-small-v2
+    - sentence-transformers/paraphrase-distilroberta-base-v1
+    - sentence-transformers/paraphrase-distilroberta-base-v2
+    - sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
+    - sentence-transformers/paraphrase-multilingual-mpnet-base-v2
+    - sentence-transformers/paraphrase-xlm-r-multilingual-v1
+    - sentence-transformers/quora-distilbert-base
+    - sentence-transformers/quora-distilbert-multilingual
+    - sentence-transformers/sentence-t5-base
+    - sentence-transformers/sentence-t5-large
+    - sentence-transformers/sentence-t5-xxl
+    - sentence-transformers/sentence-t5-xl
+    - sentence-transformers/stsb-distilroberta-base-v2
+    - sentence-transformers/stsb-mpnet-base-v2
+    - sentence-transformers/stsb-roberta-base-v2
+    - sentence-transformers/stsb-xlm-r-multilingual
+    - sentence-transformers/xlm-r-distilroberta-base-paraphrase-v1
+    - sentence-transformers/clip-ViT-L-14
+    - sentence-transformers/clip-ViT-B-16
+    - sentence-transformers/use-cmlm-multilingual
+    - sentence-transformers/all-MiniLM-L12-v1
+    ```
+
+!!! info
+    You can also load many other model architectures from the library. For example models from sources such as BAAI, nomic, salesforce research, etc.
+    See this HF hub page for all [supported models](https://huggingface.co/models?library=sentence-transformers).
+
+!!! note "BAAI Embeddings example"
+    Here is an example that uses BAAI embedding model from the HuggingFace Hub [supported models](https://huggingface.co/models?library=sentence-transformers)
+    ```python
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import get_registry
+
+    db = lancedb.connect("/tmp/db")
+    model = get_registry().get("sentence-transformers").create(name="BAAI/bge-small-en-v1.5", device="cpu")
+
+    class Words(LanceModel):
+        text: str = model.SourceField()
+        vector: Vector(model.ndims()) = model.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)
+    ```
+Visit sentence-transformers [HuggingFace HUB](https://huggingface.co/sentence-transformers) page for more information on the available models.
+
+
+### Huggingface embedding models
+We offer support for all huggingface models (which can be loaded via [transformers](https://huggingface.co/docs/transformers/en/index) library). The default model is `colbert-ir/colbertv2.0` which also has its own special callout - `registry.get("colbert")` 
+
+Example usage - 
+```python
+import lancedb
+import pandas as pd
+
+from lancedb.embeddings import get_registry
+from lancedb.pydantic import LanceModel, Vector
+
+model = get_registry().get("huggingface").create(name='facebook/bart-base')
+
+class Words(LanceModel):
+    text: str = model.SourceField()
+    vector: Vector(model.ndims()) = model.VectorField()
+
+df = pd.DataFrame({"text": ["hi hello sayonara", "goodbye world"]})
+table = db.create_table("greets", schema=Words)
+table.add(df)
+query = "old greeting"
+actual = table.search(query).limit(1).to_pydantic(Words)[0]
+print(actual.text)
+```
+
+
+### Ollama embeddings
+Generate embeddings via the [ollama](https://github.com/ollama/ollama-python) python library. More details:
+
+- [Ollama docs on embeddings](https://github.com/ollama/ollama/blob/main/docs/api.md#generate-embeddings)
+- [Ollama blog on embeddings](https://ollama.com/blog/embedding-models)
+
+| Parameter              | Type                       | Default Value            | Description                                                                                                                                    |
+|------------------------|----------------------------|--------------------------|------------------------------------------------------------------------------------------------------------------------------------------------|
+| `name`                 | `str`                      | `nomic-embed-text`       | The name of the model.                                                                                                                         |
+| `host`                 | `str`                      | `http://localhost:11434` | The Ollama host to connect to.                                                                                                                 |
+| `options`              | `ollama.Options` or `dict` | `None`                   | Additional model parameters listed in the documentation for the Modelfile such as `temperature`. |
+| `keep_alive`           | `float` or `str`           | `"5m"`                   | Controls how long the model will stay loaded into memory following the request.                                                                |
+| `ollama_client_kwargs` | `dict`                     | `{}`                     | kwargs that can be past to the `ollama.Client`.                                                                                                |
+
 ```python
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
+
 db = lancedb.connect("/tmp/db")
-registry = EmbeddingFunctionRegistry.get_instance()
-func = registry.get("sentence-transformers").create(device="cpu")
+func = get_registry().get("ollama").create(name="nomic-embed-text")
 
 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"}
-    ]
-)
+table = db.create_table("words", schema=Words, mode="overwrite")
+table.add([
+    {"text": "hello world"},
+    {"text": "goodbye world"}
+])
 
 query = "greetings"
 actual = table.search(query).limit(1).to_pydantic(Words)[0]
 print(actual.text)
 ```
 
+
 ### OpenAI embeddings
 LanceDB registers the OpenAI embeddings function in the registry by default, as `openai`. Below 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. |
+| `dim` | `int` |  Model default   | For OpenAI's newer text-embedding-3 model, we can specify a dimensionality that is smaller than the 1536 size. This feature supports it |
 
 
 ```python
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
+
 db = lancedb.connect("/tmp/db")
-registry = EmbeddingFunctionRegistry.get_instance()
-func = registry.get("openai").create()
+func = get_registry().get("openai").create(name="text-embedding-ada-002")
 
 class Words(LanceModel):
     text: str = func.SourceField()
     vector: Vector(func.ndims()) = func.VectorField()
 
-table = db.create_table("words", schema=Words)
+table = db.create_table("words", schema=Words, mode="overwrite")
 table.add(
     [
-        {"text": "hello world"}
+        {"text": "hello world"},
         {"text": "goodbye world"}
     ]
     )
@@ -155,6 +366,108 @@ tbl.add(df)
 rs = tbl.search("hello").limit(1).to_pandas()
 ```
 
+### Cohere Embeddings
+Using cohere API requires cohere package, which can be installed using `pip install cohere`. Cohere embeddings are used to generate embeddings for text data. The embeddings can be used for various tasks like semantic search, clustering, and classification.
+You also need to set the `COHERE_API_KEY` environment variable to use the Cohere API.
+
+Supported models are:
+```
+    * embed-english-v3.0
+    * embed-multilingual-v3.0
+    * embed-english-light-v3.0
+    * embed-multilingual-light-v3.0
+    * embed-english-v2.0
+    * embed-english-light-v2.0
+    * embed-multilingual-v2.0
+```
+
+Supported parameters (to be passed in `create` method) are:
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"embed-english-v2.0"` | The model ID of the cohere model to use. Supported base models for Text Embeddings: embed-english-v3.0, embed-multilingual-v3.0, embed-english-light-v3.0, embed-multilingual-light-v3.0, embed-english-v2.0, embed-english-light-v2.0, embed-multilingual-v2.0 |
+| `source_input_type` | `str` | `"search_document"` | The type of input data to be used for the source column. |
+| `query_input_type` | `str` | `"search_query"` | The type of input data to be used for the query. |
+
+Cohere supports following input types:
+
+| Input Type               | Description                          |
+|-------------------------|---------------------------------------|
+| "`search_document`"     | Used for embeddings stored in a vector|
+|                         | database for search use-cases.        |
+| "`search_query`"        | Used for embeddings of search queries |
+|                         | run against a vector DB               |
+| "`semantic_similarity`" | Specifies the given text will be used |
+|                         | for Semantic Textual Similarity (STS) |
+| "`classification`"      | Used for embeddings passed through a  |
+|                         | text classifier.                      |
+| "`clustering`"          | Used for the embeddings run through a |
+|                         | clustering algorithm                  |
+
+Usage Example:
+    
+    ```python
+        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)
+    ```
+
+### Jina Embeddings
+Jina embeddings are used to generate embeddings for text and image data.
+You also need to set the `JINA_API_KEY` environment variable to use the Jina API.
+
+You can find a list of supported models under [https://jina.ai/embeddings/](https://jina.ai/embeddings/)
+
+Supported parameters (to be passed in `create` method) are:
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"jina-clip-v1"` | The model ID of the jina model to use |
+
+Usage Example:
+
+```python
+    import os
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import EmbeddingFunctionRegistry
+
+    os.environ['JINA_API_KEY'] = 'jina_*'
+
+    jina_embed = EmbeddingFunctionRegistry.get_instance().get("jina").create(name="jina-embeddings-v2-base-en")
+
+
+    class TextModel(LanceModel):
+        text: str = jina_embed.SourceField()
+        vector: Vector(jina_embed.ndims()) = jina_embed.VectorField()
+
+
+    data = [{"text": "hello world"},
+            {"text": "goodbye world"}]
+
+    db = lancedb.connect("~/.lancedb-2")
+    tbl = db.create_table("test", schema=TextModel, mode="overwrite")
+
+    tbl.add(data)
+```
+
 ### AWS Bedrock Text Embedding Functions
 AWS Bedrock supports multiple base models for generating text embeddings. You need to setup the AWS credentials to use this embedding function.
 You can do so by using `awscli` and also add your session_token:
@@ -172,7 +485,8 @@ Supported Embedding modelIDs are:
 * `cohere.embed-english-v3`
 * `cohere.embed-multilingual-v3`
 
-Supported paramters (to be passed in `create` method) are:
+Supported parameters (to be passed in `create` method) are:
+
 | Parameter | Type | Default Value | Description |
 |---|---|---|---|
 | **name** | str | "amazon.titan-embed-text-v1" | The model ID of the bedrock model to use. Supported base models for Text Embeddings: amazon.titan-embed-text-v1, cohere.embed-english-v3, cohere.embed-multilingual-v3 |
@@ -186,6 +500,10 @@ Supported paramters (to be passed in `create` method) are:
 Usage Example:
 
 ```python
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
+
 model = get_registry().get("bedrock-text").create()
 
 class TextModel(LanceModel):
@@ -219,12 +537,13 @@ This embedding function supports ingesting images as both bytes and urls. You ca
 !!! info
     LanceDB supports ingesting images directly from accessible links.
 
-
 ```python
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
 
 db = lancedb.connect(tmp_path)
-registry = EmbeddingFunctionRegistry.get_instance()
-func = registry.get("open-clip").create()
+func = get_registry.get("open-clip").create()
 
 class Images(LanceModel):
     label: str
@@ -246,7 +565,7 @@ uris = [
 # 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}]
+    pd.DataFrame({"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
@@ -285,4 +604,121 @@ print(actual.label)
 
 ```
 
+### Imagebind embeddings
+We have support for [imagebind](https://github.com/facebookresearch/ImageBind) model embeddings. You can download our version of the packaged model via - `pip install imagebind-packaged==0.1.2`.
+
+This function is registered as `imagebind` and supports Audio, Video and Text modalities(extending to Thermal,Depth,IMU data):
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"imagebind_huge"` | Name of the model. |
+| `device` | `str` | `"cpu"` | The device to run the model on. Can be `"cpu"` or `"gpu"`. |
+| `normalize` | `bool` | `False` | set to `True` to normalize your inputs before model ingestion. |
+
+Below is an example demonstrating how the API works:
+
+```python
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
+
+db = lancedb.connect(tmp_path)
+func = get_registry.get("imagebind").create()
+
+class ImageBindModel(LanceModel):
+    text: str
+    image_uri: str = func.SourceField()
+    audio_path: str
+    vector: Vector(func.ndims()) = func.VectorField()
+
+# add locally accessible image paths
+text_list=["A dog.", "A car", "A bird"]
+image_paths=[".assets/dog_image.jpg", ".assets/car_image.jpg", ".assets/bird_image.jpg"]
+audio_paths=[".assets/dog_audio.wav", ".assets/car_audio.wav", ".assets/bird_audio.wav"]
+
+# Load data
+inputs = [
+    {"text": a, "audio_path": b, "image_uri": c}
+    for a, b, c in zip(text_list, audio_paths, image_paths)
+]
+
+#create table and add data
+table = db.create_table("img_bind", schema=ImageBindModel)
+table.add(inputs)
+```
+
+Now, we can search using any modality:
+
+#### image search
+```python
+query_image = "./assets/dog_image2.jpg" #download an image and enter that path here
+actual = table.search(query_image).limit(1).to_pydantic(ImageBindModel)[0]
+print(actual.text == "dog")
+```
+#### audio search
+
+```python
+query_audio = "./assets/car_audio2.wav" #download an audio clip and enter path here
+actual = table.search(query_audio).limit(1).to_pydantic(ImageBindModel)[0]
+print(actual.text == "car")
+```
+#### Text search
+You can add any input query and fetch the result as follows:
+```python
+query = "an animal which flies and tweets" 
+actual = table.search(query).limit(1).to_pydantic(ImageBindModel)[0]
+print(actual.text == "bird")
+```
+
 If you have any questions about the embeddings API, supported models, or see a relevant model missing, please raise an issue [on GitHub](https://github.com/lancedb/lancedb/issues).
+
+### Jina Embeddings
+Jina embeddings can also be used to embed both text and image data, only some of the models support image data and you can check the list
+under [https://jina.ai/embeddings/](https://jina.ai/embeddings/)
+
+Supported parameters (to be passed in `create` method) are:
+
+| Parameter | Type | Default Value | Description |
+|---|---|---|---|
+| `name` | `str` | `"jina-clip-v1"` | The model ID of the jina model to use |
+
+Usage Example:
+
+```python
+    import os
+    import requests
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import get_registry
+    import pandas as pd
+
+    os.environ['JINA_API_KEY'] = 'jina_*'
+
+    db = lancedb.connect("~/.lancedb")
+    func = get_registry().get("jina").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(
+      pd.DataFrame({"label": labels, "image_uri": uris, "image_bytes": image_bytes})
+    )
+```

docs/src/embeddings/embedding_explicit.md

@@ -1,141 +0,0 @@
-In this workflow, you define your own embedding function and pass it as a callable to LanceDB, invoking it in your code to generate the embeddings. Let's look at some examples.
-
-### Hugging Face
-
-!!! note
-    Currently, the Hugging Face method is only supported in the Python SDK.
-
-=== "Python"
-    The most popular open source option is to use the [sentence-transformers](https://www.sbert.net/) 
-    library, which can be installed via pip.
-
-    ```bash
-    pip install sentence-transformers
-    ```
-
-    The example below shows how to use the `paraphrase-albert-small-v2` model to generate embeddings 
-    for a given document.
-
-    ```python
-    from sentence_transformers import SentenceTransformer
-
-    name="paraphrase-albert-small-v2"
-    model = SentenceTransformer(name)
-
-    # used for both training and querying
-    def embed_func(batch):
-        return [model.encode(sentence) for sentence in batch]
-    ```
-
-### OpenAI
-
-Another popular alternative is to use an external API like OpenAI's [embeddings API](https://platform.openai.com/docs/guides/embeddings/what-are-embeddings).
-
-=== "Python"
-      ```python
-        import openai
-        import os
-
-        # Configuring the environment variable OPENAI_API_KEY
-        if "OPENAI_API_KEY" not in os.environ:
-        # OR set the key here as a variable
-        openai.api_key = "sk-..."
-
-        # verify that the API key is working
-        assert len(openai.Model.list()["data"]) > 0
-
-        def embed_func(c):
-            rs = openai.Embedding.create(input=c, engine="text-embedding-ada-002")
-            return [record["embedding"] for record in rs["data"]]
-      ```
-
-=== "JavaScript"
-      ```javascript
-        const lancedb = require("vectordb");
-
-        // You need to provide an OpenAI API key
-        const apiKey = "sk-..."
-        // The embedding function will create embeddings for the 'text' column
-        const embedding = new lancedb.OpenAIEmbeddingFunction('text', apiKey)
-      ```
-
-## Applying an embedding function to data
-
-=== "Python"
-    Using an embedding function, you can apply it to raw data
-    to generate embeddings for each record.
-
-    Say you have a pandas DataFrame with a `text` column that you want embedded,
-    you can use the `with_embeddings` function to generate embeddings and add them to 
-    an existing table.
-
-    ```python
-     import pandas as pd
-     from lancedb.embeddings import with_embeddings
-
-     df = pd.DataFrame(
-        [
-            {"text": "pepperoni"},
-            {"text": "pineapple"}
-        ]
-    )
-     data = with_embeddings(embed_func, df)
-
-     # The output is used to create / append to a table
-     # db.create_table("my_table", data=data)
-    ```
-
-    If your data is in a different column, you can specify the `column` kwarg to `with_embeddings`.
-
-    By default, LanceDB calls the function with batches of 1000 rows. This can be configured
-    using the `batch_size` parameter to `with_embeddings`.
-
-    LanceDB automatically wraps the function with retry and rate-limit logic to ensure the OpenAI
-    API call is reliable.
-
-=== "JavaScript"
-    Using an embedding function, you can apply it to raw data
-    to generate embeddings for each record.
-
-    Simply pass the embedding function created above and LanceDB will use it to generate
-    embeddings for your data.
-
-    ```javascript
-    const db = await lancedb.connect("data/sample-lancedb");
-    const data = [
-    { text: "pepperoni"},
-    { text: "pineapple"}
-    ]
-
-    const table = await db.createTable("vectors", data, embedding)
-    ```
-
-## Querying using an embedding function
-
-!!! warning
-    At query time, you **must** use the same embedding function you used to vectorize your data.
-    If you use a different embedding function, the embeddings will not reside in the same vector
-    space and the results will be nonsensical.
-
-=== "Python"
-     ```python
-     query = "What's the best pizza topping?"
-     query_vector = embed_func([query])[0]
-     results = (
-        tbl.search(query_vector)
-        .limit(10)
-        .to_pandas()
-     )
-     ```
-
-     The above snippet returns a pandas DataFrame with the 10 closest vectors to the query.
-
-=== "JavaScript"
-     ```javascript
-      const results = await table
-        .search("What's the best pizza topping?")
-        .limit(10)
-        .execute()
-     ```
-
-     The above snippet returns an array of records with the top 10 nearest neighbors to the query.

docs/src/embeddings/embedding_functions.md

@@ -2,62 +2,164 @@ Representing multi-modal data as vector embeddings is becoming a standard practi
 
 For this purpose, LanceDB introduces an **embedding functions API**, that allow you simply set up once, during the configuration stage of your project. After this, the table remembers it, effectively making the embedding functions *disappear in the background* so you don't have to worry about manually passing callables, and instead, simply focus on the rest of your data engineering pipeline.
 
+!!! Note "LanceDB cloud doesn't support embedding functions yet"
+    LanceDB Cloud does not support embedding functions yet. You need to generate embeddings before ingesting into the table or querying.
+
 !!! warning
-    Using the implicit embeddings management approach means that you can forget about the manually passing around embedding
-    functions in your code, as long as you don't intend to change it at a later time. If your embedding function changes,
-    you'll have to re-configure your table with the new embedding function and regenerate the embeddings.
+    Using the embedding function registry means that you don't have to explicitly generate the embeddings yourself.
+    However, if your embedding function changes, you'll have to re-configure your table with the new embedding function
+    and regenerate the embeddings. In the future, we plan to support the ability to change the embedding function via
+    table metadata and have LanceDB automatically take care of regenerating the embeddings.
+
 
 ## 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!
+=== "Python"
+    In the LanceDB python SDK, we define a global embedding function registry with
+    many different embedding models and even more coming soon.
+    Here's let's an implementation of CLIP as example.
+
+    ```python
+    from lancedb.embeddings import get_registry
+
+    registry = get_registry()
+    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!
+
+=== "TypeScript"
+    In the TypeScript SDK, the choices are more limited. For now, only the OpenAI
+    embedding function is available.
+
+    ```javascript
+    import * as lancedb from '@lancedb/lancedb'
+    import { getRegistry } from '@lancedb/lancedb/embeddings'
+
+    // You need to provide an OpenAI API key
+    const apiKey = "sk-..."
+    // The embedding function will create embeddings for the 'text' column
+    const func = getRegistry().get("openai").create({apiKey})
+    ```
+=== "Rust"
+    In the Rust SDK, the choices are more limited. For now, only the OpenAI
+    embedding function is available. But unlike the Python and TypeScript SDKs, you need manually register the OpenAI embedding function.
+
+    ```toml
+    // Make sure to include the `openai` feature
+    [dependencies]
+    lancedb = {version = "*", features = ["openai"]}
+    ```
+
+    ```rust
+    --8<-- "rust/lancedb/examples/openai.rs:imports"
+    --8<-- "rust/lancedb/examples/openai.rs:openai_embeddings"
+    ```
 
 ## 2. Define the data model or schema
-The embedding function defined above abstracts away all the details about the models and dimensions required to define the schema. You can simply set a field as **source** or **vector** column. Here's how:
 
-```python
-class Pets(LanceModel):
-    vector: Vector(clip.ndims) = clip.VectorField()
-    image_uri: str = clip.SourceField()
-```
+=== "Python"
+    The embedding function defined above abstracts away all the details about the models and dimensions required to define the schema. You can simply set a field as **source** or **vector** column. Here's how:
 
-`VectorField` tells LanceDB to use the clip embedding function to generate query embeddings for the `vector` column and `SourceField` ensures that when adding data, we automatically use the specified embedding function to encode `image_uri`.
+    ```python
+    class Pets(LanceModel):
+        vector: Vector(clip.ndims()) = clip.VectorField()
+        image_uri: str = clip.SourceField()
+    ```
 
-## 3. Create LanceDB table
-Now that we have chosen/defined our embedding function and the schema, we can create the table:
+    `VectorField` tells LanceDB to use the clip embedding function to generate query embeddings for the `vector` column and `SourceField` ensures that when adding data, we automatically use the specified embedding function to encode `image_uri`.
 
-```python
-db = lancedb.connect("~/lancedb")
-table = db.create_table("pets", schema=Pets)
+=== "TypeScript"
 
-```
+    For the TypeScript SDK, a schema can be inferred from input data, or an explicit
+    Arrow schema can be provided.
 
-That's it! We've provided all the information needed to embed the source and query inputs. We can now forget about the model and dimension details and start to build our VectorDB pipeline.
+## 3. Create table and add data
 
-## 4. Ingest lots of data and query your table
-Any new or incoming data can just be added and it'll be vectorized automatically.
+Now that we have chosen/defined our embedding function and the schema,
+we can create the table and ingest data without needing to explicitly generate
+the embeddings at all:
 
-```python
-table.add([{"image_uri": u} for u in uris])
-```
+=== "Python"
+    ```python
+    db = lancedb.connect("~/lancedb")
+    table = db.create_table("pets", schema=Pets)
 
-Our OpenCLIP query embedding function supports querying via both text and images:
+    table.add([{"image_uri": u} for u in uris])
+    ```
 
-```python
-result = table.search("dog")
-```
+=== "TypeScript"
 
-Let's query an image:
+    === "@lancedb/lancedb"
 
-```python
-p = Path("path/to/images/samoyed_100.jpg")
-query_image = Image.open(p)
-table.search(query_image)
-```
+        ```ts
+        --8<-- "nodejs/examples/embedding.ts:imports"
+        --8<-- "nodejs/examples/embedding.ts:embedding_function"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const db = await lancedb.connect("data/sample-lancedb");
+        const data = [
+            { text: "pepperoni"},
+            { text: "pineapple"}
+        ]
+
+        const table = await db.createTable("vectors", data, embedding)
+        ```
+
+## 4. Querying your table
+Not only can you forget about the embeddings during ingestion, you also don't
+need to worry about it when you query the table:
+
+=== "Python"
+
+    Our OpenCLIP query embedding function supports querying via both text and images:
+
+    ```python
+    results = (
+        table.search("dog")
+            .limit(10)
+            .to_pandas()
+    )
+    ```
+
+    Or we can search using an image:
+
+    ```python
+    p = Path("path/to/images/samoyed_100.jpg")
+    query_image = Image.open(p)
+    results = (
+        table.search(query_image)
+            .limit(10)
+            .to_pandas()
+    )
+    ```
+
+    Both of the above snippet returns a pandas DataFrame with the 10 closest vectors to the query.
+
+=== "TypeScript"
+
+    === "@lancedb/lancedb"
+
+        ```ts
+        const results = await table.search("What's the best pizza topping?")
+            .limit(10)
+            .toArray()
+        ```
+
+    === "vectordb (deprecated)
+
+        ```ts
+        const results = await table
+            .search("What's the best pizza topping?")
+            .limit(10)
+            .execute()
+        ```
+
+    The above snippet returns an array of records with the top 10 nearest neighbors to the query.
 
 ---
 
@@ -84,7 +186,7 @@ You can also use the integration for adding utility operations in the schema. Fo
 
 ```python
 class Pets(LanceModel):
-    vector: Vector(clip.ndims) = clip.VectorField()
+    vector: Vector(clip.ndims()) = clip.VectorField()
     image_uri: str = clip.SourceField()
 
     @property
@@ -100,4 +202,5 @@ rs[2].image
 
 ![](../assets/dog_clip_output.png)
 
-Now that you have the basic idea about implicit management via embedding functions, let's dive deeper into a [custom API](./api.md) that you can use to implement your own embedding functions.
+Now that you have the basic idea about LanceDB embedding functions and the embedding function registry,
+let's dive deeper into defining your own [custom functions](./custom_embedding_function.md).

docs/src/embeddings/index.md

@@ -1,8 +1,134 @@
-Due to the nature of vector embeddings, they can be used to represent any kind of data, from text to images to audio. This makes them a very powerful tool for machine learning practitioners. However, there's no one-size-fits-all solution for generating embeddings - there are many different libraries and APIs (both commercial and open source) that can be used to generate embeddings from structured/unstructured data.
+Due to the nature of vector embeddings, they can be used to represent any kind of data, from text to images to audio.
+This makes them a very powerful tool for machine learning practitioners.
+However, there's no one-size-fits-all solution for generating embeddings - there are many different libraries and APIs
+(both commercial and open source) that can be used to generate embeddings from structured/unstructured data.
 
-LanceDB supports 2 methods of vectorizing your raw data into embeddings.
+LanceDB supports 3 methods of working with embeddings.
 
-1. **Explicit**: By manually calling LanceDB's `with_embedding` function to vectorize your data via an `embed_func` of your choice
-2. **Implicit**: Allow LanceDB to embed the data and queries in the background as they come in, by using the table's `EmbeddingRegistry` information
+1. You can manually generate embeddings for the data and queries. This is done outside of LanceDB.
+2. You can use the built-in [embedding functions](./embedding_functions.md) to embed the data and queries in the background.
+3. You can define your own [custom embedding function](./custom_embedding_function.md)
+   that extends the default embedding functions.
 
-See the [explicit](embedding_explicit.md) and [implicit](embedding_functions.md) embedding sections for more details.
+For python users, there is also a legacy [with_embeddings API](./legacy.md).
+It is retained for compatibility and will be removed in a future version.
+
+## Quickstart
+
+To get started with embeddings, you can use the built-in embedding functions.
+
+### OpenAI Embedding function
+
+LanceDB registers the OpenAI embeddings function in the registry as `openai`. You can pass any supported model name to the `create`. By default it uses `"text-embedding-ada-002"`.
+
+=== "Python"
+
+    ```python
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import get_registry
+
+    db = lancedb.connect("/tmp/db")
+    func = get_registry().get("openai").create(name="text-embedding-ada-002")
+
+    class Words(LanceModel):
+        text: str = func.SourceField()
+        vector: Vector(func.ndims()) = func.VectorField()
+
+    table = db.create_table("words", schema=Words, mode="overwrite")
+    table.add(
+        [
+            {"text": "hello world"},
+            {"text": "goodbye world"}
+        ]
+        )
+
+    query = "greetings"
+    actual = table.search(query).limit(1).to_pydantic(Words)[0]
+    print(actual.text)
+    ```
+
+=== "TypeScript"
+
+    ```typescript
+    --8<--- "nodejs/examples/embedding.ts:imports"
+    --8<--- "nodejs/examples/embedding.ts:openai_embeddings"
+    ```
+
+=== "Rust"
+
+    ```rust
+    --8<--- "rust/lancedb/examples/openai.rs:imports"
+    --8<--- "rust/lancedb/examples/openai.rs:openai_embeddings"
+    ```
+
+### Sentence Transformers Embedding function
+LanceDB registers the Sentence Transformers embeddings function in the registry as `sentence-transformers`. You can pass any supported model name to the `create`. By default it uses `"sentence-transformers/paraphrase-MiniLM-L6-v2"`.
+
+=== "Python"
+    ```python
+    import lancedb
+    from lancedb.pydantic import LanceModel, Vector
+    from lancedb.embeddings import get_registry
+
+    db = lancedb.connect("/tmp/db")
+    model = get_registry().get("sentence-transformers").create(name="BAAI/bge-small-en-v1.5", device="cpu")
+
+    class Words(LanceModel):
+        text: str = model.SourceField()
+        vector: Vector(model.ndims()) = model.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)
+    ```
+
+=== "TypeScript"
+
+    Coming Soon!
+
+=== "Rust"
+
+    Coming Soon!
+
+### Jina Embeddings
+
+LanceDB registers the JinaAI embeddings function in the registry as `jina`. You can pass any supported model name to the `create`. By default it uses `"jina-clip-v1"`.
+`jina-clip-v1` can handle both text and images and other models only support `text`.
+
+You need to pass `JINA_API_KEY` in the environment variable or pass it as `api_key` to `create` method.
+
+```python
+import os
+import lancedb
+from lancedb.pydantic import LanceModel, Vector
+from lancedb.embeddings import get_registry
+os.environ['JINA_API_KEY'] = "jina_*"
+
+db = lancedb.connect("/tmp/db")
+func = get_registry().get("jina").create(name="jina-clip-v1")
+
+class Words(LanceModel):
+    text: str = func.SourceField()
+    vector: Vector(func.ndims()) = func.VectorField()
+
+table = db.create_table("words", schema=Words, mode="overwrite")
+table.add(
+    [
+        {"text": "hello world"},
+        {"text": "goodbye world"}
+    ]
+    )
+
+query = "greetings"
+actual = table.search(query).limit(1).to_pydantic(Words)[0]
+print(actual.text)
+```

docs/src/embeddings/legacy.md

@@ -0,0 +1,99 @@
+The legacy `with_embeddings` API is for Python only and is deprecated.
+
+### Hugging Face
+
+The most popular open source option is to use the [sentence-transformers](https://www.sbert.net/) 
+library, which can be installed via pip.
+
+```bash
+pip install sentence-transformers
+```
+
+The example below shows how to use the `paraphrase-albert-small-v2` model to generate embeddings 
+for a given document.
+
+```python
+from sentence_transformers import SentenceTransformer
+
+name="paraphrase-albert-small-v2"
+model = SentenceTransformer(name)
+
+# used for both training and querying
+def embed_func(batch):
+    return [model.encode(sentence) for sentence in batch]
+```
+
+
+### OpenAI
+
+Another popular alternative is to use an external API like OpenAI's [embeddings API](https://platform.openai.com/docs/guides/embeddings/what-are-embeddings).
+
+```python
+import openai
+import os
+
+# Configuring the environment variable OPENAI_API_KEY
+if "OPENAI_API_KEY" not in os.environ:
+# OR set the key here as a variable
+openai.api_key = "sk-..."
+
+client = openai.OpenAI()
+
+def embed_func(c):    
+    rs = client.embeddings.create(input=c, model="text-embedding-ada-002")
+    return [record.embedding for record in rs["data"]]
+```
+
+
+## Applying an embedding function to data
+
+Using an embedding function, you can apply it to raw data
+to generate embeddings for each record.
+
+Say you have a pandas DataFrame with a `text` column that you want embedded,
+you can use the `with_embeddings` function to generate embeddings and add them to 
+an existing table.
+
+```python
+    import pandas as pd
+    from lancedb.embeddings import with_embeddings
+
+    df = pd.DataFrame(
+        [
+            {"text": "pepperoni"},
+            {"text": "pineapple"}
+        ]
+    )
+    data = with_embeddings(embed_func, df)
+
+    # The output is used to create / append to a table
+    tbl = db.create_table("my_table", data=data)
+```
+
+If your data is in a different column, you can specify the `column` kwarg to `with_embeddings`.
+
+By default, LanceDB calls the function with batches of 1000 rows. This can be configured
+using the `batch_size` parameter to `with_embeddings`.
+
+LanceDB automatically wraps the function with retry and rate-limit logic to ensure the OpenAI
+API call is reliable.
+
+## Querying using an embedding function
+
+!!! warning
+    At query time, you **must** use the same embedding function you used to vectorize your data.
+    If you use a different embedding function, the embeddings will not reside in the same vector
+    space and the results will be nonsensical.
+
+=== "Python"
+     ```python
+     query = "What's the best pizza topping?"
+     query_vector = embed_func([query])[0]
+     results = (
+        tbl.search(query_vector)
+        .limit(10)
+        .to_pandas()
+     )
+     ```
+
+     The above snippet returns a pandas DataFrame with the 10 closest vectors to the query.

docs/src/examples/examples_rust.md

@@ -0,0 +1,3 @@
+# Examples: Rust
+
+Our Rust SDK is now stable. Examples are coming soon.

docs/src/examples/image_embeddings_roboflow.md

@@ -43,7 +43,7 @@ pip install lancedb
 We also need to install a specific commit of `tantivy`, a dependency of the LanceDB full text search engine we will use later in this guide:
 
 ```
-pip install tantivy@git+https://github.com/quickwit-oss/tantivy-py#164adc87e1a033117001cf70e38c82a53014d985
+pip install tantivy
 ```
 
 Create a new Python file and add the following code:

docs/src/examples/index.md

@@ -2,10 +2,11 @@
 
 ## Recipes and example code
 
-LanceDB provides language APIs, allowing you to embed a database in your language of choice. We currently provide Python and Javascript APIs, with the Rust API and examples actively being worked on and will be available soon.
+LanceDB provides language APIs, allowing you to embed a database in your language of choice.
 
 * 🐍 [Python](examples_python.md) examples
-* 👾 [JavaScript](exampled_js.md) examples
+* 👾 [JavaScript](examples_js.md) examples
+* 🦀 Rust examples (coming soon)
 
 ## Applications powered by LanceDB
 

docs/src/examples/modal_langchain.py

@@ -1,6 +1,5 @@
 import pickle
 import re
-import sys
 import zipfile
 from pathlib import Path
 

docs/src/extra_js/init_ask_ai_widget.js

@@ -1,11 +1,79 @@
-document.addEventListener("DOMContentLoaded", function () {
-    var script = document.createElement("script");
-    script.src = "https://widget.kapa.ai/kapa-widget.bundle.js";
-    script.setAttribute("data-website-id", "c5881fae-cec0-490b-b45e-d83d131d4f25");
-    script.setAttribute("data-project-name", "LanceDB");
-    script.setAttribute("data-project-color", "#000000");
-    script.setAttribute("data-project-logo", "https://avatars.githubusercontent.com/u/108903835?s=200&v=4");
-  script.setAttribute("data-modal-example-questions","Help me create an IVF_PQ index,How do I do an exhaustive search?,How do I create a LanceDB table?,Can I use my own embedding function?");
-    script.async = true;
-    document.head.appendChild(script);
-  });
+// Creates an SVG robot icon (from Lucide)
+function robotSVG() {
+  var svg = document.createElementNS("http://www.w3.org/2000/svg", "svg");
+  svg.setAttribute("width", "24");
+  svg.setAttribute("height", "24");
+  svg.setAttribute("viewBox", "0 0 24 24");
+  svg.setAttribute("fill", "none");
+  svg.setAttribute("stroke", "currentColor");
+  svg.setAttribute("stroke-width", "2");
+  svg.setAttribute("stroke-linecap", "round");
+  svg.setAttribute("stroke-linejoin", "round");
+  svg.setAttribute("class", "lucide lucide-bot-message-square");
+
+  var path1 = document.createElementNS("http://www.w3.org/2000/svg", "path");
+  path1.setAttribute("d", "M12 6V2H8");
+  svg.appendChild(path1);
+
+  var path2 = document.createElementNS("http://www.w3.org/2000/svg", "path");
+  path2.setAttribute("d", "m8 18-4 4V8a2 2 0 0 1 2-2h12a2 2 0 0 1 2 2v8a2 2 0 0 1-2 2Z");
+  svg.appendChild(path2);
+
+  var path3 = document.createElementNS("http://www.w3.org/2000/svg", "path");
+  path3.setAttribute("d", "M2 12h2");
+  svg.appendChild(path3);
+
+  var path4 = document.createElementNS("http://www.w3.org/2000/svg", "path");
+  path4.setAttribute("d", "M9 11v2");
+  svg.appendChild(path4);
+
+  var path5 = document.createElementNS("http://www.w3.org/2000/svg", "path");
+  path5.setAttribute("d", "M15 11v2");
+  svg.appendChild(path5);
+
+  var path6 = document.createElementNS("http://www.w3.org/2000/svg", "path");
+  path6.setAttribute("d", "M20 12h2");
+  svg.appendChild(path6);
+
+  return svg
+}
+
+// Creates the Fluidic Chatbot buttom
+function fluidicButton() {
+  var btn = document.createElement("a");
+  btn.href = "https://asklancedb.com";
+  btn.target = "_blank";
+  btn.style.position = "fixed";
+  btn.style.fontWeight = "bold";
+  btn.style.fontSize = ".8rem";
+  btn.style.right = "10px";
+  btn.style.bottom = "10px";
+  btn.style.width = "80px";
+  btn.style.height = "80px";
+  btn.style.background = "linear-gradient(135deg, #7C5EFF 0%, #625eff 100%)";
+  btn.style.color = "white";
+  btn.style.borderRadius = "5px";
+  btn.style.display = "flex";
+  btn.style.flexDirection = "column";
+  btn.style.justifyContent = "center";
+  btn.style.alignItems = "center";
+  btn.style.zIndex = "1000";
+  btn.style.opacity = "0";
+  btn.style.boxShadow = "0 0 0 rgba(0, 0, 0, 0)";
+  btn.style.transition = "opacity 0.2s ease-in, box-shadow 0.2s ease-in";
+
+  setTimeout(function() {
+      btn.style.opacity = "1";
+      btn.style.boxShadow = "0 0 .2rem #0000001a,0 .2rem .4rem #0003"
+  }, 0);
+
+  return btn
+}
+
+document.addEventListener("DOMContentLoaded", function() {
+  var btn = fluidicButton()
+  btn.appendChild(robotSVG());
+  var text = document.createTextNode("Ask AI");
+  btn.appendChild(text);
+  document.body.appendChild(btn);
+});

docs/src/faq.md

@@ -16,7 +16,7 @@ As we mention in our talk titled “[Lance, a modern columnar data format](https
 
 ### Why build in Rust? 🦀
 
-We believe that the Rust ecosystem has attained mainstream maturity and that Rust will form the underpinnings of large parts of the data and ML landscape in a few years. Performance, latency and reliability are paramount to a vector DB, and building in Rust allows us to iterate and release updates more rapidly due to Rust’s safety guarantees. Both Lance (the data format) and LanceDB (the database) are written entirely in Rust. We also provide Python and JavaScript client libraries to interact with the database. Our Rust API is a little rough around the edges right now, but is fast becoming on par with the Python and JS APIs.
+We believe that the Rust ecosystem has attained mainstream maturity and that Rust will form the underpinnings of large parts of the data and ML landscape in a few years. Performance, latency and reliability are paramount to a vector DB, and building in Rust allows us to iterate and release updates more rapidly due to Rust’s safety guarantees. Both Lance (the data format) and LanceDB (the database) are written entirely in Rust. We also provide Python, JavaScript, and Rust client libraries to interact with the database.
 
 ### What is the difference between LanceDB OSS and LanceDB Cloud?
 
@@ -40,11 +40,11 @@ LanceDB and its underlying data format, Lance, are built to scale to really larg
 
 No. LanceDB is blazing fast (due to its disk-based index) for even brute force kNN search, within reason. In our benchmarks, computing 100K pairs of 1000-dimension vectors takes less than 20ms. For small datasets of ~100K records or applications that can accept ~100ms latency, an ANN index is usually not necessary.
 
-For large-scale (>1M) or higher dimension vectors, it is beneficial to create an ANN index.
+For large-scale (>1M) or higher dimension vectors, it is beneficial to create an ANN index. See the [ANN indexes](ann_indexes.md) section for more details.
 
 ### Does LanceDB support full-text search?
 
-Yes, LanceDB supports full-text search (FTS) via [Tantivy](https://github.com/quickwit-oss/tantivy). Our current FTS integration is Python-only, and our goal is to push it down to the Rust level in future versions to enable much more powerful search capabilities available to our Python, JavaScript and Rust clients.
+Yes, LanceDB supports full-text search (FTS) via [Tantivy](https://github.com/quickwit-oss/tantivy). Our current FTS integration is Python-only, and our goal is to push it down to the Rust level in future versions to enable much more powerful search capabilities available to our Python, JavaScript and Rust clients.  Follow along in the [Github issue](https://github.com/lancedb/lance/issues/1195)
 
 ### How can I speed up data inserts?
 
@@ -69,3 +69,19 @@ MinIO supports an S3 compatible API. In order to connect to a MinIO instance, yo
 - Set the envvar `AWS_ENDPOINT` to the URL of your MinIO API
 - Set the envvars `AWS_ACCESS_KEY_ID` and `AWS_SECRET_ACCESS_KEY` with your MinIO credential
 - Call `lancedb.connect("s3://minio_bucket_name")`
+
+### Where can I find benchmarks for LanceDB?
+
+Refer to this [post](https://blog.lancedb.com/benchmarking-lancedb-92b01032874a) for recent benchmarks.
+
+### How much data can LanceDB practically manage without effecting performance?
+
+We target good performance on ~10-50 billion rows and ~10-30 TB of data.
+
+### Does LanceDB support concurrent operations?
+
+LanceDB can handle concurrent reads very well, and can scale horizontally. The main constraint is how well the [storage layer](https://lancedb.github.io/lancedb/concepts/storage/) you've chosen scales. For writes, we support concurrent writing, though too many concurrent writers can lead to failing writes as there is a limited number of times a writer retries a commit
+
+!!! info "Multiprocessing with LanceDB"
+
+    For multiprocessing you should probably not use ```fork``` as lance is multi-threaded internally and ```fork``` and multi-thread do not work well.[Refer to this discussion](https://discuss.python.org/t/concerns-regarding-deprecation-of-fork-with-alive-threads/33555)

docs/src/fts.md

@@ -1,8 +1,7 @@
 # Full-text search
 
-LanceDB provides support for full-text search via [Tantivy](https://github.com/quickwit-oss/tantivy) (currently Python only), allowing you to incorporate keyword-based search (based on BM25) in your retrieval solutions. Our goal is to push the FTS integration down to the Rust level in the future, so that it's available for JavaScript users as well.
+LanceDB provides support for full-text search via [Tantivy](https://github.com/quickwit-oss/tantivy) (currently Python only), allowing you to incorporate keyword-based search (based on BM25) in your retrieval solutions. Our goal is to push the FTS integration down to the Rust level in the future, so that it's available for Rust and JavaScript users as well.  Follow along at [this Github issue](https://github.com/lancedb/lance/issues/1195)
 
-A hybrid search solution combining vector and full-text search is also on the way.
 
 ## Installation
 
@@ -55,6 +54,16 @@ This returns the result as a list of dictionaries as follows.
 !!! note
     LanceDB automatically searches on the existing FTS index if the input to the search is of type `str`. If you provide a vector as input, LanceDB will search the ANN index instead.
 
+## Tokenization
+By default the text is tokenized by splitting on punctuation and whitespaces and then removing tokens that are longer than 40 chars. For more language specific tokenization then provide the argument tokenizer_name with the 2 letter language code followed by "_stem". So for english it would be "en_stem".
+
+```python
+table.create_fts_index("text", tokenizer_name="en_stem")
+```
+
+The following [languages](https://docs.rs/tantivy/latest/tantivy/tokenizer/enum.Language.html) are currently supported.
+
+
 ## Index multiple columns
 
 If you have multiple string columns to index, there's no need to combine them manually -- simply pass them all as a list to `create_fts_index`:
@@ -75,25 +84,75 @@ applied on top of the full text search results. This can be invoked via the fami
 table.search("puppy").limit(10).where("meta='foo'").to_list()
 ```
 
-## Syntax
+## Sorting
 
-For full-text search you can perform either a phrase query like "the old man and the sea", 
-or a structured search query like "(Old AND Man) AND Sea".
-Double quotes are used to disambiguate.
+You can pre-sort the documents by specifying `ordering_field_names` when
+creating the full-text search index. Once pre-sorted, you can then specify
+`ordering_field_name` while searching to return results sorted by the given
+field. For example, 
 
-For example: 
+```
+table.create_fts_index(["text_field"], ordering_field_names=["sort_by_field"])
 
-If you intended "they could have been dogs OR cats" as a phrase query, this actually
-raises a syntax error since `OR` is a recognized operator. If you make `or` lower case,
-this avoids the syntax error. However, it is cumbersome to have to remember what will
-conflict with the query syntax. Instead, if you search using 
-`table.search('"they could have been dogs OR cats"')`, then the syntax checker avoids
-checking inside the quotes.
+(table.search("terms", ordering_field_name="sort_by_field")
+ .limit(20)
+ .to_list())
+```
+
+!!! note
+    If you wish to specify an ordering field at query time, you must also
+    have specified it during indexing time. Otherwise at query time, an
+    error will be raised that looks like `ValueError: The field does not exist: xxx`
+
+!!! note
+    The fields to sort on must be of typed unsigned integer, or else you will see 
+    an error during indexing that looks like 
+    `TypeError: argument 'value': 'float' object cannot be interpreted as an integer`.
+
+!!! note
+    You can specify multiple fields for ordering at indexing time.
+    But at query time only one ordering field is supported.
+
+
+## Phrase queries vs. terms queries
+
+For full-text search you can specify either a **phrase** query like `"the old man and the sea"`,
+or a **terms** search query like `"(Old AND Man) AND Sea"`. For more details on the terms
+query syntax, see Tantivy's [query parser rules](https://docs.rs/tantivy/latest/tantivy/query/struct.QueryParser.html).
+
+!!! tip "Note"
+    The query parser will raise an exception on queries that are ambiguous. For example, in the query `they could have been dogs OR cats`, `OR` is capitalized so it's considered a keyword query operator. But it's ambiguous how the left part should be treated. So if you submit this search query as is, you'll get `Syntax Error: they could have been dogs OR cats`.
+
+    ```py
+    # This raises a syntax error
+    table.search("they could have been dogs OR cats")
+    ```
+
+    On the other hand, lowercasing `OR` to `or` will work, because there are no capitalized logical operators and
+    the query is treated as a phrase query.
+
+    ```py
+    # This works!
+    table.search("they could have been dogs or cats")
+    ```
+
+It can be cumbersome to have to remember what will cause a syntax error depending on the type of
+query you want to perform. To make this simpler, when you want to perform a phrase query, you can
+enforce it in one of two ways:
+
+1. Place the double-quoted query inside single quotes. For example, `table.search('"they could have been dogs OR cats"')` is treated as
+a phrase query.
+2. Explicitly declare the `phrase_query()` method. This is useful when you have a phrase query that
+itself contains double quotes. For example, `table.search('the cats OR dogs were not really "pets" at all').phrase_query()`
+is treated as a phrase query.
+
+In general, a query that's declared as a phrase query will be wrapped in double quotes during parsing, with nested
+double quotes replaced by single quotes.
 
 
 ## Configurations
 
-By default, LanceDB configures a 1GB heap size limit for creating the index. You can 
+By default, LanceDB configures a 1GB heap size limit for creating the index. You can
 reduce this if running on a smaller node, or increase this for faster performance while
 indexing a larger corpus.
 
@@ -109,7 +168,6 @@ table.create_fts_index(["text1", "text2"], writer_heap_size=heap, replace=True)
    If you add data after FTS index creation, it won't be reflected
    in search results until you do a full reindex.
 
-2. We currently only support local filesystem paths for the FTS index. 
+2. We currently only support local filesystem paths for the FTS index.
    This is a tantivy limitation. We've implemented an object store plugin
    but there's no way in tantivy-py to specify to use it.
-

docs/src/guides/storage.md

@@ -32,41 +32,232 @@ LanceDB OSS supports object stores such as AWS S3 (and compatible stores), Azure
     db = lancedb.connect("az://bucket/path")
     ```
 
-=== "JavaScript"
+=== "TypeScript"
 
-    AWS S3:
+    === "@lancedb/lancedb"
 
-    ```javascript
-    const lancedb = require("lancedb");
-    const db = await lancedb.connect("s3://bucket/path");
+        AWS S3:
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect("s3://bucket/path");
+        ```
+
+        Google Cloud Storage:
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect("gs://bucket/path");
+        ```
+
+        Azure Blob Storage:
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect("az://bucket/path");
+        ```
+
+
+    === "vectordb (deprecated)"
+
+        AWS S3:
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect("s3://bucket/path");
+        ```
+
+        Google Cloud Storage:
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect("gs://bucket/path");
+        ```
+
+        Azure Blob Storage:
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect("az://bucket/path");
+        ```
+
+In most cases, when running in the respective cloud and permissions are set up correctly, no additional configuration is required. When running outside of the respective cloud, authentication credentials must be provided. Credentials and other configuration options can be set in two ways: first, by setting environment variables. And second, by passing a `storage_options` object to the `connect` function. For example, to increase the request timeout to 60 seconds, you can set the `TIMEOUT` environment variable to `60s`:
+
+```bash
+export TIMEOUT=60s
+```
+
+!!! note "`storage_options` availability"
+
+    The `storage_options` parameter is only available in Python *async* API and JavaScript API.
+    It is not yet supported in the Python synchronous API.
+
+If you only want this to apply to one particular connection, you can pass the `storage_options` argument when opening the connection:
+
+=== "Python"
+
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "s3://bucket/path",
+        storage_options={"timeout": "60s"}
+    )
     ```
 
-    Google Cloud Storage:
+=== "TypeScript"
 
-    ```javascript
-    const lancedb = require("lancedb");
-    const db = await lancedb.connect("gs://bucket/path");
+    === "@lancedb/lancedb"
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+
+        const db = await lancedb.connect("s3://bucket/path", {
+            storageOptions: {timeout: "60s"}
+        });
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect("s3://bucket/path", {
+            storageOptions: {timeout: "60s"}
+        });
+        ```
+
+Getting even more specific, you can set the `timeout` for only a particular table:
+
+=== "Python"
+
+    <!-- skip-test -->
+    ```python
+    import lancedb
+    db = await lancedb.connect_async("s3://bucket/path")
+    table = await db.create_table(
+        "table",
+        [{"a": 1, "b": 2}],
+        storage_options={"timeout": "60s"}
+    )
     ```
 
-    Azure Blob Storage:
+=== "TypeScript"
 
-    ```javascript
-    const lancedb = require("lancedb");
-    const db = await lancedb.connect("az://bucket/path");
-    ```
+    === "@lancedb/lancedb"
 
-In most cases, when running in the respective cloud and permissions are set up correctly, no additional configuration is required. When running outside of the respective cloud, authentication credentials must be provided using environment variables. In general, these environment variables are the same as those used by the respective cloud SDKs. The sections below describe the environment variables that can be used to configure each object store.
+        <!-- skip-test -->
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect("s3://bucket/path");
+        const table = db.createTable(
+            "table",
+            [{ a: 1, b: 2}],
+            {storageOptions: {timeout: "60s"}}
+        );
+        ```
 
-LanceDB OSS uses the [object-store](https://docs.rs/object_store/latest/object_store/) Rust crate for object store access. There are general environment variables that can be used to configure the object store, such as the request timeout and proxy configuration. See the [object_store ClientConfigKey](https://docs.rs/object_store/latest/object_store/enum.ClientConfigKey.html) doc for available configuration options. The environment variables that can be set are the snake-cased versions of these variable names. For example, to set `ProxyUrl` use the environment variable `PROXY_URL`. (Don't let the Rust docs intimidate you! We link to them so you can see an up-to-date list of the available options.)
+    === "vectordb (deprecated)"
 
+        <!-- skip-test -->
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect("s3://bucket/path");
+        const table = db.createTable(
+            "table",
+            [{ a: 1, b: 2}],
+            {storageOptions: {timeout: "60s"}}
+        );
+        ```
+
+!!! info "Storage option casing"
+
+    The storage option keys are case-insensitive. So `connect_timeout` and `CONNECT_TIMEOUT` are the same setting. Usually lowercase is used in the `storage_options` argument and uppercase is used for environment variables. In the `lancedb` Node package, the keys can also be provided in `camelCase` capitalization. For example, `connectTimeout` is equivalent to `connect_timeout`.
+
+### General configuration
+
+There are several options that can be set for all object stores, mostly related to network client configuration.
+
+<!-- from here: https://docs.rs/object_store/latest/object_store/enum.ClientConfigKey.html -->
+
+| Key                        | Description                                                                                      |
+|----------------------------|--------------------------------------------------------------------------------------------------|
+| `allow_http`               | Allow non-TLS, i.e. non-HTTPS connections. Default: `False`.                                      |
+| `allow_invalid_certificates`| Skip certificate validation on HTTPS connections. Default: `False`.                               |
+| `connect_timeout`          | Timeout for only the connect phase of a Client. Default: `5s`.                                    |
+| `timeout`                  | Timeout for the entire request, from connection until the response body has finished. Default: `30s`. |
+| `user_agent`               | User agent string to use in requests.                                                             |
+| `proxy_url`                | URL of a proxy server to use for requests. Default: `None`.                                       |
+| `proxy_ca_certificate`     | PEM-formatted CA certificate for proxy connections.                                                |
+| `proxy_excludes`           | List of hosts that bypass the proxy. This is a comma-separated list of domains and IP masks. Any subdomain of the provided domain will be bypassed. For example, `example.com, 192.168.1.0/24` would bypass `https://api.example.com`, `https://www.example.com`, and any IP in the range `192.168.1.0/24`. |
 
 ### AWS S3
 
-To configure credentials for AWS S3, you can use the `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`, and `AWS_SESSION_TOKEN` environment variables.
+To configure credentials for AWS S3, you can use the `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`, and `AWS_SESSION_TOKEN` keys. Region can also be set, but it is not mandatory when using AWS.
+These can be set as environment variables or passed in the `storage_options` parameter:
+
+=== "Python"
+
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "s3://bucket/path",
+        storage_options={
+            "aws_access_key_id": "my-access-key",
+            "aws_secret_access_key": "my-secret-key",
+            "aws_session_token": "my-session-token",
+        }
+    )
+    ```
+
+=== "TypeScript"
+
+    === "@lancedb/lancedb"
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect(
+            "s3://bucket/path",
+            {
+                storageOptions: {
+                    awsAccessKeyId: "my-access-key",
+                    awsSecretAccessKey: "my-secret-key",
+                    awsSessionToken: "my-session-token",
+                }
+            }
+        );
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect(
+            "s3://bucket/path",
+            {
+                storageOptions: {
+                    awsAccessKeyId: "my-access-key",
+                    awsSecretAccessKey: "my-secret-key",
+                    awsSessionToken: "my-session-token",
+                }
+            }
+        );
+        ```
 
 Alternatively, if you are using AWS SSO, you can use the `AWS_PROFILE` and `AWS_DEFAULT_REGION` environment variables.
 
-You can see a full list of environment variables [here](https://docs.rs/object_store/latest/object_store/aws/struct.AmazonS3Builder.html#method.from_env).
+The following keys can be used as both environment variables or keys in the `storage_options` parameter:
+
+| Key                                | Description                                                                                          |
+|------------------------------------|------------------------------------------------------------------------------------------------------|
+| `aws_region` / `region`             | The AWS region the bucket is in. This can be automatically detected when using AWS S3, but must be specified for S3-compatible stores. |
+| `aws_access_key_id` / `access_key_id` | The AWS access key ID to use.                                                                       |
+| `aws_secret_access_key` / `secret_access_key` | The AWS secret access key to use.                                                               |
+| `aws_session_token` / `session_token` | The AWS session token to use.                                                                     |
+| `aws_endpoint` / `endpoint`         | The endpoint to use for S3-compatible stores.                                                       |
+| `aws_virtual_hosted_style_request` / `virtual_hosted_style_request` | Whether to use virtual hosted-style requests, where the bucket name is part of the endpoint. Meant to be used with `aws_endpoint`. Default: `False`. |
+| `aws_s3_express` / `s3_express`     | Whether to use S3 Express One Zone endpoints. Default: `False`. See more details below.             |
+| `aws_server_side_encryption`        | The server-side encryption algorithm to use. Must be one of `"AES256"`, `"aws:kms"`, or `"aws:kms:dsse"`. Default: `None`. |
+| `aws_sse_kms_key_id`                | The KMS key ID to use for server-side encryption. If set, `aws_server_side_encryption` must be `"aws:kms"` or `"aws:kms:dsse"`. |
+| `aws_sse_bucket_key_enabled`        | Whether to use bucket keys for server-side encryption.                                               |
 
 !!! tip "Automatic cleanup for failed writes"
 
@@ -144,24 +335,349 @@ For **read-only access**, LanceDB will need a policy such as:
 }
 ```
 
+#### DynamoDB Commit Store for concurrent writes
+
+By default, S3 does not support concurrent writes. Having two or more processes
+writing to the same table at the same time can lead to data corruption. This is
+because S3, unlike other object stores, does not have any atomic put or copy
+operation.
+
+To enable concurrent writes, you can configure LanceDB to use a DynamoDB table
+as a commit store. This table will be used to coordinate writes between
+different processes. To enable this feature, you must modify your connection
+URI to use the `s3+ddb` scheme and add a query parameter `ddbTableName` with the
+name of the table to use.
+
+=== "Python"
+
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "s3+ddb://bucket/path?ddbTableName=my-dynamodb-table",
+    )
+    ```
+
+=== "JavaScript"
+
+    ```javascript
+    const lancedb = require("lancedb");
+
+    const db = await lancedb.connect(
+        "s3+ddb://bucket/path?ddbTableName=my-dynamodb-table",
+    );
+    ```
+
+The DynamoDB table must be created with the following schema:
+
+- Hash key: `base_uri` (string)
+- Range key: `version` (number)
+
+You can create this programmatically with:
+
+=== "Python"
+
+    <!-- skip-test -->
+    ```python
+    import boto3
+
+    dynamodb = boto3.client("dynamodb")
+    table = dynamodb.create_table(
+        TableName=table_name,
+        KeySchema=[
+            {"AttributeName": "base_uri", "KeyType": "HASH"},
+            {"AttributeName": "version", "KeyType": "RANGE"},
+        ],
+        AttributeDefinitions=[
+            {"AttributeName": "base_uri", "AttributeType": "S"},
+            {"AttributeName": "version", "AttributeType": "N"},
+        ],
+        ProvisionedThroughput={"ReadCapacityUnits": 1, "WriteCapacityUnits": 1},
+    )
+    ```
+
+=== "JavaScript"
+
+    <!-- skip-test -->
+    ```javascript
+    import {
+      CreateTableCommand,
+      DynamoDBClient,
+    } from "@aws-sdk/client-dynamodb";
+
+    const dynamodb = new DynamoDBClient({
+      region: CONFIG.awsRegion,
+      credentials: {
+        accessKeyId: CONFIG.awsAccessKeyId,
+        secretAccessKey: CONFIG.awsSecretAccessKey,
+      },
+      endpoint: CONFIG.awsEndpoint,
+    });
+    const command = new CreateTableCommand({
+      TableName: table_name,
+      AttributeDefinitions: [
+        {
+          AttributeName: "base_uri",
+          AttributeType: "S",
+        },
+        {
+          AttributeName: "version",
+          AttributeType: "N",
+        },
+      ],
+      KeySchema: [
+        { AttributeName: "base_uri", KeyType: "HASH" },
+        { AttributeName: "version", KeyType: "RANGE" },
+      ],
+      ProvisionedThroughput: {
+        ReadCapacityUnits: 1,
+        WriteCapacityUnits: 1,
+      },
+    });
+    await client.send(command);
+    ```
+
+
 #### S3-compatible stores
 
-LanceDB can also connect to S3-compatible stores, such as MinIO. To do so, you must specify two environment variables: `AWS_ENDPOINT` and `AWS_DEFAULT_REGION`. `AWS_ENDPOINT` should be the URL of the S3-compatible store, and `AWS_DEFAULT_REGION` should be the region to use.
+LanceDB can also connect to S3-compatible stores, such as MinIO. To do so, you must specify both region and endpoint:
 
-<!-- TODO: we should also document the use of S3 Express once we fully support it -->
+=== "Python"
+
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "s3://bucket/path",
+        storage_options={
+            "region": "us-east-1",
+            "endpoint": "http://minio:9000",
+        }
+    )
+    ```
+
+=== "TypeScript"
+
+    === "@lancedb/lancedb"
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect(
+            "s3://bucket/path",
+            {
+                storageOptions: {
+                    region: "us-east-1",
+                    endpoint: "http://minio:9000",
+                }
+            }
+        );
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect(
+            "s3://bucket/path",
+            {
+                storageOptions: {
+                    region: "us-east-1",
+                    endpoint: "http://minio:9000",
+                }
+            }
+        );
+        ```
+
+This can also be done with the ``AWS_ENDPOINT`` and ``AWS_DEFAULT_REGION`` environment variables.
+
+!!! tip "Local servers"
+
+    For local development, the server often has a `http` endpoint rather than a
+    secure `https` endpoint. In this case, you must also set the `ALLOW_HTTP`
+    environment variable to `true` to allow non-TLS connections, or pass the
+    storage option `allow_http` as `true`. If you do not do this, you will get
+    an error like `URL scheme is not allowed`.
+
+#### S3 Express
+
+LanceDB supports [S3 Express One Zone](https://aws.amazon.com/s3/storage-classes/express-one-zone/) endpoints, but requires additional configuration. Also, S3 Express endpoints only support connecting from an EC2 instance within the same region.
+
+To configure LanceDB to use an S3 Express endpoint, you must set the storage option `s3_express`. The bucket name in your table URI should **include the suffix**.
+
+=== "Python"
+
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "s3://my-bucket--use1-az4--x-s3/path",
+        storage_options={
+            "region": "us-east-1",
+            "s3_express": "true",
+        }
+    )
+    ```
+
+=== "TypeScript"
+
+    === "@lancedb/lancedb"
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect(
+            "s3://my-bucket--use1-az4--x-s3/path",
+            {
+                storageOptions: {
+                    region: "us-east-1",
+                    s3Express: "true",
+                }
+            }
+        );
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect(
+            "s3://my-bucket--use1-az4--x-s3/path",
+            {
+                storageOptions: {
+                    region: "us-east-1",
+                    s3Express: "true",
+                }
+            }
+        );
+        ```
 
 ### Google Cloud Storage
 
-GCS credentials are configured by setting the `GOOGLE_SERVICE_ACCOUNT` environment variable to the path of a JSON file containing the service account credentials. There are several aliases for this environment variable, documented [here](https://docs.rs/object_store/latest/object_store/gcp/struct.GoogleCloudStorageBuilder.html#method.from_env).
+GCS credentials are configured by setting the `GOOGLE_SERVICE_ACCOUNT` environment variable to the path of a JSON file containing the service account credentials. Alternatively, you can pass the path to the JSON file in the `storage_options`:
 
+=== "Python"
+
+    <!-- skip-test -->
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "gs://my-bucket/my-database",
+        storage_options={
+            "service_account": "path/to/service-account.json",
+        }
+    )
+    ```
+
+=== "TypeScript"
+
+    === "@lancedb/lancedb"
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect(
+            "gs://my-bucket/my-database",
+            {
+                storageOptions: {
+                    serviceAccount: "path/to/service-account.json",
+                }
+            }
+        );
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect(
+            "gs://my-bucket/my-database",
+            {
+                storageOptions: {
+                    serviceAccount: "path/to/service-account.json",
+                }
+            }
+        );
+        ```
 
 !!! info "HTTP/2 support"
 
     By default, GCS uses HTTP/1 for communication, as opposed to HTTP/2. This improves maximum throughput significantly. However, if you wish to use HTTP/2 for some reason, you can set the environment variable `HTTP1_ONLY` to `false`.
 
+The following keys can be used as both environment variables or keys in the `storage_options` parameter:
+<!-- source: https://docs.rs/object_store/latest/object_store/gcp/enum.GoogleConfigKey.html -->
+
+| Key                                   | Description                                  |
+|---------------------------------------|----------------------------------------------|
+| ``google_service_account`` / `service_account` | Path to the service account JSON file.       |
+| ``google_service_account_key``        | The serialized service account key.          |
+| ``google_application_credentials``    | Path to the application credentials.         |
+
 ### Azure Blob Storage
 
-Azure Blob Storage credentials can be configured by setting the `AZURE_STORAGE_ACCOUNT_NAME` and ``AZURE_STORAGE_ACCOUNT_KEY`` environment variables. The full list of environment variables that can be set are documented [here](https://docs.rs/object_store/latest/object_store/azure/struct.MicrosoftAzureBuilder.html#method.from_env).
+Azure Blob Storage credentials can be configured by setting the `AZURE_STORAGE_ACCOUNT_NAME`and `AZURE_STORAGE_ACCOUNT_KEY` environment variables. Alternatively, you can pass the account name and key in the `storage_options` parameter:
 
+=== "Python"
 
-<!-- TODO: demonstrate how to configure networked file systems for optimal performance -->
+    <!-- skip-test -->
+    ```python
+    import lancedb
+    db = await lancedb.connect_async(
+        "az://my-container/my-database",
+        storage_options={
+            account_name: "some-account",
+            account_key: "some-key",
+        }
+    )
+    ```
+
+=== "TypeScript"
+
+    === "@lancedb/lancedb"
+
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
+        const db = await lancedb.connect(
+            "az://my-container/my-database",
+            {
+                storageOptions: {
+                    accountName: "some-account",
+                    accountKey: "some-key",
+                }
+            }
+        );
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        const lancedb = require("lancedb");
+        const db = await lancedb.connect(
+            "az://my-container/my-database",
+            {
+                storageOptions: {
+                    accountName: "some-account",
+                    accountKey: "some-key",
+                }
+            }
+        );
+        ```
+
+These keys can be used as both environment variables or keys in the `storage_options` parameter:
+
+<!-- source: https://docs.rs/object_store/latest/object_store/azure/enum.AzureConfigKey.html -->
+
+| Key                                   | Description                                                                                      |
+|---------------------------------------|--------------------------------------------------------------------------------------------------|
+| ``azure_storage_account_name``        | The name of the azure storage account.                                                           |
+| ``azure_storage_account_key``         | The serialized service account key.                                                              |
+| ``azure_client_id``                   | Service principal client id for authorizing requests.                                            |
+| ``azure_client_secret``               | Service principal client secret for authorizing requests.                                        |
+| ``azure_tenant_id``                   | Tenant id used in oauth flows.                                                                   |
+| ``azure_storage_sas_key``             | Shared access signature. The signature is expected to be percent-encoded, much like they are provided in the azure storage explorer or azure portal. |
+| ``azure_storage_token``               | Bearer token.                                                                                    |
+| ``azure_storage_use_emulator``        | Use object store with azurite storage emulator.                                                  |
+| ``azure_endpoint``                    | Override the endpoint used to communicate with blob storage.                                      |
+| ``azure_use_fabric_endpoint``         | Use object store with url scheme account.dfs.fabric.microsoft.com.                               |
+| ``azure_msi_endpoint``                | Endpoint to request a imds managed identity token.                                               |
+| ``azure_object_id``                   | Object id for use with managed identity authentication.                                          |
+| ``azure_msi_resource_id``             | Msi resource id for use with managed identity authentication.                                    |
+| ``azure_federated_token_file``        | File containing token for Azure AD workload identity federation.                                 |
+| ``azure_use_azure_cli``               | Use azure cli for acquiring access token.                                                        |
+| ``azure_disable_tagging``             | Disables tagging objects. This can be desirable if not supported by the backing store.           |
+
+<!-- TODO: demonstrate how to configure networked file systems for optimal performance -->

docs/src/guides/tables.md

@@ -3,32 +3,46 @@
 
 A Table is a collection of Records in a LanceDB Database. Tables in Lance have a schema that defines the columns and their types. These schemas can include nested columns and can evolve over time.
 
-This guide will show how to create tables, insert data into them, and update the data.  
+This guide will show how to create tables, insert data into them, and update the data.
 
 
 ## Creating a LanceDB Table
 
+Initialize a LanceDB connection and create a table
+
 === "Python"
-    Initialize a LanceDB connection and create a table using one of the many methods listed below.
 
     ```python
     import lancedb
     db = lancedb.connect("./.lancedb")
     ```
 
-=== "Javascript"
-
-    Initialize a VectorDB connection and create a table using one of the many methods listed below.
-
-    ```javascript
-    const lancedb = require("vectordb");
-
-    const uri = "data/sample-lancedb";
-    const db = await lancedb.connect(uri);
-    ```
-
     LanceDB allows ingesting data from various sources - `dict`, `list[dict]`, `pd.DataFrame`, `pa.Table` or a `Iterator[pa.RecordBatch]`. Let's take a look at some of the these.
 
+=== "Typescript[^1]"
+
+    === "@lancedb/lancedb"
+
+        ```typescript
+        import * as lancedb from "@lancedb/lancedb";
+        import * as arrow from "apache-arrow";
+
+        const uri = "data/sample-lancedb";
+        const db = await lancedb.connect(uri);
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        const lancedb = require("vectordb");
+        const arrow = require("apache-arrow");
+
+        const uri = "data/sample-lancedb";
+        const db = await lancedb.connect(uri);
+        ```
+
+
+
 ### From list of tuples or dictionaries
 
 === "Python"
@@ -45,102 +59,150 @@ This guide will show how to create tables, insert data into them, and update the
 
     db["my_table"].head()
     ```
+
     !!! info "Note"
-        If the table already exists, LanceDB will raise an error by default. 
+        If the table already exists, LanceDB will raise an error by default.
 
         `create_table` supports an optional `exist_ok` parameter. When set to True
         and the table exists, then it simply opens the existing table. The data you
         passed in will NOT be appended to the table in that case.
 
-        ```python
-        db.create_table("name", data, exist_ok=True)
-        ```
-        
-        Sometimes you want to make sure that you start fresh. If you want to 
-        overwrite the table, you can pass in mode="overwrite" to the createTable function.
-
-        ```python
-        db.create_table("name", data, mode="overwrite")
-        ```
-
-=== "Javascript"
-    You can create a LanceDB table in JavaScript using an array of JSON records as follows.
-
-    ```javascript
-    const tb = await db.createTable("my_table", [{
-        "vector": [3.1, 4.1],
-        "item": "foo",
-        "price": 10.0
-    }, {
-        "vector": [5.9, 26.5],
-        "item": "bar",
-        "price": 20.0
-    }]);
-    ```
-    !!! info "Note"
-    If the table already exists, LanceDB will raise an error by default. If you want to overwrite the table, you need to specify the `WriteMode` in the createTable function.
-
-    ```javascript
-    const table = await con.createTable(tableName, data, { writeMode: WriteMode.Overwrite })
+    ```python
+    db.create_table("name", data, exist_ok=True)
     ```
 
-    ### From a Pandas DataFrame
+    Sometimes you want to make sure that you start fresh. If you want to
+    overwrite the table, you can pass in mode="overwrite" to the createTable function.
 
     ```python
-    import pandas as pd
-
-    data = pd.DataFrame({
-        "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]
-    })
-
-    db.create_table("my_table", data)
-
-    db["my_table"].head()
+    db.create_table("name", data, mode="overwrite")
     ```
-    !!! info "Note"
+
+=== "Typescript[^1]"
+    You can create a LanceDB table in JavaScript using an array of records as follows.
+
+    === "@lancedb/lancedb"
+
+
+        ```ts
+        --8<-- "nodejs/examples/basic.ts:create_table"
+        ```
+
+        This will infer the schema from the provided data. If you want to explicitly provide a schema, you can use `apache-arrow` to declare a schema
+
+        ```ts
+        --8<-- "nodejs/examples/basic.ts:create_table_with_schema"
+        ```
+
+        !!! info "Note"
+            `createTable` supports an optional `existsOk` parameter. When set to true
+            and the table exists, then it simply opens the existing table. The data you
+            passed in will NOT be appended to the table in that case.
+
+        ```ts
+        --8<-- "nodejs/examples/basic.ts:create_table_exists_ok"
+        ```
+
+        Sometimes you want to make sure that you start fresh. If you want to
+        overwrite the table, you can pass in mode: "overwrite" to the createTable function.
+
+        ```ts
+        --8<-- "nodejs/examples/basic.ts:create_table_overwrite"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        --8<-- "docs/src/basic_legacy.ts:create_table"
+        ```
+
+        This will infer the schema from the provided data. If you want to explicitly provide a schema, you can use apache-arrow to declare a schema
+
+
+
+        ```ts
+        --8<-- "docs/src/basic_legacy.ts:create_table_with_schema"
+        ```
+
+        !!! warning
+            `existsOk` is not available in `vectordb`
+
+
+
+            If the table already exists, vectordb will raise an error by default.
+            You can use `writeMode: WriteMode.Overwrite` to overwrite the table.
+            But this will delete the existing table and create a new one with the same name.
+
+
+        Sometimes you want to make sure that you start fresh.
+
+        If you want to overwrite the table, you can pass in `writeMode: lancedb.WriteMode.Overwrite` to the createTable function.
+
+        ```ts
+        const table = await con.createTable(tableName, data, {
+            writeMode: WriteMode.Overwrite
+        })
+        ```
+
+### From a Pandas DataFrame
+
+```python
+import pandas as pd
+
+data = pd.DataFrame({
+    "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]
+})
+
+db.create_table("my_table", data)
+
+db["my_table"].head()
+```
+
+!!! info "Note"
     Data is converted to Arrow before being written to disk. For maximum control over how data is saved, either provide the PyArrow schema to convert to or else provide a PyArrow Table directly.
 
-    The **`vector`** column needs to be a [Vector](../python/pydantic.md#vector-field) (defined as [pyarrow.FixedSizeList](https://arrow.apache.org/docs/python/generated/pyarrow.list_.html)) type. 
+The **`vector`** column needs to be a [Vector](../python/pydantic.md#vector-field) (defined as [pyarrow.FixedSizeList](https://arrow.apache.org/docs/python/generated/pyarrow.list_.html)) type.
 
-    ```python
-    custom_schema = pa.schema([
-    pa.field("vector", pa.list_(pa.float32(), 4)),
-    pa.field("lat", pa.float32()),
-    pa.field("long", pa.float32())
-    ])
+```python
+custom_schema = pa.schema([
+pa.field("vector", pa.list_(pa.float32(), 4)),
+pa.field("lat", pa.float32()),
+pa.field("long", pa.float32())
+])
 
-    table = db.create_table("my_table", data, schema=custom_schema)
-    ```
+table = db.create_table("my_table", data, schema=custom_schema)
+```
 
 ### From a Polars DataFrame
 
-    LanceDB supports [Polars](https://pola.rs/), a modern, fast DataFrame library
-    written in Rust. Just like in Pandas, the Polars integration is enabled by PyArrow
-    under the hood. A deeper integration between LanceDB Tables and Polars DataFrames
-    is on the way.
+LanceDB supports [Polars](https://pola.rs/), a modern, fast DataFrame library
+written in Rust. Just like in Pandas, the Polars integration is enabled by PyArrow
+under the hood. A deeper integration between LanceDB Tables and Polars DataFrames
+is on the way.
 
-    ```python
-    import polars as pl
+```python
+import polars as pl
 
-    data = pl.DataFrame({
-        "vector": [[3.1, 4.1], [5.9, 26.5]],
-        "item": ["foo", "bar"],
-        "price": [10.0, 20.0]
-    })
-    table = db.create_table("pl_table", data=data)
-    ```
+data = pl.DataFrame({
+    "vector": [[3.1, 4.1], [5.9, 26.5]],
+    "item": ["foo", "bar"],
+    "price": [10.0, 20.0]
+})
+table = db.create_table("pl_table", data=data)
+```
 
 ### From an Arrow Table
+You can also create LanceDB tables directly from Arrow tables.
+LanceDB supports float16 data type!
+
 === "Python"
-    You can also create LanceDB tables directly from Arrow tables. 
-    LanceDB supports float16 data type!
 
     ```python
     import pyarrows as pa
     import numpy as np
-    
+
     dim = 16
     total = 2
     schema = pa.schema(
@@ -160,159 +222,165 @@ This guide will show how to create tables, insert data into them, and update the
     tbl = db.create_table("f16_tbl", data, schema=schema)
     ```
 
-=== "Javascript"
-    You can also create LanceDB tables directly from Arrow tables. 
-    LanceDB supports Float16 data type!
+=== "Typescript[^1]"
 
-    ```javascript
-    --8<-- "docs/src/basic_legacy.ts:create_f16_table"
-    ```
+    === "@lancedb/lancedb"
 
-    ### From Pydantic Models
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:create_f16_table"
+        ```
 
-    When you create an empty table without data, you must specify the table schema.
-    LanceDB supports creating tables by specifying a PyArrow schema or a specialized
-    Pydantic model called `LanceModel`.
+    === "vectordb (deprecated)"
 
-    For example, the following Content model specifies a table with 5 columns:
-    `movie_id`, `vector`, `genres`, `title`, and `imdb_id`. When you create a table, you can
-    pass the class as the value of the `schema` parameter to `create_table`.
-    The `vector` column is a `Vector` type, which is a specialized Pydantic type that
-    can be configured with the vector dimensions. It is also important to note that
-    LanceDB only understands subclasses of `lancedb.pydantic.LanceModel`
-    (which itself derives from `pydantic.BaseModel`).
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:create_f16_table"
+        ```
 
-    ```python
-    from lancedb.pydantic import Vector, LanceModel
+### From Pydantic Models
 
-    class Content(LanceModel):
-        movie_id: int
-        vector: Vector(128)
-        genres: str
-        title: str
-        imdb_id: int
+When you create an empty table without data, you must specify the table schema.
+LanceDB supports creating tables by specifying a PyArrow schema or a specialized
+Pydantic model called `LanceModel`.
 
-        @property
-        def imdb_url(self) -> str:
-            return f"https://www.imdb.com/title/tt{self.imdb_id}"
+For example, the following Content model specifies a table with 5 columns:
+`movie_id`, `vector`, `genres`, `title`, and `imdb_id`. When you create a table, you can
+pass the class as the value of the `schema` parameter to `create_table`.
+The `vector` column is a `Vector` type, which is a specialized Pydantic type that
+can be configured with the vector dimensions. It is also important to note that
+LanceDB only understands subclasses of `lancedb.pydantic.LanceModel`
+(which itself derives from `pydantic.BaseModel`).
 
-    import pyarrow as pa
-    db = lancedb.connect("~/.lancedb")
-    table_name = "movielens_small"
-    table = db.create_table(table_name, schema=Content)
-    ```
+```python
+from lancedb.pydantic import Vector, LanceModel
 
-    #### Nested schemas
+class Content(LanceModel):
+    movie_id: int
+    vector: Vector(128)
+    genres: str
+    title: str
+    imdb_id: int
 
-    Sometimes your data model may contain nested objects.
-    For example, you may want to store the document string
-    and the document soure name as a nested Document object:
+    @property
+    def imdb_url(self) -> str:
+        return f"https://www.imdb.com/title/tt{self.imdb_id}"
 
-    ```python
-    class Document(BaseModel):
-        content: str
-        source: str
-    ```
+import pyarrow as pa
+db = lancedb.connect("~/.lancedb")
+table_name = "movielens_small"
+table = db.create_table(table_name, schema=Content)
+```
 
-    This can be used as the type of a LanceDB table column:
+#### Nested schemas
 
-    ```python
-    class NestedSchema(LanceModel):
-        id: str
-        vector: Vector(1536)
-        document: Document
+Sometimes your data model may contain nested objects.
+For example, you may want to store the document string
+and the document soure name as a nested Document object:
 
-    tbl = db.create_table("nested_table", schema=NestedSchema, mode="overwrite")
-    ```
+```python
+class Document(BaseModel):
+    content: str
+    source: str
+```
 
-    This creates a struct column called "document" that has two subfields 
-    called "content" and "source":
+This can be used as the type of a LanceDB table column:
 
-    ```
-    In [28]: tbl.schema
-    Out[28]:
-    id: string not null
-    vector: fixed_size_list<item: float>[1536] not null
-        child 0, item: float
-    document: struct<content: string not null, source: string not null> not null
-        child 0, content: string not null
-        child 1, source: string not null    
-    ```
+```python
+class NestedSchema(LanceModel):
+    id: str
+    vector: Vector(1536)
+    document: Document
 
-    #### Validators
+tbl = db.create_table("nested_table", schema=NestedSchema, mode="overwrite")
+```
 
-    Note that neither Pydantic nor PyArrow automatically validates that input data
-    is of the correct timezone, but this is easy to add as a custom field validator:
+This creates a struct column called "document" that has two subfields
+called "content" and "source":
 
-    ```python
-    from datetime import datetime
-    from zoneinfo import ZoneInfo
+```
+In [28]: tbl.schema
+Out[28]:
+id: string not null
+vector: fixed_size_list<item: float>[1536] not null
+    child 0, item: float
+document: struct<content: string not null, source: string not null> not null
+    child 0, content: string not null
+    child 1, source: string not null
+```
 
-    from lancedb.pydantic import LanceModel
-    from pydantic import Field, field_validator, ValidationError, ValidationInfo
+#### Validators
 
-    tzname = "America/New_York"
-    tz = ZoneInfo(tzname)
+Note that neither Pydantic nor PyArrow automatically validates that input data
+is of the correct timezone, but this is easy to add as a custom field validator:
 
-    class TestModel(LanceModel):
-        dt_with_tz: datetime = Field(json_schema_extra={"tz": tzname})
+```python
+from datetime import datetime
+from zoneinfo import ZoneInfo
 
-        @field_validator('dt_with_tz')
-        @classmethod
-        def tz_must_match(cls, dt: datetime) -> datetime:
-            assert dt.tzinfo == tz
-            return dt        
+from lancedb.pydantic import LanceModel
+from pydantic import Field, field_validator, ValidationError, ValidationInfo
 
-    ok = TestModel(dt_with_tz=datetime.now(tz))
+tzname = "America/New_York"
+tz = ZoneInfo(tzname)
 
-    try:
-        TestModel(dt_with_tz=datetime.now(ZoneInfo("Asia/Shanghai")))
-        assert 0 == 1, "this should raise ValidationError"
-    except ValidationError:
-        print("A ValidationError was raised.")
-        pass
-    ```
+class TestModel(LanceModel):
+    dt_with_tz: datetime = Field(json_schema_extra={"tz": tzname})
 
-    When you run this code it should print "A ValidationError was raised."
+    @field_validator('dt_with_tz')
+    @classmethod
+    def tz_must_match(cls, dt: datetime) -> datetime:
+        assert dt.tzinfo == tz
+        return dt
 
-    #### Pydantic custom types
+ok = TestModel(dt_with_tz=datetime.now(tz))
 
-    LanceDB does NOT yet support converting pydantic custom types. If this is something you need,
-    please file a feature request on the [LanceDB Github repo](https://github.com/lancedb/lancedb/issues/new).
+try:
+    TestModel(dt_with_tz=datetime.now(ZoneInfo("Asia/Shanghai")))
+    assert 0 == 1, "this should raise ValidationError"
+except ValidationError:
+    print("A ValidationError was raised.")
+    pass
+```
 
-    ### Using Iterators / Writing Large Datasets
+When you run this code it should print "A ValidationError was raised."
 
-    It is recommended to use iterators to add large datasets in batches when creating your table in one go. This does not create multiple versions of your dataset unlike manually adding batches using `table.add()`
+#### Pydantic custom types
 
-    LanceDB additionally supports PyArrow's `RecordBatch` Iterators or other generators producing supported data types.
+LanceDB does NOT yet support converting pydantic custom types. If this is something you need,
+please file a feature request on the [LanceDB Github repo](https://github.com/lancedb/lancedb/issues/new).
 
-    Here's an example using using `RecordBatch` iterator for creating tables.
+### Using Iterators / Writing Large Datasets
 
-    ```python
-    import pyarrow as pa
+It is recommended to use iterators to add large datasets in batches when creating your table in one go. This does not create multiple versions of your dataset unlike manually adding batches using `table.add()`
 
-    def make_batches():
-        for i in range(5):
-            yield pa.RecordBatch.from_arrays(
-                [
-                    pa.array([[3.1, 4.1, 5.1, 6.1], [5.9, 26.5, 4.7, 32.8]],
-                            pa.list_(pa.float32(), 4)),
-                    pa.array(["foo", "bar"]),
-                    pa.array([10.0, 20.0]),
-                ],
-                ["vector", "item", "price"],
-            )
+LanceDB additionally supports PyArrow's `RecordBatch` Iterators or other generators producing supported data types.
 
-    schema = pa.schema([
-        pa.field("vector", pa.list_(pa.float32(), 4)),
-        pa.field("item", pa.utf8()),
-        pa.field("price", pa.float32()),
-    ])
+Here's an example using using `RecordBatch` iterator for creating tables.
 
-    db.create_table("batched_tale", make_batches(), schema=schema)
-    ```
+```python
+import pyarrow as pa
 
-    You can also use iterators of other types like Pandas DataFrame or Pylists directly in the above example.
+def make_batches():
+    for i in range(5):
+        yield pa.RecordBatch.from_arrays(
+            [
+                pa.array([[3.1, 4.1, 5.1, 6.1], [5.9, 26.5, 4.7, 32.8]],
+                        pa.list_(pa.float32(), 4)),
+                pa.array(["foo", "bar"]),
+                pa.array([10.0, 20.0]),
+            ],
+            ["vector", "item", "price"],
+        )
+
+schema = pa.schema([
+    pa.field("vector", pa.list_(pa.float32(), 4)),
+    pa.field("item", pa.utf8()),
+    pa.field("price", pa.float32()),
+])
+
+db.create_table("batched_tale", make_batches(), schema=schema)
+```
+
+You can also use iterators of other types like Pandas DataFrame or Pylists directly in the above example.
 
 ## Open existing tables
 
@@ -329,23 +397,24 @@ This guide will show how to create tables, insert data into them, and update the
     tbl = db.open_table("my_table")
     ```
 
-=== "JavaScript"
+=== "Typescript[^1]"
+
     If you forget the name of your table, you can always get a listing of all table names.
 
-    ```javascript
+    ```typescript
     console.log(await db.tableNames());
     ```
 
     Then, you can open any existing tables.
 
-    ```javascript
+    ```typescript
     const tbl = await db.openTable("my_table");
     ```
 
 ## Creating empty table
+You can create an empty table for scenarios where you want to add data to the table later. An example would be when you want to collect data from a stream/external file and then add it to a table in batches.
 
 === "Python"
-    In Python, you can create an empty table for scenarios where you want to add data to the table later. An example would be when you want to collect data from a stream/external file and then add it to a table in batches.
 
     ```python
 
@@ -364,8 +433,8 @@ This guide will show how to create tables, insert data into them, and update the
     tbl = db.create_table("empty_table_add", schema=schema)
     ```
 
-    Alternatively, you can also use Pydantic to specify the schema for the empty table. Note that we do not 
-    directly import `pydantic` but instead use `lancedb.pydantic` which is a subclass of `pydantic.BaseModel` 
+    Alternatively, you can also use Pydantic to specify the schema for the empty table. Note that we do not
+    directly import `pydantic` but instead use `lancedb.pydantic` which is a subclass of `pydantic.BaseModel`
     that has been extended to support LanceDB specific types like `Vector`.
 
     ```python
@@ -382,9 +451,23 @@ This guide will show how to create tables, insert data into them, and update the
 
     Once the empty table has been created, you can add data to it via the various methods listed in the [Adding to a table](#adding-to-a-table) section.
 
+=== "Typescript[^1]"
+
+    === "@lancedb/lancedb"
+
+        ```typescript
+        --8<-- "nodejs/examples/basic.ts:create_empty_table"
+        ```
+
+    === "vectordb (deprecated)"
+
+        ```typescript
+        --8<-- "docs/src/basic_legacy.ts:create_empty_table"
+        ```
+
 ## Adding to a table
 
-After a table has been created, you can always add more data to it using the various methods available.
+After a table has been created, you can always add more data to it usind the `add` method
 
 === "Python"
     You can add any of the valid data structures accepted by LanceDB table, i.e, `dict`, `list[dict]`, `pd.DataFrame`, or `Iterator[pa.RecordBatch]`. Below are some examples.
@@ -452,8 +535,27 @@ After a table has been created, you can always add more data to it using the var
     tbl.add(pydantic_model_items)
     ```
 
+    ??? "Ingesting Pydantic models with LanceDB embedding API"
+        When using LanceDB's embedding API, you can add Pydantic models directly to the table. LanceDB will automatically convert the `vector` field to a vector before adding it to the table. You need to specify the default value of `vector` feild as None to allow LanceDB to automatically vectorize the data.
 
-=== "JavaScript"
+        ```python
+        import lancedb
+        from lancedb.pydantic import LanceModel, Vector
+        from lancedb.embeddings import get_registry
+
+        db = lancedb.connect("~/tmp")
+        embed_fcn = get_registry().get("huggingface").create(name="BAAI/bge-small-en-v1.5")
+
+        class Schema(LanceModel):
+            text: str = embed_fcn.SourceField()
+            vector: Vector(embed_fcn.ndims()) = embed_fcn.VectorField(default=None)
+
+        tbl = db.create_table("my_table", schema=Schema, mode="overwrite")
+        models = [Schema(text="hello"), Schema(text="world")]
+        tbl.add(models)
+        ```
+
+=== "Typescript[^1]"
 
     ```javascript
     await tbl.add(
@@ -509,15 +611,15 @@ Use the `delete()` method on tables to delete rows from a table. To choose which
     # 0  3  [5.0, 6.0]
     ```
 
-=== "JavaScript"
+=== "Typescript[^1]"
 
-    ```javascript
+    ```ts
     await tbl.delete('item = "fizz"')
     ```
 
     ### Deleting row with specific column value
 
-    ```javascript
+    ```ts
     const con = await lancedb.connect("./.lancedb")
     const data = [
       {id: 1, vector: [1, 2]},
@@ -531,7 +633,7 @@ Use the `delete()` method on tables to delete rows from a table. To choose which
 
     ### Delete from a list of values
 
-    ```javascript
+    ```ts
     const to_remove = [1, 5];
     await tbl.delete(`id IN (${to_remove.join(",")})`)
     await tbl.countRows() // Returns 1
@@ -588,26 +690,49 @@ This can be used to update zero to all rows depending on how many rows match the
     2  2  [10.0, 10.0]
     ```
 
-=== "JavaScript/Typescript"
+=== "Typescript[^1]"
 
-    API Reference: [vectordb.Table.update](../javascript/interfaces/Table.md/#update)
+    === "@lancedb/lancedb"
 
-    ```javascript
-    const lancedb = require("vectordb");
+        API Reference: [lancedb.Table.update](../js/classes/Table.md/#update)
 
-    const db = await lancedb.connect("./.lancedb");
+        ```ts
+        import * as lancedb from "@lancedb/lancedb";
 
-    const data = [
-      {x: 1, vector: [1, 2]},
-      {x: 2, vector: [3, 4]},
-      {x: 3, vector: [5, 6]},
-    ];
-    const tbl = await db.createTable("my_table", data)
+        const db = await lancedb.connect("./.lancedb");
 
-    await tbl.update({ where: "x = 2", values: {vector: [10, 10]} })
-    ```
+        const data = [
+            {x: 1, vector: [1, 2]},
+            {x: 2, vector: [3, 4]},
+            {x: 3, vector: [5, 6]},
+        ];
+        const tbl = await db.createTable("my_table", data)
 
-The `values` parameter is used to provide the new values for the columns as literal values. You can also use the `values_sql` / `valuesSql` parameter to provide SQL expressions for the new values. For example, you can use `values_sql="x + 1"` to increment the value of the `x` column by 1.
+        await tbl.update({vector: [10, 10]}, { where: "x = 2"})
+        ```
+
+    === "vectordb (deprecated)"
+
+        API Reference: [vectordb.Table.update](../javascript/interfaces/Table.md/#update)
+
+        ```ts
+        const lancedb = require("vectordb");
+
+        const db = await lancedb.connect("./.lancedb");
+
+        const data = [
+            {x: 1, vector: [1, 2]},
+            {x: 2, vector: [3, 4]},
+            {x: 3, vector: [5, 6]},
+        ];
+        const tbl = await db.createTable("my_table", data)
+
+        await tbl.update({ where: "x = 2", values: {vector: [10, 10]} })
+        ```
+
+#### Updating using a sql query
+
+  The `values` parameter is used to provide the new values for the columns as literal values. You can also use the `values_sql` / `valuesSql` parameter to provide SQL expressions for the new values. For example, you can use `values_sql="x + 1"` to increment the value of the `x` column by 1.
 
 === "Python"
 
@@ -626,16 +751,113 @@ The `values` parameter is used to provide the new values for the columns as lite
     2  3  [10.0, 10.0]
     ```
 
-=== "JavaScript/Typescript"
+=== "Typescript[^1]"
 
-    ```javascript
-    await tbl.update({ valuesSql: { x: "x + 1" } })
-    ```
+    === "@lancedb/lancedb"
+
+        Coming Soon!
+
+    === "vectordb (deprecated)"
+
+        ```ts
+        await tbl.update({ valuesSql: { x: "x + 1" } })
+        ```
 
 !!! info "Note"
 
     When rows are updated, they are moved out of the index. The row will still show up in ANN queries, but the query will not be as fast as it would be if the row was in the index. If you update a large proportion of rows, consider rebuilding the index afterwards.
 
+## Drop a table
+
+Use the `drop_table()` method on the database to remove a table.
+
+=== "Python"
+
+      ```python
+      --8<-- "python/python/tests/docs/test_basic.py:drop_table"
+      --8<-- "python/python/tests/docs/test_basic.py:drop_table_async"
+      ```
+
+      This permanently removes the table and is not recoverable, unlike deleting rows.
+      By default, if the table does not exist an exception is raised. To suppress this,
+      you can pass in `ignore_missing=True`.
+
+=== "TypeScript"
+
+      ```typescript
+      --8<-- "docs/src/basic_legacy.ts:drop_table"
+      ```
+
+      This permanently removes the table and is not recoverable, unlike deleting rows.
+      If the table does not exist an exception is raised.
+
+
+## Consistency
+
+In LanceDB OSS, users can set the `read_consistency_interval` parameter on connections to achieve different levels of read consistency. This parameter determines how frequently the database synchronizes with the underlying storage system to check for updates made by other processes. If another process updates a table, the database will not see the changes until the next synchronization.
+
+There are three possible settings for `read_consistency_interval`:
+
+1. **Unset (default)**: The database does not check for updates to tables made by other processes. This provides the best query performance, but means that clients may not see the most up-to-date data. This setting is suitable for applications where the data does not change during the lifetime of the table reference.
+2. **Zero seconds (Strong consistency)**: The database checks for updates on every read. This provides the strongest consistency guarantees, ensuring that all clients see the latest committed data. However, it has the most overhead. This setting is suitable when consistency matters more than having high QPS.
+3. **Custom interval (Eventual consistency)**: The database checks for updates at a custom interval, such as every 5 seconds. This provides eventual consistency, allowing for some lag between write and read operations. Performance wise, this is a middle ground between strong consistency and no consistency check. This setting is suitable for applications where immediate consistency is not critical, but clients should see updated data eventually.
+
+!!! tip "Consistency in LanceDB Cloud"
+
+    This is only tune-able in LanceDB OSS. In LanceDB Cloud, readers are always eventually consistent.
+
+=== "Python"
+
+    To set strong consistency, use `timedelta(0)`:
+
+    ```python
+    from datetime import timedelta
+    db = lancedb.connect("./.lancedb",. read_consistency_interval=timedelta(0))
+    table = db.open_table("my_table")
+    ```
+
+    For eventual consistency, use a custom `timedelta`:
+
+    ```python
+    from datetime import timedelta
+    db = lancedb.connect("./.lancedb", read_consistency_interval=timedelta(seconds=5))
+    table = db.open_table("my_table")
+    ```
+
+    By default, a `Table` will never check for updates from other writers. To manually check for updates you can use `checkout_latest`:
+
+    ```python
+    db = lancedb.connect("./.lancedb")
+    table = db.open_table("my_table")
+
+    # (Other writes happen to my_table from another process)
+
+    # Check for updates
+    table.checkout_latest()
+    ```
+
+=== "Typescript[^1]"
+
+    To set strong consistency, use `0`:
+
+    ```ts
+    const db = await lancedb.connect({ uri: "./.lancedb", readConsistencyInterval: 0 });
+    const table = await db.openTable("my_table");
+    ```
+
+    For eventual consistency, specify the update interval as seconds:
+
+    ```ts
+    const db = await lancedb.connect({ uri: "./.lancedb", readConsistencyInterval: 5 });
+    const table = await db.openTable("my_table");
+    ```
+
+<!-- Node doesn't yet support the version time travel: https://github.com/lancedb/lancedb/issues/1007
+    Once it does, we can show manual consistency check for Node as well.
+-->
+
 ## What's next?
 
-Learn the best practices on creating an ANN index and getting the most out of it.
+Learn the best practices on creating an ANN index and getting the most out of it.
+
+[^1]: The `vectordb` package is a legacy package that is  deprecated in favor of `@lancedb/lancedb`.  The `vectordb` package will continue to receive bug fixes and security updates until September 2024.  We recommend all new projects use `@lancedb/lancedb`.  See the [migration guide](migration.md) for more information.

docs/src/guides/tuning_retrievers/1_query_types.md

@@ -0,0 +1,131 @@
+## Improving retriever performance
+
+Try it yourself - <a href="https://colab.research.google.com/github/lancedb/lancedb/blob/main/docs/src/notebooks/lancedb_reranking.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"></a><br/>
+
+VectorDBs are used as retreivers in recommender or chatbot-based systems for retrieving relevant data based on user queries. For example, retriever is a critical component of Retrieval Augmented Generation (RAG) acrhitectures. In this section, we will discuss how to improve the performance of retrievers.
+
+There are serveral ways to improve the performance of retrievers. Some of the common techniques are:
+
+* Using different query types
+* Using hybrid search
+* Fine-tuning the embedding models
+* Using different embedding models
+
+Using different embedding models is something that's very specific to the use case and the data. So we will not discuss it here. In this section, we will discuss the first three techniques.
+
+
+!!! note "Note"
+    We'll be using a simple metric called "hit-rate" for evaluating the performance of the retriever across this guide. Hit-rate is the percentage of queries for which the retriever returned the correct answer in the top-k results. For example, if the retriever returned the correct answer in the top-3 results for 70% of the queries, then the hit-rate@3 is 0.7.
+
+
+## The dataset
+We'll be using a QA dataset generated using a LLama2 review paper. The dataset contains 221 query, context and answer triplets. The queries and answers are generated using GPT-4 based on a given query. Full script used to generate the dataset can be found on this [repo](https://github.com/lancedb/ragged). It can be downloaded from [here](https://github.com/AyushExel/assets/blob/main/data_qa.csv)
+
+### Using different query types
+Let's setup the embeddings and the dataset first. We'll use the LanceDB's `huggingface` embeddings integration for this guide. 
+
+```python
+import lancedb
+import pandas as pd
+from lancedb.embeddings import get_registry
+from lancedb.pydantic import Vector, LanceModel
+
+db = lancedb.connect("~/lancedb/query_types")
+df = pd.read_csv("data_qa.csv")
+
+embed_fcn = get_registry().get("huggingface").create(name="BAAI/bge-small-en-v1.")
+
+class Schema(LanceModel):
+    context: str = embed_fcn.SourceField()
+    vector: Vector(embed_fcn.ndims()) = embed_fcn.VectorField()
+
+table = db.create_table("qa", schema=Schema)
+table.add(df[["context"]].to_dict(orient="records"))
+
+queries = df["query"].tolist()
+```
+
+Now that we have the dataset and embeddings table set up, here's how you can run different query types on the dataset.
+
+* <b> Vector Search: </b>
+
+    ```python
+    table.search(quries[0], query_type="vector").limit(5).to_pandas()
+    ```
+    By default, LanceDB uses vector search query type for searching and it automatically converts the input query to a vector before searching when using embedding API. So, the following statement is equivalent to the above statement.
+
+    ```python
+    table.search(quries[0]).limit(5).to_pandas()
+    ```
+
+    Vector or semantic search is useful when you want to find documents that are similar to the query in terms of meaning.
+
+---
+
+* <b> Full-text Search: </b>
+    
+    FTS requires creating an index on the column you want to search on. `replace=True` will replace the existing index if it exists.
+    Once the index is created, you can search using the `fts` query type.
+    ```python
+    table.create_fts_index("context", replace=True)
+    table.search(quries[0], query_type="fts").limit(5).to_pandas()
+    ```
+
+    Full-text search is useful when you want to find documents that contain the query terms.
+
+---
+
+* <b> Hybrid Search: </b>
+
+    Hybrid search is a combination of vector and full-text search. Here's how you can run a hybrid search query on the dataset.
+    ```python
+    table.search(quries[0], query_type="hybrid").limit(5).to_pandas()
+    ```
+    Hybrid search requires a reranker to combine and rank the results from vector and full-text search. We'll cover reranking as a concept in the next section.
+
+    Hybrid search is useful when you want to combine the benefits of both vector and full-text search.
+
+    !!! note "Note"
+        By default, it uses `LinearCombinationReranker` that combines the scores from vector and full-text search using a weighted linear combination. It is the simplest reranker implementation available in LanceDB. You can also use other rerankers like `CrossEncoderReranker` or `CohereReranker` for reranking the results.
+        Learn more about rerankers [here](https://lancedb.github.io/lancedb/reranking/)
+
+    
+
+### Hit rate evaluation results
+
+Now that we have seen how to run different query types on the dataset, let's evaluate the hit-rate of each query type on the dataset.
+For brevity, the entire evaluation script is not shown here. You can find the complete evaluation and benchmarking utility scripts [here](https://github.com/lancedb/ragged).
+
+Here are the hit-rate results for the dataset:
+
+| Query Type | Hit-rate@5 |
+| --- | --- |
+| Vector Search | 0.640 |
+| Full-text Search | 0.595 |
+| Hybrid Search (w/ LinearCombinationReranker) | 0.645 |
+
+**Choosing query type** is very specific to the use case and the data. This synthetic dataset has been generated to be semantically challenging, i.e, the queries don't have a lot of keywords in common with the context. So, vector search performs better than full-text search. However, in real-world scenarios, full-text search might perform better than vector search. Hybrid search is a good choice when you want to combine the benefits of both vector and full-text search.
+
+### Evaluation results on other datasets
+
+The hit-rate results can vary based on the dataset and the query type. Here are the hit-rate results for the other datasets using the same embedding function.
+
+* <b> SQuAD Dataset: </b>
+
+    | Query Type | Hit-rate@5 |
+    | --- | --- |
+    | Vector Search | 0.822 |
+    | Full-text Search | 0.835 |
+    | Hybrid Search (w/ LinearCombinationReranker) | 0.8874 |
+
+* <b> Uber10K sec filing Dataset: </b>
+
+    | Query Type | Hit-rate@5 |
+    | --- | --- |
+    | Vector Search | 0.608 |
+    | Full-text Search | 0.82 |
+    | Hybrid Search (w/ LinearCombinationReranker) | 0.80 |
+
+In these standard datasets, FTS seems to perform much better than vector search because the queries have a lot of keywords in common with the context. So, in general choosing the query type is very specific to the use case and the data.
+
+

docs/src/guides/tuning_retrievers/2_reranking.md

@@ -0,0 +1,80 @@
+Continuing from the previous section, we can now rerank the results using more complex rerankers.
+
+Try it yourself - <a href="https://colab.research.google.com/github/lancedb/lancedb/blob/main/docs/src/notebooks/lancedb_reranking.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"></a><br/>
+
+## Reranking search results
+You can rerank any search results using a reranker. The syntax for reranking is as follows:
+
+```python
+from lancedb.rerankers import LinearCombinationReranker
+
+reranker = LinearCombinationReranker()
+table.search(quries[0], query_type="hybrid").rerank(reranker=reranker).limit(5).to_pandas()
+```
+Based on the `query_type`, the `rerank()` function can accept other arguments as well. For example, hybrid search accepts a `normalize` param to determine the score normalization method.
+
+!!! note "Note"
+    LanceDB provides a `Reranker` base class that can be extended to implement custom rerankers. Each reranker must implement the `rerank_hybrid` method. `rerank_vector` and `rerank_fts` methods are optional. For example, the `LinearCombinationReranker` only implements the `rerank_hybrid` method and so it can only be used for reranking hybrid search results.
+
+## Choosing a Reranker
+There are many rerankers available in LanceDB like `CrossEncoderReranker`, `CohereReranker`, and `ColBERT`. The choice of reranker depends on the dataset and the application. You can even implement you own custom reranker by extending the `Reranker` class. For more details about each available reranker and performance comparison, refer to the [rerankers](https://lancedb.github.io/lancedb/reranking/) documentation.
+
+In this example, we'll use the `CohereReranker` to rerank the search results. It requires  `cohere` to be installed and `COHERE_API_KEY` to be set in the environment. To get your API key, sign up on [Cohere](https://cohere.ai/).
+
+```python
+from lancedb.rerankers import CohereReranker
+
+# use Cohere reranker v3
+reranker = CohereReranker(model_name="rerank-english-v3.0") # default model is "rerank-english-v2.0"
+```
+
+### Reranking search results
+Now we can rerank all query type results using the `CohereReranker`:
+
+```python
+
+# rerank hybrid search results
+table.search(quries[0], query_type="hybrid").rerank(reranker=reranker).limit(5).to_pandas()
+
+# rerank vector search results
+table.search(quries[0], query_type="vector").rerank(reranker=reranker).limit(5).to_pandas()
+
+# rerank fts search results
+table.search(quries[0], query_type="fts").rerank(reranker=reranker).limit(5).to_pandas()
+```
+
+Each reranker can accept additional arguments. For example, `CohereReranker` accepts `top_k` and `batch_size` params to control the number of documents to rerank and the batch size for reranking respectively. Similarly, a custom reranker can accept any number of arguments based on the implementation. For example, a reranker can accept a `filter` that implements some custom logic to filter out documents before reranking.
+
+## Results
+
+Let us take a look at the same datasets from the previous sections, using the same embedding table but with Cohere reranker applied to all query types.
+
+!!! note "Note"
+    When reranking fts or vector search results, the search results are over-fetched by a factor of 2 and then reranked. From the reranked set, `top_k` (5 in this case) results are taken. This is done because reranking will have no effect on the hit-rate if we only fetch the `top_k` results.
+
+### Synthetic LLama2 paper dataset
+
+| Query Type | Hit-rate@5 |
+| --- | --- |
+| Vector |  0.640 |
+| FTS   |  0.595  |
+| Reranked vector | 0.677    |
+| Reranked fts  | 0.672    |
+| Hybrid | 0.759 |
+
+### SQuAD Dataset
+
+
+### Uber10K sec filing Dataset
+
+| Query Type | Hit-rate@5 |
+| --- | --- |
+| Vector |  0.608 |
+| FTS   |  0.824  |
+| Reranked vector | 0.671    |
+| Reranked fts  | 0.843    |
+| Hybrid | 0.849 |
+
+
+
+

docs/src/guides/tuning_retrievers/3_embed_tuning.md

@@ -0,0 +1,82 @@
+## Finetuning the Embedding Model
+Try it yourself - <a href="https://colab.research.google.com/github/lancedb/lancedb/blob/main/docs/src/notebooks/embedding_tuner.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"></a><br/>
+
+Another way to improve retriever performance is to fine-tune the embedding model itself. Fine-tuning the embedding model can help in learning better representations for the documents and queries in the dataset. This can be particularly useful when the dataset is very different from the pre-trained data used to train the embedding model.
+
+We'll use the same dataset as in the previous sections. Start off by splitting the dataset into training and validation sets:
+```python
+from sklearn.model_selection import train_test_split
+
+train_df, validation_df = train_test_split("data_qa.csv", test_size=0.2, random_state=42)
+
+train_df.to_csv("data_train.csv", index=False)
+validation_df.to_csv("data_val.csv", index=False)
+```
+
+You can use any tuning API to fine-tune embedding models. In this example, we'll utilise Llama-index as it also comes with utilities for synthetic data generation and training the model. 
+
+
+Then parse the dataset as llama-index text nodes and generate synthetic QA pairs from each node.
+```python
+from llama_index.core.node_parser import SentenceSplitter
+from llama_index.readers.file import PagedCSVReader
+from llama_index.finetuning import generate_qa_embedding_pairs
+from llama_index.core.evaluation import EmbeddingQAFinetuneDataset
+
+def load_corpus(file):
+    loader = PagedCSVReader(encoding="utf-8")
+    docs = loader.load_data(file=Path(file))
+
+    parser = SentenceSplitter()
+    nodes = parser.get_nodes_from_documents(docs)
+
+    return nodes
+
+from llama_index.llms.openai import OpenAI
+
+
+train_dataset = generate_qa_embedding_pairs(
+    llm=OpenAI(model="gpt-3.5-turbo"), nodes=train_nodes, verbose=False
+)
+val_dataset = generate_qa_embedding_pairs(
+    llm=OpenAI(model="gpt-3.5-turbo"), nodes=val_nodes, verbose=False
+)
+```
+
+Now we'll use `SentenceTransformersFinetuneEngine` engine to fine-tune the model. You can also use `sentence-transformers` or `transformers` library to fine-tune the model. 
+
+```python
+from llama_index.finetuning import SentenceTransformersFinetuneEngine
+
+finetune_engine = SentenceTransformersFinetuneEngine(
+    train_dataset,
+    model_id="BAAI/bge-small-en-v1.5",
+    model_output_path="tuned_model",
+    val_dataset=val_dataset,
+)
+finetune_engine.finetune()
+embed_model = finetune_engine.get_finetuned_model()
+```
+This saves the fine tuned embedding model in `tuned_model` folder. This al
+
+# Evaluation results
+In order to eval the retriever, you can either use this model to ingest the data into LanceDB directly or llama-index's LanceDB integration to create a `VectorStoreIndex` and use it as a retriever. 
+On performing the same hit-rate evaluation as before, we see a significant improvement in the hit-rate across all query types.
+
+### Baseline
+| Query Type | Hit-rate@5 |
+| --- | --- |
+| Vector Search | 0.640 |
+| Full-text Search | 0.595 |
+| Reranked Vector Search | 0.677 |
+| Reranked Full-text Search | 0.672 |
+| Hybrid Search (w/ CohereReranker) | 0.759|
+
+### Fine-tuned model ( 2 iterations )
+| Query Type | Hit-rate@5 |
+| --- | --- |
+| Vector Search | 0.672 |
+| Full-text Search | 0.595 |
+| Reranked Vector Search | 0.754 |
+| Reranked Full-text Search | 0.672|
+| Hybrid Search (w/ CohereReranker) | 0.768 |

docs/src/hybrid_search/eval.md

@@ -0,0 +1,51 @@
+# Hybrid Search
+
+Hybrid Search is a broad (often misused) term. It can mean anything from combining multiple methods for searching, to applying ranking methods to better sort the results. In this blog, we use the definition of "hybrid search" to mean using a combination of keyword-based and vector search.
+
+## The challenge of (re)ranking search results
+Once you have a group of the most relevant search results from multiple search sources, you'd likely standardize the score and rank them accordingly. This process can also be seen as another independent step - reranking.
+There are two approaches for reranking search results from multiple sources.
+
+* <b>Score-based</b>: Calculate final relevance scores based on a weighted linear combination of individual search algorithm scores. Example - Weighted linear combination of semantic search & keyword-based search results.
+
+* <b>Relevance-based</b>: Discards the existing scores and calculates the relevance of each search result - query pair. Example - Cross Encoder models
+
+Even though there are many strategies for reranking search results, none works for all cases. Moreover, evaluating them itself is a challenge. Also, reranking can be dataset, application specific so it's hard to generalize.
+
+### Example evaluation of hybrid search with Reranking
+
+Here's some evaluation numbers from experiment comparing these re-rankers on about 800 queries. It is modified version of an evaluation script from [llama-index](https://github.com/run-llama/finetune-embedding/blob/main/evaluate.ipynb) that measures hit-rate at top-k.
+
+<b> With OpenAI ada2 embedding </b>
+
+Vector Search baseline - `0.64`
+
+| Reranker | Top-3 | Top-5 | Top-10 |
+| --- | --- | --- | --- |
+| Linear Combination | `0.73` | `0.74` | `0.85` |
+| Cross Encoder | `0.71` | `0.70` | `0.77` |
+| Cohere | `0.81` | `0.81` | `0.85` |
+| ColBERT | `0.68` | `0.68` | `0.73` |
+
+<p>
+<img src="https://github.com/AyushExel/assets/assets/15766192/d57b1780-ef27-414c-a5c3-73bee7808a45">
+</p>
+
+<b> With OpenAI embedding-v3-small </b>
+
+Vector Search baseline - `0.59`
+
+| Reranker | Top-3 | Top-5 | Top-10 |
+| --- | --- | --- | --- |
+| Linear Combination | `0.68` | `0.70` | `0.84` |
+| Cross Encoder | `0.72` | `0.72` | `0.79` |
+| Cohere | `0.79` | `0.79` | `0.84` |
+| ColBERT | `0.70` | `0.70` | `0.76` |
+
+<p>
+<img src="https://github.com/AyushExel/assets/assets/15766192/259adfd2-6ec6-4df6-a77d-1456598970dd">
+</p>
+
+### Conclusion
+
+The results show that the reranking methods are able to improve the search results. However, the improvement is not consistent across all rerankers. The choice of reranker depends on the dataset and the application. It is also important to note that the reranking methods are not a replacement for the search methods. They are complementary and should be used together to get the best results. The speed to recall tradeoff is also an important factor to consider when choosing the reranker.

docs/src/hybrid_search/hybrid_search.md

@@ -1,22 +1,29 @@
 # Hybrid Search
 
-LanceDB supports both semantic and keyword-based search. In real world applications, it is often useful to combine these two approaches to get the best best results. For example, you may want to search for a document that is semantically similar to a query document, but also contains a specific keyword. This is an example of *hybrid search*, a search algorithm that combines multiple search techniques.
+LanceDB supports both semantic and keyword-based search (also termed full-text search, or FTS). In real world applications, it is often useful to combine these two approaches to get the best best results. For example, you may want to search for a document that is semantically similar to a query document, but also contains a specific keyword. This is an example of *hybrid search*, a search algorithm that combines multiple search techniques.
 
 ## Hybrid search in LanceDB
 You can perform hybrid search in LanceDB by combining the results of semantic and full-text search via a reranking algorithm of your choice. LanceDB provides multiple rerankers out of the box. However, you can always write a custom reranker if your use case need more sophisticated logic .
 
 ```python
+import os
+
 import lancedb
+import openai
 from lancedb.embeddings import get_registry
-from lancedb.pydanatic import LanceModel, Vector
+from lancedb.pydantic import LanceModel, Vector
 
 db = lancedb.connect("~/.lancedb")
 
 # Ingest embedding function in LanceDB table
+# Configuring the environment variable OPENAI_API_KEY
+if "OPENAI_API_KEY" not in os.environ:
+# OR set the key here as a variable
+    openai.api_key = "sk-..."
 embeddings = get_registry().get("openai").create()
 
 class Documents(LanceModel):
-    vector: Vector(embeddings.ndims) = embeddings.VectorField()
+    vector: Vector(embeddings.ndims()) = embeddings.VectorField()
     text: str = embeddings.SourceField()
 
 table = db.create_table("documents", schema=Documents)
@@ -31,17 +38,19 @@ data = [
 # ingest docs with auto-vectorization
 table.add(data)
 
+# Create a fts index before the hybrid search
+table.create_fts_index("text")
 # hybrid search with default re-ranker
 results = table.search("flower moon", query_type="hybrid").to_pandas()
 ```
 
-By default, LanceDB uses `LinearCombinationReranker(weights=0.7)` to combine and rerank the results of semantic and full-text search. You can customize the hyperparameters as needed or write your own custom reranker. Here's how you can use any of the available rerankers:
+By default, LanceDB uses `LinearCombinationReranker(weight=0.7)` to combine and rerank the results of semantic and full-text search. You can customize the hyperparameters as needed or write your own custom reranker. Here's how you can use any of the available rerankers:
 
 
 ### `rerank()` arguments
 * `normalize`: `str`, default `"score"`:
     The method to normalize the scores. Can be "rank" or "score". If "rank", the scores are converted to ranks and then normalized. If "score", the scores are normalized directly.
-* `reranker`: `Reranker`, default `LinearCombinationReranker(weights=0.7)`.
+* `reranker`: `Reranker`, default `LinearCombinationReranker(weight=0.7)`.
     The reranker to use. If not specified, the default reranker is used.
 
 
@@ -55,12 +64,12 @@ This is the default re-ranker used by LanceDB. It combines the results of semant
 ```python
 from lancedb.rerankers import LinearCombinationReranker
 
-reranker = LinearCombinationReranker(weights=0.3) # Use 0.3 as the weight for vector search
+reranker = LinearCombinationReranker(weight=0.3) # Use 0.3 as the weight for vector search
 
 results = table.search("rebel", query_type="hybrid").rerank(reranker=reranker).to_pandas()
 ```
 
-Arguments
+### Arguments
 ----------------
 * `weight`: `float`, default `0.7`:
     The weight to use for the semantic search score. The weight for the full-text search score is `1 - weights`.
@@ -82,9 +91,9 @@ reranker = CohereReranker()
 results = table.search("vampire weekend", query_type="hybrid").rerank(reranker=reranker).to_pandas()
 ```
 
-Arguments
+### Arguments
 ----------------
-* `model_name`` : str, default `"rerank-english-v2.0"``
+* `model_name` : str, default `"rerank-english-v2.0"`
         The name of the cross encoder model to use. Available cohere models are:
         - rerank-english-v2.0
         - rerank-multilingual-v2.0
@@ -108,7 +117,7 @@ results = table.search("harmony hall", query_type="hybrid").rerank(reranker=rera
 ```
 
 
-Arguments
+### Arguments
 ----------------
 * `model` : str, default `"cross-encoder/ms-marco-TinyBERT-L-6"`
         The name of the cross encoder model to use. Available cross encoder models can be found [here](https://www.sbert.net/docs/pretrained_cross-encoders.html)
@@ -121,6 +130,61 @@ Arguments
     Only returns `_relevance_score`. Does not support `return_score = "all"`.
 
 
+### ColBERT Reranker
+This reranker uses the ColBERT model to combine the results of semantic and full-text search. You can use it by passing `ColbertrReranker()` to the `rerank()` method. 
+
+ColBERT reranker model calculates relevance of given docs against the query and don't take existing fts and vector search scores into account, so it currently only supports `return_score="relevance"`. By default, it looks for `text` column to rerank the results. But you can specify the column name to use as input to the cross encoder model as described below.
+
+```python
+from lancedb.rerankers import ColbertReranker
+
+reranker = ColbertReranker()
+
+results = table.search("harmony hall", query_type="hybrid").rerank(reranker=reranker).to_pandas()
+```
+
+### Arguments
+----------------
+* `model_name` : `str`, default `"colbert-ir/colbertv2.0"`
+        The name of the cross encoder model to use.
+* `column` : `str`, default `"text"`
+        The name of the column to use as input to the cross encoder model.
+* `return_score` : `str`, default `"relevance"`
+        options are `"relevance"` or `"all"`. Only `"relevance"` is supported for now.
+
+!!! Note
+    Only returns `_relevance_score`. Does not support `return_score = "all"`.
+
+### OpenAI Reranker
+This reranker uses the OpenAI API to combine the results of semantic and full-text search. You can use it by passing `OpenaiReranker()` to the `rerank()` method.
+
+!!! Note
+    This prompts chat model to rerank results which is not a dedicated reranker model. This should be treated as experimental.
+
+!!! Tip
+    - You might run out of token limit so set the search `limits` based on your token limit.
+    - It is recommended to use gpt-4-turbo-preview, the default model, older models might lead to undesired behaviour
+
+```python
+from lancedb.rerankers import OpenaiReranker
+
+reranker = OpenaiReranker()
+
+results = table.search("harmony hall", query_type="hybrid").rerank(reranker=reranker).to_pandas()
+```
+
+### Arguments
+----------------
+* `model_name` : `str`, default `"gpt-4-turbo-preview"`
+    The name of the cross encoder model to use.
+* `column` : `str`, default `"text"`
+    The name of the column to use as input to the cross encoder model.
+* `return_score` : `str`, default `"relevance"`
+    options are "relevance" or "all". Only "relevance" is supported for now.
+* `api_key` : `str`, default `None`
+    The API key to use. If None, will use the OPENAI_API_KEY environment variable.
+
+
 ## Building Custom Rerankers
 You can build your own custom reranker by subclassing the `Reranker` class and implementing the `rerank_hybrid()` method. Here's an example of a custom reranker that combines the results of semantic and full-text search using a linear combination of the scores.
 
@@ -137,7 +201,7 @@ class MyReranker(Reranker):
         self.param1 = param1
         self.param2 = param2
     
-    def rerank_hybrid(self, vector_results: pa.Table, fts_results: pa.Table):
+    def rerank_hybrid(self, query: str, vector_results: pa.Table, fts_results: pa.Table):
         # Use the built-in merging function
         combined_result = self.merge_results(vector_results, fts_results)
         
@@ -149,24 +213,30 @@ class MyReranker(Reranker):
 
 ```
 
-You can also accept additional arguments like a filter along with fts and vector search results
+### Example of a Custom Reranker
+For the sake of simplicity let's build custom reranker that just enchances the Cohere Reranker by accepting a filter query, and accept other CohereReranker params as kwags.
 
 ```python
 
-from lancedb.rerankers import Reranker
-import pyarrow as pa
+from typing import List, Union
+import pandas as pd
+from lancedb.rerankers import CohereReranker
 
-class MyReranker(Reranker):
-    ...
-    
-    def rerank_hybrid(self, vector_results: pa.Table, fts_results: pa.Table, filter: str):
-        # Use the built-in merging function
-        combined_result = self.merge_results(vector_results, fts_results)
-        
-        # Do something with the combined results & filter
-        # ...
+class MofidifiedCohereReranker(CohereReranker):
+    def __init__(self, filters: Union[str, List[str]], **kwargs):
+        super().__init__(**kwargs)
+        filters = filters if isinstance(filters, list) else [filters]
+        self.filters = filters
 
-        # Return the combined results
-        return combined_result
+    def rerank_hybrid(self, query: str, vector_results: pa.Table, fts_results: pa.Table)-> pa.Table:
+        combined_result = super().rerank_hybrid(query, vector_results, fts_results)
+        df = combined_result.to_pandas()
+        for filter in self.filters:
+            df = df.query("not text.str.contains(@filter)")
+
+        return pa.Table.from_pandas(df)
 
 ```
+
+!!! tip
+    The `vector_results` and `fts_results` are pyarrow tables. You can convert them to pandas dataframes using `to_pandas()` method and perform any operations you want. After you are done, you can convert the dataframe back to pyarrow table using `pa.Table.from_pandas()` method and return it.

docs/src/index.md

@@ -28,7 +28,7 @@ LanceDB **Cloud** is a SaaS (software-as-a-service) solution that runs serverles
 
 * Fast production-scale vector similarity, full-text & hybrid search and a SQL query interface (via [DataFusion](https://github.com/apache/arrow-datafusion))
 
-* Native Python and Javascript/Typescript support
+* Python, Javascript/Typescript, and Rust support
 
 * Store, query & manage multi-modal data (text, images, videos, point clouds, etc.), not just the embeddings and metadata
 
@@ -54,3 +54,4 @@ The following pages go deeper into the internal of LanceDB and how to use it.
 * [Ecosystem Integrations](integrations/index.md): Integrate LanceDB with other tools in the data ecosystem
 * [Python API Reference](python/python.md): Python OSS and Cloud API references
 * [JavaScript API Reference](javascript/modules.md): JavaScript OSS and Cloud API references
+* [Rust API Reference](https://docs.rs/lancedb/latest/lancedb/index.html): Rust API reference

docs/src/integrations/index.md

@@ -13,7 +13,7 @@ Get started using these examples and quick links.
 | Integrations | |
 |---|---:|
 | <h3> LlamaIndex </h3>LlamaIndex is a simple, flexible data framework for connecting custom data sources to large language models. Llama index integrates with LanceDB as the serverless VectorDB. <h3>[Lean More](https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/LanceDBIndexDemo.html) </h3> |<img src="../assets/llama-index.jpg" alt="image" width="150" height="auto">|
-| <h3>Langchain</h3>Langchain allows building applications with LLMs through composability <h3>[Lean More](https://python.langchain.com/en/latest/modules/indexes/vectorstores/examples/lancedb.html) | <img src="../assets/langchain.png" alt="image" width="150" height="auto">|
+| <h3>Langchain</h3>Langchain allows building applications with LLMs through composability <h3>[Lean More](https://lancedb.github.io/lancedb/integrations/langchain/) | <img src="../assets/langchain.png" alt="image" width="150" height="auto">|
 | <h3>Langchain TS</h3> Javascript bindings for Langchain. It integrates with LanceDB's serverless vectordb allowing you to build powerful AI applications through composibility using only serverless functions. <h3>[Learn More]( https://js.langchain.com/docs/modules/data_connection/vectorstores/integrations/lancedb) | <img src="../assets/langchain.png" alt="image" width="150" height="auto">|
 | <h3>Voxel51</h3>  It is an open source toolkit that enables you to build better computer vision workflows by improving the quality of your datasets and delivering insights about your models.<h3>[Learn More](./voxel51.md) | <img src="../assets/voxel.gif" alt="image" width="150" height="auto">|
 | <h3>PromptTools</h3>  Offers a set of free, open-source tools for testing and experimenting with models, prompts, and configurations. The core idea is to enable developers to evaluate prompts using familiar interfaces like code and notebooks. You can use it to experiment with different configurations of LanceDB, and test how LanceDB integrates with the LLM of your choice.<h3>[Learn More](./prompttools.md) | <img src="../assets/prompttools.jpeg" alt="image" width="150" height="auto">|

docs/src/integrations/langchain.md

@@ -0,0 +1,201 @@
+# Langchain
+![Illustration](../assets/langchain.png)
+
+## Quick Start
+You can load your document data using langchain's loaders, for this example we are using `TextLoader` and `OpenAIEmbeddings` as the embedding model. Checkout Complete example here - [LangChain demo](../notebooks/langchain_example.ipynb)
+```python
+import os
+from langchain.document_loaders import TextLoader
+from langchain.vectorstores import LanceDB
+from langchain_openai import OpenAIEmbeddings
+from langchain_text_splitters import CharacterTextSplitter
+
+os.environ["OPENAI_API_KEY"] = "sk-..."
+
+loader = TextLoader("../../modules/state_of_the_union.txt") # Replace with your data path
+documents = loader.load()
+
+documents = CharacterTextSplitter().split_documents(documents)
+embeddings = OpenAIEmbeddings()
+
+docsearch = LanceDB.from_documents(documents, embeddings)
+query = "What did the president say about Ketanji Brown Jackson"
+docs = docsearch.similarity_search(query)
+print(docs[0].page_content)
+```
+
+## Documentation
+In the above example `LanceDB` vector store class object is created using `from_documents()` method  which is a `classmethod` and returns the initialized class object. 
+You can also use `LanceDB.from_texts(texts: List[str],embedding: Embeddings)` class method.  
+
+The exhaustive list of parameters for `LanceDB` vector store are :  
+- `connection`: (Optional) `lancedb.db.LanceDBConnection` connection object to use.  If not provided, a new connection will be created.  
+- `embedding`: Langchain embedding model.  
+- `vector_key`: (Optional) Column name to use for vector's in the table. Defaults to `'vector'`.   
+- `id_key`: (Optional) Column name to use for id's in the table. Defaults to `'id'`.  
+- `text_key`: (Optional) Column name to use for text in the table. Defaults to `'text'`.  
+- `table_name`: (Optional) Name of your table in the database. Defaults to `'vectorstore'`.  
+- `api_key`: (Optional) API key to use for LanceDB cloud database. Defaults to `None`.  
+- `region`: (Optional) Region to use for LanceDB cloud database. Only for LanceDB Cloud, defaults to `None`.  
+- `mode`: (Optional) Mode to use for adding data to the table. Defaults to `'overwrite'`.  
+- `reranker`: (Optional) The reranker to use for LanceDB.
+- `relevance_score_fn`: (Optional[Callable[[float], float]]) Langchain relevance score function to be used. Defaults to `None`. 
+
+```python
+db_url = "db://lang_test" # url of db you created
+api_key = "xxxxx" # your API key
+region="us-east-1-dev"  # your selected region
+
+vector_store = LanceDB(
+    uri=db_url,
+    api_key=api_key, #(dont include for local API)
+    region=region, #(dont include for local API)
+    embedding=embeddings,
+    table_name='langchain_test' #Optional
+    )
+```
+
+### Methods 
+
+##### add_texts()
+- `texts`: `Iterable` of strings to add to the vectorstore.
+- `metadatas`: Optional `list[dict()]` of metadatas associated with the texts.
+- `ids`: Optional `list` of ids to associate with the texts. 
+- `kwargs`: `Any`
+
+This method adds texts and stores respective embeddings automatically.
+
+```python
+vector_store.add_texts(texts = ['test_123'], metadatas =[{'source' :'wiki'}]) 
+
+#Additionaly, to explore the table you can load it into a df or save it in a csv file:
+
+tbl = vector_store.get_table()
+print("tbl:", tbl)
+pd_df = tbl.to_pandas()
+pd_df.to_csv("docsearch.csv", index=False)
+
+# you can also create a new vector store object using an older connection object:
+vector_store = LanceDB(connection=tbl, embedding=embeddings)
+```
+##### create_index() 
+- `col_name`: `Optional[str] = None`
+- `vector_col`: `Optional[str] = None`
+- `num_partitions`: `Optional[int] = 256`
+- `num_sub_vectors`: `Optional[int] = 96`
+- `index_cache_size`: `Optional[int] = None`
+
+This method creates an index for the vector store. For index creation make sure your table has enough data in it. An ANN index is ususally not needed for datasets ~100K vectors. For large-scale (>1M) or higher dimension vectors, it is beneficial to create an ANN index.
+
+```python
+# for creating vector index
+vector_store.create_index(vector_col='vector', metric = 'cosine')
+
+# for creating scalar index(for non-vector columns)
+vector_store.create_index(col_name='text')
+
+```
+
+##### similarity_search()
+- `query`: `str`
+- `k`: `Optional[int] = None`
+- `filter`: `Optional[Dict[str, str]] = None`
+- `fts`: `Optional[bool] = False`
+- `name`: `Optional[str] = None`
+- `kwargs`: `Any`
+
+Return documents most similar to the query without relevance scores
+
+```python
+docs = docsearch.similarity_search(query)
+print(docs[0].page_content)
+```
+
+##### similarity_search_by_vector()
+- `embedding`: `List[float]`
+- `k`: `Optional[int] = None`
+- `filter`: `Optional[Dict[str, str]] = None`
+- `name`: `Optional[str] = None`
+- `kwargs`: `Any`
+
+Returns documents most similar to the query vector.
+
+```python
+docs = docsearch.similarity_search_by_vector(query)
+print(docs[0].page_content)
+```
+
+##### similarity_search_with_score()
+- `query`: `str`
+- `k`: `Optional[int] = None`
+- `filter`: `Optional[Dict[str, str]] = None`
+- `kwargs`: `Any`
+
+Returns documents most similar to the query string with relevance scores, gets called by base class's `similarity_search_with_relevance_scores` which selects relevance score based on our `_select_relevance_score_fn`.
+
+```python
+docs = docsearch.similarity_search_with_relevance_scores(query)
+print("relevance score - ", docs[0][1])
+print("text- ", docs[0][0].page_content[:1000])
+```
+
+##### similarity_search_by_vector_with_relevance_scores()
+- `embedding`: `List[float]`
+- `k`: `Optional[int] = None`
+- `filter`: `Optional[Dict[str, str]] = None`
+- `name`: `Optional[str] = None`
+- `kwargs`: `Any`
+
+Return documents most similar to the query vector with relevance scores.
+Relevance score 
+
+```python
+docs = docsearch.similarity_search_by_vector_with_relevance_scores(query_embedding)
+print("relevance score - ", docs[0][1])
+print("text- ", docs[0][0].page_content[:1000])
+```
+
+##### max_marginal_relevance_search()
+- `query`: `str`
+- `k`: `Optional[int] = None`
+- `fetch_k` : Number of Documents to fetch to pass to MMR algorithm, `Optional[int] = None`
+- `lambda_mult`: Number between 0 and 1 that determines the degree
+                        of diversity among the results with 0 corresponding
+                        to maximum diversity and 1 to minimum diversity.
+                        Defaults to 0.5. `float = 0.5`
+- `filter`: `Optional[Dict[str, str]] = None`
+- `kwargs`: `Any`
+
+Returns docs selected using the maximal marginal relevance(MMR).
+Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents.
+
+Similarly, `max_marginal_relevance_search_by_vector()` function returns docs most similar to the embedding passed to the function using MMR. instead of a string query you need to pass the embedding to be searched for. 
+
+```python
+result = docsearch.max_marginal_relevance_search(
+        query="text"
+    )
+result_texts = [doc.page_content for doc in result]
+print(result_texts)
+
+## search by vector :
+result = docsearch.max_marginal_relevance_search_by_vector(
+        embeddings.embed_query("text")
+    )
+result_texts = [doc.page_content for doc in result]
+print(result_texts)
+```
+
+##### add_images()
+- `uris` : File path to the image. `List[str]`.
+- `metadatas` : Optional list of metadatas. `(Optional[List[dict]], optional)`
+- `ids` : Optional list of IDs. `(Optional[List[str]], optional)`
+
+Adds images by automatically creating their embeddings and adds them to the vectorstore.
+
+```python
+vec_store.add_images(uris=image_uris) 
+# here image_uris are local fs paths to the images.
+```
+
+

docs/src/integrations/llamaIndex.md

@@ -0,0 +1,142 @@
+# Llama-Index
+![Illustration](../assets/llama-index.jpg)
+
+## Quick start
+You would need to install the integration via `pip install llama-index-vector-stores-lancedb` in order to use it. 
+You can run the below script to try it out :
+```python
+import logging
+import sys
+
+# Uncomment to see debug logs
+# logging.basicConfig(stream=sys.stdout, level=logging.DEBUG)
+# logging.getLogger().addHandler(logging.StreamHandler(stream=sys.stdout))
+
+from llama_index.core import SimpleDirectoryReader, Document, StorageContext
+from llama_index.core import VectorStoreIndex
+from llama_index.vector_stores.lancedb import LanceDBVectorStore
+import textwrap
+import openai
+
+openai.api_key = "sk-..."
+
+documents = SimpleDirectoryReader("./data/your-data-dir/").load_data()
+print("Document ID:", documents[0].doc_id, "Document Hash:", documents[0].hash)
+
+## For LanceDB cloud :
+# vector_store = LanceDBVectorStore( 
+#     uri="db://db_name", # your remote DB URI
+#     api_key="sk_..", # lancedb cloud api key
+#     region="your-region" # the region you configured
+#     ...
+# )
+
+vector_store = LanceDBVectorStore(
+    uri="./lancedb", mode="overwrite", query_type="vector"
+)
+storage_context = StorageContext.from_defaults(vector_store=vector_store)
+
+index = VectorStoreIndex.from_documents(
+    documents, storage_context=storage_context
+)
+lance_filter = "metadata.file_name = 'paul_graham_essay.txt' "
+retriever = index.as_retriever(vector_store_kwargs={"where": lance_filter})
+response = retriever.retrieve("What did the author do growing up?")
+```
+
+Checkout Complete example here - [LlamaIndex demo](../notebooks/LlamaIndex_example.ipynb)
+
+### Filtering
+For metadata filtering, you can use a Lance SQL-like string filter as demonstrated in the example above. Additionally, you can also filter using the `MetadataFilters` class from LlamaIndex:
+```python
+from llama_index.core.vector_stores import (
+    MetadataFilters,
+    FilterOperator,
+    FilterCondition,
+    MetadataFilter,
+)
+
+query_filters = MetadataFilters(
+    filters=[
+        MetadataFilter(
+            key="creation_date", operator=FilterOperator.EQ, value="2024-05-23"
+        ),
+        MetadataFilter(
+            key="file_size", value=75040, operator=FilterOperator.GT
+        ),
+    ],
+    condition=FilterCondition.AND,
+)
+```
+
+### Hybrid Search
+For complete documentation, refer [here](https://lancedb.github.io/lancedb/hybrid_search/hybrid_search/). This example uses the `colbert` reranker. Make sure to install necessary dependencies for the reranker you choose.
+```python
+from lancedb.rerankers import ColbertReranker
+
+reranker = ColbertReranker()
+vector_store._add_reranker(reranker)
+
+query_engine = index.as_query_engine(
+    filters=query_filters,
+    vector_store_kwargs={
+        "query_type": "hybrid",
+    }
+)
+
+response = query_engine.query("How much did Viaweb charge per month?")
+```
+
+In the above snippet, you can change/specify query_type again when creating the engine/retriever.
+
+## API reference
+The exhaustive list of parameters for `LanceDBVectorStore` vector store are :  
+- `connection`: Optional, `lancedb.db.LanceDBConnection` connection object to use. If not provided, a new connection will be created.
+- `uri`: Optional[str], the uri of your database. Defaults to `"/tmp/lancedb"`.
+- `table_name` : Optional[str], Name of your table in the database. Defaults to `"vectors"`.
+- `table`: Optional[Any], `lancedb.db.LanceTable` object to be passed. Defaults to `None`. 
+- `vector_column_name`: Optional[Any], Column name to use for vector's in the table. Defaults to `'vector'`.   
+- `doc_id_key`: Optional[str], Column name to use for document id's in the table. Defaults to `'doc_id'`.  
+- `text_key`: Optional[str], Column name to use for text in the table. Defaults to `'text'`.  
+- `api_key`: Optional[str], API key to use for LanceDB cloud database. Defaults to `None`.  
+- `region`: Optional[str], Region to use for LanceDB cloud database. Only for LanceDB Cloud, defaults to `None`.  
+- `nprobes` : Optional[int], Set the number of probes to use. Only applicable if ANN index is created on the table else its ignored. Defaults to `20`.
+- `refine_factor` : Optional[int], Refine the results by reading extra elements and re-ranking them in memory. Defaults to `None`.
+- `reranker`: Optional[Any], The reranker to use for LanceDB.
+        Defaults to `None`.
+- `overfetch_factor`: Optional[int], The factor by which to fetch more results.
+        Defaults to `1`.
+- `mode`: Optional[str], The mode to use for LanceDB.
+            Defaults to `"overwrite"`.
+- `query_type`:Optional[str], The type of query to use for LanceDB.
+            Defaults to `"vector"`.
+
+
+### Methods 
+
+- __from_table(cls, table: lancedb.db.LanceTable) -> `LanceDBVectorStore`__ : (class method) Creates instance from lancedb table. 
+
+- **_add_reranker(self, reranker: lancedb.rerankers.Reranker) -> `None`** : Add a reranker to an existing vector store. 
+    - Usage :
+        ```python
+        from lancedb.rerankers import ColbertReranker
+        reranker = ColbertReranker()
+        vector_store._add_reranker(reranker)
+        ```               
+- **_table_exists(self, tbl_name: `Optional[str]` = `None`) -> `bool`** : Returns `True` if `tbl_name` exists in database.
+- __create_index(  
+  self, scalar: `Optional[bool]` = False, col_name: `Optional[str]` = None, num_partitions: `Optional[int]` = 256, num_sub_vectors: `Optional[int]` = 96, index_cache_size: `Optional[int]` = None, metric: `Optional[str]` = "L2",  
+) -> `None`__ : Creates a scalar(for non-vector cols) or a vector index on a table.
+        Make sure your vector column has enough data before creating an index on it.
+
+- __add(self, nodes: `List[BaseNode]`, **add_kwargs: `Any`, ) -> `List[str]`__ :
+adds Nodes to the table
+
+- **delete(self, ref_doc_id: `str`) -> `None`**: Delete nodes using with node_ids.
+- **delete_nodes(self, node_ids: `List[str]`) -> `None`** : Delete nodes using with node_ids.
+- __query(
+        self,
+        query: `VectorStoreQuery`,
+        **kwargs: `Any`,
+    ) -> `VectorStoreQueryResult`__:
+        Query index(`VectorStoreIndex`) for top k most similar nodes. Accepts llamaIndex `VectorStoreQuery` object.

docs/src/js/.nojekyll

@@ -0,0 +1 @@
+TypeDoc added this file to prevent GitHub Pages from using Jekyll. You can turn off this behavior by setting the `githubPages` option to false.

docs/src/js/README.md

@@ -0,0 +1,76 @@
+**@lancedb/lancedb** • [**Docs**](globals.md)
+
+***
+
+# LanceDB JavaScript SDK
+
+A JavaScript library for [LanceDB](https://github.com/lancedb/lancedb).
+
+## Installation
+
+```bash
+npm install @lancedb/lancedb
+```
+
+This will download the appropriate native library for your platform. We currently
+support:
+
+- Linux (x86_64 and aarch64)
+- MacOS (Intel and ARM/M1/M2)
+- Windows (x86_64 only)
+
+We do not yet support musl-based Linux (such as Alpine Linux) or aarch64 Windows.
+
+## Usage
+
+### Basic Example
+
+```javascript
+import * as lancedb from "@lancedb/lancedb";
+const db = await lancedb.connect("data/sample-lancedb");
+const table = await db.createTable("my_table", [
+  { id: 1, vector: [0.1, 1.0], item: "foo", price: 10.0 },
+  { id: 2, vector: [3.9, 0.5], item: "bar", price: 20.0 },
+]);
+const results = await table.vectorSearch([0.1, 0.3]).limit(20).toArray();
+console.log(results);
+```
+
+The [quickstart](../basic.md) contains a more complete example.
+
+## Development
+
+```sh
+npm run build
+npm run test
+```
+
+### Running lint / format
+
+LanceDb uses [biome](https://biomejs.dev/) for linting and formatting. if you are using VSCode you will need to install the official [Biome](https://marketplace.visualstudio.com/items?itemName=biomejs.biome) extension.
+To manually lint your code you can run:
+
+```sh
+npm run lint
+```
+
+to automatically fix all fixable issues:
+
+```sh
+npm run lint-fix
+```
+
+If you do not have your workspace root set to the `nodejs` directory, unfortunately the extension will not work. You can still run the linting and formatting commands manually.
+
+### Generating docs
+
+```sh
+npm run docs
+
+cd ../docs
+# Asssume the virtual environment was created
+# python3 -m venv venv
+# pip install -r requirements.txt
+. ./venv/bin/activate
+mkdocs build
+```

docs/src/js/classes/Connection.md

@@ -0,0 +1,200 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / Connection
+
+# Class: `abstract` Connection
+
+A LanceDB Connection that allows you to open tables and create new ones.
+
+Connection could be local against filesystem or remote against a server.
+
+A Connection is intended to be a long lived object and may hold open
+resources such as HTTP connection pools.  This is generally fine and
+a single connection should be shared if it is going to be used many
+times. However, if you are finished with a connection, you may call
+close to eagerly free these resources.  Any call to a Connection
+method after it has been closed will result in an error.
+
+Closing a connection is optional.  Connections will automatically
+be closed when they are garbage collected.
+
+Any created tables are independent and will continue to work even if
+the underlying connection has been closed.
+
+## Constructors
+
+### new Connection()
+
+> **new Connection**(): [`Connection`](Connection.md)
+
+#### Returns
+
+[`Connection`](Connection.md)
+
+## Methods
+
+### close()
+
+> `abstract` **close**(): `void`
+
+Close the connection, releasing any underlying resources.
+
+It is safe to call this method multiple times.
+
+Any attempt to use the connection after it is closed will result in an error.
+
+#### Returns
+
+`void`
+
+***
+
+### createEmptyTable()
+
+> `abstract` **createEmptyTable**(`name`, `schema`, `options`?): `Promise`<[`Table`](Table.md)>
+
+Creates a new empty Table
+
+#### Parameters
+
+• **name**: `string`
+
+The name of the table.
+
+• **schema**: `SchemaLike`
+
+The schema of the table
+
+• **options?**: `Partial`<[`CreateTableOptions`](../interfaces/CreateTableOptions.md)>
+
+#### Returns
+
+`Promise`<[`Table`](Table.md)>
+
+***
+
+### createTable()
+
+#### createTable(options)
+
+> `abstract` **createTable**(`options`): `Promise`<[`Table`](Table.md)>
+
+Creates a new Table and initialize it with new data.
+
+##### Parameters
+
+• **options**: `object` & `Partial`<[`CreateTableOptions`](../interfaces/CreateTableOptions.md)>
+
+The options object.
+
+##### Returns
+
+`Promise`<[`Table`](Table.md)>
+
+#### createTable(name, data, options)
+
+> `abstract` **createTable**(`name`, `data`, `options`?): `Promise`<[`Table`](Table.md)>
+
+Creates a new Table and initialize it with new data.
+
+##### Parameters
+
+• **name**: `string`
+
+The name of the table.
+
+• **data**: `TableLike` \| `Record`<`string`, `unknown`>[]
+
+Non-empty Array of Records
+to be inserted into the table
+
+• **options?**: `Partial`<[`CreateTableOptions`](../interfaces/CreateTableOptions.md)>
+
+##### Returns
+
+`Promise`<[`Table`](Table.md)>
+
+***
+
+### display()
+
+> `abstract` **display**(): `string`
+
+Return a brief description of the connection
+
+#### Returns
+
+`string`
+
+***
+
+### dropTable()
+
+> `abstract` **dropTable**(`name`): `Promise`<`void`>
+
+Drop an existing table.
+
+#### Parameters
+
+• **name**: `string`
+
+The name of the table to drop.
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### isOpen()
+
+> `abstract` **isOpen**(): `boolean`
+
+Return true if the connection has not been closed
+
+#### Returns
+
+`boolean`
+
+***
+
+### openTable()
+
+> `abstract` **openTable**(`name`, `options`?): `Promise`<[`Table`](Table.md)>
+
+Open a table in the database.
+
+#### Parameters
+
+• **name**: `string`
+
+The name of the table
+
+• **options?**: `Partial`<`OpenTableOptions`>
+
+#### Returns
+
+`Promise`<[`Table`](Table.md)>
+
+***
+
+### tableNames()
+
+> `abstract` **tableNames**(`options`?): `Promise`<`string`[]>
+
+List all the table names in this database.
+
+Tables will be returned in lexicographical order.
+
+#### Parameters
+
+• **options?**: `Partial`<[`TableNamesOptions`](../interfaces/TableNamesOptions.md)>
+
+options to control the
+paging / start point
+
+#### Returns
+
+`Promise`<`string`[]>

docs/src/js/classes/Index.md

@@ -0,0 +1,70 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / Index
+
+# Class: Index
+
+## Methods
+
+### btree()
+
+> `static` **btree**(): [`Index`](Index.md)
+
+Create a btree index
+
+A btree index is an index on a scalar columns.  The index stores a copy of the column
+in sorted order.  A header entry is created for each block of rows (currently the
+block size is fixed at 4096).  These header entries are stored in a separate
+cacheable structure (a btree).  To search for data the header is used to determine
+which blocks need to be read from disk.
+
+For example, a btree index in a table with 1Bi rows requires sizeof(Scalar) * 256Ki
+bytes of memory and will generally need to read sizeof(Scalar) * 4096 bytes to find
+the correct row ids.
+
+This index is good for scalar columns with mostly distinct values and does best when
+the query is highly selective.
+
+The btree index does not currently have any parameters though parameters such as the
+block size may be added in the future.
+
+#### Returns
+
+[`Index`](Index.md)
+
+***
+
+### ivfPq()
+
+> `static` **ivfPq**(`options`?): [`Index`](Index.md)
+
+Create an IvfPq index
+
+This index stores a compressed (quantized) copy of every vector.  These vectors
+are grouped into partitions of similar vectors.  Each partition keeps track of
+a centroid which is the average value of all vectors in the group.
+
+During a query the centroids are compared with the query vector to find the closest
+partitions.  The compressed vectors in these partitions are then searched to find
+the closest vectors.
+
+The compression scheme is called product quantization.  Each vector is divided into
+subvectors and then each subvector is quantized into a small number of bits.  the
+parameters `num_bits` and `num_subvectors` control this process, providing a tradeoff
+between index size (and thus search speed) and index accuracy.
+
+The partitioning process is called IVF and the `num_partitions` parameter controls how
+many groups to create.
+
+Note that training an IVF PQ index on a large dataset is a slow operation and
+currently is also a memory intensive operation.
+
+#### Parameters
+
+• **options?**: `Partial`<[`IvfPqOptions`](../interfaces/IvfPqOptions.md)>
+
+#### Returns
+
+[`Index`](Index.md)

docs/src/js/classes/MakeArrowTableOptions.md

@@ -0,0 +1,61 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / MakeArrowTableOptions
+
+# Class: MakeArrowTableOptions
+
+Options to control the makeArrowTable call.
+
+## Constructors
+
+### new MakeArrowTableOptions()
+
+> **new MakeArrowTableOptions**(`values`?): [`MakeArrowTableOptions`](MakeArrowTableOptions.md)
+
+#### Parameters
+
+• **values?**: `Partial`<[`MakeArrowTableOptions`](MakeArrowTableOptions.md)>
+
+#### Returns
+
+[`MakeArrowTableOptions`](MakeArrowTableOptions.md)
+
+## Properties
+
+### dictionaryEncodeStrings
+
+> **dictionaryEncodeStrings**: `boolean` = `false`
+
+If true then string columns will be encoded with dictionary encoding
+
+Set this to true if your string columns tend to repeat the same values
+often.  For more precise control use the `schema` property to specify the
+data type for individual columns.
+
+If `schema` is provided then this property is ignored.
+
+***
+
+### embeddingFunction?
+
+> `optional` **embeddingFunction**: [`EmbeddingFunctionConfig`](../namespaces/embedding/interfaces/EmbeddingFunctionConfig.md)
+
+***
+
+### embeddings?
+
+> `optional` **embeddings**: [`EmbeddingFunction`](../namespaces/embedding/classes/EmbeddingFunction.md)<`unknown`, `FunctionOptions`>
+
+***
+
+### schema?
+
+> `optional` **schema**: `SchemaLike`
+
+***
+
+### vectorColumns
+
+> **vectorColumns**: `Record`<`string`, [`VectorColumnOptions`](VectorColumnOptions.md)>

docs/src/js/classes/Query.md

@@ -0,0 +1,391 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / Query
+
+# Class: Query
+
+A builder for LanceDB queries.
+
+## Extends
+
+- [`QueryBase`](QueryBase.md)<`NativeQuery`>
+
+## Constructors
+
+### new Query()
+
+> **new Query**(`tbl`): [`Query`](Query.md)
+
+#### Parameters
+
+• **tbl**: `Table`
+
+#### Returns
+
+[`Query`](Query.md)
+
+#### Overrides
+
+[`QueryBase`](QueryBase.md).[`constructor`](QueryBase.md#constructors)
+
+## Properties
+
+### inner
+
+> `protected` **inner**: `Query` \| `Promise`<`Query`>
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`inner`](QueryBase.md#inner)
+
+## Methods
+
+### \[asyncIterator\]()
+
+> **\[asyncIterator\]**(): `AsyncIterator`<`RecordBatch`<`any`>, `any`, `undefined`>
+
+#### Returns
+
+`AsyncIterator`<`RecordBatch`<`any`>, `any`, `undefined`>
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`[asyncIterator]`](QueryBase.md#%5Basynciterator%5D)
+
+***
+
+### doCall()
+
+> `protected` **doCall**(`fn`): `void`
+
+#### Parameters
+
+• **fn**
+
+#### Returns
+
+`void`
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`doCall`](QueryBase.md#docall)
+
+***
+
+### execute()
+
+> `protected` **execute**(`options`?): [`RecordBatchIterator`](RecordBatchIterator.md)
+
+Execute the query and return the results as an
+
+#### Parameters
+
+• **options?**: `Partial`<`QueryExecutionOptions`>
+
+#### Returns
+
+[`RecordBatchIterator`](RecordBatchIterator.md)
+
+#### See
+
+ - AsyncIterator
+of
+ - RecordBatch.
+
+By default, LanceDb will use many threads to calculate results and, when
+the result set is large, multiple batches will be processed at one time.
+This readahead is limited however and backpressure will be applied if this
+stream is consumed slowly (this constrains the maximum memory used by a
+single query)
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`execute`](QueryBase.md#execute)
+
+***
+
+### explainPlan()
+
+> **explainPlan**(`verbose`): `Promise`<`string`>
+
+Generates an explanation of the query execution plan.
+
+#### Parameters
+
+• **verbose**: `boolean` = `false`
+
+If true, provides a more detailed explanation. Defaults to false.
+
+#### Returns
+
+`Promise`<`string`>
+
+A Promise that resolves to a string containing the query execution plan explanation.
+
+#### Example
+
+```ts
+import * as lancedb from "@lancedb/lancedb"
+const db = await lancedb.connect("./.lancedb");
+const table = await db.createTable("my_table", [
+  { vector: [1.1, 0.9], id: "1" },
+]);
+const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`explainPlan`](QueryBase.md#explainplan)
+
+***
+
+### ~~filter()~~
+
+> **filter**(`predicate`): `this`
+
+A filter statement to be applied to this query.
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`this`
+
+#### Alias
+
+where
+
+#### Deprecated
+
+Use `where` instead
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`filter`](QueryBase.md#filter)
+
+***
+
+### limit()
+
+> **limit**(`limit`): `this`
+
+Set the maximum number of results to return.
+
+By default, a plain search has no limit.  If this method is not
+called then every valid row from the table will be returned.
+
+#### Parameters
+
+• **limit**: `number`
+
+#### Returns
+
+`this`
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`limit`](QueryBase.md#limit)
+
+***
+
+### nativeExecute()
+
+> `protected` **nativeExecute**(`options`?): `Promise`<`RecordBatchIterator`>
+
+#### Parameters
+
+• **options?**: `Partial`<`QueryExecutionOptions`>
+
+#### Returns
+
+`Promise`<`RecordBatchIterator`>
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`nativeExecute`](QueryBase.md#nativeexecute)
+
+***
+
+### nearestTo()
+
+> **nearestTo**(`vector`): [`VectorQuery`](VectorQuery.md)
+
+Find the nearest vectors to the given query vector.
+
+This converts the query from a plain query to a vector query.
+
+This method will attempt to convert the input to the query vector
+expected by the embedding model.  If the input cannot be converted
+then an error will be thrown.
+
+By default, there is no embedding model, and the input should be
+an array-like object of numbers (something that can be used as input
+to Float32Array.from)
+
+If there is only one vector column (a column whose data type is a
+fixed size list of floats) then the column does not need to be specified.
+If there is more than one vector column you must use
+
+#### Parameters
+
+• **vector**: `IntoVector`
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+#### See
+
+ - [VectorQuery#column](VectorQuery.md#column)  to specify which column you would like
+to compare with.
+
+If no index has been created on the vector column then a vector query
+will perform a distance comparison between the query vector and every
+vector in the database and then sort the results.  This is sometimes
+called a "flat search"
+
+For small databases, with a few hundred thousand vectors or less, this can
+be reasonably fast.  In larger databases you should create a vector index
+on the column.  If there is a vector index then an "approximate" nearest
+neighbor search (frequently called an ANN search) will be performed.  This
+search is much faster, but the results will be approximate.
+
+The query can be further parameterized using the returned builder.  There
+are various ANN search parameters that will let you fine tune your recall
+accuracy vs search latency.
+
+Vector searches always have a `limit`.  If `limit` has not been called then
+a default `limit` of 10 will be used.
+ - [Query#limit](Query.md#limit)
+
+***
+
+### select()
+
+> **select**(`columns`): `this`
+
+Return only the specified columns.
+
+By default a query will return all columns from the table.  However, this can have
+a very significant impact on latency.  LanceDb stores data in a columnar fashion.  This
+means we can finely tune our I/O to select exactly the columns we need.
+
+As a best practice you should always limit queries to the columns that you need.  If you
+pass in an array of column names then only those columns will be returned.
+
+You can also use this method to create new "dynamic" columns based on your existing columns.
+For example, you may not care about "a" or "b" but instead simply want "a + b".  This is often
+seen in the SELECT clause of an SQL query (e.g. `SELECT a+b FROM my_table`).
+
+To create dynamic columns you can pass in a Map<string, string>.  A column will be returned
+for each entry in the map.  The key provides the name of the column.  The value is
+an SQL string used to specify how the column is calculated.
+
+For example, an SQL query might state `SELECT a + b AS combined, c`.  The equivalent
+input to this method would be:
+
+#### Parameters
+
+• **columns**: `string` \| `string`[] \| `Record`&lt;`string`, `string`&gt; \| `Map`&lt;`string`, `string`&gt;
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+new Map([["combined", "a + b"], ["c", "c"]])
+
+Columns will always be returned in the order given, even if that order is different than
+the order used when adding the data.
+
+Note that you can pass in a `Record<string, string>` (e.g. an object literal). This method
+uses `Object.entries` which should preserve the insertion order of the object.  However,
+object insertion order is easy to get wrong and `Map` is more foolproof.
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`select`](QueryBase.md#select)
+
+***
+
+### toArray()
+
+> **toArray**(`options`?): `Promise`&lt;`any`[]&gt;
+
+Collect the results as an array of objects.
+
+#### Parameters
+
+• **options?**: `Partial`&lt;`QueryExecutionOptions`&gt;
+
+#### Returns
+
+`Promise`&lt;`any`[]&gt;
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`toArray`](QueryBase.md#toarray)
+
+***
+
+### toArrow()
+
+> **toArrow**(`options`?): `Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+Collect the results as an Arrow
+
+#### Parameters
+
+• **options?**: `Partial`&lt;`QueryExecutionOptions`&gt;
+
+#### Returns
+
+`Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+#### See
+
+ArrowTable.
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`toArrow`](QueryBase.md#toarrow)
+
+***
+
+### where()
+
+> **where**(`predicate`): `this`
+
+A filter statement to be applied to this query.
+
+The filter should be supplied as an SQL query string.  For example:
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+x > 10
+y > 0 AND y < 100
+x > 5 OR y = 'test'
+
+Filtering performance can often be improved by creating a scalar index
+on the filter column(s).
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`where`](QueryBase.md#where)

docs/src/js/classes/QueryBase.md

@@ -0,0 +1,298 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / QueryBase
+
+# Class: QueryBase<NativeQueryType>
+
+Common methods supported by all query types
+
+## Extended by
+
+- [`Query`](Query.md)
+- [`VectorQuery`](VectorQuery.md)
+
+## Type Parameters
+
+• **NativeQueryType** *extends* `NativeQuery` \| `NativeVectorQuery`
+
+## Implements
+
+- `AsyncIterable`<`RecordBatch`>
+
+## Constructors
+
+### new QueryBase()
+
+> `protected` **new QueryBase**<`NativeQueryType`>(`inner`): [`QueryBase`](QueryBase.md)<`NativeQueryType`>
+
+#### Parameters
+
+• **inner**: `NativeQueryType` \| `Promise`<`NativeQueryType`>
+
+#### Returns
+
+[`QueryBase`](QueryBase.md)<`NativeQueryType`>
+
+## Properties
+
+### inner
+
+> `protected` **inner**: `NativeQueryType` \| `Promise`<`NativeQueryType`>
+
+## Methods
+
+### \[asyncIterator\]()
+
+> **\[asyncIterator\]**(): `AsyncIterator`<`RecordBatch`<`any`>, `any`, `undefined`>
+
+#### Returns
+
+`AsyncIterator`<`RecordBatch`<`any`>, `any`, `undefined`>
+
+#### Implementation of
+
+`AsyncIterable.[asyncIterator]`
+
+***
+
+### doCall()
+
+> `protected` **doCall**(`fn`): `void`
+
+#### Parameters
+
+• **fn**
+
+#### Returns
+
+`void`
+
+***
+
+### execute()
+
+> `protected` **execute**(`options`?): [`RecordBatchIterator`](RecordBatchIterator.md)
+
+Execute the query and return the results as an
+
+#### Parameters
+
+• **options?**: `Partial`<`QueryExecutionOptions`>
+
+#### Returns
+
+[`RecordBatchIterator`](RecordBatchIterator.md)
+
+#### See
+
+ - AsyncIterator
+of
+ - RecordBatch.
+
+By default, LanceDb will use many threads to calculate results and, when
+the result set is large, multiple batches will be processed at one time.
+This readahead is limited however and backpressure will be applied if this
+stream is consumed slowly (this constrains the maximum memory used by a
+single query)
+
+***
+
+### explainPlan()
+
+> **explainPlan**(`verbose`): `Promise`<`string`>
+
+Generates an explanation of the query execution plan.
+
+#### Parameters
+
+• **verbose**: `boolean` = `false`
+
+If true, provides a more detailed explanation. Defaults to false.
+
+#### Returns
+
+`Promise`<`string`>
+
+A Promise that resolves to a string containing the query execution plan explanation.
+
+#### Example
+
+```ts
+import * as lancedb from "@lancedb/lancedb"
+const db = await lancedb.connect("./.lancedb");
+const table = await db.createTable("my_table", [
+  { vector: [1.1, 0.9], id: "1" },
+]);
+const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
+```
+
+***
+
+### ~~filter()~~
+
+> **filter**(`predicate`): `this`
+
+A filter statement to be applied to this query.
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`this`
+
+#### Alias
+
+where
+
+#### Deprecated
+
+Use `where` instead
+
+***
+
+### limit()
+
+> **limit**(`limit`): `this`
+
+Set the maximum number of results to return.
+
+By default, a plain search has no limit.  If this method is not
+called then every valid row from the table will be returned.
+
+#### Parameters
+
+• **limit**: `number`
+
+#### Returns
+
+`this`
+
+***
+
+### nativeExecute()
+
+> `protected` **nativeExecute**(`options`?): `Promise`<`RecordBatchIterator`>
+
+#### Parameters
+
+• **options?**: `Partial`<`QueryExecutionOptions`>
+
+#### Returns
+
+`Promise`<`RecordBatchIterator`>
+
+***
+
+### select()
+
+> **select**(`columns`): `this`
+
+Return only the specified columns.
+
+By default a query will return all columns from the table.  However, this can have
+a very significant impact on latency.  LanceDb stores data in a columnar fashion.  This
+means we can finely tune our I/O to select exactly the columns we need.
+
+As a best practice you should always limit queries to the columns that you need.  If you
+pass in an array of column names then only those columns will be returned.
+
+You can also use this method to create new "dynamic" columns based on your existing columns.
+For example, you may not care about "a" or "b" but instead simply want "a + b".  This is often
+seen in the SELECT clause of an SQL query (e.g. `SELECT a+b FROM my_table`).
+
+To create dynamic columns you can pass in a Map<string, string>.  A column will be returned
+for each entry in the map.  The key provides the name of the column.  The value is
+an SQL string used to specify how the column is calculated.
+
+For example, an SQL query might state `SELECT a + b AS combined, c`.  The equivalent
+input to this method would be:
+
+#### Parameters
+
+• **columns**: `string` \| `string`[] \| `Record`&lt;`string`, `string`&gt; \| `Map`&lt;`string`, `string`&gt;
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+new Map([["combined", "a + b"], ["c", "c"]])
+
+Columns will always be returned in the order given, even if that order is different than
+the order used when adding the data.
+
+Note that you can pass in a `Record<string, string>` (e.g. an object literal). This method
+uses `Object.entries` which should preserve the insertion order of the object.  However,
+object insertion order is easy to get wrong and `Map` is more foolproof.
+```
+
+***
+
+### toArray()
+
+> **toArray**(`options`?): `Promise`&lt;`any`[]&gt;
+
+Collect the results as an array of objects.
+
+#### Parameters
+
+• **options?**: `Partial`&lt;`QueryExecutionOptions`&gt;
+
+#### Returns
+
+`Promise`&lt;`any`[]&gt;
+
+***
+
+### toArrow()
+
+> **toArrow**(`options`?): `Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+Collect the results as an Arrow
+
+#### Parameters
+
+• **options?**: `Partial`&lt;`QueryExecutionOptions`&gt;
+
+#### Returns
+
+`Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+#### See
+
+ArrowTable.
+
+***
+
+### where()
+
+> **where**(`predicate`): `this`
+
+A filter statement to be applied to this query.
+
+The filter should be supplied as an SQL query string.  For example:
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+x > 10
+y > 0 AND y < 100
+x > 5 OR y = 'test'
+
+Filtering performance can often be improved by creating a scalar index
+on the filter column(s).
+```

docs/src/js/classes/RecordBatchIterator.md

@@ -0,0 +1,39 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / RecordBatchIterator
+
+# Class: RecordBatchIterator
+
+## Implements
+
+- `AsyncIterator`<`RecordBatch`>
+
+## Constructors
+
+### new RecordBatchIterator()
+
+> **new RecordBatchIterator**(`promise`?): [`RecordBatchIterator`](RecordBatchIterator.md)
+
+#### Parameters
+
+• **promise?**: `Promise`<`RecordBatchIterator`>
+
+#### Returns
+
+[`RecordBatchIterator`](RecordBatchIterator.md)
+
+## Methods
+
+### next()
+
+> **next**(): `Promise`<`IteratorResult`<`RecordBatch`<`any`>, `any`>>
+
+#### Returns
+
+`Promise`<`IteratorResult`<`RecordBatch`<`any`>, `any`>>
+
+#### Implementation of
+
+`AsyncIterator.next`

docs/src/js/classes/Table.md

@@ -0,0 +1,696 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / Table
+
+# Class: `abstract` Table
+
+A Table is a collection of Records in a LanceDB Database.
+
+A Table object is expected to be long lived and reused for multiple operations.
+Table objects will cache a certain amount of index data in memory.  This cache
+will be freed when the Table is garbage collected.  To eagerly free the cache you
+can call the `close` method.  Once the Table is closed, it cannot be used for any
+further operations.
+
+Closing a table is optional.  It not closed, it will be closed when it is garbage
+collected.
+
+## Constructors
+
+### new Table()
+
+> **new Table**(): [`Table`](Table.md)
+
+#### Returns
+
+[`Table`](Table.md)
+
+## Accessors
+
+### name
+
+> `get` `abstract` **name**(): `string`
+
+Returns the name of the table
+
+#### Returns
+
+`string`
+
+## Methods
+
+### add()
+
+> `abstract` **add**(`data`, `options`?): `Promise`<`void`>
+
+Insert records into this Table.
+
+#### Parameters
+
+• **data**: [`Data`](../type-aliases/Data.md)
+
+Records to be inserted into the Table
+
+• **options?**: `Partial`<[`AddDataOptions`](../interfaces/AddDataOptions.md)>
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### addColumns()
+
+> `abstract` **addColumns**(`newColumnTransforms`): `Promise`<`void`>
+
+Add new columns with defined values.
+
+#### Parameters
+
+• **newColumnTransforms**: [`AddColumnsSql`](../interfaces/AddColumnsSql.md)[]
+
+pairs of column names and
+the SQL expression to use to calculate the value of the new column. These
+expressions will be evaluated for each row in the table, and can
+reference existing columns in the table.
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### alterColumns()
+
+> `abstract` **alterColumns**(`columnAlterations`): `Promise`<`void`>
+
+Alter the name or nullability of columns.
+
+#### Parameters
+
+• **columnAlterations**: [`ColumnAlteration`](../interfaces/ColumnAlteration.md)[]
+
+One or more alterations to
+apply to columns.
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### checkout()
+
+> `abstract` **checkout**(`version`): `Promise`<`void`>
+
+Checks out a specific version of the table _This is an in-place operation._
+
+This allows viewing previous versions of the table. If you wish to
+keep writing to the dataset starting from an old version, then use
+the `restore` function.
+
+Calling this method will set the table into time-travel mode. If you
+wish to return to standard mode, call `checkoutLatest`.
+
+#### Parameters
+
+• **version**: `number`
+
+The version to checkout
+
+#### Returns
+
+`Promise`<`void`>
+
+#### Example
+
+```typescript
+import * as lancedb from "@lancedb/lancedb"
+const db = await lancedb.connect("./.lancedb");
+const table = await db.createTable("my_table", [
+  { vector: [1.1, 0.9], type: "vector" },
+]);
+
+console.log(await table.version()); // 1
+console.log(table.display());
+await table.add([{ vector: [0.5, 0.2], type: "vector" }]);
+await table.checkout(1);
+console.log(await table.version()); // 2
+```
+
+***
+
+### checkoutLatest()
+
+> `abstract` **checkoutLatest**(): `Promise`<`void`>
+
+Checkout the latest version of the table. _This is an in-place operation._
+
+The table will be set back into standard mode, and will track the latest
+version of the table.
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### close()
+
+> `abstract` **close**(): `void`
+
+Close the table, releasing any underlying resources.
+
+It is safe to call this method multiple times.
+
+Any attempt to use the table after it is closed will result in an error.
+
+#### Returns
+
+`void`
+
+***
+
+### countRows()
+
+> `abstract` **countRows**(`filter`?): `Promise`<`number`>
+
+Count the total number of rows in the dataset.
+
+#### Parameters
+
+• **filter?**: `string`
+
+#### Returns
+
+`Promise`<`number`>
+
+***
+
+### createIndex()
+
+> `abstract` **createIndex**(`column`, `options`?): `Promise`<`void`>
+
+Create an index to speed up queries.
+
+Indices can be created on vector columns or scalar columns.
+Indices on vector columns will speed up vector searches.
+Indices on scalar columns will speed up filtering (in both
+vector and non-vector searches)
+
+#### Parameters
+
+• **column**: `string`
+
+• **options?**: `Partial`<[`IndexOptions`](../interfaces/IndexOptions.md)>
+
+#### Returns
+
+`Promise`<`void`>
+
+#### Note
+
+We currently don't support custom named indexes,
+The index name will always be `${column}_idx`
+
+#### Examples
+
+```ts
+// If the column has a vector (fixed size list) data type then
+// an IvfPq vector index will be created.
+const table = await conn.openTable("my_table");
+await table.createIndex("vector");
+```
+
+```ts
+// For advanced control over vector index creation you can specify
+// the index type and options.
+const table = await conn.openTable("my_table");
+await table.createIndex("vector", {
+  config: lancedb.Index.ivfPq({
+    numPartitions: 128,
+    numSubVectors: 16,
+  }),
+});
+```
+
+```ts
+// Or create a Scalar index
+await table.createIndex("my_float_col");
+```
+
+***
+
+### delete()
+
+> `abstract` **delete**(`predicate`): `Promise`<`void`>
+
+Delete the rows that satisfy the predicate.
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### display()
+
+> `abstract` **display**(): `string`
+
+Return a brief description of the table
+
+#### Returns
+
+`string`
+
+***
+
+### dropColumns()
+
+> `abstract` **dropColumns**(`columnNames`): `Promise`<`void`>
+
+Drop one or more columns from the dataset
+
+This is a metadata-only operation and does not remove the data from the
+underlying storage. In order to remove the data, you must subsequently
+call ``compact_files`` to rewrite the data without the removed columns and
+then call ``cleanup_files`` to remove the old files.
+
+#### Parameters
+
+• **columnNames**: `string`[]
+
+The names of the columns to drop. These can
+be nested column references (e.g. "a.b.c") or top-level column names
+(e.g. "a").
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### indexStats()
+
+> `abstract` **indexStats**(`name`): `Promise`<`undefined` \| [`IndexStatistics`](../interfaces/IndexStatistics.md)>
+
+List all the stats of a specified index
+
+#### Parameters
+
+• **name**: `string`
+
+The name of the index.
+
+#### Returns
+
+`Promise`<`undefined` \| [`IndexStatistics`](../interfaces/IndexStatistics.md)>
+
+The stats of the index. If the index does not exist, it will return undefined
+
+***
+
+### isOpen()
+
+> `abstract` **isOpen**(): `boolean`
+
+Return true if the table has not been closed
+
+#### Returns
+
+`boolean`
+
+***
+
+### listIndices()
+
+> `abstract` **listIndices**(): `Promise`<[`IndexConfig`](../interfaces/IndexConfig.md)[]>
+
+List all indices that have been created with [Table.createIndex](Table.md#createindex)
+
+#### Returns
+
+`Promise`<[`IndexConfig`](../interfaces/IndexConfig.md)[]>
+
+***
+
+### mergeInsert()
+
+> `abstract` **mergeInsert**(`on`): `MergeInsertBuilder`
+
+#### Parameters
+
+• **on**: `string` \| `string`[]
+
+#### Returns
+
+`MergeInsertBuilder`
+
+***
+
+### optimize()
+
+> `abstract` **optimize**(`options`?): `Promise`<`OptimizeStats`>
+
+Optimize the on-disk data and indices for better performance.
+
+Modeled after ``VACUUM`` in PostgreSQL.
+
+ Optimization covers three operations:
+
+ - Compaction: Merges small files into larger ones
+ - Prune: Removes old versions of the dataset
+ - Index: Optimizes the indices, adding new data to existing indices
+
+ Experimental API
+ ----------------
+
+ The optimization process is undergoing active development and may change.
+ Our goal with these changes is to improve the performance of optimization and
+ reduce the complexity.
+
+ That being said, it is essential today to run optimize if you want the best
+ performance.  It should be stable and safe to use in production, but it our
+ hope that the API may be simplified (or not even need to be called) in the
+ future.
+
+ The frequency an application shoudl call optimize is based on the frequency of
+ data modifications.  If data is frequently added, deleted, or updated then
+ optimize should be run frequently.  A good rule of thumb is to run optimize if
+ you have added or modified 100,000 or more records or run more than 20 data
+ modification operations.
+
+#### Parameters
+
+• **options?**: `Partial`<`OptimizeOptions`>
+
+#### Returns
+
+`Promise`<`OptimizeStats`>
+
+***
+
+### query()
+
+> `abstract` **query**(): [`Query`](Query.md)
+
+Create a [Query](Query.md) Builder.
+
+Queries allow you to search your existing data.  By default the query will
+return all the data in the table in no particular order.  The builder
+returned by this method can be used to control the query using filtering,
+vector similarity, sorting, and more.
+
+Note: By default, all columns are returned.  For best performance, you should
+only fetch the columns you need.
+
+When appropriate, various indices and statistics based pruning will be used to
+accelerate the query.
+
+#### Returns
+
+[`Query`](Query.md)
+
+A builder that can be used to parameterize the query
+
+#### Examples
+
+```ts
+// SQL-style filtering
+//
+// This query will return up to 1000 rows whose value in the `id` column
+// is greater than 5. LanceDb supports a broad set of filtering functions.
+for await (const batch of table
+  .query()
+  .where("id > 1")
+  .select(["id"])
+  .limit(20)) {
+  console.log(batch);
+}
+```
+
+```ts
+// Vector Similarity Search
+//
+// This example will find the 10 rows whose value in the "vector" column are
+// closest to the query vector [1.0, 2.0, 3.0].  If an index has been created
+// on the "vector" column then this will perform an ANN search.
+//
+// The `refineFactor` and `nprobes` methods are used to control the recall /
+// latency tradeoff of the search.
+for await (const batch of table
+  .query()
+  .where("id > 1")
+  .select(["id"])
+  .limit(20)) {
+  console.log(batch);
+}
+```
+
+```ts
+// Scan the full dataset
+//
+// This query will return everything in the table in no particular order.
+for await (const batch of table.query()) {
+  console.log(batch);
+}
+```
+
+***
+
+### restore()
+
+> `abstract` **restore**(): `Promise`<`void`>
+
+Restore the table to the currently checked out version
+
+This operation will fail if checkout has not been called previously
+
+This operation will overwrite the latest version of the table with a
+previous version.  Any changes made since the checked out version will
+no longer be visible.
+
+Once the operation concludes the table will no longer be in a checked
+out state and the read_consistency_interval, if any, will apply.
+
+#### Returns
+
+`Promise`<`void`>
+
+***
+
+### schema()
+
+> `abstract` **schema**(): `Promise`<`Schema`<`any`>>
+
+Get the schema of the table.
+
+#### Returns
+
+`Promise`<`Schema`<`any`>>
+
+***
+
+### search()
+
+#### search(query)
+
+> `abstract` **search**(`query`): [`VectorQuery`](VectorQuery.md)
+
+Create a search query to find the nearest neighbors
+of the given query vector
+
+##### Parameters
+
+• **query**: `string`
+
+the query. This will be converted to a vector using the table's provided embedding function
+
+##### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+##### Note
+
+If no embedding functions are defined in the table, this will error when collecting the results.
+
+#### search(query)
+
+> `abstract` **search**(`query`): [`VectorQuery`](VectorQuery.md)
+
+Create a search query to find the nearest neighbors
+of the given query vector
+
+##### Parameters
+
+• **query**: `IntoVector`
+
+the query vector
+
+##### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+***
+
+### toArrow()
+
+> `abstract` **toArrow**(): `Promise`<`Table`<`any`>>
+
+Return the table as an arrow table
+
+#### Returns
+
+`Promise`<`Table`<`any`>>
+
+***
+
+### update()
+
+#### update(opts)
+
+> `abstract` **update**(`opts`): `Promise`<`void`>
+
+Update existing records in the Table
+
+##### Parameters
+
+• **opts**: `object` & `Partial`<[`UpdateOptions`](../interfaces/UpdateOptions.md)>
+
+##### Returns
+
+`Promise`<`void`>
+
+##### Example
+
+```ts
+table.update({where:"x = 2", values:{"vector": [10, 10]}})
+```
+
+#### update(opts)
+
+> `abstract` **update**(`opts`): `Promise`<`void`>
+
+Update existing records in the Table
+
+##### Parameters
+
+• **opts**: `object` & `Partial`<[`UpdateOptions`](../interfaces/UpdateOptions.md)>
+
+##### Returns
+
+`Promise`<`void`>
+
+##### Example
+
+```ts
+table.update({where:"x = 2", valuesSql:{"x": "x + 1"}})
+```
+
+#### update(updates, options)
+
+> `abstract` **update**(`updates`, `options`?): `Promise`<`void`>
+
+Update existing records in the Table
+
+An update operation can be used to adjust existing values.  Use the
+returned builder to specify which columns to update.  The new value
+can be a literal value (e.g. replacing nulls with some default value)
+or an expression applied to the old value (e.g. incrementing a value)
+
+An optional condition can be specified (e.g. "only update if the old
+value is 0")
+
+Note: if your condition is something like "some_id_column == 7" and
+you are updating many rows (with different ids) then you will get
+better performance with a single [`merge_insert`] call instead of
+repeatedly calilng this method.
+
+##### Parameters
+
+• **updates**: `Record`<`string`, `string`> \| `Map`<`string`, `string`>
+
+the
+columns to update
+
+Keys in the map should specify the name of the column to update.
+Values in the map provide the new value of the column.  These can
+be SQL literal strings (e.g. "7" or "'foo'") or they can be expressions
+based on the row being updated (e.g. "my_col + 1")
+
+• **options?**: `Partial`<[`UpdateOptions`](../interfaces/UpdateOptions.md)>
+
+additional options to control
+the update behavior
+
+##### Returns
+
+`Promise`<`void`>
+
+***
+
+### vectorSearch()
+
+> `abstract` **vectorSearch**(`vector`): [`VectorQuery`](VectorQuery.md)
+
+Search the table with a given query vector.
+
+This is a convenience method for preparing a vector query and
+is the same thing as calling `nearestTo` on the builder returned
+by `query`.
+
+#### Parameters
+
+• **vector**: `IntoVector`
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+#### See
+
+[Query#nearestTo](Query.md#nearestto) for more details.
+
+***
+
+### version()
+
+> `abstract` **version**(): `Promise`<`number`>
+
+Retrieve the version of the table
+
+#### Returns
+
+`Promise`<`number`>
+
+***
+
+### parseTableData()
+
+> `static` **parseTableData**(`data`, `options`?, `streaming`?): `Promise`<`object`>
+
+#### Parameters
+
+• **data**: `TableLike` \| `Record`<`string`, `unknown`>[]
+
+• **options?**: `Partial`<[`CreateTableOptions`](../interfaces/CreateTableOptions.md)>
+
+• **streaming?**: `boolean` = `false`
+
+#### Returns
+
+`Promise`<`object`>
+
+##### buf
+
+> **buf**: `Buffer`
+
+##### mode
+
+> **mode**: `string`

docs/src/js/classes/VectorColumnOptions.md

@@ -0,0 +1,29 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / VectorColumnOptions
+
+# Class: VectorColumnOptions
+
+## Constructors
+
+### new VectorColumnOptions()
+
+> **new VectorColumnOptions**(`values`?): [`VectorColumnOptions`](VectorColumnOptions.md)
+
+#### Parameters
+
+• **values?**: `Partial`<[`VectorColumnOptions`](VectorColumnOptions.md)>
+
+#### Returns
+
+[`VectorColumnOptions`](VectorColumnOptions.md)
+
+## Properties
+
+### type
+
+> **type**: `Float`<`Floats`>
+
+Vector column type.

docs/src/js/classes/VectorQuery.md

@@ -0,0 +1,523 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / VectorQuery
+
+# Class: VectorQuery
+
+A builder used to construct a vector search
+
+This builder can be reused to execute the query many times.
+
+## Extends
+
+- [`QueryBase`](QueryBase.md)<`NativeVectorQuery`>
+
+## Constructors
+
+### new VectorQuery()
+
+> **new VectorQuery**(`inner`): [`VectorQuery`](VectorQuery.md)
+
+#### Parameters
+
+• **inner**: `VectorQuery` \| `Promise`<`VectorQuery`>
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+#### Overrides
+
+[`QueryBase`](QueryBase.md).[`constructor`](QueryBase.md#constructors)
+
+## Properties
+
+### inner
+
+> `protected` **inner**: `VectorQuery` \| `Promise`<`VectorQuery`>
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`inner`](QueryBase.md#inner)
+
+## Methods
+
+### \[asyncIterator\]()
+
+> **\[asyncIterator\]**(): `AsyncIterator`<`RecordBatch`<`any`>, `any`, `undefined`>
+
+#### Returns
+
+`AsyncIterator`<`RecordBatch`<`any`>, `any`, `undefined`>
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`[asyncIterator]`](QueryBase.md#%5Basynciterator%5D)
+
+***
+
+### bypassVectorIndex()
+
+> **bypassVectorIndex**(): [`VectorQuery`](VectorQuery.md)
+
+If this is called then any vector index is skipped
+
+An exhaustive (flat) search will be performed.  The query vector will
+be compared to every vector in the table.  At high scales this can be
+expensive.  However, this is often still useful.  For example, skipping
+the vector index can give you ground truth results which you can use to
+calculate your recall to select an appropriate value for nprobes.
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+***
+
+### column()
+
+> **column**(`column`): [`VectorQuery`](VectorQuery.md)
+
+Set the vector column to query
+
+This controls which column is compared to the query vector supplied in
+the call to
+
+#### Parameters
+
+• **column**: `string`
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+#### See
+
+[Query#nearestTo](Query.md#nearestto)
+
+This parameter must be specified if the table has more than one column
+whose data type is a fixed-size-list of floats.
+
+***
+
+### distanceType()
+
+> **distanceType**(`distanceType`): [`VectorQuery`](VectorQuery.md)
+
+Set the distance metric to use
+
+When performing a vector search we try and find the "nearest" vectors according
+to some kind of distance metric.  This parameter controls which distance metric to
+use.  See
+
+#### Parameters
+
+• **distanceType**: `"l2"` \| `"cosine"` \| `"dot"`
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+#### See
+
+[IvfPqOptions.distanceType](../interfaces/IvfPqOptions.md#distancetype) for more details on the different
+distance metrics available.
+
+Note: if there is a vector index then the distance type used MUST match the distance
+type used to train the vector index.  If this is not done then the results will be
+invalid.
+
+By default "l2" is used.
+
+***
+
+### doCall()
+
+> `protected` **doCall**(`fn`): `void`
+
+#### Parameters
+
+• **fn**
+
+#### Returns
+
+`void`
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`doCall`](QueryBase.md#docall)
+
+***
+
+### execute()
+
+> `protected` **execute**(`options`?): [`RecordBatchIterator`](RecordBatchIterator.md)
+
+Execute the query and return the results as an
+
+#### Parameters
+
+• **options?**: `Partial`<`QueryExecutionOptions`>
+
+#### Returns
+
+[`RecordBatchIterator`](RecordBatchIterator.md)
+
+#### See
+
+ - AsyncIterator
+of
+ - RecordBatch.
+
+By default, LanceDb will use many threads to calculate results and, when
+the result set is large, multiple batches will be processed at one time.
+This readahead is limited however and backpressure will be applied if this
+stream is consumed slowly (this constrains the maximum memory used by a
+single query)
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`execute`](QueryBase.md#execute)
+
+***
+
+### explainPlan()
+
+> **explainPlan**(`verbose`): `Promise`<`string`>
+
+Generates an explanation of the query execution plan.
+
+#### Parameters
+
+• **verbose**: `boolean` = `false`
+
+If true, provides a more detailed explanation. Defaults to false.
+
+#### Returns
+
+`Promise`<`string`>
+
+A Promise that resolves to a string containing the query execution plan explanation.
+
+#### Example
+
+```ts
+import * as lancedb from "@lancedb/lancedb"
+const db = await lancedb.connect("./.lancedb");
+const table = await db.createTable("my_table", [
+  { vector: [1.1, 0.9], id: "1" },
+]);
+const plan = await table.query().nearestTo([0.5, 0.2]).explainPlan();
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`explainPlan`](QueryBase.md#explainplan)
+
+***
+
+### ~~filter()~~
+
+> **filter**(`predicate`): `this`
+
+A filter statement to be applied to this query.
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`this`
+
+#### Alias
+
+where
+
+#### Deprecated
+
+Use `where` instead
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`filter`](QueryBase.md#filter)
+
+***
+
+### limit()
+
+> **limit**(`limit`): `this`
+
+Set the maximum number of results to return.
+
+By default, a plain search has no limit.  If this method is not
+called then every valid row from the table will be returned.
+
+#### Parameters
+
+• **limit**: `number`
+
+#### Returns
+
+`this`
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`limit`](QueryBase.md#limit)
+
+***
+
+### nativeExecute()
+
+> `protected` **nativeExecute**(`options`?): `Promise`<`RecordBatchIterator`>
+
+#### Parameters
+
+• **options?**: `Partial`<`QueryExecutionOptions`>
+
+#### Returns
+
+`Promise`<`RecordBatchIterator`>
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`nativeExecute`](QueryBase.md#nativeexecute)
+
+***
+
+### nprobes()
+
+> **nprobes**(`nprobes`): [`VectorQuery`](VectorQuery.md)
+
+Set the number of partitions to search (probe)
+
+This argument is only used when the vector column has an IVF PQ index.
+If there is no index then this value is ignored.
+
+The IVF stage of IVF PQ divides the input into partitions (clusters) of
+related values.
+
+The partition whose centroids are closest to the query vector will be
+exhaustiely searched to find matches.  This parameter controls how many
+partitions should be searched.
+
+Increasing this value will increase the recall of your query but will
+also increase the latency of your query.  The default value is 20.  This
+default is good for many cases but the best value to use will depend on
+your data and the recall that you need to achieve.
+
+For best results we recommend tuning this parameter with a benchmark against
+your actual data to find the smallest possible value that will still give
+you the desired recall.
+
+#### Parameters
+
+• **nprobes**: `number`
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+***
+
+### postfilter()
+
+> **postfilter**(): [`VectorQuery`](VectorQuery.md)
+
+If this is called then filtering will happen after the vector search instead of
+before.
+
+By default filtering will be performed before the vector search.  This is how
+filtering is typically understood to work.  This prefilter step does add some
+additional latency.  Creating a scalar index on the filter column(s) can
+often improve this latency.  However, sometimes a filter is too complex or scalar
+indices cannot be applied to the column.  In these cases postfiltering can be
+used instead of prefiltering to improve latency.
+
+Post filtering applies the filter to the results of the vector search.  This means
+we only run the filter on a much smaller set of data.  However, it can cause the
+query to return fewer than `limit` results (or even no results) if none of the nearest
+results match the filter.
+
+Post filtering happens during the "refine stage" (described in more detail in
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+#### See
+
+[VectorQuery#refineFactor](VectorQuery.md#refinefactor)).  This means that setting a higher refine
+factor can often help restore some of the results lost by post filtering.
+
+***
+
+### refineFactor()
+
+> **refineFactor**(`refineFactor`): [`VectorQuery`](VectorQuery.md)
+
+A multiplier to control how many additional rows are taken during the refine step
+
+This argument is only used when the vector column has an IVF PQ index.
+If there is no index then this value is ignored.
+
+An IVF PQ index stores compressed (quantized) values.  They query vector is compared
+against these values and, since they are compressed, the comparison is inaccurate.
+
+This parameter can be used to refine the results.  It can improve both improve recall
+and correct the ordering of the nearest results.
+
+To refine results LanceDb will first perform an ANN search to find the nearest
+`limit` * `refine_factor` results.  In other words, if `refine_factor` is 3 and
+`limit` is the default (10) then the first 30 results will be selected.  LanceDb
+then fetches the full, uncompressed, values for these 30 results.  The results are
+then reordered by the true distance and only the nearest 10 are kept.
+
+Note: there is a difference between calling this method with a value of 1 and never
+calling this method at all.  Calling this method with any value will have an impact
+on your search latency.  When you call this method with a `refine_factor` of 1 then
+LanceDb still needs to fetch the full, uncompressed, values so that it can potentially
+reorder the results.
+
+Note: if this method is NOT called then the distances returned in the _distance column
+will be approximate distances based on the comparison of the quantized query vector
+and the quantized result vectors.  This can be considerably different than the true
+distance between the query vector and the actual uncompressed vector.
+
+#### Parameters
+
+• **refineFactor**: `number`
+
+#### Returns
+
+[`VectorQuery`](VectorQuery.md)
+
+***
+
+### select()
+
+> **select**(`columns`): `this`
+
+Return only the specified columns.
+
+By default a query will return all columns from the table.  However, this can have
+a very significant impact on latency.  LanceDb stores data in a columnar fashion.  This
+means we can finely tune our I/O to select exactly the columns we need.
+
+As a best practice you should always limit queries to the columns that you need.  If you
+pass in an array of column names then only those columns will be returned.
+
+You can also use this method to create new "dynamic" columns based on your existing columns.
+For example, you may not care about "a" or "b" but instead simply want "a + b".  This is often
+seen in the SELECT clause of an SQL query (e.g. `SELECT a+b FROM my_table`).
+
+To create dynamic columns you can pass in a Map<string, string>.  A column will be returned
+for each entry in the map.  The key provides the name of the column.  The value is
+an SQL string used to specify how the column is calculated.
+
+For example, an SQL query might state `SELECT a + b AS combined, c`.  The equivalent
+input to this method would be:
+
+#### Parameters
+
+• **columns**: `string` \| `string`[] \| `Record`&lt;`string`, `string`&gt; \| `Map`&lt;`string`, `string`&gt;
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+new Map([["combined", "a + b"], ["c", "c"]])
+
+Columns will always be returned in the order given, even if that order is different than
+the order used when adding the data.
+
+Note that you can pass in a `Record<string, string>` (e.g. an object literal). This method
+uses `Object.entries` which should preserve the insertion order of the object.  However,
+object insertion order is easy to get wrong and `Map` is more foolproof.
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`select`](QueryBase.md#select)
+
+***
+
+### toArray()
+
+> **toArray**(`options`?): `Promise`&lt;`any`[]&gt;
+
+Collect the results as an array of objects.
+
+#### Parameters
+
+• **options?**: `Partial`&lt;`QueryExecutionOptions`&gt;
+
+#### Returns
+
+`Promise`&lt;`any`[]&gt;
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`toArray`](QueryBase.md#toarray)
+
+***
+
+### toArrow()
+
+> **toArrow**(`options`?): `Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+Collect the results as an Arrow
+
+#### Parameters
+
+• **options?**: `Partial`&lt;`QueryExecutionOptions`&gt;
+
+#### Returns
+
+`Promise`&lt;`Table`&lt;`any`&gt;&gt;
+
+#### See
+
+ArrowTable.
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`toArrow`](QueryBase.md#toarrow)
+
+***
+
+### where()
+
+> **where**(`predicate`): `this`
+
+A filter statement to be applied to this query.
+
+The filter should be supplied as an SQL query string.  For example:
+
+#### Parameters
+
+• **predicate**: `string`
+
+#### Returns
+
+`this`
+
+#### Example
+
+```ts
+x > 10
+y > 0 AND y < 100
+x > 5 OR y = 'test'
+
+Filtering performance can often be improved by creating a scalar index
+on the filter column(s).
+```
+
+#### Inherited from
+
+[`QueryBase`](QueryBase.md).[`where`](QueryBase.md#where)

docs/src/js/enumerations/WriteMode.md

@@ -0,0 +1,27 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / WriteMode
+
+# Enumeration: WriteMode
+
+Write mode for writing a table.
+
+## Enumeration Members
+
+### Append
+
+> **Append**: `"Append"`
+
+***
+
+### Create
+
+> **Create**: `"Create"`
+
+***
+
+### Overwrite
+
+> **Overwrite**: `"Overwrite"`

docs/src/js/functions/connect.md

@@ -0,0 +1,82 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / connect
+
+# Function: connect()
+
+## connect(uri, opts)
+
+> **connect**(`uri`, `opts`?): `Promise`<[`Connection`](../classes/Connection.md)>
+
+Connect to a LanceDB instance at the given URI.
+
+Accepted formats:
+
+- `/path/to/database` - local database
+- `s3://bucket/path/to/database` or `gs://bucket/path/to/database` - database on cloud storage
+- `db://host:port` - remote database (LanceDB cloud)
+
+### Parameters
+
+• **uri**: `string`
+
+The uri of the database. If the database uri starts
+with `db://` then it connects to a remote database.
+
+• **opts?**: `Partial`<[`ConnectionOptions`](../interfaces/ConnectionOptions.md) \| `RemoteConnectionOptions`>
+
+### Returns
+
+`Promise`<[`Connection`](../classes/Connection.md)>
+
+### See
+
+[ConnectionOptions](../interfaces/ConnectionOptions.md) for more details on the URI format.
+
+### Examples
+
+```ts
+const conn = await connect("/path/to/database");
+```
+
+```ts
+const conn = await connect(
+  "s3://bucket/path/to/database",
+  {storageOptions: {timeout: "60s"}
+});
+```
+
+## connect(opts)
+
+> **connect**(`opts`): `Promise`<[`Connection`](../classes/Connection.md)>
+
+Connect to a LanceDB instance at the given URI.
+
+Accepted formats:
+
+- `/path/to/database` - local database
+- `s3://bucket/path/to/database` or `gs://bucket/path/to/database` - database on cloud storage
+- `db://host:port` - remote database (LanceDB cloud)
+
+### Parameters
+
+• **opts**: `Partial`<[`ConnectionOptions`](../interfaces/ConnectionOptions.md) \| `RemoteConnectionOptions`> & `object`
+
+### Returns
+
+`Promise`<[`Connection`](../classes/Connection.md)>
+
+### See
+
+[ConnectionOptions](../interfaces/ConnectionOptions.md) for more details on the URI format.
+
+### Example
+
+```ts
+const conn = await connect({
+  uri: "/path/to/database",
+  storageOptions: {timeout: "60s"}
+});
+```

docs/src/js/functions/makeArrowTable.md

@@ -0,0 +1,114 @@
+[**@lancedb/lancedb**](../README.md) • **Docs**
+
+***
+
+[@lancedb/lancedb](../globals.md) / makeArrowTable
+
+# Function: makeArrowTable()
+
+> **makeArrowTable**(`data`, `options`?, `metadata`?): `ArrowTable`
+
+An enhanced version of the makeTable function from Apache Arrow
+that supports nested fields and embeddings columns.
+
+(typically you do not need to call this function.  It will be called automatically
+when creating a table or adding data to it)
+
+This function converts an array of Record<String, any> (row-major JS objects)
+to an Arrow Table (a columnar structure)
+
+Note that it currently does not support nulls.
+
+If a schema is provided then it will be used to determine the resulting array
+types.  Fields will also be reordered to fit the order defined by the schema.
+
+If a schema is not provided then the types will be inferred and the field order
+will be controlled by the order of properties in the first record.  If a type
+is inferred it will always be nullable.
+
+If the input is empty then a schema must be provided to create an empty table.
+
+When a schema is not specified then data types will be inferred.  The inference
+rules are as follows:
+
+ - boolean => Bool
+ - number => Float64
+ - String => Utf8
+ - Buffer => Binary
+ - Record<String, any> => Struct
+ - Array<any> => List
+
+## Parameters
+
+• **data**: `Record`&lt;`string`, `unknown`&gt;[]
+
+• **options?**: `Partial`&lt;[`MakeArrowTableOptions`](../classes/MakeArrowTableOptions.md)&gt;
+
+• **metadata?**: `Map`&lt;`string`, `string`&gt;
+
+## Returns
+
+`ArrowTable`
+
+## Example
+
+import { fromTableToBuffer, makeArrowTable } from "../arrow";
+import { Field, FixedSizeList, Float16, Float32, Int32, Schema } from "apache-arrow";
+
+const schema = new Schema([
+  new Field("a", new Int32()),
+  new Field("b", new Float32()),
+  new Field("c", new FixedSizeList(3, new Field("item", new Float16()))),
+ ]);
+ const table = makeArrowTable([
+   { a: 1, b: 2, c: [1, 2, 3] },
+   { a: 4, b: 5, c: [4, 5, 6] },
+   { a: 7, b: 8, c: [7, 8, 9] },
+ ], { schema });
+```
+
+By default it assumes that the column named `vector` is a vector column
+and it will be converted into a fixed size list array of type float32.
+The `vectorColumns` option can be used to support other vector column
+names and data types.
+
+```ts
+
+const schema = new Schema([
+   new Field("a", new Float64()),
+   new Field("b", new Float64()),
+   new Field(
+     "vector",
+     new FixedSizeList(3, new Field("item", new Float32()))
+   ),
+ ]);
+ const table = makeArrowTable([
+   { a: 1, b: 2, vector: [1, 2, 3] },
+   { a: 4, b: 5, vector: [4, 5, 6] },
+   { a: 7, b: 8, vector: [7, 8, 9] },
+ ]);
+ assert.deepEqual(table.schema, schema);
+```
+
+You can specify the vector column types and names using the options as well
+
+```typescript
+
+const schema = new Schema([
+   new Field('a', new Float64()),
+   new Field('b', new Float64()),
+   new Field('vec1', new FixedSizeList(3, new Field('item', new Float16()))),
+   new Field('vec2', new FixedSizeList(3, new Field('item', new Float16())))
+ ]);
+const table = makeArrowTable([
+   { a: 1, b: 2, vec1: [1, 2, 3], vec2: [2, 4, 6] },
+   { a: 4, b: 5, vec1: [4, 5, 6], vec2: [8, 10, 12] },
+   { a: 7, b: 8, vec1: [7, 8, 9], vec2: [14, 16, 18] }
+ ], {
+   vectorColumns: {
+     vec1: { type: new Float16() },
+     vec2: { type: new Float16() }
+   }
+ }
+assert.deepEqual(table.schema, schema)
+```