feat(extra-backends): Improvements, adding mamba example (#1618)

* feat(extra-backends): Improvements

vllm: add max_tokens, wire up stream event
mamba: fixups, adding examples for mamba-chat

* examples(mamba-chat): add

* docs: update

.env

@@ -23,7 +23,16 @@ MODELS_PATH=/models
 ## Enable debug mode
 # DEBUG=true
 
+## Disables COMPEL (Diffusers)
+# COMPEL=0
+
+## Enable/Disable single backend (useful if only one GPU is available)
+# SINGLE_ACTIVE_BACKEND=true
+
 ## Specify a build type. Available: cublas, openblas, clblas.
+## cuBLAS: This is a GPU-accelerated version of the complete standard BLAS (Basic Linear Algebra Subprograms) library. It's provided by Nvidia and is part of their CUDA toolkit.
+## OpenBLAS: This is an open-source implementation of the BLAS library that aims to provide highly optimized code for various platforms. It includes support for multi-threading and can be compiled to use hardware-specific features for additional performance. OpenBLAS can run on many kinds of hardware, including CPUs from Intel, AMD, and ARM.
+## clBLAS:   This is an open-source implementation of the BLAS library that uses OpenCL, a framework for writing programs that execute across heterogeneous platforms consisting of CPUs, GPUs, and other processors. clBLAS is designed to take advantage of the parallel computing power of GPUs but can also run on any hardware that supports OpenCL. This includes hardware from different vendors like Nvidia, AMD, and Intel.
 # BUILD_TYPE=openblas
 
 ## Uncomment and set to true to enable rebuilding from source
@@ -41,3 +50,40 @@ MODELS_PATH=/models
 
 ## Specify a default upload limit in MB (whisper)
 # UPLOAD_LIMIT
+
+## List of external GRPC backends (note on the container image this variable is already set to use extra backends available in extra/)
+# EXTERNAL_GRPC_BACKENDS=my-backend:127.0.0.1:9000,my-backend2:/usr/bin/backend.py
+
+### Advanced settings ###
+### Those are not really used by LocalAI, but from components in the stack ###
+##
+### Preload libraries
+# LD_PRELOAD=
+
+### Huggingface cache for models
+# HUGGINGFACE_HUB_CACHE=/usr/local/huggingface
+
+### Python backends GRPC max workers
+### Default number of workers for GRPC Python backends.
+### This actually controls wether a backend can process multiple requests or not.
+# PYTHON_GRPC_MAX_WORKERS=1
+
+### Define the number of parallel LLAMA.cpp workers (Defaults to 1)
+# LLAMACPP_PARALLEL=1
+
+### Enable to run parallel requests
+# PARALLEL_REQUESTS=true
+
+### Watchdog settings
+###
+# Enables watchdog to kill backends that are inactive for too much time
+# WATCHDOG_IDLE=true
+#
+# Enables watchdog to kill backends that are busy for too much time
+# WATCHDOG_BUSY=true
+#
+# Time in duration format (e.g. 1h30m) after which a backend is considered idle
+# WATCHDOG_IDLE_TIMEOUT=5m
+#
+# Time in duration format (e.g. 1h30m) after which a backend is considered busy
+# WATCHDOG_BUSY_TIMEOUT=5m

.gitattributes

@@ -0,0 +1 @@
+*.sh text eol=lf

.github/ISSUE_TEMPLATE/bug_report.md

@@ -2,9 +2,7 @@
 name: Bug report
 about: Create a report to help us improve
 title: ''
-labels: bug
-assignees: mudler
-
+labels: bug, unconfirmed, up-for-grabs
 ---
 
 <!-- Thanks for helping us to improve LocalAI! We welcome all bug reports. Please fill out each area of the template so we can better help you. Comments like this will be hidden when you post but you can delete them if you wish. -->

.github/ISSUE_TEMPLATE/feature_request.md

@@ -2,9 +2,7 @@
 name: Feature request
 about: Suggest an idea for this project
 title: ''
-labels: enhancement
-assignees: mudler
-
+labels: enhancement, up-for-grabs
 ---
 
 <!-- Thanks for helping us to improve LocalAI! We welcome all feature requests. Please fill out each area of the template so we can better help you. Comments like this will be hidden when you post but you can delete them if you wish. -->

.github/PULL_REQUEST_TEMPLATE.md

@@ -8,16 +8,24 @@ This PR fixes #
 **[Signed commits](../CONTRIBUTING.md#signing-off-on-commits-developer-certificate-of-origin)**
 - [ ] Yes, I signed my commits.
  
-
 <!--
 Thank you for contributing to LocalAI! 
 
-Contributing Conventions:
+Contributing Conventions
+-------------------------
 
-1. Include descriptive PR titles with [<component-name>] prepended.
-2. Build and test your changes before submitting a PR. 
+The draft above helps to give a quick overview of your PR.
+
+Remember to remove this comment and to at least:
+
+1. Include descriptive PR titles with [<component-name>] prepended. We use [conventional commits](https://www.conventionalcommits.org/en/v1.0.0/).
+2. Build and test your changes before submitting a PR (`make build`). 
 3. Sign your commits
+4. **Tag maintainer:** for a quicker response, tag the relevant maintainer (see below).
+5. **X/Twitter handle:** we announce bigger features on X/Twitter. If your PR gets announced, and you'd like a mention, we'll gladly shout you out!
 
 By following the community's contribution conventions upfront, the review process will 
 be accelerated and your PR merged more quickly.
+
+If no one reviews your PR within a few days, please @-mention @mudler.
 -->

.github/bump_docs.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+set -xe
+REPO=$1
+
+LATEST_TAG=$(curl -s "https://api.github.com/repos/$REPO/releases/latest" | jq -r '.name')
+
+cat <<< $(jq ".version = \"$LATEST_TAG\"" docs/data/version.json) > docs/data/version.json

.github/workflows/bump_deps.yaml

@@ -12,6 +12,9 @@ jobs:
           - repository: "go-skynet/go-llama.cpp"
             variable: "GOLLAMA_VERSION"
             branch: "master"
+          - repository: "ggerganov/llama.cpp"
+            variable: "CPPLLAMA_VERSION"
+            branch: "master"
           - repository: "go-skynet/go-ggml-transformers.cpp"
             variable: "GOGGMLTRANSFORMERS_VERSION"
             branch: "master"
@@ -41,7 +44,7 @@ jobs:
             branch: "master"
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - name: Bump dependencies 🔧
         run: |
           bash .github/bump_deps.sh ${{ matrix.repository }} ${{ matrix.branch }} ${{ matrix.variable }}

.github/workflows/bump_docs.yaml

@@ -0,0 +1,31 @@
+name: Bump dependencies
+on:
+  schedule:
+    - cron: 0 20 * * *
+  workflow_dispatch:
+jobs:
+  bump:
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - repository: "mudler/LocalAI"
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Bump dependencies 🔧
+        run: |
+          bash .github/bump_docs.sh ${{ matrix.repository }}
+      - name: Create Pull Request
+        uses: peter-evans/create-pull-request@v5
+        with:
+          token: ${{ secrets.UPDATE_BOT_TOKEN }}
+          push-to-fork: ci-forks/LocalAI
+          commit-message: ':arrow_up: Update docs version ${{ matrix.repository }}'
+          title: ':arrow_up: Update docs version ${{ matrix.repository }}'
+          branch: "update/docs"
+          body: Bump of ${{ matrix.repository }} version inside docs
+          signoff: true
+
+
+

.github/workflows/disabled/test-gpu.yml

@@ -0,0 +1,63 @@
+---
+name: 'GPU tests'
+
+on:
+  pull_request:
+  push:
+    branches:
+      - master
+    tags:
+      - '*'
+
+concurrency:
+  group: ci-gpu-tests-${{ github.head_ref || github.ref }}-${{ github.repository }}
+  cancel-in-progress: true
+
+jobs:
+  ubuntu-latest:
+    runs-on: gpu
+    strategy:
+      matrix:
+        go-version: ['1.21.x']
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Setup Go ${{ matrix.go-version }}
+        uses: actions/setup-go@v4
+        with:
+          go-version: ${{ matrix.go-version }}
+      # You can test your matrix by printing the current Go version
+      - name: Display Go version
+        run: go version
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo DEBIAN_FRONTEND=noninteractive apt-get install -y make wget
+      - name: Build
+        run: |
+          if [ ! -e /run/systemd/system ]; then
+            sudo mkdir /run/systemd/system
+          fi
+          sudo mkdir -p /host/tests/${{ github.head_ref || github.ref }}
+          sudo chmod -R 777 /host/tests/${{ github.head_ref || github.ref }}
+          make \
+            TEST_DIR="/host/tests/${{ github.head_ref || github.ref }}" \
+            BUILD_TYPE=cublas \
+            prepare-e2e run-e2e-image test-e2e
+      - name: Release space from worker ♻
+        if: always()
+        run: |
+          sudo rm -rf build || true
+          sudo rm -rf bin || true
+          sudo rm -rf dist || true
+          sudo docker logs $(sudo docker ps -q --filter ancestor=localai-tests) > logs.txt
+          sudo cat logs.txt || true
+          sudo rm -rf logs.txt
+          make clean || true
+          make \
+            TEST_DIR="/host/tests/${{ github.head_ref || github.ref }}" \
+            teardown-e2e || true
+          sudo rm -rf /host/tests/${{ github.head_ref || github.ref }} || true
+          docker system prune -f -a --volumes || true

.github/workflows/image-pr.yml

@@ -0,0 +1,86 @@
+---
+name: 'build container images tests'
+
+on:
+  pull_request:
+
+concurrency:
+  group: ci-${{ github.head_ref || github.ref }}-${{ github.repository }}
+  cancel-in-progress: true
+
+jobs:
+  extras-image-build:
+    uses: ./.github/workflows/image_build.yml
+    with:
+      tag-latest: ${{ matrix.tag-latest }}
+      tag-suffix: ${{ matrix.tag-suffix }}
+      ffmpeg: ${{ matrix.ffmpeg }}
+      image-type: ${{ matrix.image-type }}
+      build-type: ${{ matrix.build-type }}
+      cuda-major-version: ${{ matrix.cuda-major-version }}
+      cuda-minor-version: ${{ matrix.cuda-minor-version }}
+      platforms: ${{ matrix.platforms }}
+      runs-on: ${{ matrix.runs-on }}
+    secrets:
+      dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
+      dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
+      quayUsername: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
+      quayPassword: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
+    strategy:
+      # Pushing with all jobs in parallel
+      # eats the bandwidth of all the nodes
+      max-parallel: ${{ github.event_name != 'pull_request' && 2 || 4 }}
+      matrix:
+        include:
+          - build-type: ''
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-ffmpeg'
+            ffmpeg: 'true'
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "1"
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-cublas-cuda12-ffmpeg'
+            ffmpeg: 'true'
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
+  core-image-build:
+    uses: ./.github/workflows/image_build.yml
+    with:
+      tag-latest: ${{ matrix.tag-latest }}
+      tag-suffix: ${{ matrix.tag-suffix }}
+      ffmpeg: ${{ matrix.ffmpeg }}
+      image-type: ${{ matrix.image-type }}
+      build-type: ${{ matrix.build-type }}
+      cuda-major-version: ${{ matrix.cuda-major-version }}
+      cuda-minor-version: ${{ matrix.cuda-minor-version }}
+      platforms: ${{ matrix.platforms }}
+      runs-on: ${{ matrix.runs-on }}
+    secrets:
+      dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
+      dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
+      quayUsername: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
+      quayPassword: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
+    strategy:
+      matrix:
+        include:
+          - build-type: ''
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "1"
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-cublas-cuda12-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'

.github/workflows/image.yml

@@ -2,7 +2,6 @@
 name: 'build container images'
 
 on:
-  pull_request:
   push:
     branches:
       - master
@@ -14,96 +13,148 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-  docker:
+  extras-image-build:
+    uses: ./.github/workflows/image_build.yml
+    with:
+      tag-latest: ${{ matrix.tag-latest }}
+      tag-suffix: ${{ matrix.tag-suffix }}
+      ffmpeg: ${{ matrix.ffmpeg }}
+      image-type: ${{ matrix.image-type }}
+      build-type: ${{ matrix.build-type }}
+      cuda-major-version: ${{ matrix.cuda-major-version }}
+      cuda-minor-version: ${{ matrix.cuda-minor-version }}
+      platforms: ${{ matrix.platforms }}
+      runs-on: ${{ matrix.runs-on }}
+    secrets:
+      dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
+      dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
+      quayUsername: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
+      quayPassword: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
     strategy:
+      # Pushing with all jobs in parallel
+      # eats the bandwidth of all the nodes
+      max-parallel: ${{ github.event_name != 'pull_request' && 2 || 4 }}
       matrix:
         include:
           - build-type: ''
-            platforms: 'linux/amd64,linux/arm64'
+            #platforms: 'linux/amd64,linux/arm64'
+            platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: ''
             ffmpeg: ''
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
+          - build-type: ''
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-ffmpeg'
+            ffmpeg: 'true'
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
           - build-type: 'cublas'
-            cuda-major-version: 11
-            cuda-minor-version: 7
+            cuda-major-version: "11"
+            cuda-minor-version: "7"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda11'
             ffmpeg: ''
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
           - build-type: 'cublas'
-            cuda-major-version: 12
-            cuda-minor-version: 1
+            cuda-major-version: "12"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda12'
             ffmpeg: ''
-          - build-type: ''
-            platforms: 'linux/amd64,linux/arm64'
-            tag-latest: 'false'
-            tag-suffix: '-ffmpeg'
-            ffmpeg: 'true'
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
           - build-type: 'cublas'
-            cuda-major-version: 11
-            cuda-minor-version: 7
+            cuda-major-version: "11"
+            cuda-minor-version: "7"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda11-ffmpeg'
             ffmpeg: 'true'
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
           - build-type: 'cublas'
-            cuda-major-version: 12
-            cuda-minor-version: 1
+            cuda-major-version: "12"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda12-ffmpeg'
             ffmpeg: 'true'
-
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v3
-
-      - name: Docker meta
-        id: meta
-        uses: docker/metadata-action@v4
-        with:
-          images: quay.io/go-skynet/local-ai
-          tags: |
-            type=ref,event=branch
-            type=semver,pattern={{raw}}
-            type=sha
-          flavor: |
-            latest=${{ matrix.tag-latest }}
-            suffix=${{ matrix.tag-suffix }}
-
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@master
-        with:
-          platforms: all
-
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@master
-
-      - name: Login to DockerHub
-        if: github.event_name != 'pull_request'
-        uses: docker/login-action@v2
-        with:
-          registry: quay.io
-          username: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
-          password: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
-
-      - name: Build and push
-        uses: docker/build-push-action@v4
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          build-args: |
-            BUILD_TYPE=${{ matrix.build-type }}
-            CUDA_MAJOR_VERSION=${{ matrix.cuda-major-version }}
-            CUDA_MINOR_VERSION=${{ matrix.cuda-minor-version }}
-            FFMPEG=${{ matrix.ffmpeg }}
-          context: .
-          file: ./Dockerfile
-          platforms: ${{ matrix.platforms }}
-          push: ${{ github.event_name != 'pull_request' }}
-          tags: ${{ steps.meta.outputs.tags }}
-          labels: ${{ steps.meta.outputs.labels }}
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
+          - build-type: ''
+            #platforms: 'linux/amd64,linux/arm64'
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: ''
+            ffmpeg: ''
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
+  core-image-build:
+    uses: ./.github/workflows/image_build.yml
+    with:
+      tag-latest: ${{ matrix.tag-latest }}
+      tag-suffix: ${{ matrix.tag-suffix }}
+      ffmpeg: ${{ matrix.ffmpeg }}
+      image-type: ${{ matrix.image-type }}
+      build-type: ${{ matrix.build-type }}
+      cuda-major-version: ${{ matrix.cuda-major-version }}
+      cuda-minor-version: ${{ matrix.cuda-minor-version }}
+      platforms: ${{ matrix.platforms }}
+      runs-on: ${{ matrix.runs-on }}
+    secrets:
+      dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
+      dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
+      quayUsername: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
+      quayPassword: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
+    strategy:
+      matrix:
+        include:
+          - build-type: ''
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'
+          - build-type: 'cublas'
+            cuda-major-version: "11"
+            cuda-minor-version: "7"
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-cublas-cuda11-core'
+            ffmpeg: ''
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "1"
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-cublas-cuda12-core'
+            ffmpeg: ''
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'
+          - build-type: 'cublas'
+            cuda-major-version: "11"
+            cuda-minor-version: "7"
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-cublas-cuda11-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "1"
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-cublas-cuda12-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'

.github/workflows/image_build.yml

@@ -0,0 +1,160 @@
+---
+name: 'build container images (reusable)'
+
+on:
+  workflow_call:
+    inputs:
+      build-type:
+        description: 'Build type'
+        default: ''
+        type: string
+      cuda-major-version:
+        description: 'CUDA major version'
+        default: "11"
+        type: string
+      cuda-minor-version:
+        description: 'CUDA minor version'
+        default: "7"
+        type: string
+      platforms:
+        description: 'Platforms'
+        default: ''
+        type: string
+      tag-latest:
+        description: 'Tag latest'
+        default: ''
+        type: string
+      tag-suffix:
+        description: 'Tag suffix'
+        default: ''
+        type: string
+      ffmpeg:
+        description: 'FFMPEG'
+        default: ''
+        type: string
+      image-type:
+        description: 'Image type'
+        default: ''
+        type: string
+      runs-on:
+        description: 'Runs on'
+        required: true
+        default: ''
+        type: string
+    secrets:
+      dockerUsername:
+        required: true
+      dockerPassword:
+        required: true
+      quayUsername:
+        required: true
+      quayPassword:
+        required: true
+jobs:
+  reusable_image-build:
+    runs-on: ${{ inputs.runs-on }}
+    steps:
+      - name: Force Install GIT latest
+        run: |
+          sudo apt-get update \
+          && sudo apt-get install -y software-properties-common \
+          && sudo apt-get update \
+          && sudo add-apt-repository -y ppa:git-core/ppa \
+          && sudo apt-get update \
+          && sudo apt-get install -y git
+      - name: Checkout
+        uses: actions/checkout@v4
+      # - name: Release space from worker
+      #   run: |
+      #     echo "Listing top largest packages"
+      #     pkgs=$(dpkg-query -Wf '${Installed-Size}\t${Package}\t${Status}\n' | awk '$NF == "installed"{print $1 "\t" $2}' | sort -nr)
+      #     head -n 30 <<< "${pkgs}"
+      #     echo
+      #     df -h
+      #     echo
+      #     sudo apt-get remove -y '^llvm-.*|^libllvm.*' || true
+      #     sudo apt-get remove --auto-remove android-sdk-platform-tools || true
+      #     sudo apt-get purge --auto-remove android-sdk-platform-tools || true
+      #     sudo rm -rf /usr/local/lib/android
+      #     sudo apt-get remove -y '^dotnet-.*|^aspnetcore-.*' || true
+      #     sudo rm -rf /usr/share/dotnet
+      #     sudo apt-get remove -y '^mono-.*' || true
+      #     sudo apt-get remove -y '^ghc-.*' || true
+      #     sudo apt-get remove -y '.*jdk.*|.*jre.*' || true
+      #     sudo apt-get remove -y 'php.*' || true
+      #     sudo apt-get remove -y hhvm powershell firefox monodoc-manual msbuild || true
+      #     sudo apt-get remove -y '^google-.*' || true
+      #     sudo apt-get remove -y azure-cli || true
+      #     sudo apt-get remove -y '^mongo.*-.*|^postgresql-.*|^mysql-.*|^mssql-.*' || true
+      #     sudo apt-get remove -y '^gfortran-.*' || true
+      #     sudo apt-get remove -y microsoft-edge-stable || true
+      #     sudo apt-get remove -y firefox || true
+      #     sudo apt-get remove -y powershell || true
+      #     sudo apt-get remove -y r-base-core || true
+      #     sudo apt-get autoremove -y
+      #     sudo apt-get clean
+      #     echo
+      #     echo "Listing top largest packages"
+      #     pkgs=$(dpkg-query -Wf '${Installed-Size}\t${Package}\t${Status}\n' | awk '$NF == "installed"{print $1 "\t" $2}' | sort -nr)
+      #     head -n 30 <<< "${pkgs}"
+      #     echo
+      #     sudo rm -rfv build || true
+      #     df -h
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            quay.io/go-skynet/local-ai
+            localai/localai
+          tags: |
+            type=ref,event=branch
+            type=semver,pattern={{raw}}
+            type=sha
+          flavor: |
+            latest=${{ inputs.tag-latest }}
+            suffix=${{ inputs.tag-suffix }}
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@master
+        with:
+          platforms: all
+
+      - name: Set up Docker Buildx
+        id: buildx
+        uses: docker/setup-buildx-action@master
+
+      - name: Login to DockerHub
+        if: github.event_name != 'pull_request'
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.dockerUsername }}
+          password: ${{ secrets.dockerPassword }}
+
+      - name: Login to DockerHub
+        if: github.event_name != 'pull_request'
+        uses: docker/login-action@v3
+        with:
+          registry: quay.io
+          username: ${{ secrets.quayUsername }}
+          password: ${{ secrets.quayPassword }}
+
+      - name: Build and push
+        uses: docker/build-push-action@v5
+        with:
+          builder: ${{ steps.buildx.outputs.name }}
+          build-args: |
+            BUILD_TYPE=${{ inputs.build-type }}
+            CUDA_MAJOR_VERSION=${{ inputs.cuda-major-version }}
+            CUDA_MINOR_VERSION=${{ inputs.cuda-minor-version }}
+            FFMPEG=${{ inputs.ffmpeg }}
+            IMAGE_TYPE=${{ inputs.image-type }}
+          context: .
+          file: ./Dockerfile
+          platforms: ${{ inputs.platforms }}
+          push: ${{ github.event_name != 'pull_request' }}
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+      - name: job summary
+        run: |
+          echo "Built image: ${{ steps.meta.outputs.labels }}" >> $GITHUB_STEP_SUMMARY

.github/workflows/release.yaml

@@ -5,6 +5,10 @@ on: push
 permissions:
   contents: write
 
+concurrency:
+  group: ci-releases-${{ github.head_ref || github.ref }}-${{ github.repository }}
+  cancel-in-progress: true
+
 jobs:
   build-linux:
     strategy:
@@ -19,13 +23,34 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Clone
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
         with:
           submodules: true
+      - uses: actions/setup-go@v4
+        with:
+          go-version: '>=1.21.0'
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
+
+      - name: Cache grpc
+        id: cache-grpc
+        uses: actions/cache@v3
+        with:
+          path: grpc
+          key: ${{ runner.os }}-grpc
+      - name: Build grpc
+        if: steps.cache-grpc.outputs.cache-hit != 'true'
+        run: |
+          git clone --recurse-submodules -b v1.58.0 --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
+          cd grpc && mkdir -p cmake/build && cd cmake/build && cmake -DgRPC_INSTALL=ON \
+            -DgRPC_BUILD_TESTS=OFF \
+            ../.. && sudo make -j12
+      - name: Install gRPC
+        run: |
+          cd grpc && cd cmake/build && sudo make -j12 install
+
       - name: Build
         id: build
         env:
@@ -57,15 +82,23 @@ jobs:
     runs-on: macOS-latest
     steps:
       - name: Clone
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
         with:
           submodules: true
+      - uses: actions/setup-go@v4
+        with:
+          go-version: '>=1.21.0'
+      - name: Dependencies
+        run: |
+          brew install protobuf grpc
       - name: Build
         id: build
         env:
           CMAKE_ARGS: "${{ matrix.defines }}"
           BUILD_ID: "${{ matrix.build }}"
         run: |
+          export C_INCLUDE_PATH=/usr/local/include
+          export CPLUS_INCLUDE_PATH=/usr/local/include
           make dist
       - uses: actions/upload-artifact@v3
         with:

.github/workflows/test-extra.yml

@@ -0,0 +1,277 @@
+---
+name: 'Tests extras backends'
+
+on:
+  pull_request:
+  push:
+    branches:
+      - master
+    tags:
+      - '*'
+
+concurrency:
+  group: ci-tests-extra-${{ github.head_ref || github.ref }}-${{ github.repository }}
+  cancel-in-progress: true
+
+jobs:
+  tests-transformers:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test transformers
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/transformers
+           make -C backend/python/transformers test
+
+  tests-sentencetransformers:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test sentencetransformers
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/sentencetransformers
+           make -C backend/python/sentencetransformers test
+
+  tests-diffusers:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test diffusers
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/diffusers
+           make -C backend/python/diffusers test
+
+
+  tests-transformers-musicgen:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test transformers-musicgen
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/transformers-musicgen
+           make -C backend/python/transformers-musicgen test
+
+
+
+  tests-petals:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test petals
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/petals
+           make -C backend/python/petals test
+
+           
+
+  tests-bark:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test bark
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/bark
+           make -C backend/python/bark test
+
+           
+  # Below tests needs GPU. Commented out for now
+  # TODO: Re-enable as soon as we have GPU nodes
+  # tests-vllm:
+  #   runs-on: ubuntu-latest
+  #   steps:
+  #     - name: Clone
+  #       uses: actions/checkout@v4
+  #       with: 
+  #         submodules: true
+  #     - name: Dependencies
+  #       run: |
+  #         sudo apt-get update
+  #         sudo apt-get install build-essential ffmpeg
+  #         curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+  #            sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+  #             gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+  #            sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+  #            sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+  #            sudo apt-get update && \
+  #            sudo apt-get install -y conda
+  #         sudo apt-get install -y ca-certificates cmake curl patch
+  #         sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+  #         sudo rm -rfv /usr/bin/conda || true
+  #     - name: Test vllm
+  #       run: |
+  #          export PATH=$PATH:/opt/conda/bin
+  #          make -C backend/python/vllm
+  #          make -C backend/python/vllm test
+  tests-vallex:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2    
+          sudo rm -rfv /usr/bin/conda || true
+      - name: Test vall-e-x
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/vall-e-x
+           make -C backend/python/vall-e-x test
+
+  tests-coqui:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with: 
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential ffmpeg
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
+          sudo apt-get install -y ca-certificates cmake curl patch espeak espeak-ng          
+          sudo rm -rfv /usr/bin/conda || true
+
+      - name: Test coqui
+        run: |
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/coqui
+           make -C backend/python/coqui test

.github/workflows/test.yml

@@ -14,51 +14,120 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-  ubuntu-latest:
+  tests-linux:
     runs-on: ubuntu-latest
-
+    strategy:
+      matrix:
+        go-version: ['1.21.x']
     steps:
+      - name: Release space from worker
+        run: |
+          echo "Listing top largest packages"
+          pkgs=$(dpkg-query -Wf '${Installed-Size}\t${Package}\t${Status}\n' | awk '$NF == "installed"{print $1 "\t" $2}' | sort -nr)
+          head -n 30 <<< "${pkgs}"
+          echo
+          df -h
+          echo
+          sudo apt-get remove -y '^llvm-.*|^libllvm.*' || true
+          sudo apt-get remove --auto-remove android-sdk-platform-tools || true
+          sudo apt-get purge --auto-remove android-sdk-platform-tools || true
+          sudo rm -rf /usr/local/lib/android
+          sudo apt-get remove -y '^dotnet-.*|^aspnetcore-.*' || true
+          sudo rm -rf /usr/share/dotnet
+          sudo apt-get remove -y '^mono-.*' || true
+          sudo apt-get remove -y '^ghc-.*' || true
+          sudo apt-get remove -y '.*jdk.*|.*jre.*' || true
+          sudo apt-get remove -y 'php.*' || true
+          sudo apt-get remove -y hhvm powershell firefox monodoc-manual msbuild || true
+          sudo apt-get remove -y '^google-.*' || true
+          sudo apt-get remove -y azure-cli || true
+          sudo apt-get remove -y '^mongo.*-.*|^postgresql-.*|^mysql-.*|^mssql-.*' || true
+          sudo apt-get remove -y '^gfortran-.*' || true
+          sudo apt-get autoremove -y
+          sudo apt-get clean
+          echo
+          echo "Listing top largest packages"
+          pkgs=$(dpkg-query -Wf '${Installed-Size}\t${Package}\t${Status}\n' | awk '$NF == "installed"{print $1 "\t" $2}' | sort -nr)
+          head -n 30 <<< "${pkgs}"
+          echo
+          sudo rm -rfv build || true
+          df -h
       - name: Clone
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
         with: 
           submodules: true
+      - name: Setup Go ${{ matrix.go-version }}
+        uses: actions/setup-go@v4
+        with:
+          go-version: ${{ matrix.go-version }}
+      # You can test your matrix by printing the current Go version
+      - name: Display Go version
+        run: go version
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
-          
+          curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+             sudo install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+              gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list' && \
+             sudo /bin/bash -c 'echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list' && \
+             sudo apt-get update && \
+             sudo apt-get install -y conda
           sudo apt-get install -y ca-certificates cmake curl patch
           sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
+          sudo rm -rfv /usr/bin/conda || true
+          PATH=$PATH:/opt/conda/bin make -C backend/python/sentencetransformers
 
-          sudo mkdir /build && sudo chmod -R 777 /build && cd /build && \
-          curl -L "https://github.com/gabime/spdlog/archive/refs/tags/v1.11.0.tar.gz" | \
-          tar -xzvf - && \
-          mkdir -p "spdlog-1.11.0/build" && \
-          cd "spdlog-1.11.0/build" && \
-          cmake ..  && \
-          make -j8 && \
-          sudo cmake --install . --prefix /usr && mkdir -p "lib/Linux-$(uname -m)" && \
-          cd /build && \
-          mkdir -p "lib/Linux-$(uname -m)/piper_phonemize" && \
-          curl -L "https://github.com/rhasspy/piper-phonemize/releases/download/v1.0.0/libpiper_phonemize-amd64.tar.gz" | \
-          tar -C "lib/Linux-$(uname -m)/piper_phonemize" -xzvf - && ls -liah /build/lib/Linux-$(uname -m)/piper_phonemize/ && \
-          sudo cp -rfv /build/lib/Linux-$(uname -m)/piper_phonemize/lib/. /lib64/ && \
-          sudo cp -rfv /build/lib/Linux-$(uname -m)/piper_phonemize/lib/. /usr/lib/ && \
-          sudo cp -rfv /build/lib/Linux-$(uname -m)/piper_phonemize/include/. /usr/include/
-    
+          # Pre-build piper before we start tests in order to have shared libraries in place
+          make sources/go-piper && \
+          GO_TAGS="tts" make -C sources/go-piper piper.o && \
+          sudo cp -rfv sources/go-piper/piper-phonemize/pi/lib/. /usr/lib/ && \
+          # Pre-build stable diffusion before we install a newer version of abseil (not compatible with stablediffusion-ncn)
+          GO_TAGS="stablediffusion tts" GRPC_BACKENDS=backend-assets/grpc/stablediffusion make build
+      - name: Cache grpc
+        id: cache-grpc
+        uses: actions/cache@v3
+        with:
+          path: grpc
+          key: ${{ runner.os }}-grpc
+      - name: Build grpc
+        if: steps.cache-grpc.outputs.cache-hit != 'true'
+        run: |
+          git clone --recurse-submodules -b v1.58.0 --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
+          cd grpc && mkdir -p cmake/build && cd cmake/build && cmake -DgRPC_INSTALL=ON \
+            -DgRPC_BUILD_TESTS=OFF \
+            ../.. && sudo make -j12
+      - name: Install gRPC
+        run: |
+          cd grpc && cd cmake/build && sudo make -j12 install
       - name: Test
         run: |
-          ESPEAK_DATA="/build/lib/Linux-$(uname -m)/piper_phonemize/lib/espeak-ng-data" GO_TAGS="tts stablediffusion" make test
+          GO_TAGS="stablediffusion tts" make test
 
-  macOS-latest:
+  tests-apple:
     runs-on: macOS-latest
-
+    strategy:
+      matrix:
+        go-version: ['1.21.x']
     steps:
       - name: Clone
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
         with: 
           submodules: true
-
+      - name: Setup Go ${{ matrix.go-version }}
+        uses: actions/setup-go@v4
+        with:
+          go-version: ${{ matrix.go-version }}
+      # You can test your matrix by printing the current Go version
+      - name: Display Go version
+        run: go version
+      - name: Dependencies
+        run: |
+          brew install protobuf grpc
       - name: Test
         run: |
+          export C_INCLUDE_PATH=/usr/local/include
+          export CPLUS_INCLUDE_PATH=/usr/local/include
           CMAKE_ARGS="-DLLAMA_F16C=OFF -DLLAMA_AVX512=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF" make test

.gitignore

@@ -1,13 +1,11 @@
 # go-llama build artifacts
-go-llama
-/gpt4all
-go-stable-diffusion
-go-piper
-go-ggllm
-/piper
-
+/sources/
+__pycache__/
 *.a
 get-sources
+prepare-sources
+/backend/cpp/llama/grpc-server
+/backend/cpp/llama/llama.cpp
 
 go-ggml-transformers
 go-gpt2
@@ -21,6 +19,8 @@ LocalAI
 local-ai
 # prevent above rules from omitting the helm chart
 !charts/*
+# prevent above rules from omitting the api/localai folder
+!api/localai
 
 # Ignore models
 models/*
@@ -35,5 +35,5 @@ release/
 
 # Generated during build
 backend-assets/
-
+prepare
 /ggml-metal.metal

.gitmodules

@@ -0,0 +1,6 @@
+[submodule "docs/themes/hugo-theme-relearn"]
+	path = docs/themes/hugo-theme-relearn
+	url = https://github.com/McShelby/hugo-theme-relearn.git
+[submodule "docs/themes/lotusdocs"]
+	path = docs/themes/lotusdocs
+	url = https://github.com/colinwilson/lotusdocs

CONTRIBUTING.md

@@ -0,0 +1,72 @@
+# Contributing to localAI
+
+Thank you for your interest in contributing to LocalAI! We appreciate your time and effort in helping to improve our project. Before you get started, please take a moment to review these guidelines.
+
+## Table of Contents
+
+- [Getting Started](#getting-started)
+  - [Prerequisites](#prerequisites)
+  - [Setting up the Development Environment](#setting-up-the-development-environment)
+- [Contributing](#contributing)
+  - [Submitting an Issue](#submitting-an-issue)
+  - [Creating a Pull Request (PR)](#creating-a-pull-request-pr)
+- [Coding Guidelines](#coding-guidelines)
+- [Testing](#testing)
+- [Documentation](#documentation)
+- [Community and Communication](#community-and-communication)
+
+
+
+## Getting Started
+
+### Prerequisites
+
+- Golang [1.21]
+- Git
+- macOS/Linux
+
+### Setting up the Development Environment and running localAI in the local environment
+
+1. Clone the repository: `git clone https://github.com/go-skynet/LocalAI.git`
+2. Navigate to the project directory: `cd LocalAI`
+3. Install the required dependencies: `make prepare`
+4. Run LocalAI: `make run`
+
+## Contributing
+
+We welcome contributions from everyone! To get started, follow these steps:
+
+### Submitting an Issue
+
+If you find a bug, have a feature request, or encounter any issues, please check the [issue tracker](https://github.com/go-skynet/LocalAI/issues) to see if a similar issue has already been reported. If not, feel free to [create a new issue](https://github.com/go-skynet/LocalAI/issues/new) and provide as much detail as possible.
+
+### Creating a Pull Request (PR)
+
+1. Fork the repository.
+2. Create a new branch with a descriptive name: `git checkout -b [branch name]`
+3. Make your changes and commit them.
+4. Push the changes to your fork: `git push origin [branch name]`
+5. Create a new pull request from your branch to the main project's `main` or `master` branch.
+6. Provide a clear description of your changes in the pull request.
+7. Make any requested changes during the review process.
+8. Once your PR is approved, it will be merged into the main project.
+
+## Coding Guidelines
+
+- No specific coding guidelines at the moment. Please make sure the code can be tested. The most popular lint tools like []`golangci-lint`](https://golangci-lint.run) can help you here.
+
+## Testing
+
+`make test` cannot handle all the model now. Please be sure to add a test case for the new features or the part was changed.
+
+## Documentation
+
+- We are welcome the contribution of the documents, please open new PR in the official document repo [localai-website](https://github.com/go-skynet/localai-website)
+
+## Community and Communication
+
+- You can reach out via the Github issue tracker.
+- Open a new discussion at [Discussion](https://github.com/go-skynet/LocalAI/discussions)
+- Join the Discord channel [Discord](https://discord.gg/uJAeKSAGDy)
+
+---

Dockerfile

@@ -1,20 +1,32 @@
-ARG GO_VERSION=1.20-bullseye
+ARG GO_VERSION=1.21-bullseye
+ARG IMAGE_TYPE=extras
+# extras or core
 
-FROM golang:$GO_VERSION as requirements
+
+FROM golang:$GO_VERSION as requirements-core
 
 ARG BUILD_TYPE
 ARG CUDA_MAJOR_VERSION=11
 ARG CUDA_MINOR_VERSION=7
-ARG SPDLOG_VERSION="1.11.0"
-ARG PIPER_PHONEMIZE_VERSION='1.0.0'
 ARG TARGETARCH
 ARG TARGETVARIANT
 
 ENV BUILD_TYPE=${BUILD_TYPE}
-ARG GO_TAGS="stablediffusion tts"
+
+ENV EXTERNAL_GRPC_BACKENDS="coqui:/build/backend/python/coqui/run.sh,huggingface-embeddings:/build/backend/python/sentencetransformers/run.sh,petals:/build/backend/python/petals/run.sh,transformers:/build/backend/python/transformers/run.sh,sentencetransformers:/build/backend/python/sentencetransformers/run.sh,autogptq:/build/backend/python/autogptq/run.sh,bark:/build/backend/python/bark/run.sh,diffusers:/build/backend/python/diffusers/run.sh,exllama:/build/backend/python/exllama/run.sh,vall-e-x:/build/backend/python/vall-e-x/run.sh,vllm:/build/backend/python/vllm/run.sh,mamba:/build/backend/python/mamba/run.sh,exllama2:/build/backend/python/exllama2/run.sh,transformers-musicgen:/build/backend/python/transformers-musicgen/run.sh"
+
+ARG GO_TAGS="stablediffusion tinydream tts"
 
 RUN apt-get update && \
-    apt-get install -y ca-certificates cmake curl patch
+    apt-get install -y ca-certificates curl patch pip cmake && apt-get clean
+
+
+COPY --chmod=644 custom-ca-certs/* /usr/local/share/ca-certificates/
+RUN update-ca-certificates
+
+# Use the variables in subsequent instructions
+RUN echo "Target Architecture: $TARGETARCH"
+RUN echo "Target Variant: $TARGETVARIANT"
 
 # CuBLAS requirements
 RUN if [ "${BUILD_TYPE}" = "cublas" ]; then \
@@ -24,53 +36,49 @@ RUN if [ "${BUILD_TYPE}" = "cublas" ]; then \
     dpkg -i cuda-keyring_1.0-1_all.deb && \
     rm -f cuda-keyring_1.0-1_all.deb && \
     apt-get update && \
-    apt-get install -y cuda-nvcc-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} libcublas-dev-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} \
+    apt-get install -y cuda-nvcc-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} libcublas-dev-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} libcusparse-dev-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} libcusolver-dev-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION}  && apt-get clean \
     ; fi
 ENV PATH /usr/local/cuda/bin:${PATH}
 
+# OpenBLAS requirements and stable diffusion
+RUN apt-get install -y \
+    libopenblas-dev \
+    libopencv-dev \ 
+    && apt-get clean
+
+# Set up OpenCV
+RUN ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+
 WORKDIR /build
 
-# OpenBLAS requirements
-RUN apt-get install -y libopenblas-dev
-
-# Stable Diffusion requirements
-RUN apt-get install -y libopencv-dev && \
-    ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
-
-# Use the variables in subsequent instructions
-RUN echo "Target Architecture: $TARGETARCH"
-RUN echo "Target Variant: $TARGETVARIANT"
-
-# piper requirements
-# Use pre-compiled Piper phonemization library (includes onnxruntime)
-#RUN if echo "${GO_TAGS}" | grep -q "tts"; then \
 RUN test -n "$TARGETARCH" \
     || (echo 'warn: missing $TARGETARCH, either set this `ARG` manually, or run using `docker buildkit`')
 
-RUN curl -L "https://github.com/gabime/spdlog/archive/refs/tags/v${SPDLOG_VERSION}.tar.gz" | \
-    tar -xzvf - && \
-    mkdir -p "spdlog-${SPDLOG_VERSION}/build" && \
-    cd "spdlog-${SPDLOG_VERSION}/build" && \
-    cmake ..  && \
-    make -j8 && \
-    cmake --install . --prefix /usr && mkdir -p "lib/Linux-$(uname -m)" && \
-    cd /build && \
-    mkdir -p "lib/Linux-$(uname -m)/piper_phonemize" && \
-    curl -L "https://github.com/rhasspy/piper-phonemize/releases/download/v${PIPER_PHONEMIZE_VERSION}/libpiper_phonemize-${TARGETARCH:-$(go env GOARCH)}${TARGETVARIANT}.tar.gz" | \
-    tar -C "lib/Linux-$(uname -m)/piper_phonemize" -xzvf - && ls -liah /build/lib/Linux-$(uname -m)/piper_phonemize/ && \
-    cp -rfv /build/lib/Linux-$(uname -m)/piper_phonemize/lib/. /lib64/ && \
-    cp -rfv /build/lib/Linux-$(uname -m)/piper_phonemize/lib/. /usr/lib/ && \
-    cp -rfv /build/lib/Linux-$(uname -m)/piper_phonemize/include/. /usr/include/
-# \
-#    ; fi
+# Extras requirements
+FROM requirements-core as requirements-extras
+
+RUN curl https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | gpg --dearmor > conda.gpg && \
+    install -o root -g root -m 644 conda.gpg /usr/share/keyrings/conda-archive-keyring.gpg && \
+    gpg --keyring /usr/share/keyrings/conda-archive-keyring.gpg --no-default-keyring --fingerprint 34161F5BF5EB1D4BFBBB8F0A8AEB4F8B29D82806 && \
+    echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" > /etc/apt/sources.list.d/conda.list && \
+    echo "deb [arch=amd64 signed-by=/usr/share/keyrings/conda-archive-keyring.gpg] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | tee -a /etc/apt/sources.list.d/conda.list && \
+    apt-get update && \
+    apt-get install -y conda && apt-get clean
+
+ENV PATH="/root/.cargo/bin:${PATH}"
+RUN pip install --upgrade pip
+RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y
+RUN apt-get install -y espeak-ng espeak && apt-get clean
 
 ###################################
 ###################################
 
-FROM requirements as builder
+FROM requirements-${IMAGE_TYPE} as builder
 
 ARG GO_TAGS="stablediffusion tts"
-
+ARG GRPC_BACKENDS
+ARG BUILD_GRPC=true
+ENV GRPC_BACKENDS=${GRPC_BACKENDS}
 ENV GO_TAGS=${GO_TAGS}
 ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
 ENV NVIDIA_REQUIRE_CUDA="cuda>=${CUDA_MAJOR_VERSION}.0"
@@ -78,28 +86,51 @@ ENV NVIDIA_VISIBLE_DEVICES=all
 
 WORKDIR /build
 
-COPY Makefile .
-RUN make get-sources
-COPY go.mod .
-RUN make prepare
 COPY . .
 COPY .git .
+RUN make prepare
 
-RUN ESPEAK_DATA=/build/lib/Linux-$(uname -m)/piper_phonemize/lib/espeak-ng-data make build
+# stablediffusion does not tolerate a newer version of abseil, build it first
+RUN GRPC_BACKENDS=backend-assets/grpc/stablediffusion make build
+
+RUN if [ "${BUILD_GRPC}" = "true" ]; then \
+    git clone --recurse-submodules -b v1.58.0 --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
+    cd grpc && mkdir -p cmake/build && cd cmake/build && cmake -DgRPC_INSTALL=ON \
+      -DgRPC_BUILD_TESTS=OFF \
+       ../.. && make -j12 install \
+    ; fi
+
+# Rebuild with defaults backends
+RUN make build
+
+RUN if [ ! -d "/build/sources/go-piper/piper-phonemize/pi/lib/" ]; then \
+    mkdir -p /build/sources/go-piper/piper-phonemize/pi/lib/ \
+    touch /build/sources/go-piper/piper-phonemize/pi/lib/keep \
+    ; fi
 
 ###################################
 ###################################
 
-FROM requirements
+FROM requirements-${IMAGE_TYPE}
 
 ARG FFMPEG
+ARG BUILD_TYPE
+ARG TARGETARCH
+ARG IMAGE_TYPE=extras
 
+ENV BUILD_TYPE=${BUILD_TYPE}
 ENV REBUILD=false
 ENV HEALTHCHECK_ENDPOINT=http://localhost:8080/readyz
 
+ARG CUDA_MAJOR_VERSION=11
+ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
+ENV NVIDIA_REQUIRE_CUDA="cuda>=${CUDA_MAJOR_VERSION}.0"
+ENV NVIDIA_VISIBLE_DEVICES=all
+ENV PIP_CACHE_PURGE=true
+
 # Add FFmpeg
 RUN if [ "${FFMPEG}" = "true" ]; then \
-    apt-get install -y ffmpeg \
+    apt-get install -y ffmpeg && apt-get clean \
     ; fi
 
 WORKDIR /build
@@ -109,9 +140,65 @@ WORKDIR /build
 # see https://github.com/go-skynet/LocalAI/pull/658#discussion_r1241971626 and
 # https://github.com/go-skynet/LocalAI/pull/434
 COPY . .
-RUN make prepare-sources
+
+COPY --from=builder /build/sources ./sources/
+COPY --from=builder /build/grpc ./grpc/
+
+RUN make prepare-sources && cd /build/grpc/cmake/build && make install && rm -rf grpc
+
+# Copy the binary
 COPY --from=builder /build/local-ai ./
 
+# Copy shared libraries for piper
+COPY --from=builder /build/sources/go-piper/piper-phonemize/pi/lib/* /usr/lib/
+
+# do not let stablediffusion rebuild (requires an older version of absl)
+COPY --from=builder /build/backend-assets/grpc/stablediffusion ./backend-assets/grpc/stablediffusion
+
+## Duplicated from Makefile to avoid having a big layer that's hard to push
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/autogptq \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/bark \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/diffusers \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/vllm \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/mamba \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/sentencetransformers \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/transformers \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/vall-e-x \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/exllama \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+    PATH=$PATH:/opt/conda/bin make -C backend/python/exllama2 \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/petals \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/transformers-musicgen \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	PATH=$PATH:/opt/conda/bin make -C backend/python/coqui \
+    ; fi
+
+# Make sure the models directory exists
+RUN mkdir -p /build/models
+
 # Define the health check command
 HEALTHCHECK --interval=1m --timeout=10m --retries=10 \
   CMD curl -f $HEALTHCHECK_ENDPOINT || exit 1

Entitlements.plist

@@ -0,0 +1,10 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<dict>
+    <key>com.apple.security.network.client</key>
+    <true/>
+    <key>com.apple.security.network.server</key>
+    <true/>
+</dict>
+</plist>

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2023 Ettore Di Giacinto
+Copyright (c) 2023-2024 Ettore Di Giacinto (mudler@localai.io)
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

Makefile

@@ -4,52 +4,51 @@ GOVET=$(GOCMD) vet
 BINARY_NAME=local-ai
 
 # llama.cpp versions
-# Temporarly pinned to https://github.com/go-skynet/go-llama.cpp/pull/124
-GOLLAMA_VERSION?=cb8d7cd4cb95725a04504a9e3a26dd72a12b69ac
+GOLLAMA_VERSION?=aeba71ee842819da681ea537e78846dc75949ac0
 
-# Temporary set a specific version of llama.cpp
-# containing: https://github.com/ggerganov/llama.cpp/pull/1773 and
-# rebased on top of master.
-# This pin can be dropped when the PR above is merged, and go-llama has merged changes as well
-# Set empty to use the version pinned by go-llama
-LLAMA_CPP_REPO?=https://github.com/mudler/llama.cpp
-LLAMA_CPP_VERSION?=48ce8722a05a018681634af801fd0fd45b3a87cc
+GOLLAMA_STABLE_VERSION?=50cee7712066d9e38306eccadcfbb44ea87df4b7
+
+CPPLLAMA_VERSION?=381ee195721d8e747ee31a60c0751822b3072f02
 
 # gpt4all version
 GPT4ALL_REPO?=https://github.com/nomic-ai/gpt4all
-GPT4ALL_VERSION?=cfd70b69fcf5e587b8e0e3e9b9aaa90e19cbbc51
+GPT4ALL_VERSION?=27a8b020c36b0df8f8b82a252d261cda47cf44b8
 
 # go-ggml-transformers version
 GOGGMLTRANSFORMERS_VERSION?=ffb09d7dd71e2cbc6c5d7d05357d230eea6f369a
 
 # go-rwkv version
 RWKV_REPO?=https://github.com/donomii/go-rwkv.cpp
-RWKV_VERSION?=c898cd0f62df8f2a7830e53d1d513bef4f6f792b
+RWKV_VERSION?=633c5a3485c403cb2520693dc0991a25dace9f0f
 
 # whisper.cpp version
-WHISPER_CPP_VERSION?=85ed71aaec8e0612a84c0b67804bde75aa75a273
+WHISPER_CPP_VERSION?=37a709f6558c6d9783199e2b8cbb136e1c41d346
 
 # bert.cpp version
-BERT_VERSION?=6069103f54b9969c02e789d0fb12a23bd614285f
+BERT_VERSION?=6abe312cded14042f6b7c3cd8edf082713334a4d
 
 # go-piper version
-PIPER_VERSION?=56b8a81b4760a6fbee1a82e62f007ae7e8f010a7
-
-# go-bloomz version
-BLOOMZ_VERSION?=1834e77b83faafe912ad4092ccf7f77937349e2f
+PIPER_VERSION?=d6b6275ba037dabdba4a8b65dfdf6b2a73a67f07
 
 # stablediffusion version
-STABLEDIFFUSION_VERSION?=d89260f598afb809279bc72aa0107b4292587632
+STABLEDIFFUSION_VERSION?=902db5f066fd137697e3b69d0fa10d4782bd2c2f
 
-# Go-ggllm
-GOGGLLM_VERSION?=862477d16eefb0805261c19c9b0d053e3b2b684b
+# tinydream version
+TINYDREAM_VERSION?=772a9c0d9aaf768290e63cca3c904fe69faf677a
 
 export BUILD_TYPE?=
+export STABLE_BUILD_TYPE?=$(BUILD_TYPE)
+export CMAKE_ARGS?=
+
 CGO_LDFLAGS?=
 CUDA_LIBPATH?=/usr/local/cuda/lib64/
 GO_TAGS?=
 BUILD_ID?=git
 
+TEST_DIR=/tmp/test
+
+RANDOM := $(shell bash -c 'echo $$RANDOM')
+
 VERSION?=$(shell git describe --always --tags || echo "dev" )
 # go tool nm ./local-ai | grep Commit
 LD_FLAGS?=
@@ -57,7 +56,6 @@ override LD_FLAGS += -X "github.com/go-skynet/LocalAI/internal.Version=$(VERSION
 override LD_FLAGS += -X "github.com/go-skynet/LocalAI/internal.Commit=$(shell git rev-parse HEAD)"
 
 OPTIONAL_TARGETS?=
-ESPEAK_DATA?=
 
 OS := $(shell uname -s)
 ARCH := $(shell uname -m)
@@ -67,31 +65,61 @@ WHITE  := $(shell tput -Txterm setaf 7)
 CYAN   := $(shell tput -Txterm setaf 6)
 RESET  := $(shell tput -Txterm sgr0)
 
+# Default Docker bridge IP
+E2E_BRIDGE_IP?=172.17.0.1
+
 ifndef UNAME_S
 UNAME_S := $(shell uname -s)
 endif
 
-# workaround for rwkv.cpp
-ifeq ($(UNAME_S),Darwin)
-        CGO_LDFLAGS += -lcblas -framework Accelerate 
+ifeq ($(OS),Darwin)
+	CGO_LDFLAGS += -lcblas -framework Accelerate
+	ifeq ($(OSX_SIGNING_IDENTITY),)
+		OSX_SIGNING_IDENTITY := $(shell security find-identity -v -p codesigning | grep '"' | head -n 1 | sed -E 's/.*"(.*)"/\1/')
+	endif
+
+	# on OSX, if BUILD_TYPE is blank, we should default to use Metal
+	ifeq ($(BUILD_TYPE),)
+		BUILD_TYPE=metal
+	# disable metal if on Darwin and any other value is explicitly passed.
+	else ifneq ($(BUILD_TYPE),metal)
+		CMAKE_ARGS+=-DLLAMA_METAL=OFF
+	endif
 endif
 
 ifeq ($(BUILD_TYPE),openblas)
 	CGO_LDFLAGS+=-lopenblas
+	export WHISPER_OPENBLAS=1
 endif
 
 ifeq ($(BUILD_TYPE),cublas)
 	CGO_LDFLAGS+=-lcublas -lcudart -L$(CUDA_LIBPATH)
 	export LLAMA_CUBLAS=1
+	export WHISPER_CUBLAS=1
+endif
+
+ifeq ($(BUILD_TYPE),hipblas)
+	ROCM_HOME ?= /opt/rocm
+	export CXX=$(ROCM_HOME)/llvm/bin/clang++
+	export CC=$(ROCM_HOME)/llvm/bin/clang
+	# llama-ggml has no hipblas support, so override it here.
+	export STABLE_BUILD_TYPE=
+	export WHISPER_HIPBLAS=1
+	GPU_TARGETS ?= gfx900,gfx90a,gfx1030,gfx1031,gfx1100
+	AMDGPU_TARGETS ?= "$(GPU_TARGETS)"
+	CMAKE_ARGS+=-DLLAMA_HIPBLAS=ON -DAMDGPU_TARGETS="$(AMDGPU_TARGETS)" -DGPU_TARGETS="$(GPU_TARGETS)"
+	CGO_LDFLAGS += -O3 --rtlib=compiler-rt -unwindlib=libgcc -lhipblas -lrocblas --hip-link
 endif
 
 ifeq ($(BUILD_TYPE),metal)
 	CGO_LDFLAGS+=-framework Foundation -framework Metal -framework MetalKit -framework MetalPerformanceShaders
 	export LLAMA_METAL=1
+	export WHISPER_METAL=1
 endif
 
 ifeq ($(BUILD_TYPE),clblas)
 	CGO_LDFLAGS+=-lOpenCL -lclblast
+	export WHISPER_CLBLAST=1
 endif
 
 # glibc-static or glibc-devel-static required
@@ -104,182 +132,188 @@ ifeq ($(findstring stablediffusion,$(GO_TAGS)),stablediffusion)
 	OPTIONAL_GRPC+=backend-assets/grpc/stablediffusion
 endif
 
+ifeq ($(findstring tinydream,$(GO_TAGS)),tinydream)
+#	OPTIONAL_TARGETS+=go-tiny-dream/libtinydream.a
+	OPTIONAL_GRPC+=backend-assets/grpc/tinydream
+endif
+
 ifeq ($(findstring tts,$(GO_TAGS)),tts)
 #	OPTIONAL_TARGETS+=go-piper/libpiper_binding.a
 #	OPTIONAL_TARGETS+=backend-assets/espeak-ng-data
+	PIPER_CGO_CXXFLAGS+=-I$(CURDIR)/sources/go-piper/piper/src/cpp -I$(CURDIR)/sources/go-piper/piper/build/fi/include -I$(CURDIR)/sources/go-piper/piper/build/pi/include -I$(CURDIR)/sources/go-piper/piper/build/si/include
+	PIPER_CGO_LDFLAGS+=-L$(CURDIR)/sources/go-piper/piper/build/fi/lib -L$(CURDIR)/sources/go-piper/piper/build/pi/lib -L$(CURDIR)/sources/go-piper/piper/build/si/lib -lfmt -lspdlog -lucd
 	OPTIONAL_GRPC+=backend-assets/grpc/piper
 endif
 
+ALL_GRPC_BACKENDS=backend-assets/grpc/langchain-huggingface backend-assets/grpc/falcon-ggml backend-assets/grpc/bert-embeddings backend-assets/grpc/llama backend-assets/grpc/llama-cpp backend-assets/grpc/llama-ggml backend-assets/grpc/gpt4all backend-assets/grpc/dolly backend-assets/grpc/gpt2 backend-assets/grpc/gptj backend-assets/grpc/gptneox backend-assets/grpc/mpt backend-assets/grpc/replit backend-assets/grpc/starcoder backend-assets/grpc/rwkv backend-assets/grpc/whisper $(OPTIONAL_GRPC)
+GRPC_BACKENDS?=$(ALL_GRPC_BACKENDS) $(OPTIONAL_GRPC)
+
+# If empty, then we build all
+ifeq ($(GRPC_BACKENDS),)
+	GRPC_BACKENDS=$(ALL_GRPC_BACKENDS)
+endif
+
+ifeq ($(BUILD_API_ONLY),true)
+	GRPC_BACKENDS=
+endif
+
 .PHONY: all test build vendor
 
 all: help
 
 ## GPT4ALL
-gpt4all:
-	git clone --recurse-submodules $(GPT4ALL_REPO) gpt4all
-	cd gpt4all && git checkout -b build $(GPT4ALL_VERSION) && git submodule update --init --recursive --depth 1
-
-## go-ggllm
-go-ggllm:
-	git clone --recurse-submodules https://github.com/mudler/go-ggllm.cpp go-ggllm
-	cd go-ggllm && git checkout -b build $(GOGGLLM_VERSION) && git submodule update --init --recursive --depth 1
-
-go-ggllm/libggllm.a: go-ggllm
-	$(MAKE) -C go-ggllm BUILD_TYPE=$(BUILD_TYPE) libggllm.a
+sources/gpt4all:
+	git clone --recurse-submodules $(GPT4ALL_REPO) sources/gpt4all
+	cd sources/gpt4all && git checkout -b build $(GPT4ALL_VERSION) && git submodule update --init --recursive --depth 1
 
 ## go-piper
-go-piper:
-	git clone --recurse-submodules https://github.com/mudler/go-piper go-piper
-	cd go-piper && git checkout -b build $(PIPER_VERSION) && git submodule update --init --recursive --depth 1
+sources/go-piper:
+	git clone --recurse-submodules https://github.com/mudler/go-piper sources/go-piper
+	cd sources/go-piper && git checkout -b build $(PIPER_VERSION) && git submodule update --init --recursive --depth 1
 
 ## BERT embeddings
-go-bert:
-	git clone --recurse-submodules https://github.com/go-skynet/go-bert.cpp go-bert
-	cd go-bert && git checkout -b build $(BERT_VERSION) && git submodule update --init --recursive --depth 1
+sources/go-bert:
+	git clone --recurse-submodules https://github.com/go-skynet/go-bert.cpp sources/go-bert
+	cd sources/go-bert && git checkout -b build $(BERT_VERSION) && git submodule update --init --recursive --depth 1
 
 ## stable diffusion
-go-stable-diffusion:
-	git clone --recurse-submodules https://github.com/mudler/go-stable-diffusion go-stable-diffusion
-	cd go-stable-diffusion && git checkout -b build $(STABLEDIFFUSION_VERSION) && git submodule update --init --recursive --depth 1
+sources/go-stable-diffusion:
+	git clone --recurse-submodules https://github.com/mudler/go-stable-diffusion sources/go-stable-diffusion
+	cd sources/go-stable-diffusion && git checkout -b build $(STABLEDIFFUSION_VERSION) && git submodule update --init --recursive --depth 1
 
-go-stable-diffusion/libstablediffusion.a:
-	$(MAKE) -C go-stable-diffusion libstablediffusion.a
+sources/go-stable-diffusion/libstablediffusion.a:
+	$(MAKE) -C sources/go-stable-diffusion libstablediffusion.a
+
+## tiny-dream
+sources/go-tiny-dream:
+	git clone --recurse-submodules https://github.com/M0Rf30/go-tiny-dream sources/go-tiny-dream
+	cd sources/go-tiny-dream && git checkout -b build $(TINYDREAM_VERSION) && git submodule update --init --recursive --depth 1
+
+sources/go-tiny-dream/libtinydream.a:
+	$(MAKE) -C sources/go-tiny-dream libtinydream.a
 
 ## RWKV
-go-rwkv:
-	git clone --recurse-submodules $(RWKV_REPO) go-rwkv
-	cd go-rwkv && git checkout -b build $(RWKV_VERSION) && git submodule update --init --recursive --depth 1
+sources/go-rwkv:
+	git clone --recurse-submodules $(RWKV_REPO) sources/go-rwkv
+	cd sources/go-rwkv && git checkout -b build $(RWKV_VERSION) && git submodule update --init --recursive --depth 1
 
-go-rwkv/librwkv.a: go-rwkv
-	cd go-rwkv && cd rwkv.cpp &&	cmake . -DRWKV_BUILD_SHARED_LIBRARY=OFF &&	cmake --build . && 	cp librwkv.a ..
+sources/go-rwkv/librwkv.a: sources/go-rwkv
+	cd sources/go-rwkv && cd rwkv.cpp &&	cmake . -DRWKV_BUILD_SHARED_LIBRARY=OFF &&	cmake --build . && 	cp librwkv.a ..
 
-## bloomz
-bloomz:
-	git clone --recurse-submodules https://github.com/go-skynet/bloomz.cpp bloomz
-	cd bloomz && git checkout -b build $(BLOOMZ_VERSION) && git submodule update --init --recursive --depth 1
+sources/go-bert/libgobert.a: sources/go-bert
+	$(MAKE) -C sources/go-bert libgobert.a
 
-bloomz/libbloomz.a: bloomz
-	cd bloomz && make libbloomz.a
-
-go-bert/libgobert.a: go-bert
-	$(MAKE) -C go-bert libgobert.a
-
-backend-assets/gpt4all: gpt4all/gpt4all-bindings/golang/libgpt4all.a
+backend-assets/gpt4all: sources/gpt4all/gpt4all-bindings/golang/libgpt4all.a
 	mkdir -p backend-assets/gpt4all
-	@cp gpt4all/gpt4all-bindings/golang/buildllm/*.so backend-assets/gpt4all/ || true
-	@cp gpt4all/gpt4all-bindings/golang/buildllm/*.dylib backend-assets/gpt4all/ || true
-	@cp gpt4all/gpt4all-bindings/golang/buildllm/*.dll backend-assets/gpt4all/ || true
+	@cp sources/gpt4all/gpt4all-bindings/golang/buildllm/*.so backend-assets/gpt4all/ || true
+	@cp sources/gpt4all/gpt4all-bindings/golang/buildllm/*.dylib backend-assets/gpt4all/ || true
+	@cp sources/gpt4all/gpt4all-bindings/golang/buildllm/*.dll backend-assets/gpt4all/ || true
 
-backend-assets/espeak-ng-data:
+backend-assets/espeak-ng-data: sources/go-piper
 	mkdir -p backend-assets/espeak-ng-data
-ifdef ESPEAK_DATA
-	@cp -rf $(ESPEAK_DATA)/. backend-assets/espeak-ng-data
-else
-	@echo "ESPEAK_DATA not set, skipping tts. Note that this will break the tts functionality."
-	@touch backend-assets/espeak-ng-data/keep
-endif
+	$(MAKE) -C sources/go-piper piper.o
+	@cp -rf sources/go-piper/piper-phonemize/pi/share/espeak-ng-data/. backend-assets/espeak-ng-data
 
-gpt4all/gpt4all-bindings/golang/libgpt4all.a: gpt4all
-	$(MAKE) -C gpt4all/gpt4all-bindings/golang/ libgpt4all.a
+sources/gpt4all/gpt4all-bindings/golang/libgpt4all.a: sources/gpt4all
+	$(MAKE) -C sources/gpt4all/gpt4all-bindings/golang/ libgpt4all.a
 
 ## CEREBRAS GPT
-go-ggml-transformers:
-	git clone --recurse-submodules https://github.com/go-skynet/go-ggml-transformers.cpp go-ggml-transformers
-	cd go-ggml-transformers && git checkout -b build $(GOGPT2_VERSION) && git submodule update --init --recursive --depth 1
+sources/go-ggml-transformers:
+	git clone --recurse-submodules https://github.com/go-skynet/go-ggml-transformers.cpp sources/go-ggml-transformers
+	cd sources/go-ggml-transformers && git checkout -b build $(GOGPT2_VERSION) && git submodule update --init --recursive --depth 1
 
-go-ggml-transformers/libtransformers.a: go-ggml-transformers
-	$(MAKE) -C go-ggml-transformers libtransformers.a
+sources/go-ggml-transformers/libtransformers.a: sources/go-ggml-transformers
+	$(MAKE) -C sources/go-ggml-transformers BUILD_TYPE=$(BUILD_TYPE) libtransformers.a
 
-whisper.cpp:
-	git clone https://github.com/ggerganov/whisper.cpp.git
-	cd whisper.cpp && git checkout -b build $(WHISPER_CPP_VERSION) && git submodule update --init --recursive --depth 1
+sources/whisper.cpp:
+	git clone https://github.com/ggerganov/whisper.cpp.git sources/whisper.cpp
+	cd sources/whisper.cpp && git checkout -b build $(WHISPER_CPP_VERSION) && git submodule update --init --recursive --depth 1
 
-whisper.cpp/libwhisper.a: whisper.cpp
-	cd whisper.cpp && make libwhisper.a
+sources/whisper.cpp/libwhisper.a: sources/whisper.cpp
+	cd sources/whisper.cpp && make libwhisper.a
 
-go-llama:
-	git clone --recurse-submodules https://github.com/go-skynet/go-llama.cpp go-llama
-	cd go-llama && git checkout -b build $(GOLLAMA_VERSION) && git submodule update --init --recursive --depth 1
-ifneq ($(LLAMA_CPP_REPO),)
-	cd go-llama && rm -rf llama.cpp && git clone $(LLAMA_CPP_REPO) llama.cpp && cd llama.cpp && git checkout -b build $(LLAMA_CPP_VERSION) && git submodule update --init --recursive --depth 1
-endif
+sources/go-llama:
+	git clone --recurse-submodules https://github.com/go-skynet/go-llama.cpp sources/go-llama
+	cd sources/go-llama && git checkout -b build $(GOLLAMA_VERSION) && git submodule update --init --recursive --depth 1
 
-go-llama/libbinding.a: go-llama
-	$(MAKE) -C go-llama BUILD_TYPE=$(BUILD_TYPE) libbinding.a
+sources/go-llama-ggml:
+	git clone --recurse-submodules https://github.com/go-skynet/go-llama.cpp sources/go-llama-ggml
+	cd sources/go-llama-ggml && git checkout -b build $(GOLLAMA_STABLE_VERSION) && git submodule update --init --recursive --depth 1
 
-go-piper/libpiper_binding.a:
-	$(MAKE) -C go-piper libpiper_binding.a example/main
+sources/go-llama/libbinding.a: sources/go-llama
+	$(MAKE) -C sources/go-llama BUILD_TYPE=$(BUILD_TYPE) libbinding.a
 
-get-sources: go-llama go-ggllm go-ggml-transformers gpt4all go-piper go-rwkv whisper.cpp go-bert bloomz go-stable-diffusion
+sources/go-llama-ggml/libbinding.a: sources/go-llama-ggml
+	$(MAKE) -C sources/go-llama-ggml BUILD_TYPE=$(STABLE_BUILD_TYPE) libbinding.a
+
+sources/go-piper/libpiper_binding.a: sources/go-piper
+	$(MAKE) -C sources/go-piper libpiper_binding.a example/main
+
+backend/cpp/llama/llama.cpp:
+	LLAMA_VERSION=$(CPPLLAMA_VERSION) $(MAKE) -C backend/cpp/llama llama.cpp	
+
+get-sources: backend/cpp/llama/llama.cpp sources/go-llama sources/go-llama-ggml sources/go-ggml-transformers sources/gpt4all sources/go-piper sources/go-rwkv sources/whisper.cpp sources/go-bert sources/go-stable-diffusion sources/go-tiny-dream
 	touch $@
 
 replace:
-	$(GOCMD) mod edit -replace github.com/go-skynet/go-llama.cpp=$(shell pwd)/go-llama
-	$(GOCMD) mod edit -replace github.com/nomic-ai/gpt4all/gpt4all-bindings/golang=$(shell pwd)/gpt4all/gpt4all-bindings/golang
-	$(GOCMD) mod edit -replace github.com/go-skynet/go-ggml-transformers.cpp=$(shell pwd)/go-ggml-transformers
-	$(GOCMD) mod edit -replace github.com/donomii/go-rwkv.cpp=$(shell pwd)/go-rwkv
-	$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp=$(shell pwd)/whisper.cpp
-	$(GOCMD) mod edit -replace github.com/go-skynet/go-bert.cpp=$(shell pwd)/go-bert
-	$(GOCMD) mod edit -replace github.com/go-skynet/bloomz.cpp=$(shell pwd)/bloomz
-	$(GOCMD) mod edit -replace github.com/mudler/go-stable-diffusion=$(shell pwd)/go-stable-diffusion
-	$(GOCMD) mod edit -replace github.com/mudler/go-piper=$(shell pwd)/go-piper
-	$(GOCMD) mod edit -replace github.com/mudler/go-ggllm.cpp=$(shell pwd)/go-ggllm
+	$(GOCMD) mod edit -replace github.com/nomic-ai/gpt4all/gpt4all-bindings/golang=$(CURDIR)/sources/gpt4all/gpt4all-bindings/golang
+	$(GOCMD) mod edit -replace github.com/go-skynet/go-ggml-transformers.cpp=$(CURDIR)/sources/go-ggml-transformers
+	$(GOCMD) mod edit -replace github.com/donomii/go-rwkv.cpp=$(CURDIR)/sources/go-rwkv
+	$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp=$(CURDIR)/sources/whisper.cpp
+	$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp/bindings/go=$(CURDIR)/sources/whisper.cpp/bindings/go
+	$(GOCMD) mod edit -replace github.com/go-skynet/go-bert.cpp=$(CURDIR)/sources/go-bert
+	$(GOCMD) mod edit -replace github.com/mudler/go-stable-diffusion=$(CURDIR)/sources/go-stable-diffusion
+	$(GOCMD) mod edit -replace github.com/M0Rf30/go-tiny-dream=$(CURDIR)/sources/go-tiny-dream
+	$(GOCMD) mod edit -replace github.com/mudler/go-piper=$(CURDIR)/sources/go-piper
 
 prepare-sources: get-sources replace
 	$(GOCMD) mod download
+	touch $@
 
 ## GENERIC
 rebuild: ## Rebuilds the project
 	$(GOCMD) clean -cache
-	$(MAKE) -C go-llama clean
-	$(MAKE) -C gpt4all/gpt4all-bindings/golang/ clean
-	$(MAKE) -C go-ggml-transformers clean
-	$(MAKE) -C go-rwkv clean
-	$(MAKE) -C whisper.cpp clean
-	$(MAKE) -C go-stable-diffusion clean
-	$(MAKE) -C go-bert clean
-	$(MAKE) -C bloomz clean
-	$(MAKE) -C go-piper clean
-	$(MAKE) -C go-ggllm clean
+	$(MAKE) -C sources/go-llama clean
+	$(MAKE) -C sources/go-llama-ggml clean
+	$(MAKE) -C sources/gpt4all/gpt4all-bindings/golang/ clean
+	$(MAKE) -C sources/go-ggml-transformers clean
+	$(MAKE) -C sources/go-rwkv clean
+	$(MAKE) -C sources/whisper.cpp clean
+	$(MAKE) -C sources/go-stable-diffusion clean
+	$(MAKE) -C sources/go-bert clean
+	$(MAKE) -C sources/go-piper clean
+	$(MAKE) -C sources/go-tiny-dream clean
 	$(MAKE) build
 
-prepare: prepare-sources $(OPTIONAL_TARGETS) 
+prepare: prepare-sources $(OPTIONAL_TARGETS)
 	touch $@
 
 clean: ## Remove build related file
 	$(GOCMD) clean -cache
-	rm -fr ./go-llama
-	rm -rf ./gpt4all
-	rm -rf ./go-gpt2
-	rm -rf ./go-stable-diffusion
-	rm -rf ./go-ggml-transformers
-	rm -rf ./backend-assets
-	rm -rf ./go-rwkv
-	rm -rf ./go-bert
-	rm -rf ./bloomz
-	rm -rf ./whisper.cpp
-	rm -rf ./go-piper
-	rm -rf ./go-ggllm
+	rm -f prepare
+	rm -rf ./sources
 	rm -rf $(BINARY_NAME)
 	rm -rf release/
+	rm -rf backend-assets
+	$(MAKE) -C backend/cpp/grpc clean
+	$(MAKE) -C backend/cpp/llama clean
 
 ## Build:
 
-build: grpcs prepare ## Build the project
+build: backend-assets grpcs prepare ## Build the project
 	$(info ${GREEN}I local-ai build info:${RESET})
 	$(info ${GREEN}I BUILD_TYPE: ${YELLOW}$(BUILD_TYPE)${RESET})
 	$(info ${GREEN}I GO_TAGS: ${YELLOW}$(GO_TAGS)${RESET})
 	$(info ${GREEN}I LD_FLAGS: ${YELLOW}$(LD_FLAGS)${RESET})
-
 	CGO_LDFLAGS="$(CGO_LDFLAGS)" $(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o $(BINARY_NAME) ./
-ifeq ($(BUILD_TYPE),metal)
-	cp go-llama/build/bin/ggml-metal.metal .
-endif
 
 dist: build
 	mkdir -p release
 	cp $(BINARY_NAME) release/$(BINARY_NAME)-$(BUILD_ID)-$(OS)-$(ARCH)
 
+osx-signed: build
+	codesign --deep --force --sign "$(OSX_SIGNING_IDENTITY)" --entitlements "./Entitlements.plist" "./$(BINARY_NAME)"
+
 ## Run
 run: prepare ## run local-ai
 	CGO_LDFLAGS="$(CGO_LDFLAGS)" $(GOCMD) run ./
@@ -287,12 +321,12 @@ run: prepare ## run local-ai
 test-models/testmodel:
 	mkdir test-models
 	mkdir test-dir
-	wget https://huggingface.co/nnakasato/ggml-model-test/resolve/main/ggml-model-q4.bin -O test-models/testmodel
-	wget https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.en.bin -O test-models/whisper-en
-	wget https://huggingface.co/skeskinen/ggml/resolve/main/all-MiniLM-L6-v2/ggml-model-q4_0.bin -O test-models/bert
-	wget https://cdn.openai.com/whisper/draft-20220913a/micro-machines.wav -O test-dir/audio.wav
-	wget https://huggingface.co/mudler/rwkv-4-raven-1.5B-ggml/resolve/main/RWKV-4-Raven-1B5-v11-Eng99%2525-Other1%2525-20230425-ctx4096_Q4_0.bin -O test-models/rwkv
-	wget https://raw.githubusercontent.com/saharNooby/rwkv.cpp/5eb8f09c146ea8124633ab041d9ea0b1f1db4459/rwkv/20B_tokenizer.json -O test-models/rwkv.tokenizer.json
+	wget -q https://huggingface.co/nnakasato/ggml-model-test/resolve/main/ggml-model-q4.bin -O test-models/testmodel
+	wget -q https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.en.bin -O test-models/whisper-en
+	wget -q https://huggingface.co/mudler/all-MiniLM-L6-v2/resolve/main/ggml-model-q4_0.bin -O test-models/bert
+	wget -q https://cdn.openai.com/whisper/draft-20220913a/micro-machines.wav -O test-dir/audio.wav
+	wget -q https://huggingface.co/mudler/rwkv-4-raven-1.5B-ggml/resolve/main/RWKV-4-Raven-1B5-v11-Eng99%2525-Other1%2525-20230425-ctx4096_Q4_0.bin -O test-models/rwkv
+	wget -q https://raw.githubusercontent.com/saharNooby/rwkv.cpp/5eb8f09c146ea8124633ab041d9ea0b1f1db4459/rwkv/20B_tokenizer.json -O test-models/rwkv.tokenizer.json
 	cp tests/models_fixtures/* test-models
 
 prepare-test: grpcs
@@ -303,13 +337,34 @@ test: prepare test-models/testmodel grpcs
 	@echo 'Running tests'
 	export GO_TAGS="tts stablediffusion"
 	$(MAKE) prepare-test
-	TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
-	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="!gpt4all && !llama" --flake-attempts 5 -v -r ./api ./pkg
+	HUGGINGFACE_GRPC=$(abspath ./)/backend/python/sentencetransformers/run.sh TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
+	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="!gpt4all && !llama && !llama-gguf"  --flake-attempts 5 --fail-fast -v -r ./api ./pkg
 	$(MAKE) test-gpt4all
 	$(MAKE) test-llama
+	$(MAKE) test-llama-gguf
 	$(MAKE) test-tts
 	$(MAKE) test-stablediffusion
 
+prepare-e2e:
+	mkdir -p $(TEST_DIR)
+	cp -rfv $(abspath ./tests/e2e-fixtures)/gpu.yaml $(TEST_DIR)/gpu.yaml
+	test -e $(TEST_DIR)/ggllm-test-model.bin || wget -q https://huggingface.co/TheBloke/CodeLlama-7B-Instruct-GGUF/resolve/main/codellama-7b-instruct.Q2_K.gguf -O $(TEST_DIR)/ggllm-test-model.bin
+	docker build --build-arg BUILD_GRPC=true --build-arg GRPC_BACKENDS="$(GRPC_BACKENDS)" --build-arg IMAGE_TYPE=core --build-arg BUILD_TYPE=$(BUILD_TYPE) --build-arg CUDA_MAJOR_VERSION=11 --build-arg CUDA_MINOR_VERSION=7 --build-arg FFMPEG=true -t localai-tests .
+
+run-e2e-image:
+	ls -liah $(abspath ./tests/e2e-fixtures)
+	docker run -p 5390:8080 -e MODELS_PATH=/models -e THREADS=1 -e DEBUG=true -d --rm -v $(TEST_DIR):/models --gpus all --name e2e-tests-$(RANDOM) localai-tests
+
+test-e2e:
+	@echo 'Running e2e tests'
+	BUILD_TYPE=$(BUILD_TYPE) \
+	LOCALAI_API=http://$(E2E_BRIDGE_IP):5390/v1 \
+	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --flake-attempts 5 -v -r ./tests/e2e
+
+teardown-e2e:
+	rm -rf $(TEST_DIR) || true
+	docker stop $$(docker ps -q --filter ancestor=localai-tests)
+
 test-gpt4all: prepare-test
 	TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
 	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="gpt4all" --flake-attempts 5 -v -r ./api ./pkg
@@ -318,6 +373,10 @@ test-llama: prepare-test
 	TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
 	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="llama" --flake-attempts 5 -v -r ./api ./pkg
 
+test-llama-gguf: prepare-test
+	TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
+	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="llama-gguf" --flake-attempts 5 -v -r ./api ./pkg
+
 test-tts: prepare-test
 	TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
 	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="tts" --flake-attempts 1 -v -r ./api ./pkg
@@ -328,9 +387,7 @@ test-stablediffusion: prepare-test
 
 test-container:
 	docker build --target requirements -t local-ai-test-container .
-	docker run --name localai-tests -e GO_TAGS=$(GO_TAGS) -ti -v $(abspath ./):/build local-ai-test-container make test
-	docker rm localai-tests
-	docker rmi local-ai-test-container
+	docker run -ti --rm --entrypoint /bin/bash -ti -v $(abspath ./):/build local-ai-test-container
 
 ## Help:
 help: ## Show this help.
@@ -344,84 +401,170 @@ help: ## Show this help.
 		else if (/^## .*$$/) {printf "  ${CYAN}%s${RESET}\n", substr($$1,4)} \
 		}' $(MAKEFILE_LIST)
 
-protogen:
-	protoc --go_out=. --go_opt=paths=source_relative --go-grpc_out=. --go-grpc_opt=paths=source_relative \
-    pkg/grpc/proto/backend.proto
+protogen: protogen-go protogen-python
+
+protogen-go:
+	protoc -Ibackend/ --go_out=pkg/grpc/proto/ --go_opt=paths=source_relative --go-grpc_out=pkg/grpc/proto/ --go-grpc_opt=paths=source_relative \
+    backend/backend.proto
+
+protogen-python:
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/sentencetransformers/ --grpc_python_out=backend/python/sentencetransformers/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/transformers/ --grpc_python_out=backend/python/transformers/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/transformers-musicgen/ --grpc_python_out=backend/python/transformers-musicgen/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/autogptq/ --grpc_python_out=backend/python/autogptq/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/exllama/ --grpc_python_out=backend/python/exllama/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/bark/ --grpc_python_out=backend/python/bark/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/diffusers/ --grpc_python_out=backend/python/diffusers/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/coqui/ --grpc_python_out=backend/python/coqui/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/vall-e-x/ --grpc_python_out=backend/python/vall-e-x/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/vllm/ --grpc_python_out=backend/python/vllm/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/petals/ --grpc_python_out=backend/python/petals/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/mamba/ --grpc_python_out=backend/python/mamba/ backend/backend.proto
+	python3 -m grpc_tools.protoc -Ibackend/ --python_out=backend/python/exllama2/ --grpc_python_out=backend/python/exllama2/ backend/backend.proto
 
 ## GRPC
+# Note: it is duplicated in the Dockerfile
+prepare-extra-conda-environments:
+	$(MAKE) -C backend/python/autogptq
+	$(MAKE) -C backend/python/bark
+	$(MAKE) -C backend/python/coqui
+	$(MAKE) -C backend/python/diffusers
+	$(MAKE) -C backend/python/vllm
+	$(MAKE) -C backend/python/mamba
+	$(MAKE) -C backend/python/sentencetransformers
+	$(MAKE) -C backend/python/transformers
+	$(MAKE) -C backend/python/transformers-musicgen
+	$(MAKE) -C backend/python/vall-e-x
+	$(MAKE) -C backend/python/exllama
+	$(MAKE) -C backend/python/petals
+	$(MAKE) -C backend/python/exllama2
+
+prepare-test-extra:
+	$(MAKE) -C backend/python/transformers
+	$(MAKE) -C backend/python/diffusers
+
+test-extra: prepare-test-extra
+	$(MAKE) -C backend/python/transformers test
+	$(MAKE) -C backend/python/diffusers test
+
+backend-assets:
+	mkdir -p backend-assets
+ifeq ($(BUILD_API_ONLY),true)
+	touch backend-assets/keep
+endif
 
 backend-assets/grpc:
 	mkdir -p backend-assets/grpc
 
-backend-assets/grpc/falcon: backend-assets/grpc go-ggllm/libggllm.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggllm LIBRARY_PATH=$(shell pwd)/go-ggllm \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/falcon ./cmd/grpc/falcon/
+backend-assets/grpc/llama: backend-assets/grpc sources/go-llama/libbinding.a
+	$(GOCMD) mod edit -replace github.com/go-skynet/go-llama.cpp=$(CURDIR)/sources/go-llama
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-llama LIBRARY_PATH=$(CURDIR)/sources/go-llama \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/llama ./backend/go/llm/llama/
+# TODO: every binary should have its own folder instead, so can have different  implementations
+ifeq ($(BUILD_TYPE),metal)
+	cp backend/cpp/llama/llama.cpp/ggml-metal.metal backend-assets/grpc/
+endif
 
-backend-assets/grpc/llama: backend-assets/grpc go-llama/libbinding.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-llama LIBRARY_PATH=$(shell pwd)/go-llama \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/llama ./cmd/grpc/llama/
+## BACKEND CPP LLAMA START
+# Sets the variables in case it has to build the gRPC locally.
+INSTALLED_PACKAGES=$(CURDIR)/backend/cpp/grpc/installed_packages
+INSTALLED_LIB_CMAKE=$(INSTALLED_PACKAGES)/lib/cmake
+ADDED_CMAKE_ARGS=-Dabsl_DIR=${INSTALLED_LIB_CMAKE}/absl \
+                 -DProtobuf_DIR=${INSTALLED_LIB_CMAKE}/protobuf \
+                 -Dutf8_range_DIR=${INSTALLED_LIB_CMAKE}/utf8_range \
+                 -DgRPC_DIR=${INSTALLED_LIB_CMAKE}/grpc \
+                 -DCMAKE_CXX_STANDARD_INCLUDE_DIRECTORIES=${INSTALLED_PACKAGES}/include
 
-backend-assets/grpc/gpt4all: backend-assets/grpc backend-assets/gpt4all gpt4all/gpt4all-bindings/golang/libgpt4all.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/gpt4all/gpt4all-bindings/golang/ LIBRARY_PATH=$(shell pwd)/gpt4all/gpt4all-bindings/golang/ \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gpt4all ./cmd/grpc/gpt4all/
+backend/cpp/llama/grpc-server:
+ifdef BUILD_GRPC_FOR_BACKEND_LLAMA
+	$(MAKE) -C backend/cpp/grpc build
+	export _PROTOBUF_PROTOC=${INSTALLED_PACKAGES}/bin/proto && \
+	export _GRPC_CPP_PLUGIN_EXECUTABLE=${INSTALLED_PACKAGES}/bin/grpc_cpp_plugin && \
+	export PATH="${INSTALLED_PACKAGES}/bin:${PATH}" && \
+	CMAKE_ARGS="${CMAKE_ARGS} ${ADDED_CMAKE_ARGS}" LLAMA_VERSION=$(CPPLLAMA_VERSION) $(MAKE) -C backend/cpp/llama grpc-server
+else
+	echo "BUILD_GRPC_FOR_BACKEND_LLAMA is not defined."
+	LLAMA_VERSION=$(CPPLLAMA_VERSION) $(MAKE) -C backend/cpp/llama grpc-server			
+endif
+## BACKEND CPP LLAMA END
 
-backend-assets/grpc/dolly: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/dolly ./cmd/grpc/dolly/
+##
+backend-assets/grpc/llama-cpp: backend-assets/grpc backend/cpp/llama/grpc-server
+	cp -rfv backend/cpp/llama/grpc-server backend-assets/grpc/llama-cpp
+# TODO: every binary should have its own folder instead, so can have different metal implementations
+ifeq ($(BUILD_TYPE),metal)
+	cp backend/cpp/llama/llama.cpp/build/bin/ggml-metal.metal backend-assets/grpc/
+endif
 
-backend-assets/grpc/gpt2: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gpt2 ./cmd/grpc/gpt2/
+backend-assets/grpc/llama-ggml: backend-assets/grpc sources/go-llama-ggml/libbinding.a
+	$(GOCMD) mod edit -replace github.com/go-skynet/go-llama.cpp=$(CURDIR)/sources/go-llama-ggml
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-llama-ggml LIBRARY_PATH=$(CURDIR)/sources/go-llama-ggml \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/llama-ggml ./backend/go/llm/llama-ggml/
 
-backend-assets/grpc/gptj: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gptj ./cmd/grpc/gptj/
+backend-assets/grpc/gpt4all: backend-assets/grpc backend-assets/gpt4all sources/gpt4all/gpt4all-bindings/golang/libgpt4all.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/gpt4all/gpt4all-bindings/golang/ LIBRARY_PATH=$(CURDIR)/sources/gpt4all/gpt4all-bindings/golang/ \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gpt4all ./backend/go/llm/gpt4all/
 
-backend-assets/grpc/gptneox: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gptneox ./cmd/grpc/gptneox/
+backend-assets/grpc/dolly: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/dolly ./backend/go/llm/dolly/
 
-backend-assets/grpc/mpt: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/mpt ./cmd/grpc/mpt/
+backend-assets/grpc/gpt2: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gpt2 ./backend/go/llm/gpt2/
 
-backend-assets/grpc/replit: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/replit ./cmd/grpc/replit/
+backend-assets/grpc/gptj: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gptj ./backend/go/llm/gptj/
 
-backend-assets/grpc/falcon-ggml: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/falcon-ggml ./cmd/grpc/falcon-ggml/
+backend-assets/grpc/gptneox: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/gptneox ./backend/go/llm/gptneox/
 
-backend-assets/grpc/starcoder: backend-assets/grpc go-ggml-transformers/libtransformers.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-ggml-transformers LIBRARY_PATH=$(shell pwd)/go-ggml-transformers \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/starcoder ./cmd/grpc/starcoder/
+backend-assets/grpc/mpt: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/mpt ./backend/go/llm/mpt/
 
-backend-assets/grpc/rwkv: backend-assets/grpc go-rwkv/librwkv.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-rwkv LIBRARY_PATH=$(shell pwd)/go-rwkv \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/rwkv ./cmd/grpc/rwkv/
+backend-assets/grpc/replit: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/replit ./backend/go/llm/replit/
 
-backend-assets/grpc/bloomz: backend-assets/grpc bloomz/libbloomz.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/bloomz LIBRARY_PATH=$(shell pwd)/bloomz \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/bloomz ./cmd/grpc/bloomz/
+backend-assets/grpc/falcon-ggml: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/falcon-ggml ./backend/go/llm/falcon-ggml/
 
-backend-assets/grpc/bert-embeddings: backend-assets/grpc go-bert/libgobert.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-bert LIBRARY_PATH=$(shell pwd)/go-bert \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/bert-embeddings ./cmd/grpc/bert-embeddings/
+backend-assets/grpc/starcoder: backend-assets/grpc sources/go-ggml-transformers/libtransformers.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-ggml-transformers LIBRARY_PATH=$(CURDIR)/sources/go-ggml-transformers \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/starcoder ./backend/go/llm/starcoder/
+
+backend-assets/grpc/rwkv: backend-assets/grpc sources/go-rwkv/librwkv.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-rwkv LIBRARY_PATH=$(CURDIR)/sources/go-rwkv \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/rwkv ./backend/go/llm/rwkv
+
+backend-assets/grpc/bert-embeddings: backend-assets/grpc sources/go-bert/libgobert.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-bert LIBRARY_PATH=$(CURDIR)/sources/go-bert \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/bert-embeddings ./backend/go/llm/bert/
 
 backend-assets/grpc/langchain-huggingface: backend-assets/grpc
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/langchain-huggingface ./cmd/grpc/langchain-huggingface/
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/langchain-huggingface ./backend/go/llm/langchain/
 
-backend-assets/grpc/stablediffusion: backend-assets/grpc go-stable-diffusion/libstablediffusion.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/go-stable-diffusion/ LIBRARY_PATH=$(shell pwd)/go-stable-diffusion/ \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/stablediffusion ./cmd/grpc/stablediffusion/
+backend-assets/grpc/stablediffusion: backend-assets/grpc
+	if [ ! -f backend-assets/grpc/stablediffusion ]; then \
+		$(MAKE) sources/go-stable-diffusion/libstablediffusion.a; \
+		CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-stable-diffusion/ LIBRARY_PATH=$(CURDIR)/sources/go-stable-diffusion/ \
+		$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/stablediffusion ./backend/go/image/stablediffusion; \
+	fi
 
-backend-assets/grpc/piper: backend-assets/grpc backend-assets/espeak-ng-data go-piper/libpiper_binding.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" LIBRARY_PATH=$(shell pwd)/go-piper \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/piper ./cmd/grpc/piper/
+backend-assets/grpc/tinydream: backend-assets/grpc sources/go-tiny-dream/libtinydream.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" LIBRARY_PATH=$(CURDIR)/go-tiny-dream \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/tinydream ./backend/go/image/tinydream
 
-backend-assets/grpc/whisper: backend-assets/grpc whisper.cpp/libwhisper.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(shell pwd)/whisper.cpp LIBRARY_PATH=$(shell pwd)/whisper.cpp \
-	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/whisper ./cmd/grpc/whisper/
+backend-assets/grpc/piper: backend-assets/grpc backend-assets/espeak-ng-data sources/go-piper/libpiper_binding.a
+	CGO_CXXFLAGS="$(PIPER_CGO_CXXFLAGS)" CGO_LDFLAGS="$(PIPER_CGO_LDFLAGS)" LIBRARY_PATH=$(CURDIR)/sources/go-piper \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/piper ./backend/go/tts/
 
-grpcs: prepare backend-assets/grpc/langchain-huggingface backend-assets/grpc/falcon-ggml backend-assets/grpc/bert-embeddings backend-assets/grpc/falcon backend-assets/grpc/bloomz backend-assets/grpc/llama backend-assets/grpc/gpt4all backend-assets/grpc/dolly backend-assets/grpc/gpt2 backend-assets/grpc/gptj backend-assets/grpc/gptneox backend-assets/grpc/mpt backend-assets/grpc/replit backend-assets/grpc/starcoder backend-assets/grpc/rwkv backend-assets/grpc/whisper $(OPTIONAL_GRPC)
+backend-assets/grpc/whisper: backend-assets/grpc sources/whisper.cpp/libwhisper.a
+	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/whisper.cpp LIBRARY_PATH=$(CURDIR)/sources/whisper.cpp \
+	$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/whisper ./backend/go/transcribe/
+
+grpcs: prepare $(GRPC_BACKENDS)

README.md

@@ -1,208 +1,142 @@
 <h1 align="center">
   <br>
-  <img height="300" src="https://user-images.githubusercontent.com/2420543/233147843-88697415-6dbf-4368-a862-ab217f9f7342.jpeg"> <br>
+  <img height="300" src="https://github.com/go-skynet/LocalAI/assets/2420543/0966aa2a-166e-4f99-a3e5-6c915fc997dd"> <br>
     LocalAI
 <br>
 </h1>
 
-[![tests](https://github.com/go-skynet/LocalAI/actions/workflows/test.yml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/test.yml) [![build container images](https://github.com/go-skynet/LocalAI/actions/workflows/image.yml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/image.yml)
+<p align="center">
+<a href="https://github.com/go-skynet/LocalAI/fork" target="blank">
+<img src="https://img.shields.io/github/forks/go-skynet/LocalAI?style=for-the-badge" alt="LocalAI forks"/>
+</a>
+<a href="https://github.com/go-skynet/LocalAI/stargazers" target="blank">
+<img src="https://img.shields.io/github/stars/go-skynet/LocalAI?style=for-the-badge" alt="LocalAI stars"/>
+</a>
+<a href="https://github.com/go-skynet/LocalAI/pulls" target="blank">
+<img src="https://img.shields.io/github/issues-pr/go-skynet/LocalAI?style=for-the-badge" alt="LocalAI pull-requests"/>
+</a>
+<a href='https://github.com/go-skynet/LocalAI/releases'>
+<img src='https://img.shields.io/github/release/go-skynet/LocalAI?&label=Latest&style=for-the-badge'>
+</a>
+</p>
 
-[![](https://dcbadge.vercel.app/api/server/uJAeKSAGDy?style=flat-square&theme=default-inverted)](https://discord.gg/uJAeKSAGDy) 
+[<img src="https://img.shields.io/badge/dockerhub-images-important.svg?logo=Docker">](https://hub.docker.com/r/localai/localai)
+[<img src="https://img.shields.io/badge/quay.io-images-important.svg?">](https://quay.io/repository/go-skynet/local-ai?tab=tags&tag=latest)
 
-[Documentation website](https://localai.io/)
+> :bulb: Get help - [❓FAQ](https://localai.io/faq/) [💭Discussions](https://github.com/go-skynet/LocalAI/discussions) [:speech_balloon: Discord](https://discord.gg/uJAeKSAGDy) [:book: Documentation website](https://localai.io/)
+>
+> [💻 Quickstart](https://localai.io/basics/getting_started/) [📣 News](https://localai.io/basics/news/) [ 🛫 Examples ](https://github.com/go-skynet/LocalAI/tree/master/examples/) [ 🖼️ Models ](https://localai.io/models/) [ 🚀 Roadmap ](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap)
 
-**LocalAI** is a drop-in replacement REST API that's compatible with OpenAI API specifications for local inferencing. It allows you to run LLMs (and not only) locally or on-prem with consumer grade hardware, supporting multiple model families that are compatible with the ggml format. Does not require GPU.
+[![tests](https://github.com/go-skynet/LocalAI/actions/workflows/test.yml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/test.yml)[![Build and Release](https://github.com/go-skynet/LocalAI/actions/workflows/release.yaml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/release.yaml)[![build container images](https://github.com/go-skynet/LocalAI/actions/workflows/image.yml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/image.yml)[![Bump dependencies](https://github.com/go-skynet/LocalAI/actions/workflows/bump_deps.yaml/badge.svg)](https://github.com/go-skynet/LocalAI/actions/workflows/bump_deps.yaml)[![Artifact Hub](https://img.shields.io/endpoint?url=https://artifacthub.io/badge/repository/localai)](https://artifacthub.io/packages/search?repo=localai)
 
-For a list of the supported model families, please see [the model compatibility table](https://localai.io/model-compatibility/index.html#model-compatibility-table).
+<p align="center">
+<a href="https://twitter.com/LocalAI_API" target="blank">
+<img src="https://img.shields.io/twitter/follow/LocalAI_API?label=Follow: LocalAI_API&style=social" alt="Follow LocalAI_API"/>
+</a>
+<a href="https://discord.gg/uJAeKSAGDy" target="blank">
+<img src="https://dcbadge.vercel.app/api/server/uJAeKSAGDy?style=flat-square&theme=default-inverted" alt="Join LocalAI Discord Community"/>
+</a>
 
-In a nutshell:
+**LocalAI** is the free, Open Source OpenAI alternative. LocalAI act as a drop-in replacement REST API that’s compatible with OpenAI API specifications for local inferencing. It allows you to run LLMs, generate images, audio (and not only) locally or on-prem with consumer grade hardware, supporting multiple model families. Does not require GPU.
 
-- Local, OpenAI drop-in alternative REST API. You own your data.
-- NO GPU required. NO Internet access is required either
-  - Optional, GPU Acceleration is available in `llama.cpp`-compatible LLMs. See also the [build section](https://localai.io/basics/build/index.html). 
-- Supports multiple models:
-  - 📖 Text generation with GPTs (`llama.cpp`, `gpt4all.cpp`, ... and more)
-  - 🗣 Text to Audio 🎺🆕
-  - 🔈 Audio to Text (Audio transcription with `whisper.cpp`)
-  - 🎨 Image generation with stable diffusion
-- 🏃 Once loaded the first time, it keep models loaded in memory for faster inference
-- ⚡ Doesn't shell-out, but uses C++ bindings for a faster inference and better performance. 
+## 🔥🔥 Hot topics / Roadmap
 
-LocalAI was created by [Ettore Di Giacinto](https://github.com/mudler/) and is a community-driven project, focused on making the AI accessible to anyone. Any contribution, feedback and PR is welcome! 
+[Roadmap](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap)
 
-See the [Getting started](https://localai.io/basics/getting_started/index.html) and [examples](https://github.com/go-skynet/LocalAI/tree/master/examples/) sections to learn how to use LocalAI. For a list of curated models check out the [model gallery](https://localai.io/models/).
+- Mamba support: https://github.com/mudler/LocalAI/pull/1589
+- Start and share models with config file: https://github.com/mudler/LocalAI/pull/1522
+- 🐸 Coqui: https://github.com/mudler/LocalAI/pull/1489
+- Inline templates: https://github.com/mudler/LocalAI/pull/1452
+- Mixtral: https://github.com/mudler/LocalAI/pull/1449
+- Img2vid https://github.com/mudler/LocalAI/pull/1442
+- Musicgen https://github.com/mudler/LocalAI/pull/1387
 
+Hot topics (looking for contributors):
+- Backends v2: https://github.com/mudler/LocalAI/issues/1126
+- Improving UX v2: https://github.com/mudler/LocalAI/issues/1373
 
-| [ChatGPT OSS alternative](https://github.com/go-skynet/LocalAI/tree/master/examples/chatbot-ui)                                                                                                                | [Image generation](https://localai.io/api-endpoints/index.html#image-generation)                                                                                                              |
-|------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------|
-|  ![Screenshot from 2023-04-26 23-59-55](https://user-images.githubusercontent.com/2420543/234715439-98d12e03-d3ce-4f94-ab54-2b256808e05e.png)            | ![b6441997879](https://github.com/go-skynet/LocalAI/assets/2420543/d50af51c-51b7-4f39-b6c2-bf04c403894c)                  |
+If you want to help and contribute, issues up for grabs: https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3A%22up+for+grabs%22
 
-|                                                                    [Telegram bot](https://github.com/go-skynet/LocalAI/tree/master/examples/telegram-bot)   | [Flowise](https://github.com/go-skynet/LocalAI/tree/master/examples/flowise)                                                                                                                     |
-|------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------|
-![Screenshot from 2023-06-09 00-36-26](https://github.com/go-skynet/LocalAI/assets/2420543/e98b4305-fa2d-41cf-9d2f-1bb2d75ca902)   |  ![Screenshot from 2023-05-30 18-01-03](https://github.com/go-skynet/LocalAI/assets/2420543/02458782-0549-4131-971c-95ee56ec1af8)|    |
+## 💻 [Getting started](https://localai.io/basics/getting_started/index.html)
 
-## Hot topics / Roadmap
+For a detailed step-by-step introduction, refer to the [Getting Started](https://localai.io/basics/getting_started/index.html) guide. For those in a hurry, here's a straightforward one-liner to launch a LocalAI instance with [phi-2](https://huggingface.co/microsoft/phi-2) using `docker`:
 
-- [x] Support for embeddings
-- [x] Support for audio transcription with https://github.com/ggerganov/whisper.cpp
-- [X] Support for text-to-audio
-- [x] GPU/CUDA support ( https://github.com/go-skynet/LocalAI/issues/69 )
-- [X] Enable automatic downloading of models from a curated gallery
-- [X] Enable automatic downloading of models from HuggingFace
-- [ ] Upstream our golang bindings to llama.cpp (https://github.com/ggerganov/llama.cpp/issues/351) 
-- [ ] Enable gallery management directly from the webui.
-- [ ] 🔥 OpenAI functions: https://github.com/go-skynet/LocalAI/issues/588
+```
+docker run -ti -p 8080:8080 localai/localai:v2.5.1-ffmpeg-core phi-2
+```
 
-## News
+## 🚀 [Features](https://localai.io/features/)
 
-- 🔥🔥🔥 28-06-2023: **v1.20.0**: Added text to audio and gallery huggingface repositories! [Release notes](https://localai.io/basics/news/index.html#-28-06-2023-__v1200__-) [Changelog](https://github.com/go-skynet/LocalAI/releases/tag/v1.20.0)
-- 🔥🔥🔥 19-06-2023: **v1.19.0**: CUDA support! [Release notes](https://localai.io/basics/news/index.html#-19-06-2023-__v1190__-) [Changelog](https://github.com/go-skynet/LocalAI/releases/tag/v1.19.0)
-- 🔥🔥🔥 06-06-2023: **v1.18.0**: Many updates, new features, and much more 🚀, check out the [Release notes](https://localai.io/basics/news/index.html#-06-06-2023-__v1180__-)!
-- 29-05-2023: LocalAI now has a website, [https://localai.io](https://localai.io)! check the news in the [dedicated section](https://localai.io/basics/news/index.html)!
+- 📖 [Text generation with GPTs](https://localai.io/features/text-generation/) (`llama.cpp`, `gpt4all.cpp`, ... [:book: and more](https://localai.io/model-compatibility/index.html#model-compatibility-table))
+- 🗣 [Text to Audio](https://localai.io/features/text-to-audio/)
+- 🔈 [Audio to Text](https://localai.io/features/audio-to-text/) (Audio transcription with `whisper.cpp`)
+- 🎨 [Image generation with stable diffusion](https://localai.io/features/image-generation)
+- 🔥 [OpenAI functions](https://localai.io/features/openai-functions/) 🆕
+- 🧠 [Embeddings generation for vector databases](https://localai.io/features/embeddings/)
+- ✍️ [Constrained grammars](https://localai.io/features/constrained_grammars/)
+- 🖼️ [Download Models directly from Huggingface ](https://localai.io/models/)
+- 🆕 [Vision API](https://localai.io/features/gpt-vision/)
 
-For latest news, follow also on Twitter [@LocalAI_API](https://twitter.com/LocalAI_API) and [@mudler_it](https://twitter.com/mudler_it)
+## 💻 Usage
 
-## Media, Blogs, Social
+Check out the [Getting started](https://localai.io/basics/getting_started/index.html) section in our documentation.
+
+### 🔗 Community and integrations
+
+Build and deploy custom containers:
+- https://github.com/sozercan/aikit
+
+WebUIs:
+- https://github.com/Jirubizu/localai-admin
+- https://github.com/go-skynet/LocalAI-frontend
+
+Model galleries
+- https://github.com/go-skynet/model-gallery
+  
+Auto Docker / Model setup
+- https://io.midori-ai.xyz/howtos/easy-localai-installer/
+- https://io.midori-ai.xyz/howtos/easy-model-installer/
+
+Other:
+- Helm chart https://github.com/go-skynet/helm-charts
+- VSCode extension https://github.com/badgooooor/localai-vscode-plugin
+- Local Smart assistant https://github.com/mudler/LocalAGI
+- Home Assistant https://github.com/sammcj/homeassistant-localai / https://github.com/drndos/hass-openai-custom-conversation
+- Discord bot https://github.com/mudler/LocalAGI/tree/main/examples/discord
+- Slack bot https://github.com/mudler/LocalAGI/tree/main/examples/slack
+- Telegram bot https://github.com/mudler/LocalAI/tree/master/examples/telegram-bot
+- Examples: https://github.com/mudler/LocalAI/tree/master/examples/
+
+### 🔗 Resources
+
+- 🆕 New! [LLM finetuning guide](https://localai.io/advanced/fine-tuning/)
+- [How to build locally](https://localai.io/basics/build/index.html)
+- [How to install in Kubernetes](https://localai.io/basics/getting_started/index.html#run-localai-in-kubernetes)
+- [Projects integrating LocalAI](https://localai.io/integrations/)
+- [How tos section](https://io.midori-ai.xyz/howtos/) (curated by our community)
+
+## :book: 🎥 [Media, Blogs, Social](https://localai.io/basics/news/#media-blogs-social)
 
 - [Create a slackbot for teams and OSS projects that answer to documentation](https://mudler.pm/posts/smart-slackbot-for-teams/)
 - [LocalAI meets k8sgpt](https://www.youtube.com/watch?v=PKrDNuJ_dfE)
 - [Question Answering on Documents locally with LangChain, LocalAI, Chroma, and GPT4All](https://mudler.pm/posts/localai-question-answering/)
 - [Tutorial to use k8sgpt with LocalAI](https://medium.com/@tyler_97636/k8sgpt-localai-unlock-kubernetes-superpowers-for-free-584790de9b65)
 
-## Contribute and help
+## Citation
 
-To help the project you can:
-
-- [Hacker news post](https://news.ycombinator.com/item?id=35726934) - help us out by voting if you like this project.
-
-- If you have technological skills and want to contribute to development, have a look at the open issues. If you are new you can have a look at the [good-first-issue](https://github.com/go-skynet/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3A%22good+first+issue%22) and [help-wanted](https://github.com/go-skynet/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3A%22help+wanted%22) labels.
-
-- If you don't have technological skills you can still help improving documentation or add examples or share your user-stories with our community, any help and contribution is welcome!
-
-## Usage
-
-Check out the [Getting started](https://localai.io/basics/getting_started/index.html) section. Here below you will find generic, quick instructions to get ready and use LocalAI.
-
-The easiest way to run LocalAI is by using `docker-compose` (to build locally, see [building LocalAI](https://localai.io/basics/build/index.html)):
-
-```bash
-
-git clone https://github.com/go-skynet/LocalAI
-
-cd LocalAI
-
-# (optional) Checkout a specific LocalAI tag
-# git checkout -b build <TAG>
-
-# copy your models to models/
-cp your-model.bin models/
-
-# (optional) Edit the .env file to set things like context size and threads
-# vim .env
-
-# start with docker-compose
-docker-compose up -d --pull always
-# or you can build the images with:
-# docker-compose up -d --build
-
-# Now API is accessible at localhost:8080
-curl http://localhost:8080/v1/models
-# {"object":"list","data":[{"id":"your-model.bin","object":"model"}]}
-
-curl http://localhost:8080/v1/completions -H "Content-Type: application/json" -d '{
-     "model": "your-model.bin",            
-     "prompt": "A long time ago in a galaxy far, far away",
-     "temperature": 0.7
-   }'
-```
-
-### Example: Use GPT4ALL-J model
-
-<details>
-
-```bash
-# Clone LocalAI
-git clone https://github.com/go-skynet/LocalAI
-
-cd LocalAI
-
-# (optional) Checkout a specific LocalAI tag
-# git checkout -b build <TAG>
-
-# Download gpt4all-j to models/
-wget https://gpt4all.io/models/ggml-gpt4all-j.bin -O models/ggml-gpt4all-j
-
-# Use a template from the examples
-cp -rf prompt-templates/ggml-gpt4all-j.tmpl models/
-
-# (optional) Edit the .env file to set things like context size and threads
-# vim .env
-
-# start with docker-compose
-docker-compose up -d --pull always
-# or you can build the images with:
-# docker-compose up -d --build
-# Now API is accessible at localhost:8080
-curl http://localhost:8080/v1/models
-# {"object":"list","data":[{"id":"ggml-gpt4all-j","object":"model"}]}
-
-curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
-     "model": "ggml-gpt4all-j",
-     "messages": [{"role": "user", "content": "How are you?"}],
-     "temperature": 0.9 
-   }'
-
-# {"model":"ggml-gpt4all-j","choices":[{"message":{"role":"assistant","content":"I'm doing well, thanks. How about you?"}}]}
-```
-</details>
-
-
-### Build locally
-
-<details>
-
-In order to build the `LocalAI` container image locally you can use `docker`:
+If you utilize this repository, data in a downstream project, please consider citing it with:
 
 ```
-# build the image
-docker build -t localai .
-docker run localai
+@misc{localai,
+  author = {Ettore Di Giacinto},
+  title = {LocalAI: The free, Open source OpenAI alternative},
+  year = {2023},
+  publisher = {GitHub},
+  journal = {GitHub repository},
+  howpublished = {\url{https://github.com/go-skynet/LocalAI}},
 ```
 
-Or you can build the binary with `make`:
-
-```
-make build
-```
-
-</details>
-
-See the [build section](https://localai.io/basics/build/index.html) in our documentation for detailed instructions.
-
-### Run LocalAI in Kubernetes
-
-LocalAI can be installed inside Kubernetes with helm. See [installation instructions](https://localai.io/basics/getting_started/index.html#run-localai-in-kubernetes).
-
-## Supported API endpoints
-
-See the [list of the supported API endpoints](https://localai.io/api-endpoints/index.html) and how to configure image generation and audio transcription.
-
-## Frequently asked questions
-
-See [the FAQ](https://localai.io/faq/index.html) section for a list of common questions.
-
-## Projects already using LocalAI to run local models
-
-Feel free to open up a PR to get your project listed!
-
-- [Kairos](https://github.com/kairos-io/kairos)
-- [k8sgpt](https://github.com/k8sgpt-ai/k8sgpt#running-local-models)
-- [Spark](https://github.com/cedriking/spark)
-- [autogpt4all](https://github.com/aorumbayev/autogpt4all)
-- [Mods](https://github.com/charmbracelet/mods)
-- [Flowise](https://github.com/FlowiseAI/Flowise)
-
-## Sponsors
+## ❤️ Sponsors
 
 > Do you find LocalAI useful?
 
@@ -210,26 +144,27 @@ Support the project by becoming [a backer or sponsor](https://github.com/sponsor
 
 A huge thank you to our generous sponsors who support this project:
 
-| ![Spectro Cloud logo_600x600px_transparent bg](https://github.com/go-skynet/LocalAI/assets/2420543/68a6f3cb-8a65-4a4d-99b5-6417a8905512) | 
+| ![Spectro Cloud logo_600x600px_transparent bg](https://github.com/go-skynet/LocalAI/assets/2420543/68a6f3cb-8a65-4a4d-99b5-6417a8905512) |
 |:-----------------------------------------------:|
-|  [Spectro Cloud](https://www.spectrocloud.com/)  |  
+|  [Spectro Cloud](https://www.spectrocloud.com/)  |
 |  Spectro Cloud kindly supports LocalAI by providing GPU and computing resources to run tests on lamdalabs!  |
 
-## Star history
+And a huge shout-out to individuals sponsoring the project by donating hardware or backing the project.
+
+- [Sponsor list](https://github.com/sponsors/mudler)
+- JDAM00 (donating HW for the CI)
+
+## 🌟 Star history
 
 [![LocalAI Star history Chart](https://api.star-history.com/svg?repos=go-skynet/LocalAI&type=Date)](https://star-history.com/#go-skynet/LocalAI&Date)
 
-## License
+## 📖 License
 
 LocalAI is a community-driven project created by [Ettore Di Giacinto](https://github.com/mudler/).
 
-MIT
+MIT - Author Ettore Di Giacinto
 
-## Author
-
-Ettore Di Giacinto and others
-
-## Acknowledgements
+## 🙇 Acknowledgements
 
 LocalAI couldn't have been built without the help of great software already available from the community. Thank you!
 
@@ -240,9 +175,12 @@ LocalAI couldn't have been built without the help of great software already avai
 - https://github.com/EdVince/Stable-Diffusion-NCNN
 - https://github.com/ggerganov/whisper.cpp
 - https://github.com/saharNooby/rwkv.cpp
+- https://github.com/rhasspy/piper
+- https://github.com/cmp-nct/ggllm.cpp
 
-## Contributors
+## 🤗 Contributors
 
+This is a community project, a special thanks to our contributors! 🤗
 <a href="https://github.com/go-skynet/LocalAI/graphs/contributors">
   <img src="https://contrib.rocks/image?repo=go-skynet/LocalAI" />
 </a>

api/api.go

@@ -1,14 +1,22 @@
 package api
 
 import (
+	"encoding/json"
 	"errors"
+	"fmt"
+	"os"
+	"strings"
 
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/localai"
 	"github.com/go-skynet/LocalAI/api/openai"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/api/schema"
 	"github.com/go-skynet/LocalAI/internal"
+	"github.com/go-skynet/LocalAI/metrics"
 	"github.com/go-skynet/LocalAI/pkg/assets"
+	"github.com/go-skynet/LocalAI/pkg/model"
+	"github.com/go-skynet/LocalAI/pkg/startup"
 
 	"github.com/gofiber/fiber/v2"
 	"github.com/gofiber/fiber/v2/middleware/cors"
@@ -18,7 +26,7 @@ import (
 	"github.com/rs/zerolog/log"
 )
 
-func App(opts ...options.AppOption) (*fiber.App, error) {
+func Startup(opts ...options.AppOption) (*options.Option, *config.ConfigLoader, error) {
 	options := options.NewOptions(opts...)
 
 	zerolog.SetGlobalLevel(zerolog.InfoLevel)
@@ -26,6 +34,87 @@ func App(opts ...options.AppOption) (*fiber.App, error) {
 		zerolog.SetGlobalLevel(zerolog.DebugLevel)
 	}
 
+	log.Info().Msgf("Starting LocalAI using %d threads, with models path: %s", options.Threads, options.Loader.ModelPath)
+	log.Info().Msgf("LocalAI version: %s", internal.PrintableVersion())
+
+	startup.PreloadModelsConfigurations(options.Loader.ModelPath, options.ModelsURL...)
+
+	cl := config.NewConfigLoader()
+	if err := cl.LoadConfigs(options.Loader.ModelPath); err != nil {
+		log.Error().Msgf("error loading config files: %s", err.Error())
+	}
+
+	if options.ConfigFile != "" {
+		if err := cl.LoadConfigFile(options.ConfigFile); err != nil {
+			log.Error().Msgf("error loading config file: %s", err.Error())
+		}
+	}
+
+	if err := cl.Preload(options.Loader.ModelPath); err != nil {
+		log.Error().Msgf("error downloading models: %s", err.Error())
+	}
+
+	if options.PreloadJSONModels != "" {
+		if err := localai.ApplyGalleryFromString(options.Loader.ModelPath, options.PreloadJSONModels, cl, options.Galleries); err != nil {
+			return nil, nil, err
+		}
+	}
+
+	if options.PreloadModelsFromPath != "" {
+		if err := localai.ApplyGalleryFromFile(options.Loader.ModelPath, options.PreloadModelsFromPath, cl, options.Galleries); err != nil {
+			return nil, nil, err
+		}
+	}
+
+	if options.Debug {
+		for _, v := range cl.ListConfigs() {
+			cfg, _ := cl.GetConfig(v)
+			log.Debug().Msgf("Model: %s (config: %+v)", v, cfg)
+		}
+	}
+
+	if options.AssetsDestination != "" {
+		// Extract files from the embedded FS
+		err := assets.ExtractFiles(options.BackendAssets, options.AssetsDestination)
+		log.Debug().Msgf("Extracting backend assets files to %s", options.AssetsDestination)
+		if err != nil {
+			log.Warn().Msgf("Failed extracting backend assets files: %s (might be required for some backends to work properly, like gpt4all)", err)
+		}
+	}
+
+	// turn off any process that was started by GRPC if the context is canceled
+	go func() {
+		<-options.Context.Done()
+		log.Debug().Msgf("Context canceled, shutting down")
+		options.Loader.StopAllGRPC()
+	}()
+
+	if options.WatchDog {
+		wd := model.NewWatchDog(
+			options.Loader,
+			options.WatchDogBusyTimeout,
+			options.WatchDogIdleTimeout,
+			options.WatchDogBusy,
+			options.WatchDogIdle)
+		options.Loader.SetWatchDog(wd)
+		go wd.Run()
+		go func() {
+			<-options.Context.Done()
+			log.Debug().Msgf("Context canceled, shutting down")
+			wd.Shutdown()
+		}()
+	}
+
+	return options, cl, nil
+}
+
+func App(opts ...options.AppOption) (*fiber.App, error) {
+
+	options, cl, err := Startup(opts...)
+	if err != nil {
+		return nil, fmt.Errorf("failed basic startup tasks with error %s", err.Error())
+	}
+
 	// Return errors as JSON responses
 	app := fiber.New(fiber.Config{
 		BodyLimit:             options.UploadLimitMB * 1024 * 1024, // this is the default limit of 4MB
@@ -43,8 +132,8 @@ func App(opts ...options.AppOption) (*fiber.App, error) {
 
 			// Send custom error page
 			return ctx.Status(code).JSON(
-				openai.ErrorResponse{
-					Error: &openai.APIError{Message: err.Error(), Code: code},
+				schema.ErrorResponse{
+					Error: &schema.APIError{Message: err.Error(), Code: code},
 				},
 			)
 		},
@@ -56,49 +145,54 @@ func App(opts ...options.AppOption) (*fiber.App, error) {
 		}))
 	}
 
-	log.Info().Msgf("Starting LocalAI using %d threads, with models path: %s", options.Threads, options.Loader.ModelPath)
-	log.Info().Msgf("LocalAI version: %s", internal.PrintableVersion())
-
-	cm := config.NewConfigLoader()
-	if err := cm.LoadConfigs(options.Loader.ModelPath); err != nil {
-		log.Error().Msgf("error loading config files: %s", err.Error())
-	}
-
-	if options.ConfigFile != "" {
-		if err := cm.LoadConfigFile(options.ConfigFile); err != nil {
-			log.Error().Msgf("error loading config file: %s", err.Error())
-		}
-	}
-
-	if options.Debug {
-		for _, v := range cm.ListConfigs() {
-			cfg, _ := cm.GetConfig(v)
-			log.Debug().Msgf("Model: %s (config: %+v)", v, cfg)
-		}
-	}
-
-	if options.AssetsDestination != "" {
-		// Extract files from the embedded FS
-		err := assets.ExtractFiles(options.BackendAssets, options.AssetsDestination)
-		log.Debug().Msgf("Extracting backend assets files to %s", options.AssetsDestination)
-		if err != nil {
-			log.Warn().Msgf("Failed extracting backend assets files: %s (might be required for some backends to work properly, like gpt4all)", err)
-		}
-	}
-
 	// Default middleware config
 	app.Use(recover.New())
-
-	if options.PreloadJSONModels != "" {
-		if err := localai.ApplyGalleryFromString(options.Loader.ModelPath, options.PreloadJSONModels, cm, options.Galleries); err != nil {
-			return nil, err
-		}
+	if options.Metrics != nil {
+		app.Use(metrics.APIMiddleware(options.Metrics))
 	}
 
-	if options.PreloadModelsFromPath != "" {
-		if err := localai.ApplyGalleryFromFile(options.Loader.ModelPath, options.PreloadModelsFromPath, cm, options.Galleries); err != nil {
-			return nil, err
+	// Auth middleware checking if API key is valid. If no API key is set, no auth is required.
+	auth := func(c *fiber.Ctx) error {
+		if len(options.ApiKeys) == 0 {
+			return c.Next()
 		}
+
+		// Check for api_keys.json file
+		fileContent, err := os.ReadFile("api_keys.json")
+		if err == nil {
+			// Parse JSON content from the file
+			var fileKeys []string
+			err := json.Unmarshal(fileContent, &fileKeys)
+			if err != nil {
+				return c.Status(fiber.StatusInternalServerError).JSON(fiber.Map{"message": "Error parsing api_keys.json"})
+			}
+
+			// Add file keys to options.ApiKeys
+			options.ApiKeys = append(options.ApiKeys, fileKeys...)
+		}
+
+		if len(options.ApiKeys) == 0 {
+			return c.Next()
+		}
+
+		authHeader := c.Get("Authorization")
+		if authHeader == "" {
+			return c.Status(fiber.StatusUnauthorized).JSON(fiber.Map{"message": "Authorization header missing"})
+		}
+		authHeaderParts := strings.Split(authHeader, " ")
+		if len(authHeaderParts) != 2 || authHeaderParts[0] != "Bearer" {
+			return c.Status(fiber.StatusUnauthorized).JSON(fiber.Map{"message": "Invalid Authorization header format"})
+		}
+
+		apiKey := authHeaderParts[1]
+		for _, key := range options.ApiKeys {
+			if apiKey == key {
+				return c.Next()
+			}
+		}
+
+		return c.Status(fiber.StatusUnauthorized).JSON(fiber.Map{"message": "Invalid API key"})
+
 	}
 
 	if options.CORS {
@@ -114,44 +208,49 @@ func App(opts ...options.AppOption) (*fiber.App, error) {
 
 	// LocalAI API endpoints
 	galleryService := localai.NewGalleryService(options.Loader.ModelPath)
-	galleryService.Start(options.Context, cm)
+	galleryService.Start(options.Context, cl)
 
-	app.Get("/version", func(c *fiber.Ctx) error {
+	app.Get("/version", auth, func(c *fiber.Ctx) error {
 		return c.JSON(struct {
 			Version string `json:"version"`
 		}{Version: internal.PrintableVersion()})
 	})
 
-	app.Post("/models/apply", localai.ApplyModelGalleryEndpoint(options.Loader.ModelPath, cm, galleryService.C, options.Galleries))
-	app.Get("/models/available", localai.ListModelFromGalleryEndpoint(options.Galleries, options.Loader.ModelPath))
-	app.Get("/models/jobs/:uuid", localai.GetOpStatusEndpoint(galleryService))
+	modelGalleryService := localai.CreateModelGalleryService(options.Galleries, options.Loader.ModelPath, galleryService)
+	app.Post("/models/apply", auth, modelGalleryService.ApplyModelGalleryEndpoint())
+	app.Get("/models/available", auth, modelGalleryService.ListModelFromGalleryEndpoint())
+	app.Get("/models/galleries", auth, modelGalleryService.ListModelGalleriesEndpoint())
+	app.Post("/models/galleries", auth, modelGalleryService.AddModelGalleryEndpoint())
+	app.Delete("/models/galleries", auth, modelGalleryService.RemoveModelGalleryEndpoint())
+	app.Get("/models/jobs/:uuid", auth, modelGalleryService.GetOpStatusEndpoint())
+	app.Get("/models/jobs", auth, modelGalleryService.GetAllStatusEndpoint())
 
 	// openAI compatible API endpoint
 
 	// chat
-	app.Post("/v1/chat/completions", openai.ChatEndpoint(cm, options))
-	app.Post("/chat/completions", openai.ChatEndpoint(cm, options))
+	app.Post("/v1/chat/completions", auth, openai.ChatEndpoint(cl, options))
+	app.Post("/chat/completions", auth, openai.ChatEndpoint(cl, options))
 
 	// edit
-	app.Post("/v1/edits", openai.EditEndpoint(cm, options))
-	app.Post("/edits", openai.EditEndpoint(cm, options))
+	app.Post("/v1/edits", auth, openai.EditEndpoint(cl, options))
+	app.Post("/edits", auth, openai.EditEndpoint(cl, options))
 
 	// completion
-	app.Post("/v1/completions", openai.CompletionEndpoint(cm, options))
-	app.Post("/completions", openai.CompletionEndpoint(cm, options))
-	app.Post("/v1/engines/:model/completions", openai.CompletionEndpoint(cm, options))
+	app.Post("/v1/completions", auth, openai.CompletionEndpoint(cl, options))
+	app.Post("/completions", auth, openai.CompletionEndpoint(cl, options))
+	app.Post("/v1/engines/:model/completions", auth, openai.CompletionEndpoint(cl, options))
 
 	// embeddings
-	app.Post("/v1/embeddings", openai.EmbeddingsEndpoint(cm, options))
-	app.Post("/embeddings", openai.EmbeddingsEndpoint(cm, options))
-	app.Post("/v1/engines/:model/embeddings", openai.EmbeddingsEndpoint(cm, options))
+	app.Post("/v1/embeddings", auth, openai.EmbeddingsEndpoint(cl, options))
+	app.Post("/embeddings", auth, openai.EmbeddingsEndpoint(cl, options))
+	app.Post("/v1/engines/:model/embeddings", auth, openai.EmbeddingsEndpoint(cl, options))
 
 	// audio
-	app.Post("/v1/audio/transcriptions", openai.TranscriptEndpoint(cm, options))
-	app.Post("/tts", localai.TTSEndpoint(cm, options))
+	app.Post("/v1/audio/transcriptions", auth, openai.TranscriptEndpoint(cl, options))
+	app.Post("/tts", auth, localai.TTSEndpoint(cl, options))
 
 	// images
-	app.Post("/v1/images/generations", openai.ImageEndpoint(cm, options))
+	app.Post("/v1/images/generations", auth, openai.ImageEndpoint(cl, options))
 
 	if options.ImageDir != "" {
 		app.Static("/generated-images", options.ImageDir)
@@ -169,16 +268,16 @@ func App(opts ...options.AppOption) (*fiber.App, error) {
 	app.Get("/healthz", ok)
 	app.Get("/readyz", ok)
 
-	// models
-	app.Get("/v1/models", openai.ListModelsEndpoint(options.Loader, cm))
-	app.Get("/models", openai.ListModelsEndpoint(options.Loader, cm))
+	// Experimental Backend Statistics Module
+	backendMonitor := localai.NewBackendMonitor(cl, options) // Split out for now
+	app.Get("/backend/monitor", localai.BackendMonitorEndpoint(backendMonitor))
+	app.Post("/backend/shutdown", localai.BackendShutdownEndpoint(backendMonitor))
 
-	// turn off any process that was started by GRPC if the context is canceled
-	go func() {
-		<-options.Context.Done()
-		log.Debug().Msgf("Context canceled, shutting down")
-		options.Loader.StopGRPC()
-	}()
+	// models
+	app.Get("/v1/models", auth, openai.ListModelsEndpoint(options.Loader, cl))
+	app.Get("/models", auth, openai.ListModelsEndpoint(options.Loader, cl))
+
+	app.Get("/metrics", metrics.MetricsHandler())
 
 	return app, nil
 }

api/api_test.go

@@ -8,7 +8,6 @@ import (
 	"errors"
 	"fmt"
 	"io"
-	"io/ioutil"
 	"net/http"
 	"os"
 	"path/filepath"
@@ -16,9 +15,10 @@ import (
 
 	. "github.com/go-skynet/LocalAI/api"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/metrics"
+	"github.com/go-skynet/LocalAI/pkg/downloader"
 	"github.com/go-skynet/LocalAI/pkg/gallery"
 	"github.com/go-skynet/LocalAI/pkg/model"
-	"github.com/go-skynet/LocalAI/pkg/utils"
 	"github.com/gofiber/fiber/v2"
 	. "github.com/onsi/ginkgo/v2"
 	. "github.com/onsi/gomega"
@@ -30,10 +30,10 @@ import (
 )
 
 type modelApplyRequest struct {
-	ID        string            `json:"id"`
-	URL       string            `json:"url"`
-	Name      string            `json:"name"`
-	Overrides map[string]string `json:"overrides"`
+	ID        string                 `json:"id"`
+	URL       string                 `json:"url"`
+	Name      string                 `json:"name"`
+	Overrides map[string]interface{} `json:"overrides"`
 }
 
 func getModelStatus(url string) (response map[string]interface{}) {
@@ -45,7 +45,7 @@ func getModelStatus(url string) (response map[string]interface{}) {
 	}
 	defer resp.Body.Close()
 
-	body, err := ioutil.ReadAll(resp.Body)
+	body, err := io.ReadAll(resp.Body)
 	if err != nil {
 		fmt.Println("Error reading response body:", err)
 		return
@@ -61,7 +61,7 @@ func getModelStatus(url string) (response map[string]interface{}) {
 }
 
 func getModels(url string) (response []gallery.GalleryModel) {
-	utils.GetURI(url, func(url string, i []byte) error {
+	downloader.GetURI(url, func(url string, i []byte) error {
 		// Unmarshal YAML data into a struct
 		return json.Unmarshal(i, &response)
 	})
@@ -97,7 +97,7 @@ func postModelApplyRequest(url string, request modelApplyRequest) (response map[
 	}
 	defer resp.Body.Close()
 
-	body, err := ioutil.ReadAll(resp.Body)
+	body, err := io.ReadAll(resp.Body)
 	if err != nil {
 		fmt.Println("Error reading response body:", err)
 		return
@@ -125,6 +125,11 @@ var _ = Describe("API test", func() {
 	var cancel context.CancelFunc
 	var tmpdir string
 
+	commonOpts := []options.AppOption{
+		options.WithDebug(true),
+		options.WithDisableMessage(true),
+	}
+
 	Context("API with ephemeral models", func() {
 		BeforeEach(func() {
 			var err error
@@ -143,12 +148,12 @@ var _ = Describe("API test", func() {
 					Name:            "bert2",
 					URL:             "https://raw.githubusercontent.com/go-skynet/model-gallery/main/bert-embeddings.yaml",
 					Overrides:       map[string]interface{}{"foo": "bar"},
-					AdditionalFiles: []gallery.File{gallery.File{Filename: "foo.yaml", URI: "https://raw.githubusercontent.com/go-skynet/model-gallery/main/bert-embeddings.yaml"}},
+					AdditionalFiles: []gallery.File{{Filename: "foo.yaml", URI: "https://raw.githubusercontent.com/go-skynet/model-gallery/main/bert-embeddings.yaml"}},
 				},
 			}
 			out, err := yaml.Marshal(g)
 			Expect(err).ToNot(HaveOccurred())
-			err = ioutil.WriteFile(filepath.Join(tmpdir, "gallery_simple.yaml"), out, 0644)
+			err = os.WriteFile(filepath.Join(tmpdir, "gallery_simple.yaml"), out, 0644)
 			Expect(err).ToNot(HaveOccurred())
 
 			galleries := []gallery.Gallery{
@@ -158,10 +163,15 @@ var _ = Describe("API test", func() {
 				},
 			}
 
+			metricsService, err := metrics.SetupMetrics()
+			Expect(err).ToNot(HaveOccurred())
+
 			app, err = App(
-				options.WithContext(c),
-				options.WithGalleries(galleries),
-				options.WithModelLoader(modelLoader), options.WithBackendAssets(backendAssets), options.WithBackendAssetsOutput(tmpdir))
+				append(commonOpts,
+					options.WithMetrics(metricsService),
+					options.WithContext(c),
+					options.WithGalleries(galleries),
+					options.WithModelLoader(modelLoader), options.WithBackendAssets(backendAssets), options.WithBackendAssetsOutput(tmpdir))...)
 			Expect(err).ToNot(HaveOccurred())
 			go app.Listen("127.0.0.1:9090")
 
@@ -237,7 +247,7 @@ var _ = Describe("API test", func() {
 				response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
 					URL:  "https://raw.githubusercontent.com/go-skynet/model-gallery/main/bert-embeddings.yaml",
 					Name: "bert",
-					Overrides: map[string]string{
+					Overrides: map[string]interface{}{
 						"backend": "llama",
 					},
 				})
@@ -263,7 +273,7 @@ var _ = Describe("API test", func() {
 				response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
 					URL:       "https://raw.githubusercontent.com/go-skynet/model-gallery/main/bert-embeddings.yaml",
 					Name:      "bert",
-					Overrides: map[string]string{},
+					Overrides: map[string]interface{}{},
 				})
 
 				Expect(response["uuid"]).ToNot(BeEmpty(), fmt.Sprint(response))
@@ -284,14 +294,14 @@ var _ = Describe("API test", func() {
 				Expect(content["backend"]).To(Equal("bert-embeddings"))
 			})
 
-			It("runs openllama", Label("llama"), func() {
+			It("runs openllama(llama-ggml backend)", Label("llama"), func() {
 				if runtime.GOOS != "linux" {
 					Skip("test supported only on linux")
 				}
 				response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
 					URL:       "github:go-skynet/model-gallery/openllama_3b.yaml",
 					Name:      "openllama_3b",
-					Overrides: map[string]string{},
+					Overrides: map[string]interface{}{"backend": "llama-ggml", "mmap": true, "f16": true, "context_size": 128},
 				})
 
 				Expect(response["uuid"]).ToNot(BeEmpty(), fmt.Sprint(response))
@@ -346,6 +356,83 @@ var _ = Describe("API test", func() {
 				Expect(resp2.Choices[0].Message.FunctionCall).ToNot(BeNil())
 				Expect(resp2.Choices[0].Message.FunctionCall.Name).To(Equal("get_current_weather"), resp2.Choices[0].Message.FunctionCall.Name)
 
+				var res map[string]string
+				err = json.Unmarshal([]byte(resp2.Choices[0].Message.FunctionCall.Arguments), &res)
+				Expect(err).ToNot(HaveOccurred())
+				Expect(res["location"]).To(Equal("San Francisco, California, United States"), fmt.Sprint(res))
+				Expect(res["unit"]).To(Equal("celcius"), fmt.Sprint(res))
+				Expect(string(resp2.Choices[0].FinishReason)).To(Equal("function_call"), fmt.Sprint(resp2.Choices[0].FinishReason))
+
+			})
+
+			It("runs openllama gguf(llama-cpp)", Label("llama-gguf"), func() {
+				if runtime.GOOS != "linux" {
+					Skip("test supported only on linux")
+				}
+				modelName := "codellama"
+				response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
+					URL:       "github:go-skynet/model-gallery/codellama-7b-instruct.yaml",
+					Name:      modelName,
+					Overrides: map[string]interface{}{"backend": "llama", "mmap": true, "f16": true, "context_size": 128},
+				})
+
+				Expect(response["uuid"]).ToNot(BeEmpty(), fmt.Sprint(response))
+
+				uuid := response["uuid"].(string)
+
+				Eventually(func() bool {
+					response := getModelStatus("http://127.0.0.1:9090/models/jobs/" + uuid)
+					return response["processed"].(bool)
+				}, "360s", "10s").Should(Equal(true))
+
+				By("testing chat")
+				resp, err := client.CreateChatCompletion(context.TODO(), openai.ChatCompletionRequest{Model: modelName, Messages: []openai.ChatCompletionMessage{
+					{
+						Role:    "user",
+						Content: "How much is 2+2?",
+					},
+				}})
+				Expect(err).ToNot(HaveOccurred())
+				Expect(len(resp.Choices)).To(Equal(1))
+				Expect(resp.Choices[0].Message.Content).To(Or(ContainSubstring("4"), ContainSubstring("four")))
+
+				By("testing functions")
+				resp2, err := client.CreateChatCompletion(
+					context.TODO(),
+					openai.ChatCompletionRequest{
+						Model: modelName,
+						Messages: []openai.ChatCompletionMessage{
+							{
+								Role:    "user",
+								Content: "What is the weather like in San Francisco (celsius)?",
+							},
+						},
+						Functions: []openai.FunctionDefinition{
+							openai.FunctionDefinition{
+								Name:        "get_current_weather",
+								Description: "Get the current weather",
+								Parameters: jsonschema.Definition{
+									Type: jsonschema.Object,
+									Properties: map[string]jsonschema.Definition{
+										"location": {
+											Type:        jsonschema.String,
+											Description: "The city and state, e.g. San Francisco, CA",
+										},
+										"unit": {
+											Type: jsonschema.String,
+											Enum: []string{"celcius", "fahrenheit"},
+										},
+									},
+									Required: []string{"location"},
+								},
+							},
+						},
+					})
+				Expect(err).ToNot(HaveOccurred())
+				Expect(len(resp2.Choices)).To(Equal(1))
+				Expect(resp2.Choices[0].Message.FunctionCall).ToNot(BeNil())
+				Expect(resp2.Choices[0].Message.FunctionCall.Name).To(Equal("get_current_weather"), resp2.Choices[0].Message.FunctionCall.Name)
+
 				var res map[string]string
 				err = json.Unmarshal([]byte(resp2.Choices[0].Message.FunctionCall.Arguments), &res)
 				Expect(err).ToNot(HaveOccurred())
@@ -360,9 +447,8 @@ var _ = Describe("API test", func() {
 				}
 
 				response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
-					URL:       "github:go-skynet/model-gallery/gpt4all-j.yaml",
-					Name:      "gpt4all-j",
-					Overrides: map[string]string{},
+					URL:  "github:go-skynet/model-gallery/gpt4all-j.yaml",
+					Name: "gpt4all-j",
 				})
 
 				Expect(response["uuid"]).ToNot(BeEmpty(), fmt.Sprint(response))
@@ -372,7 +458,7 @@ var _ = Describe("API test", func() {
 				Eventually(func() bool {
 					response := getModelStatus("http://127.0.0.1:9090/models/jobs/" + uuid)
 					return response["processed"].(bool)
-				}, "360s", "10s").Should(Equal(true))
+				}, "960s", "10s").Should(Equal(true))
 
 				resp, err := client.CreateChatCompletion(context.TODO(), openai.ChatCompletionRequest{Model: "gpt4all-j", Messages: []openai.ChatCompletionMessage{openai.ChatCompletionMessage{Role: "user", Content: "How are you?"}}})
 				Expect(err).ToNot(HaveOccurred())
@@ -399,14 +485,19 @@ var _ = Describe("API test", func() {
 				},
 			}
 
+			metricsService, err := metrics.SetupMetrics()
+			Expect(err).ToNot(HaveOccurred())
+
 			app, err = App(
-				options.WithContext(c),
-				options.WithAudioDir(tmpdir),
-				options.WithImageDir(tmpdir),
-				options.WithGalleries(galleries),
-				options.WithModelLoader(modelLoader),
-				options.WithBackendAssets(backendAssets),
-				options.WithBackendAssetsOutput(tmpdir),
+				append(commonOpts,
+					options.WithContext(c),
+					options.WithMetrics(metricsService),
+					options.WithAudioDir(tmpdir),
+					options.WithImageDir(tmpdir),
+					options.WithGalleries(galleries),
+					options.WithModelLoader(modelLoader),
+					options.WithBackendAssets(backendAssets),
+					options.WithBackendAssetsOutput(tmpdir))...,
 			)
 			Expect(err).ToNot(HaveOccurred())
 			go app.Listen("127.0.0.1:9090")
@@ -465,6 +556,9 @@ var _ = Describe("API test", func() {
 
 			response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
 				ID: "model-gallery@stablediffusion",
+				Overrides: map[string]interface{}{
+					"parameters": map[string]interface{}{"model": "stablediffusion_assets"},
+				},
 			})
 
 			Expect(response["uuid"]).ToNot(BeEmpty(), fmt.Sprint(response))
@@ -499,8 +593,16 @@ var _ = Describe("API test", func() {
 			modelLoader = model.NewModelLoader(os.Getenv("MODELS_PATH"))
 			c, cancel = context.WithCancel(context.Background())
 
-			var err error
-			app, err = App(options.WithContext(c), options.WithModelLoader(modelLoader))
+			metricsService, err := metrics.SetupMetrics()
+			Expect(err).ToNot(HaveOccurred())
+
+			app, err = App(
+				append(commonOpts,
+					options.WithExternalBackend("huggingface", os.Getenv("HUGGINGFACE_GRPC")),
+					options.WithContext(c),
+					options.WithModelLoader(modelLoader),
+					options.WithMetrics(metricsService),
+				)...)
 			Expect(err).ToNot(HaveOccurred())
 			go app.Listen("127.0.0.1:9090")
 
@@ -524,7 +626,7 @@ var _ = Describe("API test", func() {
 		It("returns the models list", func() {
 			models, err := client.ListModels(context.TODO())
 			Expect(err).ToNot(HaveOccurred())
-			Expect(len(models.Models)).To(Equal(10))
+			Expect(len(models.Models)).To(Equal(6)) // If "config.yaml" should be included, this should be 8?
 		})
 		It("can generate completions", func() {
 			resp, err := client.CreateCompletion(context.TODO(), openai.CompletionRequest{Model: "testmodel", Prompt: "abcdedfghikl"})
@@ -555,9 +657,10 @@ var _ = Describe("API test", func() {
 		})
 
 		It("returns errors", func() {
+			backends := len(model.AutoLoadBackends) + 1 // +1 for huggingface
 			_, err := client.CreateCompletion(context.TODO(), openai.CompletionRequest{Model: "foomodel", Prompt: "abcdedfghikl"})
 			Expect(err).To(HaveOccurred())
-			Expect(err.Error()).To(ContainSubstring("error, status code: 500, message: could not load model - all backends returned error: 12 errors occurred:"))
+			Expect(err.Error()).To(ContainSubstring(fmt.Sprintf("error, status code: 500, message: could not load model - all backends returned error: %d errors occurred:", backends)))
 		})
 		It("transcribes audio", func() {
 			if runtime.GOOS != "linux" {
@@ -585,7 +688,7 @@ var _ = Describe("API test", func() {
 					Input: []string{"sun", "cat"},
 				},
 			)
-			Expect(err).ToNot(HaveOccurred())
+			Expect(err).ToNot(HaveOccurred(), err)
 			Expect(len(resp.Data[0].Embedding)).To(BeNumerically("==", 384))
 			Expect(len(resp.Data[1].Embedding)).To(BeNumerically("==", 384))
 
@@ -601,6 +704,36 @@ var _ = Describe("API test", func() {
 			Expect(resp2.Data[0].Embedding).To(Equal(sunEmbedding))
 		})
 
+		Context("External gRPC calls", func() {
+			It("calculate embeddings with sentencetransformers", func() {
+				if runtime.GOOS != "linux" {
+					Skip("test supported only on linux")
+				}
+				resp, err := client.CreateEmbeddings(
+					context.Background(),
+					openai.EmbeddingRequest{
+						Model: openai.AdaCodeSearchCode,
+						Input: []string{"sun", "cat"},
+					},
+				)
+				Expect(err).ToNot(HaveOccurred())
+				Expect(len(resp.Data[0].Embedding)).To(BeNumerically("==", 384))
+				Expect(len(resp.Data[1].Embedding)).To(BeNumerically("==", 384))
+
+				sunEmbedding := resp.Data[0].Embedding
+				resp2, err := client.CreateEmbeddings(
+					context.Background(),
+					openai.EmbeddingRequest{
+						Model: openai.AdaCodeSearchCode,
+						Input: []string{"sun"},
+					},
+				)
+				Expect(err).ToNot(HaveOccurred())
+				Expect(resp2.Data[0].Embedding).To(Equal(sunEmbedding))
+				Expect(resp2.Data[0].Embedding).ToNot(Equal(resp.Data[1].Embedding))
+			})
+		})
+
 		Context("backends", func() {
 			It("runs rwkv completion", func() {
 				if runtime.GOOS != "linux" {
@@ -672,8 +805,16 @@ var _ = Describe("API test", func() {
 			modelLoader = model.NewModelLoader(os.Getenv("MODELS_PATH"))
 			c, cancel = context.WithCancel(context.Background())
 
-			var err error
-			app, err = App(options.WithContext(c), options.WithModelLoader(modelLoader), options.WithConfigFile(os.Getenv("CONFIG_FILE")))
+			metricsService, err := metrics.SetupMetrics()
+			Expect(err).ToNot(HaveOccurred())
+
+			app, err = App(
+				append(commonOpts,
+					options.WithContext(c),
+					options.WithMetrics(metricsService),
+					options.WithModelLoader(modelLoader),
+					options.WithConfigFile(os.Getenv("CONFIG_FILE")))...,
+			)
 			Expect(err).ToNot(HaveOccurred())
 			go app.Listen("127.0.0.1:9090")
 
@@ -692,19 +833,14 @@ var _ = Describe("API test", func() {
 			cancel()
 			app.Shutdown()
 		})
-		It("can generate chat completions from config file", func() {
-			models, err := client.ListModels(context.TODO())
-			Expect(err).ToNot(HaveOccurred())
-			Expect(len(models.Models)).To(Equal(12))
-		})
-		It("can generate chat completions from config file", func() {
-			resp, err := client.CreateChatCompletion(context.TODO(), openai.ChatCompletionRequest{Model: "list1", Messages: []openai.ChatCompletionMessage{openai.ChatCompletionMessage{Role: "user", Content: "abcdedfghikl"}}})
+		It("can generate chat completions from config file (list1)", func() {
+			resp, err := client.CreateChatCompletion(context.TODO(), openai.ChatCompletionRequest{Model: "list1", Messages: []openai.ChatCompletionMessage{{Role: "user", Content: "abcdedfghikl"}}})
 			Expect(err).ToNot(HaveOccurred())
 			Expect(len(resp.Choices)).To(Equal(1))
 			Expect(resp.Choices[0].Message.Content).ToNot(BeEmpty())
 		})
-		It("can generate chat completions from config file", func() {
-			resp, err := client.CreateChatCompletion(context.TODO(), openai.ChatCompletionRequest{Model: "list2", Messages: []openai.ChatCompletionMessage{openai.ChatCompletionMessage{Role: "user", Content: "abcdedfghikl"}}})
+		It("can generate chat completions from config file (list2)", func() {
+			resp, err := client.CreateChatCompletion(context.TODO(), openai.ChatCompletionRequest{Model: "list2", Messages: []openai.ChatCompletionMessage{{Role: "user", Content: "abcdedfghikl"}}})
 			Expect(err).ToNot(HaveOccurred())
 			Expect(len(resp.Choices)).To(Equal(1))
 			Expect(resp.Choices[0].Message.Content).ToNot(BeEmpty())

api/backend/embeddings.go

@@ -2,7 +2,6 @@ package backend
 
 import (
 	"fmt"
-	"sync"
 
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
@@ -22,13 +21,13 @@ func ModelEmbedding(s string, tokens []int, loader *model.ModelLoader, c config.
 	var inferenceModel interface{}
 	var err error
 
-	opts := []model.Option{
-		model.WithLoadGRPCLLMModelOpts(grpcOpts),
+	opts := modelOpts(c, o, []model.Option{
+		model.WithLoadGRPCLoadModelOpts(grpcOpts),
 		model.WithThreads(uint32(c.Threads)),
 		model.WithAssetDir(o.AssetsDestination),
-		model.WithModelFile(modelFile),
+		model.WithModel(modelFile),
 		model.WithContext(o.Context),
-	}
+	})
 
 	if c.Backend == "" {
 		inferenceModel, err = loader.GreedyLoader(opts...)
@@ -76,18 +75,6 @@ func ModelEmbedding(s string, tokens []int, loader *model.ModelLoader, c config.
 	}
 
 	return func() ([]float32, error) {
-		// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
-		mutexMap.Lock()
-		l, ok := mutexes[modelFile]
-		if !ok {
-			m := &sync.Mutex{}
-			mutexes[modelFile] = m
-			l = m
-		}
-		mutexMap.Unlock()
-		l.Lock()
-		defer l.Unlock()
-
 		embeds, err := fn()
 		if err != nil {
 			return embeds, err

api/backend/image.go

@@ -1,26 +1,38 @@
 package backend
 
 import (
-	"fmt"
-	"sync"
-
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
 	"github.com/go-skynet/LocalAI/pkg/grpc/proto"
 	model "github.com/go-skynet/LocalAI/pkg/model"
 )
 
-func ImageGeneration(height, width, mode, step, seed int, positive_prompt, negative_prompt, dst string, loader *model.ModelLoader, c config.Config, o *options.Option) (func() error, error) {
-	if c.Backend != model.StableDiffusionBackend {
-		return nil, fmt.Errorf("endpoint only working with stablediffusion models")
-	}
+func ImageGeneration(height, width, mode, step, seed int, positive_prompt, negative_prompt, src, dst string, loader *model.ModelLoader, c config.Config, o *options.Option) (func() error, error) {
 
-	inferenceModel, err := loader.BackendLoader(
+	opts := modelOpts(c, o, []model.Option{
 		model.WithBackendString(c.Backend),
 		model.WithAssetDir(o.AssetsDestination),
 		model.WithThreads(uint32(c.Threads)),
 		model.WithContext(o.Context),
-		model.WithModelFile(c.ImageGenerationAssets),
+		model.WithModel(c.Model),
+		model.WithLoadGRPCLoadModelOpts(&proto.ModelOptions{
+			CUDA:          c.CUDA || c.Diffusers.CUDA,
+			SchedulerType: c.Diffusers.SchedulerType,
+			PipelineType:  c.Diffusers.PipelineType,
+			CFGScale:      c.Diffusers.CFGScale,
+			LoraAdapter:   c.LoraAdapter,
+			LoraScale:     c.LoraScale,
+			LoraBase:      c.LoraBase,
+			IMG2IMG:       c.Diffusers.IMG2IMG,
+			CLIPModel:     c.Diffusers.ClipModel,
+			CLIPSubfolder: c.Diffusers.ClipSubFolder,
+			CLIPSkip:      int32(c.Diffusers.ClipSkip),
+			ControlNet:    c.Diffusers.ControlNet,
+		}),
+	})
+
+	inferenceModel, err := loader.BackendLoader(
+		opts...,
 	)
 	if err != nil {
 		return nil, err
@@ -30,31 +42,20 @@ func ImageGeneration(height, width, mode, step, seed int, positive_prompt, negat
 		_, err := inferenceModel.GenerateImage(
 			o.Context,
 			&proto.GenerateImageRequest{
-				Height:         int32(height),
-				Width:          int32(width),
-				Mode:           int32(mode),
-				Step:           int32(step),
-				Seed:           int32(seed),
-				PositivePrompt: positive_prompt,
-				NegativePrompt: negative_prompt,
-				Dst:            dst,
+				Height:           int32(height),
+				Width:            int32(width),
+				Mode:             int32(mode),
+				Step:             int32(step),
+				Seed:             int32(seed),
+				CLIPSkip:         int32(c.Diffusers.ClipSkip),
+				PositivePrompt:   positive_prompt,
+				NegativePrompt:   negative_prompt,
+				Dst:              dst,
+				Src:              src,
+				EnableParameters: c.Diffusers.EnableParameters,
 			})
 		return err
 	}
 
-	return func() error {
-		// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
-		mutexMap.Lock()
-		l, ok := mutexes[c.Backend]
-		if !ok {
-			m := &sync.Mutex{}
-			mutexes[c.Backend] = m
-			l = m
-		}
-		mutexMap.Unlock()
-		l.Lock()
-		defer l.Unlock()
-
-		return fn()
-	}, nil
+	return fn, nil
 }

api/backend/llm.go

@@ -1,17 +1,32 @@
 package backend
 
 import (
+	"context"
+	"os"
 	"regexp"
 	"strings"
 	"sync"
+	"unicode/utf8"
 
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/pkg/gallery"
 	"github.com/go-skynet/LocalAI/pkg/grpc"
 	model "github.com/go-skynet/LocalAI/pkg/model"
+	"github.com/go-skynet/LocalAI/pkg/utils"
 )
 
-func ModelInference(s string, loader *model.ModelLoader, c config.Config, o *options.Option, tokenCallback func(string) bool) (func() (string, error), error) {
+type LLMResponse struct {
+	Response string // should this be []byte?
+	Usage    TokenUsage
+}
+
+type TokenUsage struct {
+	Prompt     int
+	Completion int
+}
+
+func ModelInference(ctx context.Context, s string, images []string, loader *model.ModelLoader, c config.Config, o *options.Option, tokenCallback func(string, TokenUsage) bool) (func() (LLMResponse, error), error) {
 	modelFile := c.Model
 
 	grpcOpts := gRPCModelOpts(c)
@@ -19,56 +34,107 @@ func ModelInference(s string, loader *model.ModelLoader, c config.Config, o *opt
 	var inferenceModel *grpc.Client
 	var err error
 
-	opts := []model.Option{
-		model.WithLoadGRPCLLMModelOpts(grpcOpts),
+	opts := modelOpts(c, o, []model.Option{
+		model.WithLoadGRPCLoadModelOpts(grpcOpts),
 		model.WithThreads(uint32(c.Threads)), // some models uses this to allocate threads during startup
 		model.WithAssetDir(o.AssetsDestination),
-		model.WithModelFile(modelFile),
+		model.WithModel(modelFile),
 		model.WithContext(o.Context),
+	})
+
+	if c.Backend != "" {
+		opts = append(opts, model.WithBackendString(c.Backend))
+	}
+
+	// Check if the modelFile exists, if it doesn't try to load it from the gallery
+	if o.AutoloadGalleries { // experimental
+		if _, err := os.Stat(modelFile); os.IsNotExist(err) {
+			utils.ResetDownloadTimers()
+			// if we failed to load the model, we try to download it
+			err := gallery.InstallModelFromGalleryByName(o.Galleries, modelFile, loader.ModelPath, gallery.GalleryModel{}, utils.DisplayDownloadFunction)
+			if err != nil {
+				return nil, err
+			}
+		}
 	}
 
 	if c.Backend == "" {
 		inferenceModel, err = loader.GreedyLoader(opts...)
 	} else {
-		opts = append(opts, model.WithBackendString(c.Backend))
 		inferenceModel, err = loader.BackendLoader(opts...)
 	}
+
 	if err != nil {
 		return nil, err
 	}
 
 	// in GRPC, the backend is supposed to answer to 1 single token if stream is not supported
-	fn := func() (string, error) {
+	fn := func() (LLMResponse, error) {
 		opts := gRPCPredictOpts(c, loader.ModelPath)
 		opts.Prompt = s
+		opts.Images = images
+
+		tokenUsage := TokenUsage{}
+
+		// check the per-model feature flag for usage, since tokenCallback may have a cost.
+		// Defaults to off as for now it is still experimental
+		if c.FeatureFlag.Enabled("usage") {
+			userTokenCallback := tokenCallback
+			if userTokenCallback == nil {
+				userTokenCallback = func(token string, usage TokenUsage) bool {
+					return true
+				}
+			}
+
+			promptInfo, pErr := inferenceModel.TokenizeString(ctx, opts)
+			if pErr == nil && promptInfo.Length > 0 {
+				tokenUsage.Prompt = int(promptInfo.Length)
+			}
+
+			tokenCallback = func(token string, usage TokenUsage) bool {
+				tokenUsage.Completion++
+				return userTokenCallback(token, tokenUsage)
+			}
+		}
+
 		if tokenCallback != nil {
 			ss := ""
-			err := inferenceModel.PredictStream(o.Context, opts, func(s string) {
-				tokenCallback(s)
-				ss += s
+
+			var partialRune []byte
+			err := inferenceModel.PredictStream(ctx, opts, func(chars []byte) {
+				partialRune = append(partialRune, chars...)
+
+				for len(partialRune) > 0 {
+					r, size := utf8.DecodeRune(partialRune)
+					if r == utf8.RuneError {
+						// incomplete rune, wait for more bytes
+						break
+					}
+
+					tokenCallback(string(r), tokenUsage)
+					ss += string(r)
+
+					partialRune = partialRune[size:]
+				}
 			})
-			return ss, err
+			return LLMResponse{
+				Response: ss,
+				Usage:    tokenUsage,
+			}, err
 		} else {
-			reply, err := inferenceModel.Predict(o.Context, opts)
-			return reply.Message, err
+			// TODO: Is the chicken bit the only way to get here? is that acceptable?
+			reply, err := inferenceModel.Predict(ctx, opts)
+			if err != nil {
+				return LLMResponse{}, err
+			}
+			return LLMResponse{
+				Response: string(reply.Message),
+				Usage:    tokenUsage,
+			}, err
 		}
 	}
 
-	return func() (string, error) {
-		// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
-		mutexMap.Lock()
-		l, ok := mutexes[modelFile]
-		if !ok {
-			m := &sync.Mutex{}
-			mutexes[modelFile] = m
-			l = m
-		}
-		mutexMap.Unlock()
-		l.Lock()
-		defer l.Unlock()
-
-		return fn()
-	}, nil
+	return fn, nil
 }
 
 var cutstrings map[string]*regexp.Regexp = make(map[string]*regexp.Regexp)
@@ -93,6 +159,9 @@ func Finetune(config config.Config, input, prediction string) string {
 	for _, c := range config.TrimSpace {
 		prediction = strings.TrimSpace(strings.TrimPrefix(prediction, c))
 	}
-	return prediction
 
+	for _, c := range config.TrimSuffix {
+		prediction = strings.TrimSpace(strings.TrimSuffix(prediction, c))
+	}
+	return prediction
 }

api/backend/lock.go

@@ -1,22 +0,0 @@
-package backend
-
-import "sync"
-
-// mutex still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
-var mutexMap sync.Mutex
-var mutexes map[string]*sync.Mutex = make(map[string]*sync.Mutex)
-
-func Lock(s string) *sync.Mutex {
-	// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
-	mutexMap.Lock()
-	l, ok := mutexes[s]
-	if !ok {
-		m := &sync.Mutex{}
-		mutexes[s] = m
-		l = m
-	}
-	mutexMap.Unlock()
-	l.Lock()
-
-	return l
-}

api/backend/options.go

@@ -5,29 +5,80 @@ import (
 	"path/filepath"
 
 	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
+	model "github.com/go-skynet/LocalAI/pkg/model"
 
 	config "github.com/go-skynet/LocalAI/api/config"
+	"github.com/go-skynet/LocalAI/api/options"
 )
 
+func modelOpts(c config.Config, o *options.Option, opts []model.Option) []model.Option {
+	if o.SingleBackend {
+		opts = append(opts, model.WithSingleActiveBackend())
+	}
+
+	if o.ParallelBackendRequests {
+		opts = append(opts, model.EnableParallelRequests)
+	}
+
+	if c.GRPC.Attempts != 0 {
+		opts = append(opts, model.WithGRPCAttempts(c.GRPC.Attempts))
+	}
+
+	if c.GRPC.AttemptsSleepTime != 0 {
+		opts = append(opts, model.WithGRPCAttemptsDelay(c.GRPC.AttemptsSleepTime))
+	}
+
+	for k, v := range o.ExternalGRPCBackends {
+		opts = append(opts, model.WithExternalBackend(k, v))
+	}
+
+	return opts
+}
+
 func gRPCModelOpts(c config.Config) *pb.ModelOptions {
 	b := 512
 	if c.Batch != 0 {
 		b = c.Batch
 	}
+
 	return &pb.ModelOptions{
-		ContextSize: int32(c.ContextSize),
-		Seed:        int32(c.Seed),
-		NBatch:      int32(b),
-		F16Memory:   c.F16,
-		MLock:       c.MMlock,
-		NUMA:        c.NUMA,
-		Embeddings:  c.Embeddings,
-		LowVRAM:     c.LowVRAM,
-		NGPULayers:  int32(c.NGPULayers),
-		MMap:        c.MMap,
-		MainGPU:     c.MainGPU,
-		Threads:     int32(c.Threads),
-		TensorSplit: c.TensorSplit,
+		ContextSize:    int32(c.ContextSize),
+		Seed:           int32(c.Seed),
+		NBatch:         int32(b),
+		NoMulMatQ:      c.NoMulMatQ,
+		CUDA:           c.CUDA, // diffusers, transformers
+		DraftModel:     c.DraftModel,
+		AudioPath:      c.VallE.AudioPath,
+		Quantization:   c.Quantization,
+		MMProj:         c.MMProj,
+		YarnExtFactor:  c.YarnExtFactor,
+		YarnAttnFactor: c.YarnAttnFactor,
+		YarnBetaFast:   c.YarnBetaFast,
+		YarnBetaSlow:   c.YarnBetaSlow,
+		LoraAdapter:    c.LoraAdapter,
+		LoraBase:       c.LoraBase,
+		LoraScale:      c.LoraScale,
+		NGQA:           c.NGQA,
+		RMSNormEps:     c.RMSNormEps,
+		F16Memory:      c.F16,
+		MLock:          c.MMlock,
+		RopeFreqBase:   c.RopeFreqBase,
+		RopeFreqScale:  c.RopeFreqScale,
+		NUMA:           c.NUMA,
+		Embeddings:     c.Embeddings,
+		LowVRAM:        c.LowVRAM,
+		NGPULayers:     int32(c.NGPULayers),
+		MMap:           c.MMap,
+		MainGPU:        c.MainGPU,
+		Threads:        int32(c.Threads),
+		TensorSplit:    c.TensorSplit,
+		// AutoGPTQ
+		ModelBaseName:    c.AutoGPTQ.ModelBaseName,
+		Device:           c.AutoGPTQ.Device,
+		UseTriton:        c.AutoGPTQ.Triton,
+		UseFastTokenizer: c.AutoGPTQ.UseFastTokenizer,
+		// RWKV
+		Tokenizer: c.Tokenizer,
 	}
 }
 
@@ -39,34 +90,38 @@ func gRPCPredictOpts(c config.Config, modelPath string) *pb.PredictOptions {
 		promptCachePath = p
 	}
 	return &pb.PredictOptions{
-		Temperature:     float32(c.Temperature),
-		TopP:            float32(c.TopP),
-		TopK:            int32(c.TopK),
-		Tokens:          int32(c.Maxtokens),
-		Threads:         int32(c.Threads),
-		PromptCacheAll:  c.PromptCacheAll,
-		PromptCacheRO:   c.PromptCacheRO,
-		PromptCachePath: promptCachePath,
-		F16KV:           c.F16,
-		DebugMode:       c.Debug,
-		Grammar:         c.Grammar,
-
-		Mirostat:          int32(c.Mirostat),
-		MirostatETA:       float32(c.MirostatETA),
-		MirostatTAU:       float32(c.MirostatTAU),
-		Debug:             c.Debug,
-		StopPrompts:       c.StopWords,
-		Repeat:            int32(c.RepeatPenalty),
-		NKeep:             int32(c.Keep),
-		Batch:             int32(c.Batch),
-		IgnoreEOS:         c.IgnoreEOS,
-		Seed:              int32(c.Seed),
-		FrequencyPenalty:  float32(c.FrequencyPenalty),
-		MLock:             c.MMlock,
-		MMap:              c.MMap,
-		MainGPU:           c.MainGPU,
-		TensorSplit:       c.TensorSplit,
-		TailFreeSamplingZ: float32(c.TFZ),
-		TypicalP:          float32(c.TypicalP),
+		Temperature:         float32(c.Temperature),
+		TopP:                float32(c.TopP),
+		NDraft:              c.NDraft,
+		TopK:                int32(c.TopK),
+		Tokens:              int32(c.Maxtokens),
+		Threads:             int32(c.Threads),
+		PromptCacheAll:      c.PromptCacheAll,
+		PromptCacheRO:       c.PromptCacheRO,
+		PromptCachePath:     promptCachePath,
+		F16KV:               c.F16,
+		DebugMode:           c.Debug,
+		Grammar:             c.Grammar,
+		NegativePromptScale: c.NegativePromptScale,
+		RopeFreqBase:        c.RopeFreqBase,
+		RopeFreqScale:       c.RopeFreqScale,
+		NegativePrompt:      c.NegativePrompt,
+		Mirostat:            int32(c.LLMConfig.Mirostat),
+		MirostatETA:         float32(c.LLMConfig.MirostatETA),
+		MirostatTAU:         float32(c.LLMConfig.MirostatTAU),
+		Debug:               c.Debug,
+		StopPrompts:         c.StopWords,
+		Repeat:              int32(c.RepeatPenalty),
+		NKeep:               int32(c.Keep),
+		Batch:               int32(c.Batch),
+		IgnoreEOS:           c.IgnoreEOS,
+		Seed:                int32(c.Seed),
+		FrequencyPenalty:    float32(c.FrequencyPenalty),
+		MLock:               c.MMlock,
+		MMap:                c.MMap,
+		MainGPU:             c.MainGPU,
+		TensorSplit:         c.TensorSplit,
+		TailFreeSamplingZ:   float32(c.TFZ),
+		TypicalP:            float32(c.TypicalP),
 	}
 }

api/backend/transcript.go

@@ -0,0 +1,39 @@
+package backend
+
+import (
+	"context"
+	"fmt"
+
+	config "github.com/go-skynet/LocalAI/api/config"
+	"github.com/go-skynet/LocalAI/api/schema"
+
+	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/pkg/grpc/proto"
+	model "github.com/go-skynet/LocalAI/pkg/model"
+)
+
+func ModelTranscription(audio, language string, loader *model.ModelLoader, c config.Config, o *options.Option) (*schema.Result, error) {
+
+	opts := modelOpts(c, o, []model.Option{
+		model.WithBackendString(model.WhisperBackend),
+		model.WithModel(c.Model),
+		model.WithContext(o.Context),
+		model.WithThreads(uint32(c.Threads)),
+		model.WithAssetDir(o.AssetsDestination),
+	})
+
+	whisperModel, err := o.Loader.BackendLoader(opts...)
+	if err != nil {
+		return nil, err
+	}
+
+	if whisperModel == nil {
+		return nil, fmt.Errorf("could not load whisper model")
+	}
+
+	return whisperModel.AudioTranscription(context.Background(), &proto.TranscriptRequest{
+		Dst:      audio,
+		Language: language,
+		Threads:  uint32(c.Threads),
+	})
+}

api/backend/tts.go

@@ -0,0 +1,79 @@
+package backend
+
+import (
+	"context"
+	"fmt"
+	"os"
+	"path/filepath"
+
+	api_config "github.com/go-skynet/LocalAI/api/config"
+	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/pkg/grpc/proto"
+	model "github.com/go-skynet/LocalAI/pkg/model"
+	"github.com/go-skynet/LocalAI/pkg/utils"
+)
+
+func generateUniqueFileName(dir, baseName, ext string) string {
+	counter := 1
+	fileName := baseName + ext
+
+	for {
+		filePath := filepath.Join(dir, fileName)
+		_, err := os.Stat(filePath)
+		if os.IsNotExist(err) {
+			return fileName
+		}
+
+		counter++
+		fileName = fmt.Sprintf("%s_%d%s", baseName, counter, ext)
+	}
+}
+
+func ModelTTS(backend, text, modelFile string, loader *model.ModelLoader, o *options.Option) (string, *proto.Result, error) {
+	bb := backend
+	if bb == "" {
+		bb = model.PiperBackend
+	}
+	opts := modelOpts(api_config.Config{}, o, []model.Option{
+		model.WithBackendString(bb),
+		model.WithModel(modelFile),
+		model.WithContext(o.Context),
+		model.WithAssetDir(o.AssetsDestination),
+	})
+	piperModel, err := o.Loader.BackendLoader(opts...)
+	if err != nil {
+		return "", nil, err
+	}
+
+	if piperModel == nil {
+		return "", nil, fmt.Errorf("could not load piper model")
+	}
+
+	if err := os.MkdirAll(o.AudioDir, 0755); err != nil {
+		return "", nil, fmt.Errorf("failed creating audio directory: %s", err)
+	}
+
+	fileName := generateUniqueFileName(o.AudioDir, "piper", ".wav")
+	filePath := filepath.Join(o.AudioDir, fileName)
+
+	// If the model file is not empty, we pass it joined with the model path
+	modelPath := ""
+	if modelFile != "" {
+		if bb != model.TransformersMusicGen {
+			modelPath = filepath.Join(o.Loader.ModelPath, modelFile)
+			if err := utils.VerifyPath(modelPath, o.Loader.ModelPath); err != nil {
+				return "", nil, err
+			}
+		} else {
+			modelPath = modelFile
+		}
+	}
+
+	res, err := piperModel.TTS(context.Background(), &proto.TTSRequest{
+		Text:  text,
+		Model: modelPath,
+		Dst:   filePath,
+	})
+
+	return filePath, res, err
+}

api/config/config.go

@@ -1,6 +1,7 @@
 package api_config
 
 import (
+	"errors"
 	"fmt"
 	"io/fs"
 	"os"
@@ -8,47 +9,137 @@ import (
 	"strings"
 	"sync"
 
+	"github.com/go-skynet/LocalAI/pkg/downloader"
+	"github.com/go-skynet/LocalAI/pkg/utils"
+	"github.com/rs/zerolog/log"
 	"gopkg.in/yaml.v3"
 )
 
 type Config struct {
 	PredictionOptions `yaml:"parameters"`
-	Name              string            `yaml:"name"`
-	StopWords         []string          `yaml:"stopwords"`
-	Cutstrings        []string          `yaml:"cutstrings"`
-	TrimSpace         []string          `yaml:"trimspace"`
-	ContextSize       int               `yaml:"context_size"`
-	F16               bool              `yaml:"f16"`
-	NUMA              bool              `yaml:"numa"`
-	Threads           int               `yaml:"threads"`
-	Debug             bool              `yaml:"debug"`
-	Roles             map[string]string `yaml:"roles"`
-	Embeddings        bool              `yaml:"embeddings"`
-	Backend           string            `yaml:"backend"`
-	TemplateConfig    TemplateConfig    `yaml:"template"`
-	MirostatETA       float64           `yaml:"mirostat_eta"`
-	MirostatTAU       float64           `yaml:"mirostat_tau"`
-	Mirostat          int               `yaml:"mirostat"`
-	NGPULayers        int               `yaml:"gpu_layers"`
-	MMap              bool              `yaml:"mmap"`
-	MMlock            bool              `yaml:"mmlock"`
-	LowVRAM           bool              `yaml:"low_vram"`
+	Name              string `yaml:"name"`
 
-	TensorSplit           string `yaml:"tensor_split"`
-	MainGPU               string `yaml:"main_gpu"`
-	ImageGenerationAssets string `yaml:"asset_dir"`
+	F16            bool              `yaml:"f16"`
+	Threads        int               `yaml:"threads"`
+	Debug          bool              `yaml:"debug"`
+	Roles          map[string]string `yaml:"roles"`
+	Embeddings     bool              `yaml:"embeddings"`
+	Backend        string            `yaml:"backend"`
+	TemplateConfig TemplateConfig    `yaml:"template"`
 
-	PromptCachePath string `yaml:"prompt_cache_path"`
-	PromptCacheAll  bool   `yaml:"prompt_cache_all"`
-	PromptCacheRO   bool   `yaml:"prompt_cache_ro"`
-
-	Grammar string `yaml:"grammar"`
-
-	PromptStrings, InputStrings                []string
-	InputToken                                 [][]int
-	functionCallString, functionCallNameString string
+	PromptStrings, InputStrings                []string `yaml:"-"`
+	InputToken                                 [][]int  `yaml:"-"`
+	functionCallString, functionCallNameString string   `yaml:"-"`
 
 	FunctionsConfig Functions `yaml:"function"`
+
+	FeatureFlag FeatureFlag `yaml:"feature_flags"` // Feature Flag registry. We move fast, and features may break on a per model/backend basis. Registry for (usually temporary) flags that indicate aborting something early.
+	// LLM configs (GPT4ALL, Llama.cpp, ...)
+	LLMConfig `yaml:",inline"`
+
+	// AutoGPTQ specifics
+	AutoGPTQ AutoGPTQ `yaml:"autogptq"`
+
+	// Diffusers
+	Diffusers Diffusers `yaml:"diffusers"`
+	Step      int       `yaml:"step"`
+
+	// GRPC Options
+	GRPC GRPC `yaml:"grpc"`
+
+	// Vall-e-x
+	VallE VallE `yaml:"vall-e"`
+
+	// CUDA
+	// Explicitly enable CUDA or not (some backends might need it)
+	CUDA bool `yaml:"cuda"`
+
+	DownloadFiles []File `yaml:"download_files"`
+
+	Description string `yaml:"description"`
+	Usage       string `yaml:"usage"`
+}
+
+type File struct {
+	Filename string `yaml:"filename" json:"filename"`
+	SHA256   string `yaml:"sha256" json:"sha256"`
+	URI      string `yaml:"uri" json:"uri"`
+}
+
+type VallE struct {
+	AudioPath string `yaml:"audio_path"`
+}
+
+type FeatureFlag map[string]*bool
+
+func (ff FeatureFlag) Enabled(s string) bool {
+	v, exist := ff[s]
+	return exist && v != nil && *v
+}
+
+type GRPC struct {
+	Attempts          int `yaml:"attempts"`
+	AttemptsSleepTime int `yaml:"attempts_sleep_time"`
+}
+
+type Diffusers struct {
+	CUDA             bool    `yaml:"cuda"`
+	PipelineType     string  `yaml:"pipeline_type"`
+	SchedulerType    string  `yaml:"scheduler_type"`
+	EnableParameters string  `yaml:"enable_parameters"` // A list of comma separated parameters to specify
+	CFGScale         float32 `yaml:"cfg_scale"`         // Classifier-Free Guidance Scale
+	IMG2IMG          bool    `yaml:"img2img"`           // Image to Image Diffuser
+	ClipSkip         int     `yaml:"clip_skip"`         // Skip every N frames
+	ClipModel        string  `yaml:"clip_model"`        // Clip model to use
+	ClipSubFolder    string  `yaml:"clip_subfolder"`    // Subfolder to use for clip model
+	ControlNet       string  `yaml:"control_net"`
+}
+
+type LLMConfig struct {
+	SystemPrompt    string   `yaml:"system_prompt"`
+	TensorSplit     string   `yaml:"tensor_split"`
+	MainGPU         string   `yaml:"main_gpu"`
+	RMSNormEps      float32  `yaml:"rms_norm_eps"`
+	NGQA            int32    `yaml:"ngqa"`
+	PromptCachePath string   `yaml:"prompt_cache_path"`
+	PromptCacheAll  bool     `yaml:"prompt_cache_all"`
+	PromptCacheRO   bool     `yaml:"prompt_cache_ro"`
+	MirostatETA     float64  `yaml:"mirostat_eta"`
+	MirostatTAU     float64  `yaml:"mirostat_tau"`
+	Mirostat        int      `yaml:"mirostat"`
+	NGPULayers      int      `yaml:"gpu_layers"`
+	MMap            bool     `yaml:"mmap"`
+	MMlock          bool     `yaml:"mmlock"`
+	LowVRAM         bool     `yaml:"low_vram"`
+	Grammar         string   `yaml:"grammar"`
+	StopWords       []string `yaml:"stopwords"`
+	Cutstrings      []string `yaml:"cutstrings"`
+	TrimSpace       []string `yaml:"trimspace"`
+	TrimSuffix      []string `yaml:"trimsuffix"`
+
+	ContextSize  int     `yaml:"context_size"`
+	NUMA         bool    `yaml:"numa"`
+	LoraAdapter  string  `yaml:"lora_adapter"`
+	LoraBase     string  `yaml:"lora_base"`
+	LoraScale    float32 `yaml:"lora_scale"`
+	NoMulMatQ    bool    `yaml:"no_mulmatq"`
+	DraftModel   string  `yaml:"draft_model"`
+	NDraft       int32   `yaml:"n_draft"`
+	Quantization string  `yaml:"quantization"`
+	MMProj       string  `yaml:"mmproj"`
+
+	RopeScaling    string  `yaml:"rope_scaling"`
+	YarnExtFactor  float32 `yaml:"yarn_ext_factor"`
+	YarnAttnFactor float32 `yaml:"yarn_attn_factor"`
+	YarnBetaFast   float32 `yaml:"yarn_beta_fast"`
+	YarnBetaSlow   float32 `yaml:"yarn_beta_slow"`
+}
+
+type AutoGPTQ struct {
+	ModelBaseName    string `yaml:"model_base_name"`
+	Device           string `yaml:"device"`
+	Triton           bool   `yaml:"triton"`
+	UseFastTokenizer bool   `yaml:"use_fast_tokenizer"`
 }
 
 type Functions struct {
@@ -58,10 +149,11 @@ type Functions struct {
 }
 
 type TemplateConfig struct {
-	Completion string `yaml:"completion"`
-	Functions  string `yaml:"function"`
-	Chat       string `yaml:"chat"`
-	Edit       string `yaml:"edit"`
+	Chat        string `yaml:"chat"`
+	ChatMessage string `yaml:"chat_message"`
+	Completion  string `yaml:"completion"`
+	Edit        string `yaml:"edit"`
+	Functions   string `yaml:"function"`
 }
 
 type ConfigLoader struct {
@@ -169,6 +261,16 @@ func (cm *ConfigLoader) GetConfig(m string) (Config, bool) {
 	return v, exists
 }
 
+func (cm *ConfigLoader) GetAllConfigs() []Config {
+	cm.Lock()
+	defer cm.Unlock()
+	var res []Config
+	for _, v := range cm.configs {
+		res = append(res, v)
+	}
+	return res
+}
+
 func (cm *ConfigLoader) ListConfigs() []string {
 	cm.Lock()
 	defer cm.Unlock()
@@ -179,6 +281,67 @@ func (cm *ConfigLoader) ListConfigs() []string {
 	return res
 }
 
+// Preload prepare models if they are not local but url or huggingface repositories
+func (cm *ConfigLoader) Preload(modelPath string) error {
+	cm.Lock()
+	defer cm.Unlock()
+
+	status := func(fileName, current, total string, percent float64) {
+		utils.DisplayDownloadFunction(fileName, current, total, percent)
+	}
+
+	log.Info().Msgf("Preloading models from %s", modelPath)
+
+	for i, config := range cm.configs {
+
+		// Download files and verify their SHA
+		for _, file := range config.DownloadFiles {
+			log.Debug().Msgf("Checking %q exists and matches SHA", file.Filename)
+
+			if err := utils.VerifyPath(file.Filename, modelPath); err != nil {
+				return err
+			}
+			// Create file path
+			filePath := filepath.Join(modelPath, file.Filename)
+
+			if err := downloader.DownloadFile(file.URI, filePath, file.SHA256, status); err != nil {
+				return err
+			}
+		}
+
+		modelURL := config.PredictionOptions.Model
+		modelURL = downloader.ConvertURL(modelURL)
+
+		if downloader.LooksLikeURL(modelURL) {
+			// md5 of model name
+			md5Name := utils.MD5(modelURL)
+
+			// check if file exists
+			if _, err := os.Stat(filepath.Join(modelPath, md5Name)); errors.Is(err, os.ErrNotExist) {
+				err := downloader.DownloadFile(modelURL, filepath.Join(modelPath, md5Name), "", status)
+				if err != nil {
+					return err
+				}
+			}
+
+			cc := cm.configs[i]
+			c := &cc
+			c.PredictionOptions.Model = md5Name
+			cm.configs[i] = *c
+		}
+		if cm.configs[i].Name != "" {
+			log.Info().Msgf("Model name: %s", cm.configs[i].Name)
+		}
+		if cm.configs[i].Description != "" {
+			log.Info().Msgf("Model description: %s", cm.configs[i].Description)
+		}
+		if cm.configs[i].Usage != "" {
+			log.Info().Msgf("Model usage: \n%s", cm.configs[i].Usage)
+		}
+	}
+	return nil
+}
+
 func (cm *ConfigLoader) LoadConfigs(path string) error {
 	cm.Lock()
 	defer cm.Unlock()
@@ -196,7 +359,7 @@ func (cm *ConfigLoader) LoadConfigs(path string) error {
 	}
 	for _, file := range files {
 		// Skip templates, YAML and .keep files
-		if !strings.Contains(file.Name(), ".yaml") {
+		if !strings.Contains(file.Name(), ".yaml") && !strings.Contains(file.Name(), ".yml") {
 			continue
 		}
 		c, err := ReadConfig(filepath.Join(path, file.Name()))

api/config/prediction.go

@@ -34,4 +34,17 @@ type PredictionOptions struct {
 
 	TypicalP float64 `json:"typical_p" yaml:"typical_p"`
 	Seed     int     `json:"seed" yaml:"seed"`
+
+	NegativePrompt      string  `json:"negative_prompt" yaml:"negative_prompt"`
+	RopeFreqBase        float32 `json:"rope_freq_base" yaml:"rope_freq_base"`
+	RopeFreqScale       float32 `json:"rope_freq_scale" yaml:"rope_freq_scale"`
+	NegativePromptScale float32 `json:"negative_prompt_scale" yaml:"negative_prompt_scale"`
+	// AutoGPTQ
+	UseFastTokenizer bool `json:"use_fast_tokenizer" yaml:"use_fast_tokenizer"`
+
+	// Diffusers
+	ClipSkip int `json:"clip_skip" yaml:"clip_skip"`
+
+	// RWKV (?)
+	Tokenizer string `json:"tokenizer" yaml:"tokenizer"`
 }

api/localai/backend_monitor.go

@@ -0,0 +1,162 @@
+package localai
+
+import (
+	"context"
+	"fmt"
+	"strings"
+
+	config "github.com/go-skynet/LocalAI/api/config"
+	"github.com/go-skynet/LocalAI/pkg/grpc/proto"
+
+	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/gofiber/fiber/v2"
+	"github.com/rs/zerolog/log"
+
+	gopsutil "github.com/shirou/gopsutil/v3/process"
+)
+
+type BackendMonitorRequest struct {
+	Model string `json:"model" yaml:"model"`
+}
+
+type BackendMonitorResponse struct {
+	MemoryInfo    *gopsutil.MemoryInfoStat
+	MemoryPercent float32
+	CPUPercent    float64
+}
+
+type BackendMonitor struct {
+	configLoader *config.ConfigLoader
+	options      *options.Option // Taking options in case we need to inspect ExternalGRPCBackends, though that's out of scope for now, hence the name.
+}
+
+func NewBackendMonitor(configLoader *config.ConfigLoader, options *options.Option) BackendMonitor {
+	return BackendMonitor{
+		configLoader: configLoader,
+		options:      options,
+	}
+}
+
+func (bm *BackendMonitor) SampleLocalBackendProcess(model string) (*BackendMonitorResponse, error) {
+	config, exists := bm.configLoader.GetConfig(model)
+	var backend string
+	if exists {
+		backend = config.Model
+	} else {
+		// Last ditch effort: use it raw, see if a backend happens to match.
+		backend = model
+	}
+
+	if !strings.HasSuffix(backend, ".bin") {
+		backend = fmt.Sprintf("%s.bin", backend)
+	}
+
+	pid, err := bm.options.Loader.GetGRPCPID(backend)
+
+	if err != nil {
+		log.Error().Msgf("model %s : failed to find pid %+v", model, err)
+		return nil, err
+	}
+
+	// Name is slightly frightening but this does _not_ create a new process, rather it looks up an existing process by PID.
+	backendProcess, err := gopsutil.NewProcess(int32(pid))
+
+	if err != nil {
+		log.Error().Msgf("model %s [PID %d] : error getting process info %+v", model, pid, err)
+		return nil, err
+	}
+
+	memInfo, err := backendProcess.MemoryInfo()
+
+	if err != nil {
+		log.Error().Msgf("model %s [PID %d] : error getting memory info %+v", model, pid, err)
+		return nil, err
+	}
+
+	memPercent, err := backendProcess.MemoryPercent()
+	if err != nil {
+		log.Error().Msgf("model %s [PID %d] : error getting memory percent %+v", model, pid, err)
+		return nil, err
+	}
+
+	cpuPercent, err := backendProcess.CPUPercent()
+	if err != nil {
+		log.Error().Msgf("model %s [PID %d] : error getting cpu percent %+v", model, pid, err)
+		return nil, err
+	}
+
+	return &BackendMonitorResponse{
+		MemoryInfo:    memInfo,
+		MemoryPercent: memPercent,
+		CPUPercent:    cpuPercent,
+	}, nil
+}
+
+func (bm BackendMonitor) getModelLoaderIDFromCtx(c *fiber.Ctx) (string, error) {
+	input := new(BackendMonitorRequest)
+	// Get input data from the request body
+	if err := c.BodyParser(input); err != nil {
+		return "", err
+	}
+
+	config, exists := bm.configLoader.GetConfig(input.Model)
+	var backendId string
+	if exists {
+		backendId = config.Model
+	} else {
+		// Last ditch effort: use it raw, see if a backend happens to match.
+		backendId = input.Model
+	}
+
+	if !strings.HasSuffix(backendId, ".bin") {
+		backendId = fmt.Sprintf("%s.bin", backendId)
+	}
+
+	return backendId, nil
+}
+
+func BackendMonitorEndpoint(bm BackendMonitor) func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+
+		backendId, err := bm.getModelLoaderIDFromCtx(c)
+		if err != nil {
+			return err
+		}
+
+		model := bm.options.Loader.CheckIsLoaded(backendId)
+		if model == "" {
+			return fmt.Errorf("backend %s is not currently loaded", backendId)
+		}
+
+		status, rpcErr := model.GRPC(false, nil).Status(context.TODO())
+		if rpcErr != nil {
+			log.Warn().Msgf("backend %s experienced an error retrieving status info: %s", backendId, rpcErr.Error())
+			val, slbErr := bm.SampleLocalBackendProcess(backendId)
+			if slbErr != nil {
+				return fmt.Errorf("backend %s experienced an error retrieving status info via rpc: %s, then failed local node process sample: %s", backendId, rpcErr.Error(), slbErr.Error())
+			}
+			return c.JSON(proto.StatusResponse{
+				State: proto.StatusResponse_ERROR,
+				Memory: &proto.MemoryUsageData{
+					Total: val.MemoryInfo.VMS,
+					Breakdown: map[string]uint64{
+						"gopsutil-RSS": val.MemoryInfo.RSS,
+					},
+				},
+			})
+		}
+
+		return c.JSON(status)
+	}
+}
+
+func BackendShutdownEndpoint(bm BackendMonitor) func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+		backendId, err := bm.getModelLoaderIDFromCtx(c)
+		if err != nil {
+			return err
+		}
+
+		return bm.options.Loader.ShutdownModel(backendId)
+	}
+}

api/localai/gallery.go

@@ -4,13 +4,17 @@ import (
 	"context"
 	"fmt"
 	"os"
+	"slices"
+	"strings"
 	"sync"
-	"time"
 
 	json "github.com/json-iterator/go"
+	"gopkg.in/yaml.v3"
 
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/pkg/gallery"
+	"github.com/go-skynet/LocalAI/pkg/utils"
+
 	"github.com/gofiber/fiber/v2"
 	"github.com/google/uuid"
 	"github.com/rs/zerolog/log"
@@ -24,6 +28,7 @@ type galleryOp struct {
 }
 
 type galleryOpStatus struct {
+	FileName           string  `json:"file_name"`
 	Error              error   `json:"error"`
 	Processed          bool    `json:"processed"`
 	Message            string  `json:"message"`
@@ -47,7 +52,6 @@ func NewGalleryService(modelPath string) *galleryApplier {
 	}
 }
 
-// prepareModel applies a
 func prepareModel(modelPath string, req gallery.GalleryModel, cm *config.ConfigLoader, downloadStatus func(string, string, string, float64)) error {
 
 	config, err := gallery.GetGalleryConfigFromURL(req.URL)
@@ -73,6 +77,13 @@ func (g *galleryApplier) getStatus(s string) *galleryOpStatus {
 	return g.statuses[s]
 }
 
+func (g *galleryApplier) getAllStatus() map[string]*galleryOpStatus {
+	g.Lock()
+	defer g.Unlock()
+
+	return g.statuses
+}
+
 func (g *galleryApplier) Start(c context.Context, cm *config.ConfigLoader) {
 	go func() {
 		for {
@@ -80,6 +91,8 @@ func (g *galleryApplier) Start(c context.Context, cm *config.ConfigLoader) {
 			case <-c.Done():
 				return
 			case op := <-g.C:
+				utils.ResetDownloadTimers()
+
 				g.updateStatus(op.id, &galleryOpStatus{Message: "processing", Progress: 0})
 
 				// updates the status with an error
@@ -89,14 +102,18 @@ func (g *galleryApplier) Start(c context.Context, cm *config.ConfigLoader) {
 
 				// displayDownload displays the download progress
 				progressCallback := func(fileName string, current string, total string, percentage float64) {
-					g.updateStatus(op.id, &galleryOpStatus{Message: "processing", Progress: percentage, TotalFileSize: total, DownloadedFileSize: current})
-					displayDownload(fileName, current, total, percentage)
+					g.updateStatus(op.id, &galleryOpStatus{Message: "processing", FileName: fileName, Progress: percentage, TotalFileSize: total, DownloadedFileSize: current})
+					utils.DisplayDownloadFunction(fileName, current, total, percentage)
 				}
 
 				var err error
 				// if the request contains a gallery name, we apply the gallery from the gallery list
 				if op.galleryName != "" {
-					err = gallery.InstallModelFromGallery(op.galleries, op.galleryName, g.modelPath, op.req, progressCallback)
+					if strings.Contains(op.galleryName, "@") {
+						err = gallery.InstallModelFromGallery(op.galleries, op.galleryName, g.modelPath, op.req, progressCallback)
+					} else {
+						err = gallery.InstallModelFromGalleryByName(op.galleries, op.galleryName, g.modelPath, op.req, progressCallback)
+					}
 				} else {
 					err = prepareModel(g.modelPath, op.req, cm, progressCallback)
 				}
@@ -113,40 +130,40 @@ func (g *galleryApplier) Start(c context.Context, cm *config.ConfigLoader) {
 					continue
 				}
 
+				err = cm.Preload(g.modelPath)
+				if err != nil {
+					updateError(err)
+					continue
+				}
+
 				g.updateStatus(op.id, &galleryOpStatus{Processed: true, Message: "completed", Progress: 100})
 			}
 		}
 	}()
 }
 
-var lastProgress time.Time = time.Now()
-var startTime time.Time = time.Now()
-
-func displayDownload(fileName string, current string, total string, percentage float64) {
-	currentTime := time.Now()
-
-	if currentTime.Sub(lastProgress) >= 5*time.Second {
-
-		lastProgress = currentTime
-
-		// calculate ETA based on percentage and elapsed time
-		var eta time.Duration
-		if percentage > 0 {
-			elapsed := currentTime.Sub(startTime)
-			eta = time.Duration(float64(elapsed)*(100/percentage) - float64(elapsed))
-		}
-
-		if total != "" {
-			log.Debug().Msgf("Downloading %s: %s/%s (%.2f%%) ETA: %s", fileName, current, total, percentage, eta)
-		} else {
-			log.Debug().Msgf("Downloading: %s", current)
-		}
-	}
+type galleryModel struct {
+	gallery.GalleryModel `yaml:",inline"` // https://github.com/go-yaml/yaml/issues/63
+	ID                   string           `json:"id"`
 }
 
-type galleryModel struct {
-	gallery.GalleryModel
-	ID string `json:"id"`
+func processRequests(modelPath, s string, cm *config.ConfigLoader, galleries []gallery.Gallery, requests []galleryModel) error {
+	var err error
+	for _, r := range requests {
+		utils.ResetDownloadTimers()
+		if r.ID == "" {
+			err = prepareModel(modelPath, r.GalleryModel, cm, utils.DisplayDownloadFunction)
+		} else {
+			if strings.Contains(r.ID, "@") {
+				err = gallery.InstallModelFromGallery(
+					galleries, r.ID, modelPath, r.GalleryModel, utils.DisplayDownloadFunction)
+			} else {
+				err = gallery.InstallModelFromGalleryByName(
+					galleries, r.ID, modelPath, r.GalleryModel, utils.DisplayDownloadFunction)
+			}
+		}
+	}
+	return err
 }
 
 func ApplyGalleryFromFile(modelPath, s string, cm *config.ConfigLoader, galleries []gallery.Gallery) error {
@@ -154,7 +171,13 @@ func ApplyGalleryFromFile(modelPath, s string, cm *config.ConfigLoader, gallerie
 	if err != nil {
 		return err
 	}
-	return ApplyGalleryFromString(modelPath, string(dat), cm, galleries)
+	var requests []galleryModel
+
+	if err := yaml.Unmarshal(dat, &requests); err != nil {
+		return err
+	}
+
+	return processRequests(modelPath, s, cm, galleries, requests)
 }
 
 func ApplyGalleryFromString(modelPath, s string, cm *config.ConfigLoader, galleries []gallery.Gallery) error {
@@ -164,29 +187,15 @@ func ApplyGalleryFromString(modelPath, s string, cm *config.ConfigLoader, galler
 		return err
 	}
 
-	for _, r := range requests {
-		if r.ID == "" {
-			err = prepareModel(modelPath, r.GalleryModel, cm, displayDownload)
-		} else {
-			err = gallery.InstallModelFromGallery(galleries, r.ID, modelPath, r.GalleryModel, displayDownload)
-		}
-	}
-
-	return err
+	return processRequests(modelPath, s, cm, galleries, requests)
 }
 
-/// Endpoints
+/// Endpoint Service
 
-func GetOpStatusEndpoint(g *galleryApplier) func(c *fiber.Ctx) error {
-	return func(c *fiber.Ctx) error {
-
-		status := g.getStatus(c.Params("uuid"))
-		if status == nil {
-			return fmt.Errorf("could not find any status for ID")
-		}
-
-		return c.JSON(status)
-	}
+type ModelGalleryService struct {
+	galleries      []gallery.Gallery
+	modelPath      string
+	galleryApplier *galleryApplier
 }
 
 type GalleryModel struct {
@@ -194,7 +203,31 @@ type GalleryModel struct {
 	gallery.GalleryModel
 }
 
-func ApplyModelGalleryEndpoint(modelPath string, cm *config.ConfigLoader, g chan galleryOp, galleries []gallery.Gallery) func(c *fiber.Ctx) error {
+func CreateModelGalleryService(galleries []gallery.Gallery, modelPath string, galleryApplier *galleryApplier) ModelGalleryService {
+	return ModelGalleryService{
+		galleries:      galleries,
+		modelPath:      modelPath,
+		galleryApplier: galleryApplier,
+	}
+}
+
+func (mgs *ModelGalleryService) GetOpStatusEndpoint() func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+		status := mgs.galleryApplier.getStatus(c.Params("uuid"))
+		if status == nil {
+			return fmt.Errorf("could not find any status for ID")
+		}
+		return c.JSON(status)
+	}
+}
+
+func (mgs *ModelGalleryService) GetAllStatusEndpoint() func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+		return c.JSON(mgs.galleryApplier.getAllStatus())
+	}
+}
+
+func (mgs *ModelGalleryService) ApplyModelGalleryEndpoint() func(c *fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
 		input := new(GalleryModel)
 		// Get input data from the request body
@@ -206,11 +239,11 @@ func ApplyModelGalleryEndpoint(modelPath string, cm *config.ConfigLoader, g chan
 		if err != nil {
 			return err
 		}
-		g <- galleryOp{
+		mgs.galleryApplier.C <- galleryOp{
 			req:         input.GalleryModel,
 			id:          uuid.String(),
 			galleryName: input.ID,
-			galleries:   galleries,
+			galleries:   mgs.galleries,
 		}
 		return c.JSON(struct {
 			ID        string `json:"uuid"`
@@ -219,11 +252,11 @@ func ApplyModelGalleryEndpoint(modelPath string, cm *config.ConfigLoader, g chan
 	}
 }
 
-func ListModelFromGalleryEndpoint(galleries []gallery.Gallery, basePath string) func(c *fiber.Ctx) error {
+func (mgs *ModelGalleryService) ListModelFromGalleryEndpoint() func(c *fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
-		log.Debug().Msgf("Listing models from galleries: %+v", galleries)
+		log.Debug().Msgf("Listing models from galleries: %+v", mgs.galleries)
 
-		models, err := gallery.AvailableGalleryModels(galleries, basePath)
+		models, err := gallery.AvailableGalleryModels(mgs.galleries, mgs.modelPath)
 		if err != nil {
 			return err
 		}
@@ -238,3 +271,56 @@ func ListModelFromGalleryEndpoint(galleries []gallery.Gallery, basePath string)
 		return c.Send(dat)
 	}
 }
+
+// NOTE: This is different (and much simpler!) than above! This JUST lists the model galleries that have been loaded, not their contents!
+func (mgs *ModelGalleryService) ListModelGalleriesEndpoint() func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+		log.Debug().Msgf("Listing model galleries %+v", mgs.galleries)
+		dat, err := json.Marshal(mgs.galleries)
+		if err != nil {
+			return err
+		}
+		return c.Send(dat)
+	}
+}
+
+func (mgs *ModelGalleryService) AddModelGalleryEndpoint() func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+		input := new(gallery.Gallery)
+		// Get input data from the request body
+		if err := c.BodyParser(input); err != nil {
+			return err
+		}
+		if slices.ContainsFunc(mgs.galleries, func(gallery gallery.Gallery) bool {
+			return gallery.Name == input.Name
+		}) {
+			return fmt.Errorf("%s already exists", input.Name)
+		}
+		dat, err := json.Marshal(mgs.galleries)
+		if err != nil {
+			return err
+		}
+		log.Debug().Msgf("Adding %+v to gallery list", *input)
+		mgs.galleries = append(mgs.galleries, *input)
+		return c.Send(dat)
+	}
+}
+
+func (mgs *ModelGalleryService) RemoveModelGalleryEndpoint() func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+		input := new(gallery.Gallery)
+		// Get input data from the request body
+		if err := c.BodyParser(input); err != nil {
+			return err
+		}
+		if !slices.ContainsFunc(mgs.galleries, func(gallery gallery.Gallery) bool {
+			return gallery.Name == input.Name
+		}) {
+			return fmt.Errorf("%s is not currently registered", input.Name)
+		}
+		mgs.galleries = slices.DeleteFunc(mgs.galleries, func(gallery gallery.Gallery) bool {
+			return gallery.Name == input.Name
+		})
+		return c.Send(nil)
+	}
+}

api/localai/localai.go

@@ -1,39 +1,17 @@
 package localai
 
 import (
-	"context"
-	"fmt"
-	"os"
-	"path/filepath"
-
+	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
 
 	"github.com/go-skynet/LocalAI/api/options"
-	"github.com/go-skynet/LocalAI/pkg/grpc/proto"
-	model "github.com/go-skynet/LocalAI/pkg/model"
-	"github.com/go-skynet/LocalAI/pkg/utils"
 	"github.com/gofiber/fiber/v2"
 )
 
 type TTSRequest struct {
-	Model string `json:"model" yaml:"model"`
-	Input string `json:"input" yaml:"input"`
-}
-
-func generateUniqueFileName(dir, baseName, ext string) string {
-	counter := 1
-	fileName := baseName + ext
-
-	for {
-		filePath := filepath.Join(dir, fileName)
-		_, err := os.Stat(filePath)
-		if os.IsNotExist(err) {
-			return fileName
-		}
-
-		counter++
-		fileName = fmt.Sprintf("%s_%d%s", baseName, counter, ext)
-	}
+	Model   string `json:"model" yaml:"model"`
+	Input   string `json:"input" yaml:"input"`
+	Backend string `json:"backend" yaml:"backend"`
 }
 
 func TTSEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
@@ -45,40 +23,10 @@ func TTSEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 			return err
 		}
 
-		piperModel, err := o.Loader.BackendLoader(
-			model.WithBackendString(model.PiperBackend),
-			model.WithModelFile(input.Model),
-			model.WithContext(o.Context),
-			model.WithAssetDir(o.AssetsDestination))
+		filePath, _, err := backend.ModelTTS(input.Backend, input.Input, input.Model, o.Loader, o)
 		if err != nil {
 			return err
 		}
-
-		if piperModel == nil {
-			return fmt.Errorf("could not load piper model")
-		}
-
-		if err := os.MkdirAll(o.AudioDir, 0755); err != nil {
-			return fmt.Errorf("failed creating audio directory: %s", err)
-		}
-
-		fileName := generateUniqueFileName(o.AudioDir, "piper", ".wav")
-		filePath := filepath.Join(o.AudioDir, fileName)
-
-		modelPath := filepath.Join(o.Loader.ModelPath, input.Model)
-
-		if err := utils.VerifyPath(modelPath, o.Loader.ModelPath); err != nil {
-			return err
-		}
-
-		if _, err := piperModel.TTS(context.Background(), &proto.TTSRequest{
-			Text:  input.Input,
-			Model: modelPath,
-			Dst:   filePath,
-		}); err != nil {
-			return err
-		}
-
 		return c.Download(filePath)
 	}
 }

api/openai/chat.go

@@ -6,33 +6,48 @@ import (
 	"encoding/json"
 	"fmt"
 	"strings"
+	"time"
 
 	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/api/schema"
 	"github.com/go-skynet/LocalAI/pkg/grammar"
 	model "github.com/go-skynet/LocalAI/pkg/model"
+	"github.com/go-skynet/LocalAI/pkg/utils"
 	"github.com/gofiber/fiber/v2"
+	"github.com/google/uuid"
 	"github.com/rs/zerolog/log"
 	"github.com/valyala/fasthttp"
 )
 
 func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
 	emptyMessage := ""
+	id := uuid.New().String()
+	created := int(time.Now().Unix())
 
-	process := func(s string, req *OpenAIRequest, config *config.Config, loader *model.ModelLoader, responses chan OpenAIResponse) {
-		initialMessage := OpenAIResponse{
+	process := func(s string, req *schema.OpenAIRequest, config *config.Config, loader *model.ModelLoader, responses chan schema.OpenAIResponse) {
+		initialMessage := schema.OpenAIResponse{
+			ID:      id,
+			Created: created,
 			Model:   req.Model, // we have to return what the user sent here, due to OpenAI spec.
-			Choices: []Choice{{Delta: &Message{Role: "assistant", Content: &emptyMessage}}},
+			Choices: []schema.Choice{{Delta: &schema.Message{Role: "assistant", Content: &emptyMessage}}},
 			Object:  "chat.completion.chunk",
 		}
 		responses <- initialMessage
 
-		ComputeChoices(s, req.N, config, o, loader, func(s string, c *[]Choice) {}, func(s string) bool {
-			resp := OpenAIResponse{
+		ComputeChoices(req, s, config, o, loader, func(s string, c *[]schema.Choice) {}, func(s string, usage backend.TokenUsage) bool {
+			resp := schema.OpenAIResponse{
+				ID:      id,
+				Created: created,
 				Model:   req.Model, // we have to return what the user sent here, due to OpenAI spec.
-				Choices: []Choice{{Delta: &Message{Content: &s}, Index: 0}},
+				Choices: []schema.Choice{{Delta: &schema.Message{Content: &s}, Index: 0}},
 				Object:  "chat.completion.chunk",
+				Usage: schema.OpenAIUsage{
+					PromptTokens:     usage.Prompt,
+					CompletionTokens: usage.Completion,
+					TotalTokens:      usage.Prompt + usage.Completion,
+				},
 			}
 
 			responses <- resp
@@ -43,12 +58,12 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 	return func(c *fiber.Ctx) error {
 		processFunctions := false
 		funcs := grammar.Functions{}
-		model, input, err := readInput(c, o.Loader, true)
+		modelFile, input, err := readInput(c, o, true)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
 
-		config, input, err := readConfig(model, input, cm, o.Loader, o.Debug, o.Threads, o.ContextSize, o.F16)
+		config, input, err := readConfig(modelFile, input, cm, o.Loader, o.Debug, o.Threads, o.ContextSize, o.F16)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
@@ -66,6 +81,10 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 			noActionDescription = config.FunctionsConfig.NoActionDescriptionName
 		}
 
+		if input.ResponseFormat.Type == "json_object" {
+			input.Grammar = grammar.JSONBNF
+		}
+
 		// process functions if we have any defined or if we have a function call string
 		if len(input.Functions) > 0 && config.ShouldUseFunctions() {
 			log.Debug().Msgf("Response needs to process functions")
@@ -109,10 +128,12 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 
 		var predInput string
 
+		suppressConfigSystemPrompt := false
 		mess := []string{}
-		for _, i := range input.Messages {
+		for messageIndex, i := range input.Messages {
 			var content string
 			role := i.Role
+
 			// if function call, we might want to customize the role so we can display better that the "assistant called a json action"
 			// if an "assistant_function_call" role is defined, we use it, otherwise we use the role that is passed by in the request
 			if i.FunctionCall != nil && i.Role == "assistant" {
@@ -123,34 +144,62 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 				}
 			}
 			r := config.Roles[role]
-			contentExists := i.Content != nil && *i.Content != ""
-			if r != "" {
-				if contentExists {
-					content = fmt.Sprint(r, " ", *i.Content)
+			contentExists := i.Content != nil && i.StringContent != ""
+			// First attempt to populate content via a chat message specific template
+			if config.TemplateConfig.ChatMessage != "" {
+				chatMessageData := model.ChatMessageTemplateData{
+					SystemPrompt: config.SystemPrompt,
+					Role:         r,
+					RoleName:     role,
+					Content:      i.StringContent,
+					MessageIndex: messageIndex,
 				}
-				if i.FunctionCall != nil {
-					j, err := json.Marshal(i.FunctionCall)
-					if err == nil {
-						if contentExists {
-							content += "\n" + fmt.Sprint(r, " ", string(j))
-						} else {
-							content = fmt.Sprint(r, " ", string(j))
+				templatedChatMessage, err := o.Loader.EvaluateTemplateForChatMessage(config.TemplateConfig.ChatMessage, chatMessageData)
+				if err != nil {
+					log.Error().Msgf("error processing message %+v using template \"%s\": %v. Skipping!", chatMessageData, config.TemplateConfig.ChatMessage, err)
+				} else {
+					if templatedChatMessage == "" {
+						log.Warn().Msgf("template \"%s\" produced blank output for %+v. Skipping!", config.TemplateConfig.ChatMessage, chatMessageData)
+						continue // TODO: This continue is here intentionally to skip over the line `mess = append(mess, content)` below, and to prevent the sprintf
+					}
+					log.Debug().Msgf("templated message for chat: %s", templatedChatMessage)
+					content = templatedChatMessage
+				}
+			}
+			// If this model doesn't have such a template, or if that template fails to return a value, template at the message level.
+			if content == "" {
+				if r != "" {
+					if contentExists {
+						content = fmt.Sprint(r, i.StringContent)
+					}
+					if i.FunctionCall != nil {
+						j, err := json.Marshal(i.FunctionCall)
+						if err == nil {
+							if contentExists {
+								content += "\n" + fmt.Sprint(r, " ", string(j))
+							} else {
+								content = fmt.Sprint(r, " ", string(j))
+							}
+						}
+					}
+				} else {
+					if contentExists {
+						content = fmt.Sprint(i.StringContent)
+					}
+					if i.FunctionCall != nil {
+						j, err := json.Marshal(i.FunctionCall)
+						if err == nil {
+							if contentExists {
+								content += "\n" + string(j)
+							} else {
+								content = string(j)
+							}
 						}
 					}
 				}
-			} else {
-				if contentExists {
-					content = fmt.Sprint(*i.Content)
-				}
-				if i.FunctionCall != nil {
-					j, err := json.Marshal(i.FunctionCall)
-					if err == nil {
-						if contentExists {
-							content += "\n" + string(j)
-						} else {
-							content = string(j)
-						}
-					}
+				// Special Handling: System. We care if it was printed at all, not the r branch, so check seperately
+				if contentExists && role == "system" {
+					suppressConfigSystemPrompt = true
 				}
 			}
 
@@ -170,7 +219,12 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 			c.Set("Transfer-Encoding", "chunked")
 		}
 
-		templateFile := config.Model
+		templateFile := ""
+
+		// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
+		if o.Loader.ExistsInModelPath(fmt.Sprintf("%s.tmpl", config.Model)) {
+			templateFile = config.Model
+		}
 
 		if config.TemplateConfig.Chat != "" && !processFunctions {
 			templateFile = config.TemplateConfig.Chat
@@ -180,19 +234,19 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 			templateFile = config.TemplateConfig.Functions
 		}
 
-		// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
-		templatedInput, err := o.Loader.TemplatePrefix(templateFile, struct {
-			Input     string
-			Functions []grammar.Function
-		}{
-			Input:     predInput,
-			Functions: funcs,
-		})
-		if err == nil {
-			predInput = templatedInput
-			log.Debug().Msgf("Template found, input modified to: %s", predInput)
-		} else {
-			log.Debug().Msgf("Template failed loading: %s", err.Error())
+		if templateFile != "" {
+			templatedInput, err := o.Loader.EvaluateTemplateForPrompt(model.ChatPromptTemplate, templateFile, model.PromptTemplateData{
+				SystemPrompt:         config.SystemPrompt,
+				SuppressSystemPrompt: suppressConfigSystemPrompt,
+				Input:                predInput,
+				Functions:            funcs,
+			})
+			if err == nil {
+				predInput = templatedInput
+				log.Debug().Msgf("Template found, input modified to: %s", predInput)
+			} else {
+				log.Debug().Msgf("Template failed loading: %s", err.Error())
+			}
 		}
 
 		log.Debug().Msgf("Prompt (after templating): %s", predInput)
@@ -201,31 +255,41 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 		}
 
 		if toStream {
-			responses := make(chan OpenAIResponse)
+			responses := make(chan schema.OpenAIResponse)
 
 			go process(predInput, input, config, o.Loader, responses)
 
 			c.Context().SetBodyStreamWriter(fasthttp.StreamWriter(func(w *bufio.Writer) {
 
+				usage := &schema.OpenAIUsage{}
+
 				for ev := range responses {
+					usage = &ev.Usage // Copy a pointer to the latest usage chunk so that the stop message can reference it
 					var buf bytes.Buffer
 					enc := json.NewEncoder(&buf)
 					enc.Encode(ev)
-
 					log.Debug().Msgf("Sending chunk: %s", buf.String())
-					fmt.Fprintf(w, "data: %v\n", buf.String())
+					_, err := fmt.Fprintf(w, "data: %v\n", buf.String())
+					if err != nil {
+						log.Debug().Msgf("Sending chunk failed: %v", err)
+						input.Cancel()
+						break
+					}
 					w.Flush()
 				}
 
-				resp := &OpenAIResponse{
-					Model: input.Model, // we have to return what the user sent here, due to OpenAI spec.
-					Choices: []Choice{
+				resp := &schema.OpenAIResponse{
+					ID:      id,
+					Created: created,
+					Model:   input.Model, // we have to return what the user sent here, due to OpenAI spec.
+					Choices: []schema.Choice{
 						{
 							FinishReason: "stop",
 							Index:        0,
-							Delta:        &Message{Content: &emptyMessage},
+							Delta:        &schema.Message{Content: &emptyMessage},
 						}},
 					Object: "chat.completion.chunk",
+					Usage:  *usage,
 				}
 				respData, _ := json.Marshal(resp)
 
@@ -236,10 +300,12 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 			return nil
 		}
 
-		result, err := ComputeChoices(predInput, input.N, config, o, o.Loader, func(s string, c *[]Choice) {
+		result, tokenUsage, err := ComputeChoices(input, predInput, config, o, o.Loader, func(s string, c *[]schema.Choice) {
 			if processFunctions {
 				// As we have to change the result before processing, we can't stream the answer (yet?)
 				ss := map[string]interface{}{}
+				// This prevent newlines to break JSON parsing for clients
+				s = utils.EscapeNewLines(s)
 				json.Unmarshal([]byte(s), &ss)
 				log.Debug().Msgf("Function return: %s %+v", s, ss)
 
@@ -268,7 +334,7 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 								message = backend.Finetune(*config, predInput, message)
 								log.Debug().Msgf("Reply received from LLM(finetuned): %s", message)
 
-								*c = append(*c, Choice{Message: &Message{Role: "assistant", Content: &message}})
+								*c = append(*c, schema.Choice{Message: &schema.Message{Role: "assistant", Content: &message}})
 								return
 							}
 						}
@@ -278,7 +344,11 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 					// Otherwise ask the LLM to understand the JSON output and the context, and return a message
 					// Note: This costs (in term of CPU) another computation
 					config.Grammar = ""
-					predFunc, err := backend.ModelInference(predInput, o.Loader, *config, o, nil)
+					images := []string{}
+					for _, m := range input.Messages {
+						images = append(images, m.StringImages...)
+					}
+					predFunc, err := backend.ModelInference(input.Context, predInput, images, o.Loader, *config, o, nil)
 					if err != nil {
 						log.Error().Msgf("inference error: %s", err.Error())
 						return
@@ -290,28 +360,35 @@ func ChatEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx)
 						return
 					}
 
-					prediction = backend.Finetune(*config, predInput, prediction)
-					*c = append(*c, Choice{Message: &Message{Role: "assistant", Content: &prediction}})
+					fineTunedResponse := backend.Finetune(*config, predInput, prediction.Response)
+					*c = append(*c, schema.Choice{Message: &schema.Message{Role: "assistant", Content: &fineTunedResponse}})
 				} else {
 					// otherwise reply with the function call
-					*c = append(*c, Choice{
+					*c = append(*c, schema.Choice{
 						FinishReason: "function_call",
-						Message:      &Message{Role: "assistant", FunctionCall: ss},
+						Message:      &schema.Message{Role: "assistant", FunctionCall: ss},
 					})
 				}
 
 				return
 			}
-			*c = append(*c, Choice{Message: &Message{Role: "assistant", Content: &s}})
+			*c = append(*c, schema.Choice{FinishReason: "stop", Index: 0, Message: &schema.Message{Role: "assistant", Content: &s}})
 		}, nil)
 		if err != nil {
 			return err
 		}
 
-		resp := &OpenAIResponse{
+		resp := &schema.OpenAIResponse{
+			ID:      id,
+			Created: created,
 			Model:   input.Model, // we have to return what the user sent here, due to OpenAI spec.
 			Choices: result,
 			Object:  "chat.completion",
+			Usage: schema.OpenAIUsage{
+				PromptTokens:     tokenUsage.Prompt,
+				CompletionTokens: tokenUsage.Completion,
+				TotalTokens:      tokenUsage.Prompt + tokenUsage.Completion,
+			},
 		}
 		respData, _ := json.Marshal(resp)
 		log.Debug().Msgf("Response: %s", respData)

api/openai/completion.go

@@ -6,28 +6,43 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"time"
 
+	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/api/schema"
+	"github.com/go-skynet/LocalAI/pkg/grammar"
 	model "github.com/go-skynet/LocalAI/pkg/model"
 	"github.com/gofiber/fiber/v2"
+	"github.com/google/uuid"
 	"github.com/rs/zerolog/log"
 	"github.com/valyala/fasthttp"
 )
 
 // https://platform.openai.com/docs/api-reference/completions
 func CompletionEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
-	process := func(s string, req *OpenAIRequest, config *config.Config, loader *model.ModelLoader, responses chan OpenAIResponse) {
-		ComputeChoices(s, req.N, config, o, loader, func(s string, c *[]Choice) {}, func(s string) bool {
-			resp := OpenAIResponse{
-				Model: req.Model, // we have to return what the user sent here, due to OpenAI spec.
-				Choices: []Choice{
+	id := uuid.New().String()
+	created := int(time.Now().Unix())
+
+	process := func(s string, req *schema.OpenAIRequest, config *config.Config, loader *model.ModelLoader, responses chan schema.OpenAIResponse) {
+		ComputeChoices(req, s, config, o, loader, func(s string, c *[]schema.Choice) {}, func(s string, usage backend.TokenUsage) bool {
+			resp := schema.OpenAIResponse{
+				ID:      id,
+				Created: created,
+				Model:   req.Model, // we have to return what the user sent here, due to OpenAI spec.
+				Choices: []schema.Choice{
 					{
 						Index: 0,
 						Text:  s,
 					},
 				},
 				Object: "text_completion",
+				Usage: schema.OpenAIUsage{
+					PromptTokens:     usage.Prompt,
+					CompletionTokens: usage.Completion,
+					TotalTokens:      usage.Prompt + usage.Completion,
+				},
 			}
 			log.Debug().Msgf("Sending goroutine: %s", s)
 
@@ -38,18 +53,22 @@ func CompletionEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 	}
 
 	return func(c *fiber.Ctx) error {
-		model, input, err := readInput(c, o.Loader, true)
+		modelFile, input, err := readInput(c, o, true)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
 
 		log.Debug().Msgf("`input`: %+v", input)
 
-		config, input, err := readConfig(model, input, cm, o.Loader, o.Debug, o.Threads, o.ContextSize, o.F16)
+		config, input, err := readConfig(modelFile, input, cm, o.Loader, o.Debug, o.Threads, o.ContextSize, o.F16)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
 
+		if input.ResponseFormat.Type == "json_object" {
+			input.Grammar = grammar.JSONBNF
+		}
+
 		log.Debug().Msgf("Parameter Config: %+v", config)
 
 		if input.Stream {
@@ -62,7 +81,12 @@ func CompletionEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 			c.Set("Transfer-Encoding", "chunked")
 		}
 
-		templateFile := config.Model
+		templateFile := ""
+
+		// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
+		if o.Loader.ExistsInModelPath(fmt.Sprintf("%s.tmpl", config.Model)) {
+			templateFile = config.Model
+		}
 
 		if config.TemplateConfig.Completion != "" {
 			templateFile = config.TemplateConfig.Completion
@@ -75,18 +99,17 @@ func CompletionEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 
 			predInput := config.PromptStrings[0]
 
-			// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
-			templatedInput, err := o.Loader.TemplatePrefix(templateFile, struct {
-				Input string
-			}{
-				Input: predInput,
-			})
-			if err == nil {
-				predInput = templatedInput
-				log.Debug().Msgf("Template found, input modified to: %s", predInput)
+			if templateFile != "" {
+				templatedInput, err := o.Loader.EvaluateTemplateForPrompt(model.CompletionPromptTemplate, templateFile, model.PromptTemplateData{
+					Input: predInput,
+				})
+				if err == nil {
+					predInput = templatedInput
+					log.Debug().Msgf("Template found, input modified to: %s", predInput)
+				}
 			}
 
-			responses := make(chan OpenAIResponse)
+			responses := make(chan schema.OpenAIResponse)
 
 			go process(predInput, input, config, o.Loader, responses)
 
@@ -102,9 +125,11 @@ func CompletionEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 					w.Flush()
 				}
 
-				resp := &OpenAIResponse{
-					Model: input.Model, // we have to return what the user sent here, due to OpenAI spec.
-					Choices: []Choice{
+				resp := &schema.OpenAIResponse{
+					ID:      id,
+					Created: created,
+					Model:   input.Model, // we have to return what the user sent here, due to OpenAI spec.
+					Choices: []schema.Choice{
 						{
 							Index:        0,
 							FinishReason: "stop",
@@ -121,33 +146,48 @@ func CompletionEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 			return nil
 		}
 
-		var result []Choice
-		for _, i := range config.PromptStrings {
-			// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
-			templatedInput, err := o.Loader.TemplatePrefix(templateFile, struct {
-				Input string
-			}{
-				Input: i,
-			})
-			if err == nil {
-				i = templatedInput
-				log.Debug().Msgf("Template found, input modified to: %s", i)
+		var result []schema.Choice
+
+		totalTokenUsage := backend.TokenUsage{}
+
+		for k, i := range config.PromptStrings {
+			if templateFile != "" {
+				// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
+				templatedInput, err := o.Loader.EvaluateTemplateForPrompt(model.CompletionPromptTemplate, templateFile, model.PromptTemplateData{
+					SystemPrompt: config.SystemPrompt,
+					Input:        i,
+				})
+				if err == nil {
+					i = templatedInput
+					log.Debug().Msgf("Template found, input modified to: %s", i)
+				}
 			}
 
-			r, err := ComputeChoices(i, input.N, config, o, o.Loader, func(s string, c *[]Choice) {
-				*c = append(*c, Choice{Text: s})
-			}, nil)
+			r, tokenUsage, err := ComputeChoices(
+				input, i, config, o, o.Loader, func(s string, c *[]schema.Choice) {
+					*c = append(*c, schema.Choice{Text: s, FinishReason: "stop", Index: k})
+				}, nil)
 			if err != nil {
 				return err
 			}
 
+			totalTokenUsage.Prompt += tokenUsage.Prompt
+			totalTokenUsage.Completion += tokenUsage.Completion
+
 			result = append(result, r...)
 		}
 
-		resp := &OpenAIResponse{
+		resp := &schema.OpenAIResponse{
+			ID:      id,
+			Created: created,
 			Model:   input.Model, // we have to return what the user sent here, due to OpenAI spec.
 			Choices: result,
 			Object:  "text_completion",
+			Usage: schema.OpenAIUsage{
+				PromptTokens:     totalTokenUsage.Prompt,
+				CompletionTokens: totalTokenUsage.Completion,
+				TotalTokens:      totalTokenUsage.Prompt + totalTokenUsage.Completion,
+			},
 		}
 
 		jsonResult, _ := json.Marshal(resp)

api/openai/edit.go

@@ -3,59 +3,86 @@ package openai
 import (
 	"encoding/json"
 	"fmt"
+	"time"
 
+	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/api/schema"
+	model "github.com/go-skynet/LocalAI/pkg/model"
 	"github.com/gofiber/fiber/v2"
+	"github.com/google/uuid"
+
 	"github.com/rs/zerolog/log"
 )
 
 func EditEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
-		model, input, err := readInput(c, o.Loader, true)
+		modelFile, input, err := readInput(c, o, true)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
 
-		config, input, err := readConfig(model, input, cm, o.Loader, o.Debug, o.Threads, o.ContextSize, o.F16)
+		config, input, err := readConfig(modelFile, input, cm, o.Loader, o.Debug, o.Threads, o.ContextSize, o.F16)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
 
 		log.Debug().Msgf("Parameter Config: %+v", config)
 
-		templateFile := config.Model
+		templateFile := ""
+
+		// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
+		if o.Loader.ExistsInModelPath(fmt.Sprintf("%s.tmpl", config.Model)) {
+			templateFile = config.Model
+		}
 
 		if config.TemplateConfig.Edit != "" {
 			templateFile = config.TemplateConfig.Edit
 		}
 
-		var result []Choice
+		var result []schema.Choice
+		totalTokenUsage := backend.TokenUsage{}
+
 		for _, i := range config.InputStrings {
-			// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
-			templatedInput, err := o.Loader.TemplatePrefix(templateFile, struct {
-				Input       string
-				Instruction string
-			}{Input: i})
-			if err == nil {
-				i = templatedInput
-				log.Debug().Msgf("Template found, input modified to: %s", i)
+			if templateFile != "" {
+				templatedInput, err := o.Loader.EvaluateTemplateForPrompt(model.EditPromptTemplate, templateFile, model.PromptTemplateData{
+					Input:        i,
+					Instruction:  input.Instruction,
+					SystemPrompt: config.SystemPrompt,
+				})
+				if err == nil {
+					i = templatedInput
+					log.Debug().Msgf("Template found, input modified to: %s", i)
+				}
 			}
 
-			r, err := ComputeChoices(i, input.N, config, o, o.Loader, func(s string, c *[]Choice) {
-				*c = append(*c, Choice{Text: s})
+			r, tokenUsage, err := ComputeChoices(input, i, config, o, o.Loader, func(s string, c *[]schema.Choice) {
+				*c = append(*c, schema.Choice{Text: s})
 			}, nil)
 			if err != nil {
 				return err
 			}
 
+			totalTokenUsage.Prompt += tokenUsage.Prompt
+			totalTokenUsage.Completion += tokenUsage.Completion
+
 			result = append(result, r...)
 		}
 
-		resp := &OpenAIResponse{
+		id := uuid.New().String()
+		created := int(time.Now().Unix())
+		resp := &schema.OpenAIResponse{
+			ID:      id,
+			Created: created,
 			Model:   input.Model, // we have to return what the user sent here, due to OpenAI spec.
 			Choices: result,
 			Object:  "edit",
+			Usage: schema.OpenAIUsage{
+				PromptTokens:     totalTokenUsage.Prompt,
+				CompletionTokens: totalTokenUsage.Completion,
+				TotalTokens:      totalTokenUsage.Prompt + totalTokenUsage.Completion,
+			},
 		}
 
 		jsonResult, _ := json.Marshal(resp)

api/openai/embeddings.go

@@ -3,9 +3,13 @@ package openai
 import (
 	"encoding/json"
 	"fmt"
+	"time"
 
 	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
+	"github.com/go-skynet/LocalAI/api/schema"
+	"github.com/google/uuid"
+
 	"github.com/go-skynet/LocalAI/api/options"
 	"github.com/gofiber/fiber/v2"
 	"github.com/rs/zerolog/log"
@@ -14,7 +18,7 @@ import (
 // https://platform.openai.com/docs/api-reference/embeddings
 func EmbeddingsEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
-		model, input, err := readInput(c, o.Loader, true)
+		model, input, err := readInput(c, o, true)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
@@ -25,7 +29,7 @@ func EmbeddingsEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 		}
 
 		log.Debug().Msgf("Parameter Config: %+v", config)
-		items := []Item{}
+		items := []schema.Item{}
 
 		for i, s := range config.InputToken {
 			// get the model function to call for the result
@@ -38,7 +42,7 @@ func EmbeddingsEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 			if err != nil {
 				return err
 			}
-			items = append(items, Item{Embedding: embeddings, Index: i, Object: "embedding"})
+			items = append(items, schema.Item{Embedding: embeddings, Index: i, Object: "embedding"})
 		}
 
 		for i, s := range config.InputStrings {
@@ -52,13 +56,17 @@ func EmbeddingsEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 			if err != nil {
 				return err
 			}
-			items = append(items, Item{Embedding: embeddings, Index: i, Object: "embedding"})
+			items = append(items, schema.Item{Embedding: embeddings, Index: i, Object: "embedding"})
 		}
 
-		resp := &OpenAIResponse{
-			Model:  input.Model, // we have to return what the user sent here, due to OpenAI spec.
-			Data:   items,
-			Object: "list",
+		id := uuid.New().String()
+		created := int(time.Now().Unix())
+		resp := &schema.OpenAIResponse{
+			ID:      id,
+			Created: created,
+			Model:   input.Model, // we have to return what the user sent here, due to OpenAI spec.
+			Data:    items,
+			Object:  "list",
 		}
 
 		jsonResult, _ := json.Marshal(resp)

api/openai/image.go

@@ -1,14 +1,20 @@
 package openai
 
 import (
+	"bufio"
 	"encoding/base64"
 	"encoding/json"
 	"fmt"
-	"io/ioutil"
+	"io"
+	"net/http"
 	"os"
 	"path/filepath"
 	"strconv"
 	"strings"
+	"time"
+
+	"github.com/go-skynet/LocalAI/api/schema"
+	"github.com/google/uuid"
 
 	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
@@ -18,6 +24,26 @@ import (
 	"github.com/rs/zerolog/log"
 )
 
+func downloadFile(url string) (string, error) {
+	// Get the data
+	resp, err := http.Get(url)
+	if err != nil {
+		return "", err
+	}
+	defer resp.Body.Close()
+
+	// Create the file
+	out, err := os.CreateTemp("", "image")
+	if err != nil {
+		return "", err
+	}
+	defer out.Close()
+
+	// Write the body to file
+	_, err = io.Copy(out, resp.Body)
+	return out.Name(), err
+}
+
 // https://platform.openai.com/docs/api-reference/images/create
 
 /*
@@ -35,7 +61,7 @@ import (
 */
 func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
-		m, input, err := readInput(c, o.Loader, false)
+		m, input, err := readInput(c, o, false)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
@@ -50,10 +76,58 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
 
+		src := ""
+		if input.File != "" {
+
+			fileData := []byte{}
+			// check if input.File is an URL, if so download it and save it
+			// to a temporary file
+			if strings.HasPrefix(input.File, "http://") || strings.HasPrefix(input.File, "https://") {
+				out, err := downloadFile(input.File)
+				if err != nil {
+					return fmt.Errorf("failed downloading file:%w", err)
+				}
+				defer os.RemoveAll(out)
+
+				fileData, err = os.ReadFile(out)
+				if err != nil {
+					return fmt.Errorf("failed reading file:%w", err)
+				}
+
+			} else {
+				// base 64 decode the file and write it somewhere
+				// that we will cleanup
+				fileData, err = base64.StdEncoding.DecodeString(input.File)
+				if err != nil {
+					return err
+				}
+			}
+
+			// Create a temporary file
+			outputFile, err := os.CreateTemp(o.ImageDir, "b64")
+			if err != nil {
+				return err
+			}
+			// write the base64 result
+			writer := bufio.NewWriter(outputFile)
+			_, err = writer.Write(fileData)
+			if err != nil {
+				outputFile.Close()
+				return err
+			}
+			outputFile.Close()
+			src = outputFile.Name()
+			defer os.RemoveAll(src)
+		}
+
 		log.Debug().Msgf("Parameter Config: %+v", config)
 
-		// XXX: Only stablediffusion is supported for now
-		if config.Backend == "" {
+		switch config.Backend {
+		case "stablediffusion":
+			config.Backend = model.StableDiffusionBackend
+		case "tinydream":
+			config.Backend = model.TinyDreamBackend
+		case "":
 			config.Backend = model.StableDiffusionBackend
 		}
 
@@ -71,11 +145,11 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 		}
 
 		b64JSON := false
-		if input.ResponseFormat == "b64_json" {
+		if input.ResponseFormat.Type == "b64_json" {
 			b64JSON = true
 		}
-
-		var result []Item
+		// src and clip_skip
+		var result []schema.Item
 		for _, i := range config.PromptStrings {
 			n := input.N
 			if input.N == 0 {
@@ -90,7 +164,10 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 				}
 
 				mode := 0
-				step := 15
+				step := config.Step
+				if step == 0 {
+					step = 15
+				}
 
 				if input.Mode != 0 {
 					mode = input.Mode
@@ -105,7 +182,7 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 					tempDir = o.ImageDir
 				}
 				// Create a temporary file
-				outputFile, err := ioutil.TempFile(tempDir, "b64")
+				outputFile, err := os.CreateTemp(tempDir, "b64")
 				if err != nil {
 					return err
 				}
@@ -119,7 +196,7 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 
 				baseURL := c.BaseURL()
 
-				fn, err := backend.ImageGeneration(height, width, mode, step, input.Seed, positive_prompt, negative_prompt, output, o.Loader, *config, o)
+				fn, err := backend.ImageGeneration(height, width, mode, step, input.Seed, positive_prompt, negative_prompt, src, output, o.Loader, *config, o)
 				if err != nil {
 					return err
 				}
@@ -127,7 +204,7 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 					return err
 				}
 
-				item := &Item{}
+				item := &schema.Item{}
 
 				if b64JSON {
 					defer os.RemoveAll(output)
@@ -145,8 +222,12 @@ func ImageEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx
 			}
 		}
 
-		resp := &OpenAIResponse{
-			Data: result,
+		id := uuid.New().String()
+		created := int(time.Now().Unix())
+		resp := &schema.OpenAIResponse{
+			ID:      id,
+			Created: created,
+			Data:    result,
 		}
 
 		jsonResult, _ := json.Marshal(resp)

api/openai/inference.go

@@ -4,33 +4,52 @@ import (
 	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/api/schema"
 	model "github.com/go-skynet/LocalAI/pkg/model"
 )
 
-func ComputeChoices(predInput string, n int, config *config.Config, o *options.Option, loader *model.ModelLoader, cb func(string, *[]Choice), tokenCallback func(string) bool) ([]Choice, error) {
-	result := []Choice{}
+func ComputeChoices(
+	req *schema.OpenAIRequest,
+	predInput string,
+	config *config.Config,
+	o *options.Option,
+	loader *model.ModelLoader,
+	cb func(string, *[]schema.Choice),
+	tokenCallback func(string, backend.TokenUsage) bool) ([]schema.Choice, backend.TokenUsage, error) {
+	n := req.N // number of completions to return
+	result := []schema.Choice{}
 
 	if n == 0 {
 		n = 1
 	}
 
-	// get the model function to call for the result
-	predFunc, err := backend.ModelInference(predInput, loader, *config, o, tokenCallback)
-	if err != nil {
-		return result, err
+	images := []string{}
+	for _, m := range req.Messages {
+		images = append(images, m.StringImages...)
 	}
 
+	// get the model function to call for the result
+	predFunc, err := backend.ModelInference(req.Context, predInput, images, loader, *config, o, tokenCallback)
+	if err != nil {
+		return result, backend.TokenUsage{}, err
+	}
+
+	tokenUsage := backend.TokenUsage{}
+
 	for i := 0; i < n; i++ {
 		prediction, err := predFunc()
 		if err != nil {
-			return result, err
+			return result, backend.TokenUsage{}, err
 		}
 
-		prediction = backend.Finetune(*config, predInput, prediction)
-		cb(prediction, &result)
+		tokenUsage.Prompt += prediction.Usage.Prompt
+		tokenUsage.Completion += prediction.Usage.Completion
+
+		finetunedResponse := backend.Finetune(*config, predInput, prediction.Response)
+		cb(finetunedResponse, &result)
 
 		//result = append(result, Choice{Text: prediction})
 
 	}
-	return result, err
+	return result, tokenUsage, err
 }

api/openai/list.go

@@ -1,7 +1,10 @@
 package openai
 
 import (
+	"regexp"
+
 	config "github.com/go-skynet/LocalAI/api/config"
+	"github.com/go-skynet/LocalAI/api/schema"
 	model "github.com/go-skynet/LocalAI/pkg/model"
 	"github.com/gofiber/fiber/v2"
 )
@@ -14,21 +17,50 @@ func ListModelsEndpoint(loader *model.ModelLoader, cm *config.ConfigLoader) func
 		}
 		var mm map[string]interface{} = map[string]interface{}{}
 
-		dataModels := []OpenAIModel{}
-		for _, m := range models {
-			mm[m] = nil
-			dataModels = append(dataModels, OpenAIModel{ID: m, Object: "model"})
+		dataModels := []schema.OpenAIModel{}
+
+		var filterFn func(name string) bool
+		filter := c.Query("filter")
+
+		// If filter is not specified, do not filter the list by model name
+		if filter == "" {
+			filterFn = func(_ string) bool { return true }
+		} else {
+			// If filter _IS_ specified, we compile it to a regex which is used to create the filterFn
+			rxp, err := regexp.Compile(filter)
+			if err != nil {
+				return err
+			}
+			filterFn = func(name string) bool {
+				return rxp.MatchString(name)
+			}
 		}
 
-		for _, k := range cm.ListConfigs() {
-			if _, exists := mm[k]; !exists {
-				dataModels = append(dataModels, OpenAIModel{ID: k, Object: "model"})
+		// By default, exclude any loose files that are already referenced by a configuration file.
+		excludeConfigured := c.QueryBool("excludeConfigured", true)
+
+		// Start with the known configurations
+		for _, c := range cm.GetAllConfigs() {
+			if excludeConfigured {
+				mm[c.Model] = nil
+			}
+
+			if filterFn(c.Name) {
+				dataModels = append(dataModels, schema.OpenAIModel{ID: c.Name, Object: "model"})
+			}
+		}
+
+		// Then iterate through the loose files:
+		for _, m := range models {
+			// And only adds them if they shouldn't be skipped.
+			if _, exists := mm[m]; !exists && filterFn(m) {
+				dataModels = append(dataModels, schema.OpenAIModel{ID: m, Object: "model"})
 			}
 		}
 
 		return c.JSON(struct {
-			Object string        `json:"object"`
-			Data   []OpenAIModel `json:"data"`
+			Object string               `json:"object"`
+			Data   []schema.OpenAIModel `json:"data"`
 		}{
 			Object: "list",
 			Data:   dataModels,

api/openai/request.go

@@ -1,23 +1,33 @@
 package openai
 
 import (
+	"context"
+	"encoding/base64"
 	"encoding/json"
 	"fmt"
+	"io/ioutil"
+	"net/http"
 	"os"
 	"path/filepath"
 	"strings"
 
 	config "github.com/go-skynet/LocalAI/api/config"
+	options "github.com/go-skynet/LocalAI/api/options"
+	"github.com/go-skynet/LocalAI/api/schema"
 	model "github.com/go-skynet/LocalAI/pkg/model"
 	"github.com/gofiber/fiber/v2"
 	"github.com/rs/zerolog/log"
 )
 
-func readInput(c *fiber.Ctx, loader *model.ModelLoader, randomModel bool) (string, *OpenAIRequest, error) {
-	input := new(OpenAIRequest)
+func readInput(c *fiber.Ctx, o *options.Option, randomModel bool) (string, *schema.OpenAIRequest, error) {
+	loader := o.Loader
+	input := new(schema.OpenAIRequest)
+	ctx, cancel := context.WithCancel(o.Context)
+	input.Context = ctx
+	input.Cancel = cancel
 	// Get input data from the request body
 	if err := c.BodyParser(input); err != nil {
-		return "", nil, err
+		return "", nil, fmt.Errorf("failed parsing request body: %w", err)
 	}
 
 	modelFile := input.Model
@@ -54,7 +64,38 @@ func readInput(c *fiber.Ctx, loader *model.ModelLoader, randomModel bool) (strin
 	return modelFile, input, nil
 }
 
-func updateConfig(config *config.Config, input *OpenAIRequest) {
+// this function check if the string is an URL, if it's an URL downloads the image in memory
+// encodes it in base64 and returns the base64 string
+func getBase64Image(s string) (string, error) {
+	if strings.HasPrefix(s, "http") {
+		// download the image
+		resp, err := http.Get(s)
+		if err != nil {
+			return "", err
+		}
+		defer resp.Body.Close()
+
+		// read the image data into memory
+		data, err := ioutil.ReadAll(resp.Body)
+		if err != nil {
+			return "", err
+		}
+
+		// encode the image data in base64
+		encoded := base64.StdEncoding.EncodeToString(data)
+
+		// return the base64 string
+		return encoded, nil
+	}
+
+	// if the string instead is prefixed with "data:image/jpeg;base64,", drop it
+	if strings.HasPrefix(s, "data:image/jpeg;base64,") {
+		return strings.ReplaceAll(s, "data:image/jpeg;base64,", ""), nil
+	}
+	return "", fmt.Errorf("not valid string")
+}
+
+func updateConfig(config *config.Config, input *schema.OpenAIRequest) {
 	if input.Echo {
 		config.Echo = input.Echo
 	}
@@ -65,6 +106,38 @@ func updateConfig(config *config.Config, input *OpenAIRequest) {
 		config.TopP = input.TopP
 	}
 
+	if input.Backend != "" {
+		config.Backend = input.Backend
+	}
+
+	if input.ClipSkip != 0 {
+		config.Diffusers.ClipSkip = input.ClipSkip
+	}
+
+	if input.ModelBaseName != "" {
+		config.AutoGPTQ.ModelBaseName = input.ModelBaseName
+	}
+
+	if input.NegativePromptScale != 0 {
+		config.NegativePromptScale = input.NegativePromptScale
+	}
+
+	if input.UseFastTokenizer {
+		config.UseFastTokenizer = input.UseFastTokenizer
+	}
+
+	if input.NegativePrompt != "" {
+		config.NegativePrompt = input.NegativePrompt
+	}
+
+	if input.RopeFreqBase != 0 {
+		config.RopeFreqBase = input.RopeFreqBase
+	}
+
+	if input.RopeFreqScale != 0 {
+		config.RopeFreqScale = input.RopeFreqScale
+	}
+
 	if input.Grammar != "" {
 		config.Grammar = input.Grammar
 	}
@@ -90,6 +163,35 @@ func updateConfig(config *config.Config, input *OpenAIRequest) {
 		}
 	}
 
+	// Decode each request's message content
+	index := 0
+	for i, m := range input.Messages {
+		switch content := m.Content.(type) {
+		case string:
+			input.Messages[i].StringContent = content
+		case []interface{}:
+			dat, _ := json.Marshal(content)
+			c := []schema.Content{}
+			json.Unmarshal(dat, &c)
+			for _, pp := range c {
+				if pp.Type == "text" {
+					input.Messages[i].StringContent = pp.Text
+				} else if pp.Type == "image_url" {
+					// Detect if pp.ImageURL is an URL, if it is download the image and encode it in base64:
+					base64, err := getBase64Image(pp.ImageURL.URL)
+					if err == nil {
+						input.Messages[i].StringImages = append(input.Messages[i].StringImages, base64) // TODO: make sure that we only return base64 stuff
+						// set a placeholder for each image
+						input.Messages[i].StringContent = fmt.Sprintf("[img-%d]", index) + input.Messages[i].StringContent
+						index++
+					} else {
+						fmt.Print("Failed encoding image", err)
+					}
+				}
+			}
+		}
+	}
+
 	if input.RepeatPenalty != 0 {
 		config.RepeatPenalty = input.RepeatPenalty
 	}
@@ -115,15 +217,15 @@ func updateConfig(config *config.Config, input *OpenAIRequest) {
 	}
 
 	if input.Mirostat != 0 {
-		config.Mirostat = input.Mirostat
+		config.LLMConfig.Mirostat = input.Mirostat
 	}
 
 	if input.MirostatETA != 0 {
-		config.MirostatETA = input.MirostatETA
+		config.LLMConfig.MirostatETA = input.MirostatETA
 	}
 
 	if input.MirostatTAU != 0 {
-		config.MirostatTAU = input.MirostatTAU
+		config.LLMConfig.MirostatTAU = input.MirostatTAU
 	}
 
 	if input.TypicalP != 0 {
@@ -161,7 +263,7 @@ func updateConfig(config *config.Config, input *OpenAIRequest) {
 		n, exists := fnc["name"]
 		if exists {
 			nn, e := n.(string)
-			if !e {
+			if e {
 				name = nn
 			}
 		}
@@ -180,7 +282,7 @@ func updateConfig(config *config.Config, input *OpenAIRequest) {
 	}
 }
 
-func readConfig(modelFile string, input *OpenAIRequest, cm *config.ConfigLoader, loader *model.ModelLoader, debug bool, threads, ctx int, f16 bool) (*config.Config, *OpenAIRequest, error) {
+func readConfig(modelFile string, input *schema.OpenAIRequest, cm *config.ConfigLoader, loader *model.ModelLoader, debug bool, threads, ctx int, f16 bool) (*config.Config, *schema.OpenAIRequest, error) {
 	// Load a config file if present after the model name
 	modelConfig := filepath.Join(loader.ModelPath, modelFile+".yaml")
 

api/openai/transcription.go

@@ -1,7 +1,6 @@
 package openai
 
 import (
-	"context"
 	"fmt"
 	"io"
 	"net/http"
@@ -9,10 +8,9 @@ import (
 	"path"
 	"path/filepath"
 
+	"github.com/go-skynet/LocalAI/api/backend"
 	config "github.com/go-skynet/LocalAI/api/config"
 	"github.com/go-skynet/LocalAI/api/options"
-	"github.com/go-skynet/LocalAI/pkg/grpc/proto"
-	model "github.com/go-skynet/LocalAI/pkg/model"
 
 	"github.com/gofiber/fiber/v2"
 	"github.com/rs/zerolog/log"
@@ -21,7 +19,7 @@ import (
 // https://platform.openai.com/docs/api-reference/audio/create
 func TranscriptEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
-		m, input, err := readInput(c, o.Loader, false)
+		m, input, err := readInput(c, o, false)
 		if err != nil {
 			return fmt.Errorf("failed reading parameters from request:%w", err)
 		}
@@ -61,25 +59,7 @@ func TranscriptEndpoint(cm *config.ConfigLoader, o *options.Option) func(c *fibe
 
 		log.Debug().Msgf("Audio file copied to: %+v", dst)
 
-		whisperModel, err := o.Loader.BackendLoader(
-			model.WithBackendString(model.WhisperBackend),
-			model.WithModelFile(config.Model),
-			model.WithContext(o.Context),
-			model.WithThreads(uint32(config.Threads)),
-			model.WithAssetDir(o.AssetsDestination))
-		if err != nil {
-			return err
-		}
-
-		if whisperModel == nil {
-			return fmt.Errorf("could not load whisper model")
-		}
-
-		tr, err := whisperModel.AudioTranscription(context.Background(), &proto.TranscriptRequest{
-			Dst:      dst,
-			Language: input.Language,
-			Threads:  uint32(config.Threads),
-		})
+		tr, err := backend.ModelTranscription(dst, input.Language, o.Loader, *config, o)
 		if err != nil {
 			return err
 		}

api/options/options.go

@@ -4,7 +4,9 @@ import (
 	"context"
 	"embed"
 	"encoding/json"
+	"time"
 
+	"github.com/go-skynet/LocalAI/metrics"
 	"github.com/go-skynet/LocalAI/pkg/gallery"
 	model "github.com/go-skynet/LocalAI/pkg/model"
 	"github.com/rs/zerolog/log"
@@ -23,11 +25,28 @@ type Option struct {
 	PreloadJSONModels                   string
 	PreloadModelsFromPath               string
 	CORSAllowOrigins                    string
+	ApiKeys                             []string
+	Metrics                             *metrics.Metrics
 
 	Galleries []gallery.Gallery
 
 	BackendAssets     embed.FS
 	AssetsDestination string
+
+	ExternalGRPCBackends map[string]string
+
+	AutoloadGalleries bool
+
+	SingleBackend           bool
+	ParallelBackendRequests bool
+
+	WatchDogIdle bool
+	WatchDogBusy bool
+	WatchDog     bool
+
+	ModelsURL []string
+
+	WatchDogBusyTimeout, WatchDogIdleTimeout time.Duration
 }
 
 type AppOption func(*Option)
@@ -47,12 +66,65 @@ func NewOptions(o ...AppOption) *Option {
 	return opt
 }
 
+func WithModelsURL(urls ...string) AppOption {
+	return func(o *Option) {
+		o.ModelsURL = urls
+	}
+}
+
 func WithCors(b bool) AppOption {
 	return func(o *Option) {
 		o.CORS = b
 	}
 }
 
+var EnableWatchDog = func(o *Option) {
+	o.WatchDog = true
+}
+
+var EnableWatchDogIdleCheck = func(o *Option) {
+	o.WatchDog = true
+	o.WatchDogIdle = true
+}
+
+var EnableWatchDogBusyCheck = func(o *Option) {
+	o.WatchDog = true
+	o.WatchDogBusy = true
+}
+
+func SetWatchDogBusyTimeout(t time.Duration) AppOption {
+	return func(o *Option) {
+		o.WatchDogBusyTimeout = t
+	}
+}
+
+func SetWatchDogIdleTimeout(t time.Duration) AppOption {
+	return func(o *Option) {
+		o.WatchDogIdleTimeout = t
+	}
+}
+
+var EnableSingleBackend = func(o *Option) {
+	o.SingleBackend = true
+}
+
+var EnableParallelBackendRequests = func(o *Option) {
+	o.ParallelBackendRequests = true
+}
+
+var EnableGalleriesAutoload = func(o *Option) {
+	o.AutoloadGalleries = true
+}
+
+func WithExternalBackend(name string, uri string) AppOption {
+	return func(o *Option) {
+		if o.ExternalGRPCBackends == nil {
+			o.ExternalGRPCBackends = make(map[string]string)
+		}
+		o.ExternalGRPCBackends[name] = uri
+	}
+}
+
 func WithCorsAllowOrigins(b string) AppOption {
 	return func(o *Option) {
 		o.CORSAllowOrigins = b
@@ -75,6 +147,7 @@ func WithStringGalleries(galls string) AppOption {
 	return func(o *Option) {
 		if galls == "" {
 			log.Debug().Msgf("no galleries to load")
+			o.Galleries = []gallery.Gallery{}
 			return
 		}
 		var galleries []gallery.Gallery
@@ -167,3 +240,15 @@ func WithImageDir(imageDir string) AppOption {
 		o.ImageDir = imageDir
 	}
 }
+
+func WithApiKeys(apiKeys []string) AppOption {
+	return func(o *Option) {
+		o.ApiKeys = apiKeys
+	}
+}
+
+func WithMetrics(meter *metrics.Metrics) AppOption {
+	return func(o *Option) {
+		o.Metrics = meter
+	}
+}

api/schema/openai.go

@@ -1,6 +1,8 @@
-package openai
+package schema
 
 import (
+	"context"
+
 	config "github.com/go-skynet/LocalAI/api/config"
 
 	"github.com/go-skynet/LocalAI/pkg/grammar"
@@ -46,18 +48,32 @@ type OpenAIResponse struct {
 }
 
 type Choice struct {
-	Index        int      `json:"index,omitempty"`
+	Index        int      `json:"index"`
 	FinishReason string   `json:"finish_reason,omitempty"`
 	Message      *Message `json:"message,omitempty"`
 	Delta        *Message `json:"delta,omitempty"`
 	Text         string   `json:"text,omitempty"`
 }
 
+type Content struct {
+	Type     string     `json:"type" yaml:"type"`
+	Text     string     `json:"text" yaml:"text"`
+	ImageURL ContentURL `json:"image_url" yaml:"image_url"`
+}
+
+type ContentURL struct {
+	URL string `json:"url" yaml:"url"`
+}
+
 type Message struct {
 	// The message role
 	Role string `json:"role,omitempty" yaml:"role"`
 	// The message content
-	Content *string `json:"content" yaml:"content"`
+	Content interface{} `json:"content" yaml:"content"`
+
+	StringContent string   `json:"string_content,omitempty" yaml:"string_content,omitempty"`
+	StringImages  []string `json:"string_images,omitempty" yaml:"string_images,omitempty"`
+
 	// A result of a function call
 	FunctionCall interface{} `json:"function_call,omitempty" yaml:"function_call,omitempty"`
 }
@@ -67,13 +83,22 @@ type OpenAIModel struct {
 	Object string `json:"object"`
 }
 
+type ChatCompletionResponseFormatType string
+
+type ChatCompletionResponseFormat struct {
+	Type ChatCompletionResponseFormatType `json:"type,omitempty"`
+}
+
 type OpenAIRequest struct {
 	config.PredictionOptions
 
+	Context context.Context
+	Cancel  context.CancelFunc
+
 	// whisper
 	File string `json:"file" validate:"required"`
 	//whisper/image
-	ResponseFormat string `json:"response_format"`
+	ResponseFormat ChatCompletionResponseFormat `json:"response_format"`
 	// image
 	Size string `json:"size"`
 	// Prompt is read only by completion/image API calls
@@ -102,4 +127,9 @@ type OpenAIRequest struct {
 	Grammar string `json:"grammar" yaml:"grammar"`
 
 	JSONFunctionGrammarObject *grammar.JSONFunctionStructure `json:"grammar_json_functions" yaml:"grammar_json_functions"`
+
+	Backend string `json:"backend" yaml:"backend"`
+
+	// AutoGPTQ
+	ModelBaseName string `json:"model_base_name" yaml:"model_base_name"`
 }

api/schema/whisper.go

@@ -1,4 +1,4 @@
-package api
+package schema
 
 import "time"
 

backend/backend.proto

@@ -16,6 +16,8 @@ service Backend {
   rpc GenerateImage(GenerateImageRequest) returns (Result) {}
   rpc AudioTranscription(TranscriptRequest) returns (TranscriptResult) {}
   rpc TTS(TTSRequest) returns (Result) {}
+  rpc TokenizeString(PredictOptions) returns (TokenizationResponse) {}
+  rpc Status(HealthMessage) returns (StatusResponse) {}
 }
 
 message HealthMessage {}
@@ -57,11 +59,17 @@ message PredictOptions {
   bool Debug = 34;
   repeated int32 EmbeddingTokens = 35;
   string Embeddings = 36;
+  float RopeFreqBase = 37;
+  float RopeFreqScale = 38;
+  float NegativePromptScale = 39;
+  string NegativePrompt = 40;
+  int32 NDraft = 41;
+  repeated string Images = 42;
 }
 
 // The response message containing the result
 message Reply {
-  string message = 1;
+  bytes message = 1;
 }
 
 message ModelOptions {
@@ -81,6 +89,51 @@ message ModelOptions {
   string TensorSplit = 14;
   int32 Threads = 15;
   string LibrarySearchPath = 16;
+  float RopeFreqBase = 17;
+  float RopeFreqScale = 18;
+  float RMSNormEps = 19;
+  int32 NGQA = 20;
+  string ModelFile = 21;
+
+  // AutoGPTQ
+  string Device = 22;
+  bool UseTriton = 23;
+  string ModelBaseName = 24;
+  bool UseFastTokenizer = 25;
+
+  // Diffusers
+  string PipelineType = 26;
+  string SchedulerType = 27;
+  bool CUDA = 28;
+  float CFGScale = 29;
+  bool IMG2IMG = 30;
+  string CLIPModel = 31;
+  string CLIPSubfolder = 32;
+  int32 CLIPSkip = 33;
+  string ControlNet = 48;
+
+  string Tokenizer = 34;
+
+  // LLM (llama.cpp)
+  string LoraBase = 35;
+  string LoraAdapter = 36;
+  float LoraScale = 42;
+
+  bool NoMulMatQ = 37;
+  string DraftModel = 39;
+  
+  string AudioPath = 38;
+
+  // vllm
+  string Quantization = 40;
+
+  string MMProj = 41;
+
+  string RopeScaling = 43;
+  float YarnExtFactor = 44;
+  float YarnAttnFactor = 45;
+  float YarnBetaFast = 46;
+  float YarnBetaSlow = 47;
 }
 
 message Result {
@@ -120,6 +173,11 @@ message GenerateImageRequest {
   string positive_prompt = 6;
   string negative_prompt = 7;
   string dst = 8;
+  string src = 9;
+
+  // Diffusers
+  string EnableParameters = 10;
+  int32 CLIPSkip = 11;
 }
 
 message TTSRequest {
@@ -127,3 +185,24 @@ message TTSRequest {
   string model = 2;
   string dst = 3;
 }
+
+message TokenizationResponse {
+  int32 length = 1;
+  repeated int32 tokens = 2;
+}
+
+message MemoryUsageData {
+  uint64 total = 1;
+  map<string, uint64> breakdown = 2;
+}
+
+message StatusResponse {
+  enum State {
+    UNINITIALIZED = 0;
+    BUSY = 1;
+    READY = 2;
+    ERROR = -1;
+  }
+  State state = 1;
+  MemoryUsageData memory = 2;
+}

backend/cpp/grpc/.gitignore

@@ -0,0 +1,3 @@
+installed_packages/
+grpc_build/
+grpc_repo/

backend/cpp/grpc/Makefile

@@ -0,0 +1,66 @@
+# Basic platform detection
+HOST_SYSTEM = $(shell uname | cut -f 1 -d_)
+SYSTEM ?= $(HOST_SYSTEM)
+
+TAG_LIB_GRPC?=v1.59.0
+GIT_REPO_LIB_GRPC?=https://github.com/grpc/grpc.git
+GIT_CLONE_DEPTH?=1
+NUM_BUILD_THREADS?=$(shell nproc --ignore=1)
+
+INSTALLED_PACKAGES=installed_packages
+GRPC_REPO=grpc_repo
+GRPC_BUILD=grpc_build
+
+export CMAKE_ARGS?=
+CMAKE_ARGS+=-DCMAKE_BUILD_TYPE=Release
+CMAKE_ARGS+=-DgRPC_INSTALL=ON
+CMAKE_ARGS+=-DEXECUTABLE_OUTPUT_PATH=../$(INSTALLED_PACKAGES)/grpc/bin
+CMAKE_ARGS+=-DLIBRARY_OUTPUT_PATH=../$(INSTALLED_PACKAGES)/grpc/lib
+CMAKE_ARGS+=-DgRPC_BUILD_TESTS=OFF
+CMAKE_ARGS+=-DgRPC_BUILD_CSHARP_EXT=OFF
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_CPP_PLUGIN=ON
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_CSHARP_PLUGIN=OFF
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_NODE_PLUGIN=OFF
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_OBJECTIVE_C_PLUGIN=OFF
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_PHP_PLUGIN=OFF
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_PYTHON_PLUGIN=ON
+CMAKE_ARGS+=-DgRPC_BUILD_GRPC_RUBY_PLUGIN=OFF
+CMAKE_ARGS+=-Dprotobuf_WITH_ZLIB=ON
+CMAKE_ARGS+=-DRE2_BUILD_TESTING=OFF
+CMAKE_ARGS+=-DCMAKE_INSTALL_PREFIX=../$(INSTALLED_PACKAGES)
+
+# windows need to set OPENSSL_NO_ASM. Results in slower crypto performance but doesn't build otherwise.
+# May be resolvable, but for now its set. More info: https://stackoverflow.com/a/75240504/480673
+ifeq ($(SYSTEM),MSYS)
+CMAKE_ARGS+=-DOPENSSL_NO_ASM=ON
+endif
+ifeq ($(SYSTEM),MINGW64)
+CMAKE_ARGS+=-DOPENSSL_NO_ASM=ON
+endif
+ifeq ($(SYSTEM),MINGW32)
+CMAKE_ARGS+=-DOPENSSL_NO_ASM=ON
+endif
+ifeq ($(SYSTEM),CYGWIN)
+CMAKE_ARGS+=-DOPENSSL_NO_ASM=ON
+endif
+
+$(INSTALLED_PACKAGES): grpc_build
+
+$(GRPC_REPO):
+	git clone --depth $(GIT_CLONE_DEPTH) -b $(TAG_LIB_GRPC) $(GIT_REPO_LIB_GRPC) $(GRPC_REPO)/grpc
+	cd $(GRPC_REPO)/grpc && git submodule update --init --recursive --depth $(GIT_CLONE_DEPTH)
+
+$(GRPC_BUILD): $(GRPC_REPO)
+	mkdir -p $(GRPC_BUILD)
+	cd $(GRPC_BUILD) && cmake $(CMAKE_ARGS) ../$(GRPC_REPO)/grpc && cmake --build . -- -j ${NUM_BUILD_THREADS} && cmake --build . --target install -- -j ${NUM_BUILD_THREADS}
+
+build: $(INSTALLED_PACKAGES)
+
+rebuild:
+	rm -rf grpc_build
+	$(MAKE) grpc_build
+
+clean:
+	rm -rf grpc_build
+	rm -rf grpc_repo
+	rm -rf installed_packages

backend/cpp/llama/CMakeLists.txt

@@ -0,0 +1,82 @@
+
+## XXX: In some versions of CMake clip wasn't being built before llama.
+## This is an hack for now, but it should be fixed in the future.
+set(TARGET myclip)
+add_library(${TARGET} clip.cpp clip.h)
+install(TARGETS ${TARGET} LIBRARY)
+target_link_libraries(${TARGET} PRIVATE common ggml ${CMAKE_THREAD_LIBS_INIT})
+target_compile_features(${TARGET} PRIVATE cxx_std_11)
+if (NOT MSVC)
+    target_compile_options(${TARGET} PRIVATE -Wno-cast-qual) # stb_image.h
+endif()
+
+set(TARGET grpc-server)
+# END CLIP hack
+set(CMAKE_CXX_STANDARD 17)
+cmake_minimum_required(VERSION 3.15)
+set(TARGET grpc-server)
+set(_PROTOBUF_LIBPROTOBUF libprotobuf)
+set(_REFLECTION grpc++_reflection)
+
+if (${CMAKE_SYSTEM_NAME} MATCHES "Darwin")
+    # Set correct Homebrew install folder for Apple Silicon and Intel Macs
+    if (CMAKE_HOST_SYSTEM_PROCESSOR MATCHES "arm64")
+        set(HOMEBREW_DEFAULT_PREFIX "/opt/homebrew")
+    else()
+        set(HOMEBREW_DEFAULT_PREFIX "/usr/local")
+    endif()
+
+    link_directories("${HOMEBREW_DEFAULT_PREFIX}/lib")
+    include_directories("${HOMEBREW_DEFAULT_PREFIX}/include")
+endif()
+
+find_package(absl CONFIG REQUIRED)
+find_package(Protobuf CONFIG REQUIRED)
+find_package(gRPC CONFIG REQUIRED)
+
+find_program(_PROTOBUF_PROTOC protoc)
+set(_GRPC_GRPCPP grpc++)
+find_program(_GRPC_CPP_PLUGIN_EXECUTABLE grpc_cpp_plugin)
+
+include_directories(${CMAKE_CURRENT_BINARY_DIR})
+include_directories(${Protobuf_INCLUDE_DIRS})
+
+message(STATUS "Using protobuf version ${Protobuf_VERSION} | Protobuf_INCLUDE_DIRS: ${Protobuf_INCLUDE_DIRS} | CMAKE_CURRENT_BINARY_DIR: ${CMAKE_CURRENT_BINARY_DIR}")
+
+# Proto file
+get_filename_component(hw_proto "../../../../../../backend/backend.proto" ABSOLUTE)
+get_filename_component(hw_proto_path "${hw_proto}" PATH)
+
+# Generated sources
+set(hw_proto_srcs "${CMAKE_CURRENT_BINARY_DIR}/backend.pb.cc")
+set(hw_proto_hdrs "${CMAKE_CURRENT_BINARY_DIR}/backend.pb.h")
+set(hw_grpc_srcs "${CMAKE_CURRENT_BINARY_DIR}/backend.grpc.pb.cc")
+set(hw_grpc_hdrs "${CMAKE_CURRENT_BINARY_DIR}/backend.grpc.pb.h")
+
+add_custom_command(
+      OUTPUT "${hw_proto_srcs}" "${hw_proto_hdrs}" "${hw_grpc_srcs}" "${hw_grpc_hdrs}"
+      COMMAND ${_PROTOBUF_PROTOC}
+      ARGS --grpc_out "${CMAKE_CURRENT_BINARY_DIR}"
+        --cpp_out "${CMAKE_CURRENT_BINARY_DIR}"
+        -I "${hw_proto_path}"
+        --plugin=protoc-gen-grpc="${_GRPC_CPP_PLUGIN_EXECUTABLE}"
+        "${hw_proto}"
+      DEPENDS "${hw_proto}")
+
+# hw_grpc_proto
+add_library(hw_grpc_proto
+  ${hw_grpc_srcs}
+  ${hw_grpc_hdrs}
+  ${hw_proto_srcs}
+  ${hw_proto_hdrs} )
+
+add_executable(${TARGET} grpc-server.cpp json.hpp )
+target_link_libraries(${TARGET} PRIVATE common llama myclip ${CMAKE_THREAD_LIBS_INIT} absl::flags hw_grpc_proto
+  absl::flags_parse
+  gRPC::${_REFLECTION}
+  gRPC::${_GRPC_GRPCPP}
+  protobuf::${_PROTOBUF_LIBPROTOBUF})
+target_compile_features(${TARGET} PRIVATE cxx_std_11)
+if(TARGET BUILD_INFO)
+  add_dependencies(${TARGET} BUILD_INFO)
+endif()

backend/cpp/llama/Makefile

@@ -0,0 +1,53 @@
+
+LLAMA_VERSION?=
+
+CMAKE_ARGS?=
+BUILD_TYPE?=
+
+# If build type is cublas, then we set -DLLAMA_CUBLAS=ON to CMAKE_ARGS automatically
+ifeq ($(BUILD_TYPE),cublas)
+	CMAKE_ARGS+=-DLLAMA_CUBLAS=ON
+# If build type is openblas then we set -DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS
+# to CMAKE_ARGS automatically
+else ifeq ($(BUILD_TYPE),openblas)
+	CMAKE_ARGS+=-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS
+# If build type is clblast (openCL) we set -DLLAMA_CLBLAST=ON -DCLBlast_DIR=/some/path
+else ifeq ($(BUILD_TYPE),clblast)
+	CMAKE_ARGS+=-DLLAMA_CLBLAST=ON -DCLBlast_DIR=/some/path
+# If it's hipblas we do have also to set CC=/opt/rocm/llvm/bin/clang CXX=/opt/rocm/llvm/bin/clang++ 
+else ifeq ($(BUILD_TYPE),hipblas)
+	CMAKE_ARGS+=-DLLAMA_HIPBLAS=ON
+endif
+
+llama.cpp:
+	git clone --recurse-submodules https://github.com/ggerganov/llama.cpp llama.cpp
+	if [ -z "$(LLAMA_VERSION)" ]; then \
+		exit 1; \
+	fi
+	cd llama.cpp && git checkout -b build $(LLAMA_VERSION) && git submodule update --init --recursive --depth 1
+
+llama.cpp/examples/grpc-server:
+	mkdir -p llama.cpp/examples/grpc-server
+	cp -r $(abspath ./)/CMakeLists.txt llama.cpp/examples/grpc-server/
+	cp -r $(abspath ./)/grpc-server.cpp llama.cpp/examples/grpc-server/
+	cp -rfv $(abspath ./)/json.hpp llama.cpp/examples/grpc-server/
+	echo "add_subdirectory(grpc-server)" >> llama.cpp/examples/CMakeLists.txt
+## XXX: In some versions of CMake clip wasn't being built before llama.
+## This is an hack for now, but it should be fixed in the future.
+	cp -rfv llama.cpp/examples/llava/clip.h llama.cpp/examples/grpc-server/clip.h
+	cp -rfv llama.cpp/examples/llava/clip.cpp llama.cpp/examples/grpc-server/clip.cpp
+
+rebuild:
+	cp -rfv $(abspath ./)/CMakeLists.txt llama.cpp/examples/grpc-server/
+	cp -rfv $(abspath ./)/grpc-server.cpp llama.cpp/examples/grpc-server/
+	cp -rfv $(abspath ./)/json.hpp llama.cpp/examples/grpc-server/
+	rm -rf grpc-server
+	$(MAKE) grpc-server
+
+clean:
+	rm -rf llama.cpp
+	rm -rf grpc-server
+
+grpc-server: llama.cpp llama.cpp/examples/grpc-server
+	cd llama.cpp && mkdir -p build && cd build && cmake .. $(CMAKE_ARGS) && cmake --build . --config Release
+	cp llama.cpp/build/bin/grpc-server .

backend/go/image/stablediffusion/main.go

@@ -5,8 +5,6 @@ package main
 import (
 	"flag"
 
-	rwkv "github.com/go-skynet/LocalAI/pkg/grpc/llm/rwkv"
-
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )
 
@@ -17,7 +15,7 @@ var (
 func main() {
 	flag.Parse()
 
-	if err := grpc.StartServer(*addr, &rwkv.LLM{}); err != nil {
+	if err := grpc.StartServer(*addr, &Image{}); err != nil {
 		panic(err)
 	}
 }

backend/go/image/stablediffusion/stablediffusion.go

@@ -1,4 +1,4 @@
-package image
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
@@ -8,20 +8,20 @@ import (
 	"github.com/go-skynet/LocalAI/pkg/stablediffusion"
 )
 
-type StableDiffusion struct {
-	base.Base
+type Image struct {
+	base.SingleThread
 	stablediffusion *stablediffusion.StableDiffusion
 }
 
-func (sd *StableDiffusion) Load(opts *pb.ModelOptions) error {
+func (image *Image) Load(opts *pb.ModelOptions) error {
 	var err error
 	// Note: the Model here is a path to a directory containing the model files
-	sd.stablediffusion, err = stablediffusion.New(opts.Model)
+	image.stablediffusion, err = stablediffusion.New(opts.ModelFile)
 	return err
 }
 
-func (sd *StableDiffusion) GenerateImage(opts *pb.GenerateImageRequest) error {
-	return sd.stablediffusion.GenerateImage(
+func (image *Image) GenerateImage(opts *pb.GenerateImageRequest) error {
+	return image.stablediffusion.GenerateImage(
 		int(opts.Height),
 		int(opts.Width),
 		int(opts.Mode),

backend/go/image/tinydream/main.go

@@ -6,7 +6,6 @@ import (
 	"flag"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
-	bert "github.com/go-skynet/LocalAI/pkg/grpc/llm/bert"
 )
 
 var (
@@ -16,7 +15,7 @@ var (
 func main() {
 	flag.Parse()
 
-	if err := grpc.StartServer(*addr, &bert.Embeddings{}); err != nil {
+	if err := grpc.StartServer(*addr, &Image{}); err != nil {
 		panic(err)
 	}
 }

backend/go/image/tinydream/tinydream.go

@@ -0,0 +1,32 @@
+package main
+
+// This is a wrapper to statisfy the GRPC service interface
+// It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
+import (
+	"github.com/go-skynet/LocalAI/pkg/grpc/base"
+	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
+	"github.com/go-skynet/LocalAI/pkg/tinydream"
+)
+
+type Image struct {
+	base.SingleThread
+	tinydream *tinydream.TinyDream
+}
+
+func (image *Image) Load(opts *pb.ModelOptions) error {
+	var err error
+	// Note: the Model here is a path to a directory containing the model files
+	image.tinydream, err = tinydream.New(opts.ModelFile)
+	return err
+}
+
+func (image *Image) GenerateImage(opts *pb.GenerateImageRequest) error {
+	return image.tinydream.GenerateImage(
+		int(opts.Height),
+		int(opts.Width),
+		int(opts.Step),
+		int(opts.Seed),
+		opts.PositivePrompt,
+		opts.NegativePrompt,
+		opts.Dst)
+}

backend/go/llm/bert/bert.go

@@ -1,4 +1,4 @@
-package bert
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
@@ -10,17 +10,18 @@ import (
 )
 
 type Embeddings struct {
-	base.Base
+	base.SingleThread
 	bert *bert.Bert
 }
 
 func (llm *Embeddings) Load(opts *pb.ModelOptions) error {
-	model, err := bert.New(opts.Model)
+	model, err := bert.New(opts.ModelFile)
 	llm.bert = model
 	return err
 }
 
 func (llm *Embeddings) Embeddings(opts *pb.PredictOptions) ([]float32, error) {
+
 	if len(opts.EmbeddingTokens) > 0 {
 		tokens := []int{}
 		for _, t := range opts.EmbeddingTokens {

backend/go/llm/bert/main.go

@@ -5,8 +5,6 @@ package main
 import (
 	"flag"
 
-	bloomz "github.com/go-skynet/LocalAI/pkg/grpc/llm/bloomz"
-
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )
 
@@ -17,7 +15,7 @@ var (
 func main() {
 	flag.Parse()
 
-	if err := grpc.StartServer(*addr, &bloomz.LLM{}); err != nil {
+	if err := grpc.StartServer(*addr, &Embeddings{}); err != nil {
 		panic(err)
 	}
 }

backend/go/llm/dolly/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/falcon-ggml/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/gpt2/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/gpt4all/gpt4all.go

@@ -1,4 +1,4 @@
-package gpt4all
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
@@ -11,13 +11,13 @@ import (
 )
 
 type LLM struct {
-	base.Base
+	base.SingleThread
 
 	gpt4all *gpt4all.Model
 }
 
 func (llm *LLM) Load(opts *pb.ModelOptions) error {
-	model, err := gpt4all.New(opts.Model,
+	model, err := gpt4all.New(opts.ModelFile,
 		gpt4all.SetThreads(int(opts.Threads)),
 		gpt4all.SetLibrarySearchPath(opts.LibrarySearchPath))
 	llm.gpt4all = model

backend/go/llm/gpt4all/main.go

@@ -0,0 +1,21 @@
+package main
+
+// Note: this is started internally by LocalAI and a server is allocated for each model
+
+import (
+	"flag"
+
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+func main() {
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &LLM{}); err != nil {
+		panic(err)
+	}
+}

backend/go/llm/gptj/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/gptneox/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/langchain/langchain.go

@@ -1,4 +1,4 @@
-package langchain
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)

backend/go/llm/langchain/main.go

@@ -0,0 +1,21 @@
+package main
+
+// Note: this is started internally by LocalAI and a server is allocated for each model
+
+import (
+	"flag"
+
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+func main() {
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &LLM{}); err != nil {
+		panic(err)
+	}
+}

backend/go/llm/llama-ggml/llama.go

@@ -1,4 +1,4 @@
-package llama
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
@@ -11,13 +11,34 @@ import (
 )
 
 type LLM struct {
-	base.Base
+	base.SingleThread
 
 	llama *llama.LLama
 }
 
 func (llm *LLM) Load(opts *pb.ModelOptions) error {
-	llamaOpts := []llama.ModelOption{}
+	ropeFreqBase := float32(10000)
+	ropeFreqScale := float32(1)
+
+	if opts.RopeFreqBase != 0 {
+		ropeFreqBase = opts.RopeFreqBase
+	}
+	if opts.RopeFreqScale != 0 {
+		ropeFreqScale = opts.RopeFreqScale
+	}
+
+	llamaOpts := []llama.ModelOption{
+		llama.WithRopeFreqBase(ropeFreqBase),
+		llama.WithRopeFreqScale(ropeFreqScale),
+	}
+
+	if opts.NGQA != 0 {
+		llamaOpts = append(llamaOpts, llama.WithGQA(int(opts.NGQA)))
+	}
+
+	if opts.RMSNormEps != 0 {
+		llamaOpts = append(llamaOpts, llama.WithRMSNormEPS(opts.RMSNormEps))
+	}
 
 	if opts.ContextSize != 0 {
 		llamaOpts = append(llamaOpts, llama.SetContext(int(opts.ContextSize)))
@@ -49,18 +70,33 @@ func (llm *LLM) Load(opts *pb.ModelOptions) error {
 		llamaOpts = append(llamaOpts, llama.EnabelLowVRAM)
 	}
 
-	model, err := llama.New(opts.Model, llamaOpts...)
+	model, err := llama.New(opts.ModelFile, llamaOpts...)
 	llm.llama = model
+
 	return err
 }
 
 func buildPredictOptions(opts *pb.PredictOptions) []llama.PredictOption {
+	ropeFreqBase := float32(10000)
+	ropeFreqScale := float32(1)
+
+	if opts.RopeFreqBase != 0 {
+		ropeFreqBase = opts.RopeFreqBase
+	}
+	if opts.RopeFreqScale != 0 {
+		ropeFreqScale = opts.RopeFreqScale
+	}
 	predictOptions := []llama.PredictOption{
-		llama.SetTemperature(float64(opts.Temperature)),
-		llama.SetTopP(float64(opts.TopP)),
+		llama.SetTemperature(opts.Temperature),
+		llama.SetTopP(opts.TopP),
 		llama.SetTopK(int(opts.TopK)),
 		llama.SetTokens(int(opts.Tokens)),
 		llama.SetThreads(int(opts.Threads)),
+		llama.WithGrammar(opts.Grammar),
+		llama.SetRopeFreqBase(ropeFreqBase),
+		llama.SetRopeFreqScale(ropeFreqScale),
+		llama.SetNegativePromptScale(opts.NegativePromptScale),
+		llama.SetNegativePrompt(opts.NegativePrompt),
 	}
 
 	if opts.PromptCacheAll {
@@ -71,8 +107,6 @@ func buildPredictOptions(opts *pb.PredictOptions) []llama.PredictOption {
 		predictOptions = append(predictOptions, llama.EnablePromptCacheRO)
 	}
 
-	predictOptions = append(predictOptions, llama.WithGrammar(opts.Grammar))
-
 	// Expected absolute path
 	if opts.PromptCachePath != "" {
 		predictOptions = append(predictOptions, llama.SetPathPromptCache(opts.PromptCachePath))
@@ -83,11 +117,11 @@ func buildPredictOptions(opts *pb.PredictOptions) []llama.PredictOption {
 	}
 
 	if opts.MirostatETA != 0 {
-		predictOptions = append(predictOptions, llama.SetMirostatETA(float64(opts.MirostatETA)))
+		predictOptions = append(predictOptions, llama.SetMirostatETA(opts.MirostatETA))
 	}
 
 	if opts.MirostatTAU != 0 {
-		predictOptions = append(predictOptions, llama.SetMirostatTAU(float64(opts.MirostatTAU)))
+		predictOptions = append(predictOptions, llama.SetMirostatTAU(opts.MirostatTAU))
 	}
 
 	if opts.Debug {
@@ -97,7 +131,7 @@ func buildPredictOptions(opts *pb.PredictOptions) []llama.PredictOption {
 	predictOptions = append(predictOptions, llama.SetStopWords(opts.StopPrompts...))
 
 	if opts.PresencePenalty != 0 {
-		predictOptions = append(predictOptions, llama.SetPenalty(float64(opts.PresencePenalty)))
+		predictOptions = append(predictOptions, llama.SetPenalty(opts.PresencePenalty))
 	}
 
 	if opts.NKeep != 0 {
@@ -122,13 +156,13 @@ func buildPredictOptions(opts *pb.PredictOptions) []llama.PredictOption {
 
 	//predictOptions = append(predictOptions, llama.SetLogitBias(c.Seed))
 
-	predictOptions = append(predictOptions, llama.SetFrequencyPenalty(float64(opts.FrequencyPenalty)))
+	predictOptions = append(predictOptions, llama.SetFrequencyPenalty(opts.FrequencyPenalty))
 	predictOptions = append(predictOptions, llama.SetMlock(opts.MLock))
 	predictOptions = append(predictOptions, llama.SetMemoryMap(opts.MMap))
 	predictOptions = append(predictOptions, llama.SetPredictionMainGPU(opts.MainGPU))
 	predictOptions = append(predictOptions, llama.SetPredictionTensorSplit(opts.TensorSplit))
-	predictOptions = append(predictOptions, llama.SetTailFreeSamplingZ(float64(opts.TailFreeSamplingZ)))
-	predictOptions = append(predictOptions, llama.SetTypicalP(float64(opts.TypicalP)))
+	predictOptions = append(predictOptions, llama.SetTailFreeSamplingZ(opts.TailFreeSamplingZ))
+	predictOptions = append(predictOptions, llama.SetTypicalP(opts.TypicalP))
 	return predictOptions
 }
 

backend/go/llm/llama-ggml/main.go

@@ -0,0 +1,19 @@
+package main
+
+import (
+	"flag"
+
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+func main() {
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &LLM{}); err != nil {
+		panic(err)
+	}
+}

backend/go/llm/llama/llama.go

@@ -0,0 +1,257 @@
+package main
+
+// This is a wrapper to statisfy the GRPC service interface
+// It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
+import (
+	"fmt"
+	"path/filepath"
+
+	"github.com/go-skynet/LocalAI/pkg/grpc/base"
+	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
+	"github.com/go-skynet/go-llama.cpp"
+)
+
+type LLM struct {
+	base.SingleThread
+
+	llama      *llama.LLama
+	draftModel *llama.LLama
+}
+
+func (llm *LLM) Load(opts *pb.ModelOptions) error {
+	ropeFreqBase := float32(10000)
+	ropeFreqScale := float32(1)
+
+	if opts.RopeFreqBase != 0 {
+		ropeFreqBase = opts.RopeFreqBase
+	}
+	if opts.RopeFreqScale != 0 {
+		ropeFreqScale = opts.RopeFreqScale
+	}
+
+	llamaOpts := []llama.ModelOption{
+		llama.WithRopeFreqBase(ropeFreqBase),
+		llama.WithRopeFreqScale(ropeFreqScale),
+	}
+
+	if opts.NoMulMatQ {
+		llamaOpts = append(llamaOpts, llama.SetMulMatQ(false))
+	}
+
+	// Get base path of opts.ModelFile and use the same for lora (assume the same path)
+	basePath := filepath.Dir(opts.ModelFile)
+
+	if opts.LoraAdapter != "" {
+		llamaOpts = append(llamaOpts, llama.SetLoraAdapter(filepath.Join(basePath, opts.LoraAdapter)))
+	}
+
+	if opts.LoraBase != "" {
+		llamaOpts = append(llamaOpts, llama.SetLoraBase(filepath.Join(basePath, opts.LoraBase)))
+	}
+
+	if opts.ContextSize != 0 {
+		llamaOpts = append(llamaOpts, llama.SetContext(int(opts.ContextSize)))
+	}
+	if opts.F16Memory {
+		llamaOpts = append(llamaOpts, llama.EnableF16Memory)
+	}
+	if opts.Embeddings {
+		llamaOpts = append(llamaOpts, llama.EnableEmbeddings)
+	}
+	if opts.NGPULayers != 0 {
+		llamaOpts = append(llamaOpts, llama.SetGPULayers(int(opts.NGPULayers)))
+	}
+
+	llamaOpts = append(llamaOpts, llama.SetMMap(opts.MMap))
+	llamaOpts = append(llamaOpts, llama.SetMainGPU(opts.MainGPU))
+	llamaOpts = append(llamaOpts, llama.SetTensorSplit(opts.TensorSplit))
+	if opts.NBatch != 0 {
+		llamaOpts = append(llamaOpts, llama.SetNBatch(int(opts.NBatch)))
+	} else {
+		llamaOpts = append(llamaOpts, llama.SetNBatch(512))
+	}
+
+	if opts.NUMA {
+		llamaOpts = append(llamaOpts, llama.EnableNUMA)
+	}
+
+	if opts.LowVRAM {
+		llamaOpts = append(llamaOpts, llama.EnabelLowVRAM)
+	}
+
+	if opts.DraftModel != "" {
+		// https://github.com/ggerganov/llama.cpp/blob/71ca2fad7d6c0ef95ef9944fb3a1a843e481f314/examples/speculative/speculative.cpp#L40
+		llamaOpts = append(llamaOpts, llama.SetPerplexity(true))
+	}
+
+	model, err := llama.New(opts.ModelFile, llamaOpts...)
+
+	if opts.DraftModel != "" {
+		// opts.DraftModel is relative to opts.ModelFile, so we need to get the basepath of opts.ModelFile
+		if !filepath.IsAbs(opts.DraftModel) {
+			dir := filepath.Dir(opts.ModelFile)
+			opts.DraftModel = filepath.Join(dir, opts.DraftModel)
+		}
+
+		draftModel, err := llama.New(opts.DraftModel, llamaOpts...)
+		if err != nil {
+			return err
+		}
+		llm.draftModel = draftModel
+	}
+
+	llm.llama = model
+
+	return err
+}
+
+func buildPredictOptions(opts *pb.PredictOptions) []llama.PredictOption {
+	ropeFreqBase := float32(10000)
+	ropeFreqScale := float32(1)
+
+	if opts.RopeFreqBase != 0 {
+		ropeFreqBase = opts.RopeFreqBase
+	}
+	if opts.RopeFreqScale != 0 {
+		ropeFreqScale = opts.RopeFreqScale
+	}
+	predictOptions := []llama.PredictOption{
+		llama.SetTemperature(opts.Temperature),
+		llama.SetTopP(opts.TopP),
+		llama.SetTopK(int(opts.TopK)),
+		llama.SetTokens(int(opts.Tokens)),
+		llama.SetThreads(int(opts.Threads)),
+		llama.WithGrammar(opts.Grammar),
+		llama.SetRopeFreqBase(ropeFreqBase),
+		llama.SetRopeFreqScale(ropeFreqScale),
+		llama.SetNegativePromptScale(opts.NegativePromptScale),
+		llama.SetNegativePrompt(opts.NegativePrompt),
+	}
+
+	if opts.PromptCacheAll {
+		predictOptions = append(predictOptions, llama.EnablePromptCacheAll)
+	}
+
+	if opts.PromptCacheRO {
+		predictOptions = append(predictOptions, llama.EnablePromptCacheRO)
+	}
+
+	// Expected absolute path
+	if opts.PromptCachePath != "" {
+		predictOptions = append(predictOptions, llama.SetPathPromptCache(opts.PromptCachePath))
+	}
+
+	if opts.Mirostat != 0 {
+		predictOptions = append(predictOptions, llama.SetMirostat(int(opts.Mirostat)))
+	}
+
+	if opts.MirostatETA != 0 {
+		predictOptions = append(predictOptions, llama.SetMirostatETA(opts.MirostatETA))
+	}
+
+	if opts.MirostatTAU != 0 {
+		predictOptions = append(predictOptions, llama.SetMirostatTAU(opts.MirostatTAU))
+	}
+
+	if opts.Debug {
+		predictOptions = append(predictOptions, llama.Debug)
+	}
+
+	predictOptions = append(predictOptions, llama.SetStopWords(opts.StopPrompts...))
+
+	if opts.PresencePenalty != 0 {
+		predictOptions = append(predictOptions, llama.SetPenalty(opts.PresencePenalty))
+	}
+
+	if opts.NKeep != 0 {
+		predictOptions = append(predictOptions, llama.SetNKeep(int(opts.NKeep)))
+	}
+
+	if opts.Batch != 0 {
+		predictOptions = append(predictOptions, llama.SetBatch(int(opts.Batch)))
+	}
+
+	if opts.F16KV {
+		predictOptions = append(predictOptions, llama.EnableF16KV)
+	}
+
+	if opts.IgnoreEOS {
+		predictOptions = append(predictOptions, llama.IgnoreEOS)
+	}
+
+	if opts.Seed != 0 {
+		predictOptions = append(predictOptions, llama.SetSeed(int(opts.Seed)))
+	}
+
+	if opts.NDraft != 0 {
+		predictOptions = append(predictOptions, llama.SetNDraft(int(opts.NDraft)))
+	}
+	//predictOptions = append(predictOptions, llama.SetLogitBias(c.Seed))
+
+	predictOptions = append(predictOptions, llama.SetFrequencyPenalty(opts.FrequencyPenalty))
+	predictOptions = append(predictOptions, llama.SetMlock(opts.MLock))
+	predictOptions = append(predictOptions, llama.SetMemoryMap(opts.MMap))
+	predictOptions = append(predictOptions, llama.SetPredictionMainGPU(opts.MainGPU))
+	predictOptions = append(predictOptions, llama.SetPredictionTensorSplit(opts.TensorSplit))
+	predictOptions = append(predictOptions, llama.SetTailFreeSamplingZ(opts.TailFreeSamplingZ))
+	predictOptions = append(predictOptions, llama.SetTypicalP(opts.TypicalP))
+	return predictOptions
+}
+
+func (llm *LLM) Predict(opts *pb.PredictOptions) (string, error) {
+	if llm.draftModel != nil {
+		return llm.llama.SpeculativeSampling(llm.draftModel, opts.Prompt, buildPredictOptions(opts)...)
+	}
+	return llm.llama.Predict(opts.Prompt, buildPredictOptions(opts)...)
+}
+
+func (llm *LLM) PredictStream(opts *pb.PredictOptions, results chan string) error {
+	predictOptions := buildPredictOptions(opts)
+
+	predictOptions = append(predictOptions, llama.SetTokenCallback(func(token string) bool {
+		results <- token
+		return true
+	}))
+
+	go func() {
+		var err error
+		if llm.draftModel != nil {
+			_, err = llm.llama.SpeculativeSampling(llm.draftModel, opts.Prompt, buildPredictOptions(opts)...)
+		} else {
+			_, err = llm.llama.Predict(opts.Prompt, predictOptions...)
+		}
+
+		if err != nil {
+			fmt.Println("err: ", err)
+		}
+		close(results)
+	}()
+
+	return nil
+}
+
+func (llm *LLM) Embeddings(opts *pb.PredictOptions) ([]float32, error) {
+	predictOptions := buildPredictOptions(opts)
+
+	if len(opts.EmbeddingTokens) > 0 {
+		tokens := []int{}
+		for _, t := range opts.EmbeddingTokens {
+			tokens = append(tokens, int(t))
+		}
+		return llm.llama.TokenEmbeddings(tokens, predictOptions...)
+	}
+
+	return llm.llama.Embeddings(opts.Embeddings, predictOptions...)
+}
+
+func (llm *LLM) TokenizeString(opts *pb.PredictOptions) (pb.TokenizationResponse, error) {
+	predictOptions := buildPredictOptions(opts)
+	l, tokens, err := llm.llama.TokenizeString(opts.Prompt, predictOptions...)
+	if err != nil {
+		return pb.TokenizationResponse{}, err
+	}
+	return pb.TokenizationResponse{
+		Length: l,
+		Tokens: tokens,
+	}, nil
+}

backend/go/llm/llama/main.go

@@ -1,12 +1,12 @@
 package main
 
+// GRPC Falcon server
+
 // Note: this is started internally by LocalAI and a server is allocated for each model
 
 import (
 	"flag"
 
-	tts "github.com/go-skynet/LocalAI/pkg/grpc/tts"
-
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )
 
@@ -17,7 +17,7 @@ var (
 func main() {
 	flag.Parse()
 
-	if err := grpc.StartServer(*addr, &tts.Piper{}); err != nil {
+	if err := grpc.StartServer(*addr, &LLM{}); err != nil {
 		panic(err)
 	}
 }

backend/go/llm/mpt/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/replit/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/rwkv/main.go

@@ -0,0 +1,21 @@
+package main
+
+// Note: this is started internally by LocalAI and a server is allocated for each model
+
+import (
+	"flag"
+
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+func main() {
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &LLM{}); err != nil {
+		panic(err)
+	}
+}

backend/go/llm/rwkv/rwkv.go

@@ -1,4 +1,4 @@
-package rwkv
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
@@ -14,15 +14,21 @@ import (
 const tokenizerSuffix = ".tokenizer.json"
 
 type LLM struct {
-	base.Base
+	base.SingleThread
 
 	rwkv *rwkv.RwkvState
 }
 
 func (llm *LLM) Load(opts *pb.ModelOptions) error {
-	modelPath := filepath.Dir(opts.Model)
-	modelFile := filepath.Base(opts.Model)
-	model := rwkv.LoadFiles(opts.Model, filepath.Join(modelPath, modelFile+tokenizerSuffix), uint32(opts.GetThreads()))
+	tokenizerFile := opts.Tokenizer
+	if tokenizerFile == "" {
+		modelFile := filepath.Base(opts.ModelFile)
+		tokenizerFile = modelFile + tokenizerSuffix
+	}
+	modelPath := filepath.Dir(opts.ModelFile)
+	tokenizerPath := filepath.Join(modelPath, tokenizerFile)
+
+	model := rwkv.LoadFiles(opts.ModelFile, tokenizerPath, uint32(opts.GetThreads()))
 
 	if model == nil {
 		return fmt.Errorf("could not load model")
@@ -32,7 +38,6 @@ func (llm *LLM) Load(opts *pb.ModelOptions) error {
 }
 
 func (llm *LLM) Predict(opts *pb.PredictOptions) (string, error) {
-
 	stopWord := "\n"
 	if len(opts.StopPrompts) > 0 {
 		stopWord = opts.StopPrompts[0]
@@ -69,3 +74,22 @@ func (llm *LLM) PredictStream(opts *pb.PredictOptions, results chan string) erro
 
 	return nil
 }
+
+func (llm *LLM) TokenizeString(opts *pb.PredictOptions) (pb.TokenizationResponse, error) {
+	tokens, err := llm.rwkv.Tokenizer.Encode(opts.Prompt)
+	if err != nil {
+		return pb.TokenizationResponse{}, err
+	}
+
+	l := len(tokens)
+	i32Tokens := make([]int32, l)
+
+	for i, t := range tokens {
+		i32Tokens[i] = int32(t.ID)
+	}
+
+	return pb.TokenizationResponse{
+		Length: int32(l),
+		Tokens: i32Tokens,
+	}, nil
+}

backend/go/llm/starcoder/main.go

@@ -5,7 +5,7 @@ package main
 import (
 	"flag"
 
-	transformers "github.com/go-skynet/LocalAI/pkg/grpc/llm/transformers"
+	transformers "github.com/go-skynet/LocalAI/backend/go/llm/transformers"
 
 	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )

backend/go/llm/transformers/dolly.go

@@ -12,13 +12,13 @@ import (
 )
 
 type Dolly struct {
-	base.Base
+	base.SingleThread
 
 	dolly *transformers.Dolly
 }
 
 func (llm *Dolly) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewDolly(opts.Model)
+	model, err := transformers.NewDolly(opts.ModelFile)
 	llm.dolly = model
 	return err
 }
@@ -29,6 +29,7 @@ func (llm *Dolly) Predict(opts *pb.PredictOptions) (string, error) {
 
 // fallback to Predict
 func (llm *Dolly) PredictStream(opts *pb.PredictOptions, results chan string) error {
+
 	go func() {
 		res, err := llm.dolly.Predict(opts.Prompt, buildPredictOptions(opts)...)
 

backend/go/llm/transformers/falcon.go

@@ -12,13 +12,13 @@ import (
 )
 
 type Falcon struct {
-	base.Base
+	base.SingleThread
 
 	falcon *transformers.Falcon
 }
 
 func (llm *Falcon) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewFalcon(opts.Model)
+	model, err := transformers.NewFalcon(opts.ModelFile)
 	llm.falcon = model
 	return err
 }

backend/go/llm/transformers/gpt2.go

@@ -12,13 +12,13 @@ import (
 )
 
 type GPT2 struct {
-	base.Base
+	base.SingleThread
 
 	gpt2 *transformers.GPT2
 }
 
 func (llm *GPT2) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.New(opts.Model)
+	model, err := transformers.New(opts.ModelFile)
 	llm.gpt2 = model
 	return err
 }

backend/go/llm/transformers/gptj.go

@@ -12,13 +12,13 @@ import (
 )
 
 type GPTJ struct {
-	base.Base
+	base.SingleThread
 
 	gptj *transformers.GPTJ
 }
 
 func (llm *GPTJ) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewGPTJ(opts.Model)
+	model, err := transformers.NewGPTJ(opts.ModelFile)
 	llm.gptj = model
 	return err
 }

backend/go/llm/transformers/gptneox.go

@@ -12,13 +12,13 @@ import (
 )
 
 type GPTNeoX struct {
-	base.Base
+	base.SingleThread
 
 	gptneox *transformers.GPTNeoX
 }
 
 func (llm *GPTNeoX) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewGPTNeoX(opts.Model)
+	model, err := transformers.NewGPTNeoX(opts.ModelFile)
 	llm.gptneox = model
 	return err
 }

backend/go/llm/transformers/mpt.go

@@ -12,13 +12,13 @@ import (
 )
 
 type MPT struct {
-	base.Base
+	base.SingleThread
 
 	mpt *transformers.MPT
 }
 
 func (llm *MPT) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewMPT(opts.Model)
+	model, err := transformers.NewMPT(opts.ModelFile)
 	llm.mpt = model
 	return err
 }

backend/go/llm/transformers/replit.go

@@ -12,13 +12,13 @@ import (
 )
 
 type Replit struct {
-	base.Base
+	base.SingleThread
 
 	replit *transformers.Replit
 }
 
 func (llm *Replit) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewReplit(opts.Model)
+	model, err := transformers.NewReplit(opts.ModelFile)
 	llm.replit = model
 	return err
 }

backend/go/llm/transformers/starcoder.go

@@ -12,13 +12,13 @@ import (
 )
 
 type Starcoder struct {
-	base.Base
+	base.SingleThread
 
 	starcoder *transformers.Starcoder
 }
 
 func (llm *Starcoder) Load(opts *pb.ModelOptions) error {
-	model, err := transformers.NewStarcoder(opts.Model)
+	model, err := transformers.NewStarcoder(opts.ModelFile)
 	llm.starcoder = model
 	return err
 }

backend/go/transcribe/main.go

@@ -0,0 +1,21 @@
+package main
+
+// Note: this is started internally by LocalAI and a server is allocated for each model
+
+import (
+	"flag"
+
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+func main() {
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &Whisper{}); err != nil {
+		panic(err)
+	}
+}

backend/go/transcribe/transcript.go

@@ -1,4 +1,4 @@
-package whisper
+package main
 
 import (
 	"fmt"
@@ -7,8 +7,8 @@ import (
 	"path/filepath"
 
 	"github.com/ggerganov/whisper.cpp/bindings/go/pkg/whisper"
-	wav "github.com/go-audio/wav"
-	"github.com/go-skynet/LocalAI/pkg/grpc/whisper/api"
+	"github.com/go-audio/wav"
+	"github.com/go-skynet/LocalAI/api/schema"
 )
 
 func sh(c string) (string, error) {
@@ -29,8 +29,8 @@ func audioToWav(src, dst string) error {
 	return nil
 }
 
-func Transcript(model whisper.Model, audiopath, language string, threads uint) (api.Result, error) {
-	res := api.Result{}
+func Transcript(model whisper.Model, audiopath, language string, threads uint) (schema.Result, error) {
+	res := schema.Result{}
 
 	dir, err := os.MkdirTemp("", "whisper")
 	if err != nil {
@@ -90,7 +90,7 @@ func Transcript(model whisper.Model, audiopath, language string, threads uint) (
 			tokens = append(tokens, t.Id)
 		}
 
-		segment := api.Segment{Id: s.Num, Text: s.Text, Start: s.Start, End: s.End, Tokens: tokens}
+		segment := schema.Segment{Id: s.Num, Text: s.Text, Start: s.Start, End: s.End, Tokens: tokens}
 		res.Segments = append(res.Segments, segment)
 
 		res.Text += s.Text

backend/go/transcribe/whisper.go

@@ -1,27 +1,26 @@
-package transcribe
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
 import (
 	"github.com/ggerganov/whisper.cpp/bindings/go/pkg/whisper"
+	"github.com/go-skynet/LocalAI/api/schema"
 	"github.com/go-skynet/LocalAI/pkg/grpc/base"
 	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
-	whisperutil "github.com/go-skynet/LocalAI/pkg/grpc/whisper"
-	"github.com/go-skynet/LocalAI/pkg/grpc/whisper/api"
 )
 
 type Whisper struct {
-	base.Base
+	base.SingleThread
 	whisper whisper.Model
 }
 
 func (sd *Whisper) Load(opts *pb.ModelOptions) error {
 	// Note: the Model here is a path to a directory containing the model files
-	w, err := whisper.New(opts.Model)
+	w, err := whisper.New(opts.ModelFile)
 	sd.whisper = w
 	return err
 }
 
-func (sd *Whisper) AudioTranscription(opts *pb.TranscriptRequest) (api.Result, error) {
-	return whisperutil.Transcript(sd.whisper, opts.Dst, opts.Language, uint(opts.Threads))
+func (sd *Whisper) AudioTranscription(opts *pb.TranscriptRequest) (schema.Result, error) {
+	return Transcript(sd.whisper, opts.Dst, opts.Language, uint(opts.Threads))
 }

backend/go/tts/main.go

@@ -0,0 +1,21 @@
+package main
+
+// Note: this is started internally by LocalAI and a server is allocated for each model
+
+import (
+	"flag"
+
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
+)
+
+var (
+	addr = flag.String("addr", "localhost:50051", "the address to connect to")
+)
+
+func main() {
+	flag.Parse()
+
+	if err := grpc.StartServer(*addr, &Piper{}); err != nil {
+		panic(err)
+	}
+}

backend/go/tts/piper.go

@@ -1,9 +1,11 @@
-package tts
+package main
 
 // This is a wrapper to statisfy the GRPC service interface
 // It is meant to be used by the main executable that is the server for the specific backend type (falcon, gpt3, etc)
 import (
+	"fmt"
 	"os"
+	"path/filepath"
 
 	"github.com/go-skynet/LocalAI/pkg/grpc/base"
 	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
@@ -11,11 +13,14 @@ import (
 )
 
 type Piper struct {
-	base.Base
+	base.SingleThread
 	piper *PiperB
 }
 
 func (sd *Piper) Load(opts *pb.ModelOptions) error {
+	if filepath.Ext(opts.ModelFile) != ".onnx" {
+		return fmt.Errorf("unsupported model type %s (should end with .onnx)", opts.ModelFile)
+	}
 	var err error
 	// Note: the Model here is a path to a directory containing the model files
 	sd.piper, err = New(opts.LibrarySearchPath)

backend/python/README.md

@@ -0,0 +1,38 @@
+# Common commands about conda environment
+
+## Create a new empty conda environment
+
+```
+conda create --name <env-name> python=<your version> -y
+
+conda create --name autogptq python=3.11 -y
+```
+
+## To activate the environment
+
+As of conda 4.4
+```
+conda activate autogptq
+```
+
+The conda version older than 4.4
+
+```
+source activate autogptq
+```
+
+## Install the packages to your environment
+
+Sometimes you need to install the packages from the conda-forge channel
+
+By using `conda`
+```
+conda install <your-package-name>
+
+conda install -c conda-forge <your package-name>
+```
+
+Or by using `pip`
+```
+pip install <your-package-name>
+```

backend/python/autogptq/Makefile

@@ -0,0 +1,4 @@
+.PHONY: autogptq
+autogptq:
+	$(MAKE) -C ../common-env/transformers
+

backend/python/autogptq/README.md

@@ -0,0 +1,5 @@
+# Creating a separate environment for the autogptq project
+
+```
+make autogptq
+```

backend/python/autogptq/autogptq.py

@@ -0,0 +1,112 @@
+#!/usr/bin/env python3
+from concurrent import futures
+import argparse
+import signal
+import sys
+import os
+import time
+
+import grpc
+import backend_pb2
+import backend_pb2_grpc
+from auto_gptq import AutoGPTQForCausalLM
+from transformers import AutoTokenizer
+from transformers import TextGenerationPipeline
+
+_ONE_DAY_IN_SECONDS = 60 * 60 * 24
+
+# If MAX_WORKERS are specified in the environment use it, otherwise default to 1
+MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1'))
+
+# Implement the BackendServicer class with the service methods
+class BackendServicer(backend_pb2_grpc.BackendServicer):
+    def Health(self, request, context):
+        return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
+    def LoadModel(self, request, context):
+        try:
+            device = "cuda:0"
+            if request.Device != "":
+                device = request.Device
+
+            tokenizer = AutoTokenizer.from_pretrained(request.Model, use_fast=request.UseFastTokenizer)
+
+            model = AutoGPTQForCausalLM.from_quantized(request.Model,
+                    model_basename=request.ModelBaseName,
+                    use_safetensors=True,
+                    trust_remote_code=True,
+                    device=device,
+                    use_triton=request.UseTriton,
+                    quantize_config=None)
+            
+            self.model = model
+            self.tokenizer = tokenizer
+        except Exception as err:
+            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
+        return backend_pb2.Result(message="Model loaded successfully", success=True)
+
+    def Predict(self, request, context):
+        penalty = 1.0
+        if request.Penalty != 0.0:
+            penalty = request.Penalty
+        tokens = 512
+        if request.Tokens != 0:
+            tokens = request.Tokens
+        top_p = 0.95
+        if request.TopP != 0.0:
+            top_p = request.TopP
+
+        # Implement Predict RPC
+        pipeline = TextGenerationPipeline(
+            model=self.model, 
+            tokenizer=self.tokenizer,
+            max_new_tokens=tokens,
+            temperature=request.Temperature,
+            top_p=top_p,
+            repetition_penalty=penalty,
+            )
+        t = pipeline(request.Prompt)[0]["generated_text"]
+        # Remove prompt from response if present
+        if request.Prompt in t:
+            t = t.replace(request.Prompt, "")
+
+        return backend_pb2.Result(message=bytes(t, encoding='utf-8'))
+
+    def PredictStream(self, request, context):
+        # Implement PredictStream RPC
+        #for reply in some_data_generator():
+        #    yield reply
+        # Not implemented yet
+        return self.Predict(request, context)
+
+
+def serve(address):
+    server = grpc.server(futures.ThreadPoolExecutor(max_workers=MAX_WORKERS))
+    backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server)
+    server.add_insecure_port(address)
+    server.start()
+    print("Server started. Listening on: " + address, file=sys.stderr)
+
+    # Define the signal handler function
+    def signal_handler(sig, frame):
+        print("Received termination signal. Shutting down...")
+        server.stop(0)
+        sys.exit(0)
+
+    # Set the signal handlers for SIGINT and SIGTERM
+    signal.signal(signal.SIGINT, signal_handler)
+    signal.signal(signal.SIGTERM, signal_handler)
+
+    try:
+        while True:
+            time.sleep(_ONE_DAY_IN_SECONDS)
+    except KeyboardInterrupt:
+        server.stop(0)
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Run the gRPC server.")
+    parser.add_argument(
+        "--addr", default="localhost:50051", help="The address to bind the server to."
+    )
+    args = parser.parse_args()
+
+    serve(args.addr)

backend/python/autogptq/autogptq.yml

@@ -0,0 +1,86 @@
+name: autogptq
+channels:
+  - defaults
+dependencies:
+  - _libgcc_mutex=0.1=main
+  - _openmp_mutex=5.1=1_gnu
+  - bzip2=1.0.8=h7b6447c_0
+  - ca-certificates=2023.08.22=h06a4308_0
+  - ld_impl_linux-64=2.38=h1181459_1
+  - libffi=3.4.4=h6a678d5_0
+  - libgcc-ng=11.2.0=h1234567_1
+  - libgomp=11.2.0=h1234567_1
+  - libstdcxx-ng=11.2.0=h1234567_1
+  - libuuid=1.41.5=h5eee18b_0
+  - ncurses=6.4=h6a678d5_0
+  - openssl=3.0.11=h7f8727e_2
+  - pip=23.2.1=py311h06a4308_0
+  - python=3.11.5=h955ad1f_0
+  - readline=8.2=h5eee18b_0
+  - setuptools=68.0.0=py311h06a4308_0
+  - sqlite=3.41.2=h5eee18b_0
+  - tk=8.6.12=h1ccaba5_0
+  - wheel=0.41.2=py311h06a4308_0
+  - xz=5.4.2=h5eee18b_0
+  - zlib=1.2.13=h5eee18b_0
+  - pip:
+      - accelerate==0.23.0
+      - aiohttp==3.8.5
+      - aiosignal==1.3.1
+      - async-timeout==4.0.3
+      - attrs==23.1.0
+      - auto-gptq==0.4.2
+      - certifi==2023.7.22
+      - charset-normalizer==3.3.0
+      - datasets==2.14.5
+      - dill==0.3.7
+      - filelock==3.12.4
+      - frozenlist==1.4.0
+      - fsspec==2023.6.0
+      - grpcio==1.59.0
+      - huggingface-hub==0.16.4
+      - idna==3.4
+      - jinja2==3.1.2
+      - markupsafe==2.1.3
+      - mpmath==1.3.0
+      - multidict==6.0.4
+      - multiprocess==0.70.15
+      - networkx==3.1
+      - numpy==1.26.0
+      - nvidia-cublas-cu12==12.1.3.1
+      - nvidia-cuda-cupti-cu12==12.1.105
+      - nvidia-cuda-nvrtc-cu12==12.1.105
+      - nvidia-cuda-runtime-cu12==12.1.105
+      - nvidia-cudnn-cu12==8.9.2.26
+      - nvidia-cufft-cu12==11.0.2.54
+      - nvidia-curand-cu12==10.3.2.106
+      - nvidia-cusolver-cu12==11.4.5.107
+      - nvidia-cusparse-cu12==12.1.0.106
+      - nvidia-nccl-cu12==2.18.1
+      - nvidia-nvjitlink-cu12==12.2.140
+      - nvidia-nvtx-cu12==12.1.105
+      - packaging==23.2
+      - pandas==2.1.1
+      - peft==0.5.0
+      - protobuf==4.24.4
+      - psutil==5.9.5
+      - pyarrow==13.0.0
+      - python-dateutil==2.8.2
+      - pytz==2023.3.post1
+      - pyyaml==6.0.1
+      - regex==2023.10.3
+      - requests==2.31.0
+      - rouge==1.0.1
+      - safetensors==0.3.3
+      - six==1.16.0
+      - sympy==1.12
+      - tokenizers==0.14.0
+      - torch==2.1.0
+      - tqdm==4.66.1
+      - transformers==4.34.0
+      - triton==2.1.0
+      - typing-extensions==4.8.0
+      - tzdata==2023.3
+      - urllib3==2.0.6
+      - xxhash==3.4.1
+      - yarl==1.9.2