fix(llama.cpp): fix eos without cache

.bruno/LocalAI Test Requests/Sound Generation/musicgen.bru

@@ -1,23 +0,0 @@
-meta {
-  name: musicgen
-  type: http
-  seq: 1
-}
-
-post {
-  url: {{PROTOCOL}}{{HOST}}:{{PORT}}/v1/sound-generation
-  body: json
-  auth: none
-}
-
-headers {
-  Content-Type: application/json
-}
-
-body:json {
-  {
-      "model_id": "facebook/musicgen-small",
-      "text": "Exciting 80s Newscast Interstitial",
-      "duration_seconds": 8
-  }
-}

.bruno/LocalAI Test Requests/model gallery/model delete.bru

@@ -1,11 +0,0 @@
-meta {
-  name: model delete
-  type: http
-  seq: 7
-}
-
-post {
-  url: {{PROTOCOL}}{{HOST}}:{{PORT}}/models/galleries
-  body: none
-  auth: none
-}

.bruno/LocalAI Test Requests/transcription/transcribe.bru

@@ -1,16 +0,0 @@
-meta {
-  name: transcribe
-  type: http
-  seq: 1
-}
-
-post {
-  url: {{PROTOCOL}}{{HOST}}:{{PORT}}/v1/audio/transcriptions
-  body: multipartForm
-  auth: none
-}
-
-body:multipart-form {
-  file: @file(transcription/gb1.ogg)
-  model: whisper-1
-}

.devcontainer-scripts/postcreate.sh

@@ -1,17 +0,0 @@
-#!/bin/bash
-
-cd /workspace
-
-# Get the files into the volume without a bind mount
-if [ ! -d ".git" ]; then
-    git clone https://github.com/mudler/LocalAI.git .
-else
-    git fetch
-fi
-
-echo "Standard Post-Create script completed."
-
-if [ -f "/devcontainer-customization/postcreate.sh" ]; then
-    echo "Launching customization postcreate.sh"
-    bash "/devcontainer-customization/postcreate.sh"
-fi

.devcontainer-scripts/poststart.sh

@@ -1,16 +0,0 @@
-#!/bin/bash
-
-cd /workspace
-
-# Grab the pre-stashed backend assets to avoid build issues
-cp -r /build/backend-assets /workspace/backend-assets
-
-# Ensures generated source files are present upon load
-make prepare
-
-echo "Standard Post-Start script completed."
-
-if [ -f "/devcontainer-customization/poststart.sh" ]; then
-    echo "Launching customization poststart.sh"
-    bash "/devcontainer-customization/poststart.sh"
-fi

.devcontainer-scripts/utils.sh

@@ -1,55 +0,0 @@
-#!/bin/bash
-
-# This file contains some really simple functions that are useful when building up customization scripts.
-
-
-# Checks if the git config has a user registered - and sets it up if not.
-#
-# Param 1: name
-# Param 2: email
-#
-config_user() {
-    echo "Configuring git for $1 <$2>"
-    local gcn=$(git config --global user.name)
-    if [ -z "${gcn}" ]; then
-        echo "Setting up git user / remote"
-        git config --global user.name "$1"
-        git config --global user.email "$2"
-        
-    fi
-}
-
-# Checks if the git remote is configured - and sets it up if not. Fetches either way.
-#
-# Param 1: remote name
-# Param 2: remote url
-#
-config_remote() {
-    echo "Adding git remote and fetching $2 as $1"
-    local gr=$(git remote -v | grep $1)
-    if [ -z "${gr}" ]; then
-        git remote add $1 $2
-    fi
-    git fetch $1
-}
-
-# Setup special .ssh files
-# Prints out lines of text to make things pretty
-# Param 1: bash array, filenames relative to the customization directory that should be copied to ~/.ssh
-setup_ssh() {
-    echo "starting ~/.ssh directory setup..."
-    mkdir -p "${HOME}.ssh"
-    chmod 0700 "${HOME}/.ssh"
-    echo "-----"
-    local files=("$@")
-    for file in "${files[@]}" ; do
-        local cfile="/devcontainer-customization/${file}"
-        local hfile="${HOME}/.ssh/${file}"
-        if [ ! -f "${hfile}" ]; then
-            echo "copying \"${file}\""
-            cp "${cfile}" "${hfile}"
-            chmod 600 "${hfile}"
-        fi
-    done
-    echo "~/.ssh directory setup complete!"
-}

.devcontainer/customization/README.md

@@ -1,25 +0,0 @@
-Place any additional resources your environment requires in this directory
-
-Script hooks are currently called for:
-`postcreate.sh` and `poststart.sh`
-
-If files with those names exist here, they will be called at the end of the normal script.
-
-This is a good place to set things like `git config --global user.name` are set - and to handle any other files that are mounted via this directory.
-
-To assist in doing so, `source /.devcontainer-scripts/utils.sh` will provide utility functions that may be useful - for example:
-
-```
-#!/bin/bash
-
-source "/.devcontainer-scripts/utils.sh"
-
-sshfiles=("config", "key.pub")
-
-setup_ssh "${sshfiles[@]}"
-
-config_user "YOUR NAME" "YOUR EMAIL"
-
-config_remote "REMOTE NAME" "REMOTE URL"
-
-```

.devcontainer/devcontainer.json

@@ -1,24 +0,0 @@
-{
-    "$schema": "https://raw.githubusercontent.com/devcontainers/spec/main/schemas/devContainer.schema.json",
-    "name": "LocalAI",
-    "workspaceFolder": "/workspace",
-    "dockerComposeFile": [ "./docker-compose-devcontainer.yml" ],
-    "service": "api",
-    "shutdownAction": "stopCompose",
-    "customizations": {
-        "vscode": {
-            "extensions": [
-                "golang.go",
-                "ms-vscode.makefile-tools",
-                "ms-azuretools.vscode-docker",
-                "ms-python.python",
-                "ms-python.debugpy",
-                "wayou.vscode-todo-highlight",
-                "waderyan.gitblame"
-            ]
-        }
-    },
-    "forwardPorts": [8080, 3000],
-    "postCreateCommand": "bash /.devcontainer-scripts/postcreate.sh",
-    "postStartCommand": "bash /.devcontainer-scripts/poststart.sh"
-}

.devcontainer/docker-compose-devcontainer.yml

@@ -1,48 +0,0 @@
-services:
-  api:
-    build:
-      context: ..
-      dockerfile: Dockerfile
-      target: devcontainer
-      args:
-      - FFMPEG=true
-      - IMAGE_TYPE=extras
-      - GO_TAGS=stablediffusion p2p tts
-    env_file:
-      - ../.env
-    ports:
-      - 8080:8080
-    volumes:
-      - localai_workspace:/workspace
-      - ../models:/host-models
-      - ./customization:/devcontainer-customization
-    command: /bin/sh -c "while sleep 1000; do :; done"
-    cap_add:
-      - SYS_PTRACE
-    security_opt:
-      - seccomp:unconfined
-  prometheus:
-    image: prom/prometheus
-    container_name: prometheus
-    command:
-      - '--config.file=/etc/prometheus/prometheus.yml'
-    ports:
-      - 9090:9090
-    restart: unless-stopped
-    volumes:
-      - ./prometheus:/etc/prometheus
-      - prom_data:/prometheus
-  grafana:
-    image: grafana/grafana
-    container_name: grafana
-    ports:
-      - 3000:3000
-    restart: unless-stopped
-    environment:
-      - GF_SECURITY_ADMIN_USER=admin
-      - GF_SECURITY_ADMIN_PASSWORD=grafana
-    volumes:
-      - ./grafana:/etc/grafana/provisioning/datasources
-volumes:
-  prom_data:
-  localai_workspace:

.devcontainer/grafana/datasource.yml

@@ -1,10 +0,0 @@
-
-apiVersion: 1
-
-datasources:
-- name: Prometheus
-  type: prometheus
-  url: http://prometheus:9090 
-  isDefault: true
-  access: proxy
-  editable: true

.devcontainer/prometheus/prometheus.yml

@@ -1,21 +0,0 @@
-global:
-  scrape_interval: 15s
-  scrape_timeout: 10s
-  evaluation_interval: 15s
-alerting:
-  alertmanagers:
-  - static_configs:
-    - targets: []
-    scheme: http
-    timeout: 10s
-    api_version: v1
-scrape_configs:
-- job_name: prometheus
-  honor_timestamps: true
-  scrape_interval: 15s
-  scrape_timeout: 10s
-  metrics_path: /metrics
-  scheme: http
-  static_configs:
-  - targets:
-    - localhost:9090

.dockerignore

@@ -1,17 +1,6 @@
 .idea
-.github
-.vscode
-.devcontainer
 models
 examples/chatbot-ui/models
 examples/rwkv/models
 examples/**/models
-Dockerfile*
-__pycache__
-
-# SonarQube
-.scannerwork
-
-# backend virtual environments
-**/venv
-backend/python/**/source
+Dockerfile

.editorconfig

@@ -1,31 +0,0 @@
-
-root = true
-
-[*]
-indent_style = space
-indent_size = 2
-end_of_line = lf
-charset = utf-8
-trim_trailing_whitespace = true
-insert_final_newline = true
-
-[*.go]
-indent_style = tab
-
-[Makefile]
-indent_style = tab
-
-[*.proto]
-indent_size = 2
-
-[*.py]
-indent_size = 4
-
-[*.js]
-indent_size = 2
-
-[*.yaml]
-indent_size = 2
-
-[*.md]
-trim_trailing_whitespace = false

.env

@@ -1,33 +1,33 @@
 ## Set number of threads.
 ## Note: prefer the number of physical cores. Overbooking the CPU degrades performance notably.
-# LOCALAI_THREADS=14
+# THREADS=14
 
 ## Specify a different bind address (defaults to ":8080")
-# LOCALAI_ADDRESS=127.0.0.1:8080
+# ADDRESS=127.0.0.1:8080
 
 ## Default models context size
-# LOCALAI_CONTEXT_SIZE=512
+# CONTEXT_SIZE=512
 #
 ## Define galleries.
 ## models will to install will be visible in `/models/available`
-# LOCALAI_GALLERIES=[{"name":"localai", "url":"github:mudler/LocalAI/gallery/index.yaml@master"}]
+# GALLERIES=[{"name":"model-gallery", "url":"github:go-skynet/model-gallery/index.yaml"}]
 
 ## CORS settings
-# LOCALAI_CORS=true
-# LOCALAI_CORS_ALLOW_ORIGINS=*
+# CORS=true
+# CORS_ALLOW_ORIGINS=*
 
 ## Default path for models
 #
-# LOCALAI_MODELS_PATH=/models
+# MODELS_PATH=/models
 
 ## Enable debug mode
-# LOCALAI_LOG_LEVEL=debug
+# DEBUG=true
 
 ## Disables COMPEL (Diffusers)
 # COMPEL=0
 
 ## Enable/Disable single backend (useful if only one GPU is available)
-# LOCALAI_SINGLE_ACTIVE_BACKEND=true
+# 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.
@@ -46,13 +46,13 @@
 # GO_TAGS=stablediffusion
 
 ## Path where to store generated images
-# LOCALAI_IMAGE_PATH=/tmp/generated/images
+# IMAGE_PATH=/tmp
 
 ## Specify a default upload limit in MB (whisper)
-# LOCALAI_UPLOAD_LIMIT=15
+# UPLOAD_LIMIT
 
 ## List of external GRPC backends (note on the container image this variable is already set to use extra backends available in extra/)
-# LOCALAI_EXTERNAL_GRPC_BACKENDS=my-backend:127.0.0.1:9000,my-backend2:/usr/bin/backend.py
+# 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 ###
@@ -71,27 +71,19 @@
 ### Define the number of parallel LLAMA.cpp workers (Defaults to 1)
 # LLAMACPP_PARALLEL=1
 
-### Define a list of GRPC Servers for llama-cpp workers to distribute the load
-# https://github.com/ggerganov/llama.cpp/pull/6829
-# https://github.com/ggerganov/llama.cpp/blob/master/examples/rpc/README.md
-# LLAMACPP_GRPC_SERVERS=""
-
 ### Enable to run parallel requests
-# LOCALAI_PARALLEL_REQUESTS=true
-
-# Enable to allow p2p mode
-# LOCALAI_P2P=true
+# PARALLEL_REQUESTS=true
 
 ### Watchdog settings
 ###
 # Enables watchdog to kill backends that are inactive for too much time
-# LOCALAI_WATCHDOG_IDLE=true
-#
-# Time in duration format (e.g. 1h30m) after which a backend is considered idle
-# LOCALAI_WATCHDOG_IDLE_TIMEOUT=5m
+# WATCHDOG_IDLE=true
 #
 # Enables watchdog to kill backends that are busy for too much time
-# LOCALAI_WATCHDOG_BUSY=true
+# 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
-# LOCALAI_WATCHDOG_BUSY_TIMEOUT=5m
+# WATCHDOG_BUSY_TIMEOUT=5m

.gitattributes

@@ -1,2 +1 @@
 *.sh text eol=lf
-backend/cpp/llama/*.hpp linguist-vendored

.github/bump_deps.sh

@@ -6,17 +6,4 @@ VAR=$3
 
 LAST_COMMIT=$(curl -s -H "Accept: application/vnd.github.VERSION.sha" "https://api.github.com/repos/$REPO/commits/$BRANCH")
 
-# Read $VAR from Makefile (only first match)
-set +e
-CURRENT_COMMIT="$(grep -m1 "^$VAR?=" Makefile | cut -d'=' -f2)"
-set -e
-
 sed -i Makefile -e "s/$VAR?=.*/$VAR?=$LAST_COMMIT/"
-
-if [ -z "$CURRENT_COMMIT" ]; then
-    echo "Could not find $VAR in Makefile."
-    exit 0
-fi
-
-echo "Changes: https://github.com/$REPO/compare/${CURRENT_COMMIT}..${LAST_COMMIT}" >> "${VAR}_message.txt"
-echo "${LAST_COMMIT}" >> "${VAR}_commit.txt"

.github/bump_docs.sh

@@ -2,6 +2,6 @@
 set -xe
 REPO=$1
 
-LATEST_TAG=$(curl -s "https://api.github.com/repos/$REPO/releases/latest" | jq -r '.tag_name')
+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/check_and_update.py

@@ -1,85 +0,0 @@
-import hashlib
-from huggingface_hub import hf_hub_download, get_paths_info
-import requests
-import sys
-import os
-
-uri = sys.argv[1]
-file_name = uri.split('/')[-1]
-
-# Function to parse the URI and determine download method
-def parse_uri(uri):
-    if uri.startswith('huggingface://'):
-        repo_id = uri.split('://')[1]
-        return 'huggingface', repo_id.rsplit('/', 1)[0]
-    elif 'huggingface.co' in uri:
-        parts = uri.split('/resolve/')
-        if len(parts) > 1:
-            repo_path = parts[0].split('https://huggingface.co/')[-1]
-            return 'huggingface', repo_path
-    return 'direct', uri
-
-def calculate_sha256(file_path):
-    sha256_hash = hashlib.sha256()
-    with open(file_path, 'rb') as f:
-        for byte_block in iter(lambda: f.read(4096), b''):
-            sha256_hash.update(byte_block)
-    return sha256_hash.hexdigest()
-
-def manual_safety_check_hf(repo_id):
-    scanResponse = requests.get('https://huggingface.co/api/models/' + repo_id + "/scan")
-    scan = scanResponse.json()
-    # Check if 'hasUnsafeFile' exists in the response
-    if 'hasUnsafeFile' in scan:
-        if scan['hasUnsafeFile']:
-            return scan
-        else:
-            return None
-    else:
-        return None
-
-download_type, repo_id_or_url = parse_uri(uri)
-
-new_checksum =  None
-file_path = None
-
-# Decide download method based on URI type
-if download_type == 'huggingface':
-    # Check if the repo is flagged as dangerous by HF
-    hazard = manual_safety_check_hf(repo_id_or_url)
-    if hazard != None:
-        print(f'Error: HuggingFace has detected security problems for {repo_id_or_url}: {str(hazard)}', filename=file_name)
-        sys.exit(5)
-    # Use HF API to pull sha
-    for file in get_paths_info(repo_id_or_url, [file_name], repo_type='model'):
-        try:
-            new_checksum = file.lfs.sha256
-            break
-        except Exception as e:
-            print(f'Error from Hugging Face Hub: {str(e)}', file=sys.stderr)
-            sys.exit(2)
-    if new_checksum is None:
-        try:
-            file_path = hf_hub_download(repo_id=repo_id_or_url, filename=file_name)
-        except Exception as e:
-            print(f'Error from Hugging Face Hub: {str(e)}', file=sys.stderr)
-            sys.exit(2)
-else:
-    response = requests.get(repo_id_or_url)
-    if response.status_code == 200:
-        with open(file_name, 'wb') as f:
-            f.write(response.content)
-        file_path = file_name
-    elif response.status_code == 404:
-        print(f'File not found: {response.status_code}', file=sys.stderr)
-        sys.exit(2)
-    else:
-        print(f'Error downloading file: {response.status_code}', file=sys.stderr)
-        sys.exit(1)
-
-if new_checksum is None:
-    new_checksum = calculate_sha256(file_path)
-    print(new_checksum)
-    os.remove(file_path)
-else:
-    print(new_checksum)

.github/checksum_checker.sh

@@ -1,63 +0,0 @@
-#!/bin/bash
-# This scripts needs yq and huggingface_hub to be installed
-# to install hugingface_hub run pip install huggingface_hub
-
-# Path to the input YAML file
-input_yaml=$1
-
-# Function to download file and check checksum using Python
-function check_and_update_checksum() {
-    model_name="$1"
-    file_name="$2"
-    uri="$3"
-    old_checksum="$4"
-    idx="$5"
-
-    # Download the file and calculate new checksum using Python
-    new_checksum=$(python3 ./.github/check_and_update.py $uri)
-    result=$?
-
-    if [[ $result -eq 5 ]]; then
-        echo "Contaminated entry detected, deleting entry for $model_name..."
-        yq eval -i "del([$idx])" "$input_yaml"
-        return
-    fi
-
-    if [[ "$new_checksum" == "" ]]; then
-        echo "Error calculating checksum for $file_name. Skipping..."
-        return
-    fi
-
-    echo "Checksum for $file_name: $new_checksum"
-
-    # Compare and update the YAML file if checksums do not match
-    
-    if [[ $result -eq 2 ]]; then
-        echo "File not found, deleting entry for $file_name..."
-        # yq eval -i "del(.[$idx].files[] | select(.filename == \"$file_name\"))" "$input_yaml"
-    elif [[ "$old_checksum" != "$new_checksum" ]]; then
-        echo "Checksum mismatch for $file_name. Updating..."
-        yq eval -i "del(.[$idx].files[] | select(.filename == \"$file_name\").sha256)" "$input_yaml"
-        yq eval -i "(.[$idx].files[] | select(.filename == \"$file_name\")).sha256 = \"$new_checksum\"" "$input_yaml"
-    elif [[ $result -ne 0 ]]; then
-        echo "Error downloading file $file_name. Skipping..."
-    else
-        echo "Checksum match for $file_name. No update needed."
-    fi
-}
-
-# Read the YAML and process each file
-len=$(yq eval '. | length' "$input_yaml")
-for ((i=0; i<$len; i++))
-do
-    name=$(yq eval ".[$i].name" "$input_yaml")
-    files_len=$(yq eval ".[$i].files | length" "$input_yaml")
-    for ((j=0; j<$files_len; j++))
-    do
-        filename=$(yq eval ".[$i].files[$j].filename" "$input_yaml")
-        uri=$(yq eval ".[$i].files[$j].uri" "$input_yaml")
-        checksum=$(yq eval ".[$i].files[$j].sha256" "$input_yaml")
-        echo "Checking model $name, file $filename. URI = $uri, Checksum = $checksum"
-        check_and_update_checksum "$name" "$filename" "$uri" "$checksum" "$i"
-    done
-done

.github/ci/modelslist.go

@@ -1,304 +0,0 @@
-package main
-
-import (
-	"fmt"
-	"html/template"
-	"io/ioutil"
-	"os"
-
-	"github.com/microcosm-cc/bluemonday"
-	"gopkg.in/yaml.v3"
-)
-
-var modelPageTemplate string = `
-<!DOCTYPE html>
-<html>
-<head>
-    <meta charset="UTF-8">
-    <meta name="viewport" content="width=device-width, initial-scale=1.0">
-    <title>LocalAI models</title>
-    <link href="https://cdnjs.cloudflare.com/ajax/libs/flowbite/2.3.0/flowbite.min.css" rel="stylesheet" />
-    <script src="https://cdn.jsdelivr.net/npm/vanilla-lazyload@19.1.3/dist/lazyload.min.js"></script>
-
-    <link
-    rel="stylesheet"
-    href="https://cdn.jsdelivr.net/gh/highlightjs/cdn-release@11.8.0/build/styles/default.min.css"
-  />
-    <script
-    defer
-    src="https://cdn.jsdelivr.net/gh/highlightjs/cdn-release@11.8.0/build/highlight.min.js"
-  ></script>
-    <script
-    defer
-    src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js"
-  ></script>
-  <script
-    defer
-    src="https://cdn.jsdelivr.net/npm/marked/marked.min.js"
-  ></script>
-  <script
-    defer
-    src="https://cdn.jsdelivr.net/npm/dompurify@3.0.6/dist/purify.min.js"
-  ></script>
-
-  <link href="/static/general.css" rel="stylesheet" />
-    <link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;600;700&family=Roboto:wght@400;500&display=swap" rel="stylesheet">
-    <link
-    href="https://fonts.googleapis.com/css?family=Roboto:300,400,500,700,900&display=swap"
-    rel="stylesheet" />
-  <link
-    rel="stylesheet"
-    href="https://cdn.jsdelivr.net/npm/tw-elements/css/tw-elements.min.css" />
-  <script src="https://cdn.tailwindcss.com/3.3.0"></script>
-  <script>
-    tailwind.config = {
-      darkMode: "class",
-      theme: {
-        fontFamily: {
-          sans: ["Roboto", "sans-serif"],
-          body: ["Roboto", "sans-serif"],
-          mono: ["ui-monospace", "monospace"],
-        },
-      },
-      corePlugins: {
-        preflight: false,
-      },
-    };
-  </script>
-    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.1.1/css/all.min.css">
-    <script src="https://unpkg.com/htmx.org@1.9.12" integrity="sha384-ujb1lZYygJmzgSwoxRggbCHcjc0rB2XoQrxeTUQyRjrOnlCoYta87iKBWq3EsdM2" crossorigin="anonymous"></script>
-</head>
-
-<body class="bg-gray-900 text-gray-200">
-<div class="flex flex-col min-h-screen">
-
-<nav class="bg-gray-800 shadow-lg">
-    <div class="container mx-auto px-4 py-4">
-        <div class="flex items-center justify-between">
-            <div class="flex items-center">
-                <a href="/" class="text-white text-xl font-bold"><img src="https://github.com/mudler/LocalAI/assets/2420543/0966aa2a-166e-4f99-a3e5-6c915fc997dd" alt="LocalAI Logo" class="h-10 mr-3 border-2 border-gray-300 shadow rounded"></a>
-                <a href="/" class="text-white text-xl font-bold">LocalAI</a>
-            </div>
-            <!-- Menu button for small screens -->
-            <div class="lg:hidden">
-                <button id="menu-toggle" class="text-gray-400 hover:text-white focus:outline-none">
-                    <i class="fas fa-bars fa-lg"></i>
-                </button>
-            </div>
-            <!-- Navigation links -->
-            <div class="hidden lg:flex lg:items-center lg:justify-end lg:flex-1 lg:w-0">
-                <a href="https://localai.io" class="text-gray-400 hover:text-white px-3 py-2 rounded" target="_blank" ><i class="fas fa-book-reader pr-2"></i> Documentation</a>
-            </div>
-        </div>
-        <!-- Collapsible menu for small screens -->
-        <div class="hidden lg:hidden" id="mobile-menu">
-            <div class="pt-4 pb-3 border-t border-gray-700">
-
-                <a href="https://localai.io" class="block text-gray-400 hover:text-white px-3 py-2 rounded mt-1" target="_blank" ><i class="fas fa-book-reader pr-2"></i> Documentation</a>
-
-            </div>
-        </div>
-    </div>
-</nav>
-
-<style>
-  .is-hidden {
-	display: none;
-	  }
-</style>
-
-<div class="container mx-auto px-4 flex-grow">
-
-<div class="models mt-12">
-	<h2 class="text-center text-3xl font-semibold text-gray-100">
-	LocalAI model gallery list </h2><br>
-
-	<h2 class="text-center text-3xl font-semibold text-gray-100">
-
-	 🖼️ Available {{.AvailableModels}} models</i> <a href="https://localai.io/models/" target="_blank" >
-			<i class="fas fa-circle-info pr-2"></i>
-		</a></h2>
-
-	<h3>
-	Refer to the Model gallery <a href="https://localai.io/models/" target="_blank" ><i class="fas fa-circle-info pr-2"></i></a> for more information on how to use the models with LocalAI.<br>
-
-	You can install models with the CLI command <code>local-ai models install <model-name></code>. or by using the WebUI.
-	</h3>
-
-	<input class="form-control appearance-none block w-full mt-5 px-3 py-2 text-base font-normal text-gray-300 pb-2 mb-5 bg-gray-800 bg-clip-padding border border-solid border-gray-600 rounded transition ease-in-out m-0 focus:text-gray-300 focus:bg-gray-900 focus:border-blue-500 focus:outline-none" type="search"
-	id="searchbox" placeholder="Live search keyword..">
-	  <div class="dark grid grid-cols-1 grid-rows-1 md:grid-cols-3 block rounded-lg shadow-secondary-1 dark:bg-surface-dark">
-		{{ range $_, $model := .Models }}
-		<div class="box me-4 mb-2 block rounded-lg bg-white shadow-secondary-1  dark:bg-gray-800 dark:bg-surface-dark dark:text-white text-surface pb-2">
-		<div>
-		    {{ $icon := "https://upload.wikimedia.org/wikipedia/commons/6/65/No-Image-Placeholder.svg" }}
-			{{ if $model.Icon }}
-	  		{{ $icon = $model.Icon }}
-	  		{{ end }}
-			<div class="flex justify-center items-center">
-				<img data-src="{{ $icon }}" alt="{{$model.Name}}" class="rounded-t-lg max-h-48 max-w-96 object-cover mt-3 lazy">
-			</div>
-	  		<div class="p-6 text-surface dark:text-white">
-				<h5 class="mb-2 text-xl font-medium leading-tight">{{$model.Name}}</h5>
-
-
-				<p class="mb-4 text-base truncate">{{ $model.Description }}</p>
-
-			</div>
-			<div class="px-6 pt-4 pb-2">
-
-      <!-- Modal toggle -->
-      <button data-modal-target="{{ $model.Name}}-modal" data-modal-toggle="{{ $model.Name }}-modal" class="block text-white bg-blue-700 hover:bg-blue-800 focus:ring-4 focus:outline-none focus:ring-blue-300 font-medium rounded-lg text-sm px-5 py-2.5 text-center dark:bg-blue-600 dark:hover:bg-blue-700 dark:focus:ring-blue-800" type="button">
-        More info
-      </button>
-
-    <!-- Main modal -->
-    <div id="{{ $model.Name}}-modal" tabindex="-1" aria-hidden="true" class="hidden overflow-y-auto overflow-x-hidden fixed top-0 right-0 left-0 z-50 justify-center items-center w-full md:inset-0 h-[calc(100%-1rem)] max-h-full">
-        <div class="relative p-4 w-full max-w-2xl max-h-full">
-            <!-- Modal content -->
-            <div class="relative bg-white rounded-lg shadow dark:bg-gray-700">
-                <!-- Modal header -->
-                <div class="flex items-center justify-between p-4 md:p-5 border-b rounded-t dark:border-gray-600">
-                    <h3 class="text-xl font-semibold text-gray-900 dark:text-white">
-                        {{ $model.Name}}
-                    </h3>
-                    <button type="button" class="text-gray-400 bg-transparent hover:bg-gray-200 hover:text-gray-900 rounded-lg text-sm w-8 h-8 ms-auto inline-flex justify-center items-center dark:hover:bg-gray-600 dark:hover:text-white" data-modal-hide="{{$model.Name}}-modal">
-                        <svg class="w-3 h-3" aria-hidden="true" xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 14 14">
-                            <path stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="m1 1 6 6m0 0 6 6M7 7l6-6M7 7l-6 6"/>
-                        </svg>
-                        <span class="sr-only">Close modal</span>
-                    </button>
-                </div>
-                <!-- Modal body -->
-                <div class="p-4 md:p-5 space-y-4">
-                    <div class="flex justify-center items-center">
-                    <img data-src="{{ $icon }}" alt="{{$model.Name}}" class="lazy rounded-t-lg max-h-48 max-w-96 object-cover mt-3">
-                  </div>
-
-                    <p class="text-base leading-relaxed text-gray-500 dark:text-gray-400">
-                    {{ $model.Description }}
-
-                    </p>
-
-                    <p class="text-base leading-relaxed text-gray-500 dark:text-gray-400">
-                    To install the model with the CLI, run: <br>
-                    <code> local-ai models install {{$model.Name}} </code> <br>
-
-                    <hr>
-                    See also <a href="https://localai.io/models/" target="_blank" >
-                    Installation <i class="fas fa-circle-info pr-2"></i>
-                    </a> to see how to install models with the REST API.
-                    </p>
-
-                    <p class="text-base leading-relaxed text-gray-500 dark:text-gray-400">
-                    <ul>
-                    {{ range $_, $u := $model.URLs }}
-                    <li><a href="{{ $u }}" target=_blank><i class="fa-solid fa-link"></i> {{ $u }}</a></li>
-                    {{ end }}
-                    </ul>
-                    </p>
-                </div>
-                <!-- Modal footer -->
-                <div class="flex items-center p-4 md:p-5 border-t border-gray-200 rounded-b dark:border-gray-600">
-                    <button data-modal-hide="{{ $model.Name}}-modal" type="button" class="py-2.5 px-5 ms-3 text-sm font-medium text-gray-900 focus:outline-none bg-white rounded-lg border border-gray-200 hover:bg-gray-100 hover:text-blue-700 focus:z-10 focus:ring-4 focus:ring-gray-100 dark:focus:ring-gray-700 dark:bg-gray-800 dark:text-gray-400 dark:border-gray-600 dark:hover:text-white dark:hover:bg-gray-700">Close</button>
-                </div>
-            </div>
-        </div>
-    </div>
-
-
-			</div>
-		</div>
-		</div>
-		{{ end }}
-
-		</div>
-  </div>
-</div>
-
-<script>
-var lazyLoadInstance = new LazyLoad({
-  // Your custom settings go here
-});
-
-let cards = document.querySelectorAll('.box')
-
-function liveSearch() {
-    let search_query = document.getElementById("searchbox").value;
-
-    //Use innerText if all contents are visible
-    //Use textContent for including hidden elements
-    for (var i = 0; i < cards.length; i++) {
-        if(cards[i].textContent.toLowerCase()
-                .includes(search_query.toLowerCase())) {
-            cards[i].classList.remove("is-hidden");
-        } else {
-            cards[i].classList.add("is-hidden");
-        }
-    }
-}
-
-//A little delay
-let typingTimer;
-let typeInterval = 500;
-let searchInput = document.getElementById('searchbox');
-
-searchInput.addEventListener('keyup', () => {
-    clearTimeout(typingTimer);
-    typingTimer = setTimeout(liveSearch, typeInterval);
-});
-</script>
-
-</div>
-
-<script src="https://cdnjs.cloudflare.com/ajax/libs/flowbite/2.3.0/flowbite.min.js"></script>
-</body>
-</html>
-`
-
-type GalleryModel struct {
-	Name        string   `json:"name" yaml:"name"`
-	URLs        []string `json:"urls" yaml:"urls"`
-	Icon        string   `json:"icon" yaml:"icon"`
-	Description string   `json:"description" yaml:"description"`
-}
-
-func main() {
-	// read the YAML file which contains the models
-
-	f, err := ioutil.ReadFile(os.Args[1])
-	if err != nil {
-		fmt.Println("Error reading file:", err)
-		return
-	}
-
-	models := []*GalleryModel{}
-	err = yaml.Unmarshal(f, &models)
-	if err != nil {
-		// write to stderr
-		os.Stderr.WriteString("Error unmarshaling YAML: " + err.Error() + "\n")
-		return
-	}
-
-	// Ensure that all arbitrary text content is sanitized before display
-	for i, m := range models {
-		models[i].Name = bluemonday.StrictPolicy().Sanitize(m.Name)
-		models[i].Description = bluemonday.StrictPolicy().Sanitize(m.Description)
-	}
-
-	// render the template
-	data := struct {
-		Models          []*GalleryModel
-		AvailableModels int
-	}{
-		Models:          models,
-		AvailableModels: len(models),
-	}
-	tmpl := template.Must(template.New("modelPage").Parse(modelPageTemplate))
-
-	err = tmpl.Execute(os.Stdout, data)
-	if err != nil {
-		fmt.Println("Error executing template:", err)
-		return
-	}
-}

.github/dependabot.yml

@@ -1,135 +0,0 @@
-# https://docs.github.com/en/code-security/dependabot/dependabot-version-updates/configuration-options-for-the-dependabot.yml-file
-version: 2
-updates:
-  - package-ecosystem: "gitsubmodule"
-    directory: "/"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "gomod"
-    directory: "/"
-    schedule:
-      interval: "weekly"
-    ignore:
-    - dependency-name: "github.com/mudler/LocalAI/pkg/grpc/proto"
-  - package-ecosystem: "github-actions"
-    # Workflow files stored in the default location of `.github/workflows`. (You don't need to specify `/.github/workflows` for `directory`. You can use `directory: "/"`.)
-    directory: "/"
-    schedule:
-      # Check for updates to GitHub Actions every weekday
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    # Workflow files stored in the default location of `.github/workflows`. (You don't need to specify `/.github/workflows` for `directory`. You can use `directory: "/"`.)
-    directory: "/"
-    schedule:
-      # Check for updates to GitHub Actions every weekday
-      interval: "weekly"
-  - package-ecosystem: "docker"
-    # Workflow files stored in the default location of `.github/workflows`. (You don't need to specify `/.github/workflows` for `directory`. You can use `directory: "/"`.)
-    directory: "/"
-    schedule:
-      # Check for updates to GitHub Actions every weekday
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/autogptq"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/bark"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/common/template"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/coqui"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/diffusers"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/exllama"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/exllama2"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/mamba"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/openvoice"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/parler-tts"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/rerankers"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/sentencetransformers"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/transformers"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/transformers-musicgen"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/vall-e-x"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/backend/python/vllm"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/examples/chainlit"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/examples/functions"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/examples/langchain/langchainpy-localai-example"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/examples/langchain-chroma"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "pip"
-    directory: "/examples/streamlit-bot"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "docker"
-    directory: "/examples/k8sgpt"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "docker"
-    directory: "/examples/kubernetes"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "docker"
-    directory: "/examples/langchain"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "gomod"
-    directory: "/examples/semantic-todo"
-    schedule:
-      interval: "weekly"
-  - package-ecosystem: "docker"
-    directory: "/examples/telegram-bot"
-    schedule:
-      interval: "weekly"

.github/labeler.yml

@@ -1,29 +0,0 @@
-enhancements:
- - head-branch: ['^feature', 'feature']
-
-dependencies:
-- any:
-  - changed-files:
-    - any-glob-to-any-file: 'Makefile'
-
-kind/documentation:
-- any:
-  - changed-files:
-    - any-glob-to-any-file: 'docs/*'
-  - changed-files:
-    - any-glob-to-any-file: '*.md'
-
-area/ai-model:
-- any:
-  - changed-files:
-    - any-glob-to-any-file: 'gallery/*'
-
-examples:
-- any:
-  - changed-files:
-    - any-glob-to-any-file: 'examples/*'
-
-ci:
-- any:
-  - changed-files:
-    - any-glob-to-any-file: '.github/*'

.github/release.yml

@@ -12,26 +12,13 @@ changelog:
     - title: "Bug fixes :bug:"
       labels:
         - bug
-        - regression
-    - title: "🖧 P2P area"
-      labels:
-         - area/p2p
     - title: Exciting New Features 🎉
       labels:
         - Semver-Minor
         - enhancement
-        - ux
-        - roadmap
-    - title: 🧠 Models
-      labels:
-        - area/ai-model
-    - title: 📖 Documentation and examples
-      labels:
-        - kind/documentation
-        - examples
     - title: 👒 Dependencies
       labels:
         - dependencies
     - title: Other Changes
       labels:
-        - "*"
+        - "*"

.github/workflows/bump_deps.yaml

@@ -9,17 +9,32 @@ jobs:
       fail-fast: false
       matrix:
         include:
+          - 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"
+          - repository: "donomii/go-rwkv.cpp"
+            variable: "RWKV_VERSION"
+            branch: "main"
           - repository: "ggerganov/whisper.cpp"
             variable: "WHISPER_CPP_VERSION"
             branch: "master"
-          - repository: "PABannier/bark.cpp"
-            variable: "BARKCPP_VERSION"
+          - repository: "go-skynet/go-bert.cpp"
+            variable: "BERT_VERSION"
+            branch: "master"
+          - repository: "go-skynet/bloomz.cpp"
+            variable: "BLOOMZ_VERSION"
             branch: "main"
-          - repository: "leejet/stable-diffusion.cpp"
-            variable: "STABLEDIFFUSION_GGML_VERSION"
+          - repository: "nomic-ai/gpt4all"
+            variable: "GPT4ALL_VERSION"
+            branch: "main"
+          - repository: "mudler/go-ggllm.cpp"
+            variable: "GOGGLLM_VERSION"
             branch: "master"
           - repository: "mudler/go-stable-diffusion"
             variable: "STABLEDIFFUSION_VERSION"
@@ -31,30 +46,17 @@ jobs:
     steps:
       - uses: actions/checkout@v4
       - name: Bump dependencies 🔧
-        id: bump
         run: |
           bash .github/bump_deps.sh ${{ matrix.repository }} ${{ matrix.branch }} ${{ matrix.variable }}
-          {
-            echo 'message<<EOF'
-            cat "${{ matrix.variable }}_message.txt"
-            echo EOF
-          } >> "$GITHUB_OUTPUT"
-          {
-            echo 'commit<<EOF'
-            cat "${{ matrix.variable }}_commit.txt"
-            echo EOF
-          } >> "$GITHUB_OUTPUT"
-          rm -rfv ${{ matrix.variable }}_message.txt
-          rm -rfv ${{ matrix.variable }}_commit.txt
       - name: Create Pull Request
-        uses: peter-evans/create-pull-request@v7
+        uses: peter-evans/create-pull-request@v5
         with:
           token: ${{ secrets.UPDATE_BOT_TOKEN }}
           push-to-fork: ci-forks/LocalAI
           commit-message: ':arrow_up: Update ${{ matrix.repository }}'
-          title: 'chore: :arrow_up: Update ${{ matrix.repository }} to `${{ steps.bump.outputs.commit }}`'
+          title: ':arrow_up: Update ${{ matrix.repository }}'
           branch: "update/${{ matrix.variable }}"
-          body: ${{ steps.bump.outputs.message }}
+          body: Bump of ${{ matrix.repository }} version
           signoff: true
 
 

.github/workflows/bump_docs.yaml

@@ -17,12 +17,12 @@ jobs:
         run: |
           bash .github/bump_docs.sh ${{ matrix.repository }}
       - name: Create Pull Request
-        uses: peter-evans/create-pull-request@v7
+        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: 'docs: :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/checksum_checker.yaml

@@ -1,47 +0,0 @@
-name: Check if checksums are up-to-date
-on:
-  schedule:
-    - cron: 0 20 * * *
-  workflow_dispatch:
-jobs:
-  checksum_check:
-    runs-on: arc-runner-set
-    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
-      - uses: actions/checkout@v4
-      - name: Install dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y pip wget
-          sudo pip install --upgrade pip
-          pip install huggingface_hub
-      - name: 'Setup yq'
-        uses: dcarbone/install-yq-action@v1.3.1
-        with:
-          version: 'v4.44.2'
-          download-compressed: true
-          force: true
-
-      - name: Checksum checker 🔧
-        run: |
-          export HF_HOME=/hf_cache
-          sudo mkdir /hf_cache
-          sudo chmod 777 /hf_cache
-          bash .github/checksum_checker.sh gallery/index.yaml
-      - name: Create Pull Request
-        uses: peter-evans/create-pull-request@v7
-        with:
-          token: ${{ secrets.UPDATE_BOT_TOKEN }}
-          push-to-fork: ci-forks/LocalAI
-          commit-message: ':arrow_up: Checksum updates in gallery/index.yaml'
-          title: 'chore(model-gallery): :arrow_up: update checksum'
-          branch: "update/checksum"
-          body: Updating checksums in gallery/index.yaml
-          signoff: true

.github/workflows/dependabot_auto.yml

@@ -1,43 +0,0 @@
-name: Dependabot auto-merge
-on:
-- pull_request_target
-
-permissions:
-  contents: write
-  pull-requests: write
-  packages: read
-
-jobs:
-  dependabot:
-    runs-on: ubuntu-latest
-    if: ${{ github.actor == 'dependabot[bot]' }}
-    steps:
-      - name: Dependabot metadata
-        id: metadata
-        uses: dependabot/fetch-metadata@v2.2.0
-        with:
-          github-token: "${{ secrets.GITHUB_TOKEN }}"
-          skip-commit-verification: true
-
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Approve a PR if not already approved
-        run: |
-          gh pr checkout "$PR_URL"
-            if [ "$(gh pr status --json reviewDecision -q .currentBranch.reviewDecision)" != "APPROVED" ];
-          then
-            gh pr review --approve "$PR_URL"
-          else
-            echo "PR already approved.";
-          fi
-        env:
-          PR_URL: ${{github.event.pull_request.html_url}}
-          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
-
-      - name: Enable auto-merge for Dependabot PRs
-        if: ${{ contains(github.event.pull_request.title, 'bump')}}
-        run: gh pr merge --auto --squash "$PR_URL"
-        env:
-          PR_URL: ${{github.event.pull_request.html_url}}
-          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}

.github/workflows/deploy-explorer.yaml

@@ -1,64 +0,0 @@
-name: Explorer deployment
-
-on:
-  push:
-    branches:
-      - master
-    tags:
-      - 'v*'
-
-concurrency:
-  group: ci-deploy-${{ github.head_ref || github.ref }}-${{ github.repository }}
-
-jobs:
-  build-linux:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Clone
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-      - uses: actions/setup-go@v5
-        with:
-          go-version: '1.21.x'
-          cache: false
-      - name: Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y wget curl build-essential ffmpeg protobuf-compiler ccache upx-ucl gawk cmake libgmock-dev
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-          make protogen-go
-      - name: Build api
-        run: |
-          CGO_ENABLED=0 make build-api
-      - name: rm
-        uses: appleboy/ssh-action@v1.2.0
-        with:
-            host: ${{ secrets.EXPLORER_SSH_HOST }}
-            username: ${{ secrets.EXPLORER_SSH_USERNAME }}
-            key: ${{ secrets.EXPLORER_SSH_KEY }}
-            port: ${{ secrets.EXPLORER_SSH_PORT }}
-            script: |
-                sudo rm -rf local-ai/ || true
-      - name: copy file via ssh
-        uses: appleboy/scp-action@v0.1.7
-        with:
-            host: ${{ secrets.EXPLORER_SSH_HOST }}
-            username: ${{ secrets.EXPLORER_SSH_USERNAME }}
-            key: ${{ secrets.EXPLORER_SSH_KEY }}
-            port: ${{ secrets.EXPLORER_SSH_PORT }}
-            source: "local-ai"
-            overwrite: true
-            rm: true
-            target: ./local-ai
-      - name: restarting
-        uses: appleboy/ssh-action@v1.2.0
-        with:
-            host: ${{ secrets.EXPLORER_SSH_HOST }}
-            username: ${{ secrets.EXPLORER_SSH_USERNAME }}
-            key: ${{ secrets.EXPLORER_SSH_KEY }}
-            port: ${{ secrets.EXPLORER_SSH_PORT }}
-            script: |
-                sudo cp -rfv local-ai/local-ai /usr/bin/local-ai
-                sudo systemctl restart local-ai

.github/workflows/disabled/comment-pr.yaml

@@ -1,83 +0,0 @@
-name: Comment PRs
-on:
-  pull_request_target:
-
-jobs:
-  comment-pr:
-    env:
-        MODEL_NAME: hermes-2-theta-llama-3-8b
-    runs-on: ubuntu-latest
-    steps:
-    - name: Checkout code
-      uses: actions/checkout@v3
-      with:
-        ref: "${{ github.event.pull_request.merge_commit_sha }}"
-        fetch-depth: 0 # needed to checkout all branches for this Action to work
-    - uses: mudler/localai-github-action@v1
-      with:
-        model: 'hermes-2-theta-llama-3-8b' # Any from models.localai.io, or from huggingface.com with: "huggingface://<repository>/file"
-      # Check the PR diff using the current branch and the base branch of the PR
-    - uses: GrantBirki/git-diff-action@v2.7.0
-      id: git-diff-action
-      with:
-            json_diff_file_output: diff.json
-            raw_diff_file_output: diff.txt
-            file_output_only: "true"
-            base_branch: ${{ github.event.pull_request.base.sha }}
-    - name: Show diff
-      env:
-        DIFF: ${{ steps.git-diff-action.outputs.raw-diff-path }}
-      run: |
-            cat $DIFF
-    - name: Summarize
-      env:
-        DIFF: ${{ steps.git-diff-action.outputs.raw-diff-path }}
-      id: summarize
-      run: |
-            input="$(cat $DIFF)"
-
-            # Define the LocalAI API endpoint
-            API_URL="http://localhost:8080/chat/completions"
-
-            # Create a JSON payload using jq to handle special characters
-            json_payload=$(jq -n --arg input "$input" '{
-            model: "'$MODEL_NAME'",
-            messages: [
-                {
-                role: "system",
-                content: "You are LocalAI-bot in Github that helps understanding PRs and assess complexity. Explain what has changed in this PR diff and why"
-                },
-                {
-                role: "user",
-                content: $input
-                }
-            ]
-            }')
-
-            # Send the request to LocalAI
-            response=$(curl -s -X POST $API_URL \
-            -H "Content-Type: application/json" \
-            -d "$json_payload")
-
-            # Extract the summary from the response
-            summary="$(echo $response | jq -r '.choices[0].message.content')"
-
-            # Print the summary
-            #  -H "Authorization: Bearer $API_KEY" \
-            echo "Summary:"
-            echo "$summary"
-            echo "payload sent"
-            echo "$json_payload"
-            {
-                echo 'message<<EOF'
-                echo "$summary"
-                echo EOF
-              } >> "$GITHUB_OUTPUT"
-            docker logs --tail 10 local-ai
-    - uses: mshick/add-pr-comment@v2
-      if: always()
-      with:
-          repo-token: ${{ secrets.UPDATE_BOT_TOKEN }}
-          message: ${{ steps.summarize.outputs.message }}
-          message-failure: |
-            Uh oh! Could not analyze this PR, maybe it's too big?

.github/workflows/generate_grpc_cache.yaml

@@ -1,94 +0,0 @@
-name: 'generate and publish GRPC docker caches'
-
-on:
-  workflow_dispatch:
-  push:
-    branches:
-      - master
-
-concurrency:
-  group: grpc-cache-${{ github.head_ref || github.ref }}-${{ github.repository }}
-  cancel-in-progress: true
-
-jobs:
-  generate_caches:
-    strategy:
-      matrix:
-        include:
-          - grpc-base-image: ubuntu:22.04
-            runs-on: 'ubuntu-latest'
-            platforms: 'linux/amd64,linux/arm64'
-    runs-on: ${{matrix.runs-on}}
-    steps:
-      - name: Release space from worker
-        if: matrix.runs-on == 'ubuntu-latest'
-        run: |
-          echo "Listing top largest packages"
-          pkgs=$(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
-          sudo rm -rf /usr/share/dotnet || true
-          sudo rm -rf /opt/ghc || true
-          sudo rm -rf "/usr/local/share/boost" || true
-          sudo rm -rf "$AGENT_TOOLSDIRECTORY" || true
-          df -h
-
-      - 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: Checkout
-        uses: actions/checkout@v4
-
-      - name: Cache GRPC
-        uses: docker/build-push-action@v6
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          # The build-args MUST be an EXACT match between the image cache and other workflow steps that want to use that cache.
-          # This means that even the MAKEFLAGS have to be an EXACT match.
-          # If the build-args are not an EXACT match, it will result in a cache miss, which will require GRPC to be built from scratch.
-          build-args: |
-            GRPC_BASE_IMAGE=${{ matrix.grpc-base-image }}
-            GRPC_MAKEFLAGS=--jobs=4 --output-sync=target
-            GRPC_VERSION=v1.65.0
-          context: .
-          file: ./Dockerfile
-          cache-to: type=gha,ignore-error=true
-          cache-from: type=gha
-          target: grpc
-          platforms: ${{ matrix.platforms }}
-          push: false

.github/workflows/generate_intel_image.yaml

@@ -1,59 +0,0 @@
-name: 'generate and publish intel docker caches'
-
-on:
-  workflow_dispatch:
-  push:
-    branches:
-      - master
-
-concurrency:
-  group: intel-cache-${{ github.head_ref || github.ref }}-${{ github.repository }}
-  cancel-in-progress: true
-
-jobs:
-  generate_caches:
-    strategy:
-      matrix:
-        include:
-          - base-image: intel/oneapi-basekit:2025.0.0-0-devel-ubuntu22.04
-            runs-on: 'ubuntu-latest'
-            platforms: 'linux/amd64'
-    runs-on: ${{matrix.runs-on}}
-    steps:
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@master
-        with:
-          platforms: all
-      - name: Login to DockerHub
-        if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.DOCKERHUB_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_PASSWORD }}
-
-      - name: Login to quay
-        if: github.event_name != 'pull_request'
-        uses: docker/login-action@v3
-        with:
-          registry: quay.io
-          username: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
-          password: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@master
-
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Cache Intel images
-        uses: docker/build-push-action@v6
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          build-args: |
-            BASE_IMAGE=${{ matrix.base-image }}
-          context: .
-          file: ./Dockerfile
-          tags: quay.io/go-skynet/intel-oneapi-base:latest
-          push: true
-          target: intel
-          platforms: ${{ matrix.platforms }}

.github/workflows/image-pr.yml

@@ -22,8 +22,6 @@ jobs:
       platforms: ${{ matrix.platforms }}
       runs-on: ${{ matrix.runs-on }}
       base-image: ${{ matrix.base-image }}
-      grpc-base-image: ${{ matrix.grpc-base-image }}
-      makeflags: ${{ matrix.makeflags }}
     secrets:
       dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
       dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
@@ -32,22 +30,20 @@ jobs:
     strategy:
       # Pushing with all jobs in parallel
       # eats the bandwidth of all the nodes
-      max-parallel: ${{ github.event_name != 'pull_request' && 4 || 8 }}
+      max-parallel: ${{ github.event_name != 'pull_request' && 2 || 4 }}
       matrix:
         include:
-          # This is basically covered by the AIO test
-          # - build-type: ''
-          #   platforms: 'linux/amd64'
-          #   tag-latest: 'false'
-          #   tag-suffix: '-ffmpeg'
-          #   ffmpeg: 'true'
-          #   image-type: 'extras'
-          #   runs-on: 'arc-runner-set'
-          #   base-image: "ubuntu:22.04"
-          #   makeflags: "--jobs=3 --output-sync=target"
+          - build-type: ''
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-ffmpeg'
+            ffmpeg: 'true'
+            image-type: 'extras'
+            runs-on: 'arc-runner-set'
+            base-image: "ubuntu:22.04"
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "0"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda12-ffmpeg'
@@ -55,86 +51,66 @@ jobs:
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=3 --output-sync=target"
-          # - build-type: 'hipblas'
-          #   platforms: 'linux/amd64'
-          #   tag-latest: 'false'
-          #   tag-suffix: '-hipblas'
-          #   ffmpeg: 'false'
-          #   image-type: 'extras'
-          #   base-image: "rocm/dev-ubuntu-22.04:6.1"
-          #   grpc-base-image: "ubuntu:22.04"
-          #   runs-on: 'arc-runner-set'
-          #   makeflags: "--jobs=3 --output-sync=target"
-          # - build-type: 'sycl_f16'
-          #   platforms: 'linux/amd64'
-          #   tag-latest: 'false'
-          #   base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-          #   grpc-base-image: "ubuntu:22.04"
-          #   tag-suffix: 'sycl-f16-ffmpeg'
-          #   ffmpeg: 'true'
-          #   image-type: 'extras'
-          #   runs-on: 'arc-runner-set'
-          #   makeflags: "--jobs=3 --output-sync=target"
-  # 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 }}
-  #     base-image: ${{ matrix.base-image }}
-  #     grpc-base-image: ${{ matrix.grpc-base-image }}
-  #     makeflags: ${{ matrix.makeflags }}
-  #   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'
-          #   base-image: "ubuntu:22.04"
-          #   makeflags: "--jobs=4 --output-sync=target"
-          # - build-type: 'sycl_f16'
-          #   platforms: 'linux/amd64'
-          #   tag-latest: 'false'
-          #   base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-          #   grpc-base-image: "ubuntu:22.04"
-          #   tag-suffix: 'sycl-f16-ffmpeg-core'
-          #   ffmpeg: 'true'
-          #   image-type: 'core'
-          #   runs-on: 'arc-runner-set'
-          #   makeflags: "--jobs=3 --output-sync=target"
-          # - build-type: 'cublas'
-          #   cuda-major-version: "12"
-          #   cuda-minor-version: "0"
-          #   platforms: 'linux/amd64'
-          #   tag-latest: 'false'
-          #   tag-suffix: '-cublas-cuda12-ffmpeg-core'
-          #   ffmpeg: 'true'
-          #   image-type: 'core'
-          #   runs-on: 'ubuntu-latest'
-          #   base-image: "ubuntu:22.04"
-          #   makeflags: "--jobs=4 --output-sync=target"
-          # - build-type: 'vulkan'
-          #   platforms: 'linux/amd64'
-          #   tag-latest: 'false'
-          #   tag-suffix: '-vulkan-ffmpeg-core'
-          #   ffmpeg: 'true'
-          #   image-type: 'core'
-          #   runs-on: 'ubuntu-latest'
-          #   base-image: "ubuntu:22.04"
-          #   makeflags: "--jobs=4 --output-sync=target"
+          - build-type: 'hipblas'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-hipblas'
+            ffmpeg: 'false'
+            image-type: 'extras'
+            base-image: "rocm/dev-ubuntu-22.04:6.0-complete"
+            runs-on: 'arc-runner-set'
+          - build-type: 'sycl_f16'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
+            tag-suffix: 'sycl-f16-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 }}
+      base-image: ${{ matrix.base-image }}
+    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'
+            base-image: "ubuntu:22.04"
+          - build-type: 'sycl_f16'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
+            tag-suffix: 'sycl-f16-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            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-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:22.04"

.github/workflows/image.yml

@@ -13,78 +13,6 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-  hipblas-jobs:
-    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 }}
-      base-image: ${{ matrix.base-image }}
-      grpc-base-image: ${{ matrix.grpc-base-image }}
-      aio: ${{ matrix.aio }}
-      makeflags: ${{ matrix.makeflags }}
-      latest-image: ${{ matrix.latest-image }}
-      latest-image-aio: ${{ matrix.latest-image-aio }}
-    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: 2
-      matrix:
-        include:
-          - build-type: 'hipblas'
-            platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            tag-suffix: '-hipblas-ffmpeg'
-            ffmpeg: 'true'
-            image-type: 'extras'
-            aio: "-aio-gpu-hipblas"
-            base-image: "rocm/dev-ubuntu-22.04:6.1"
-            grpc-base-image: "ubuntu:22.04"
-            latest-image: 'latest-gpu-hipblas'
-            latest-image-aio: 'latest-aio-gpu-hipblas'
-            runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
-          - build-type: 'hipblas'
-            platforms: 'linux/amd64'
-            tag-latest: 'false'
-            tag-suffix: '-hipblas'
-            ffmpeg: 'false'
-            image-type: 'extras'
-            base-image: "rocm/dev-ubuntu-22.04:6.1"
-            grpc-base-image: "ubuntu:22.04"
-            runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
-          - build-type: 'hipblas'
-            platforms: 'linux/amd64'
-            tag-latest: 'false'
-            tag-suffix: '-hipblas-ffmpeg-core'
-            ffmpeg: 'true'
-            image-type: 'core'
-            base-image: "rocm/dev-ubuntu-22.04:6.1"
-            grpc-base-image: "ubuntu:22.04"
-            runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
-          - build-type: 'hipblas'
-            platforms: 'linux/amd64'
-            tag-latest: 'false'
-            tag-suffix: '-hipblas-core'
-            ffmpeg: 'false'
-            image-type: 'core'
-            base-image: "rocm/dev-ubuntu-22.04:6.1"
-            grpc-base-image: "ubuntu:22.04"
-            runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
   self-hosted-jobs:
     uses: ./.github/workflows/image_build.yml
     with:
@@ -98,11 +26,6 @@ jobs:
       platforms: ${{ matrix.platforms }}
       runs-on: ${{ matrix.runs-on }}
       base-image: ${{ matrix.base-image }}
-      grpc-base-image: ${{ matrix.grpc-base-image }}
-      aio: ${{ matrix.aio }}
-      makeflags: ${{ matrix.makeflags }}
-      latest-image: ${{ matrix.latest-image }}
-      latest-image-aio: ${{ matrix.latest-image-aio }}
     secrets:
       dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
       dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
@@ -111,7 +34,7 @@ jobs:
     strategy:
       # Pushing with all jobs in parallel
       # eats the bandwidth of all the nodes
-      max-parallel: ${{ github.event_name != 'pull_request' && 5 || 8 }}
+      max-parallel: ${{ github.event_name != 'pull_request' && 2 || 4 }}
       matrix:
         include:
           # Extra images
@@ -124,16 +47,14 @@ jobs:
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: ''
             platforms: 'linux/amd64'
-            tag-latest: 'auto'
+            tag-latest: 'false'
             tag-suffix: '-ffmpeg'
             ffmpeg: 'true'
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'cublas'
             cuda-major-version: "11"
             cuda-minor-version: "7"
@@ -144,10 +65,9 @@ jobs:
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "0"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda12'
@@ -155,35 +75,26 @@ jobs:
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'cublas'
             cuda-major-version: "11"
             cuda-minor-version: "7"
             platforms: 'linux/amd64'
-            tag-latest: 'auto'
+            tag-latest: 'false'
             tag-suffix: '-cublas-cuda11-ffmpeg'
             ffmpeg: 'true'
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            aio: "-aio-gpu-nvidia-cuda-11"
-            latest-image: 'latest-gpu-nvidia-cuda-11'
-            latest-image-aio: 'latest-aio-gpu-nvidia-cuda-11'
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "0"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
-            tag-latest: 'auto'
+            tag-latest: 'false'
             tag-suffix: '-cublas-cuda12-ffmpeg'
             ffmpeg: 'true'
             image-type: 'extras'
             runs-on: 'arc-runner-set'
             base-image: "ubuntu:22.04"
-            aio: "-aio-gpu-nvidia-cuda-12"
-            latest-image: 'latest-gpu-nvidia-cuda-12'
-            latest-image-aio: 'latest-aio-gpu-nvidia-cuda-12'
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: ''
             #platforms: 'linux/amd64,linux/arm64'
             platforms: 'linux/amd64'
@@ -193,75 +104,88 @@ jobs:
             image-type: 'extras'
             base-image: "ubuntu:22.04"
             runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
+          - build-type: 'hipblas'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-hipblas-ffmpeg'
+            ffmpeg: 'true'
+            image-type: 'extras'
+            base-image: "rocm/dev-ubuntu-22.04:6.0-complete"
+            runs-on: 'arc-runner-set'
+          - build-type: 'hipblas'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-hipblas'
+            ffmpeg: 'false'
+            image-type: 'extras'
+            base-image: "rocm/dev-ubuntu-22.04:6.0-complete"
+            runs-on: 'arc-runner-set'
           - build-type: 'sycl_f16'
             platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-            grpc-base-image: "ubuntu:22.04"
+            tag-latest: 'false'
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
             tag-suffix: '-sycl-f16-ffmpeg'
             ffmpeg: 'true'
             image-type: 'extras'
             runs-on: 'arc-runner-set'
-            aio: "-aio-gpu-intel-f16"
-            latest-image: 'latest-gpu-intel-f16'
-            latest-image-aio: 'latest-aio-gpu-intel-f16'
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'sycl_f32'
             platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-            grpc-base-image: "ubuntu:22.04"
+            tag-latest: 'false'
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
             tag-suffix: '-sycl-f32-ffmpeg'
             ffmpeg: 'true'
             image-type: 'extras'
             runs-on: 'arc-runner-set'
-            aio: "-aio-gpu-intel-f32"
-            latest-image: 'latest-gpu-intel-f32'
-            latest-image-aio: 'latest-aio-gpu-intel-f32'
-            makeflags: "--jobs=3 --output-sync=target"
           # Core images
           - build-type: 'sycl_f16'
             platforms: 'linux/amd64'
             tag-latest: 'false'
-            base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-            grpc-base-image: "ubuntu:22.04"
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
             tag-suffix: '-sycl-f16-core'
             ffmpeg: 'false'
             image-type: 'core'
             runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'sycl_f32'
             platforms: 'linux/amd64'
             tag-latest: 'false'
-            base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-            grpc-base-image: "ubuntu:22.04"
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
             tag-suffix: '-sycl-f32-core'
             ffmpeg: 'false'
             image-type: 'core'
             runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'sycl_f16'
             platforms: 'linux/amd64'
             tag-latest: 'false'
-            base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-            grpc-base-image: "ubuntu:22.04"
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
             tag-suffix: '-sycl-f16-ffmpeg-core'
             ffmpeg: 'true'
             image-type: 'core'
             runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
           - build-type: 'sycl_f32'
             platforms: 'linux/amd64'
             tag-latest: 'false'
-            base-image: "quay.io/go-skynet/intel-oneapi-base:latest"
-            grpc-base-image: "ubuntu:22.04"
+            base-image: "intel/oneapi-basekit:2024.0.1-devel-ubuntu22.04"
             tag-suffix: '-sycl-f32-ffmpeg-core'
             ffmpeg: 'true'
             image-type: 'core'
             runs-on: 'arc-runner-set'
-            makeflags: "--jobs=3 --output-sync=target"
-
+          - build-type: 'hipblas'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-hipblas-ffmpeg-core'
+            ffmpeg: 'true'
+            image-type: 'core'
+            base-image: "rocm/dev-ubuntu-22.04:6.0-complete"
+            runs-on: 'arc-runner-set'
+          - build-type: 'hipblas'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
+            tag-suffix: '-hipblas-core'
+            ffmpeg: 'false'
+            image-type: 'core'
+            base-image: "rocm/dev-ubuntu-22.04:6.0-complete"
+            runs-on: 'arc-runner-set'
+  
   core-image-build:
     uses: ./.github/workflows/image_build.yml
     with:
@@ -274,33 +198,23 @@ jobs:
       cuda-minor-version: ${{ matrix.cuda-minor-version }}
       platforms: ${{ matrix.platforms }}
       runs-on: ${{ matrix.runs-on }}
-      aio: ${{ matrix.aio }}
       base-image: ${{ matrix.base-image }}
-      grpc-base-image: ${{ matrix.grpc-base-image }}
-      makeflags: ${{ matrix.makeflags }}
-      latest-image: ${{ matrix.latest-image }}
-      latest-image-aio: ${{ matrix.latest-image-aio }}
     secrets:
       dockerUsername: ${{ secrets.DOCKERHUB_USERNAME }}
       dockerPassword: ${{ secrets.DOCKERHUB_PASSWORD }}
       quayUsername: ${{ secrets.LOCALAI_REGISTRY_USERNAME }}
       quayPassword: ${{ secrets.LOCALAI_REGISTRY_PASSWORD }}
     strategy:
-      max-parallel: ${{ github.event_name != 'pull_request' && 2 || 4 }}
       matrix:
         include:
           - build-type: ''
-            platforms: 'linux/amd64,linux/arm64'
-            tag-latest: 'auto'
+            platforms: 'linux/amd64'
+            tag-latest: 'false'
             tag-suffix: '-ffmpeg-core'
             ffmpeg: 'true'
             image-type: 'core'
             base-image: "ubuntu:22.04"
-            runs-on: 'arc-runner-set'
-            aio: "-aio-cpu"
-            latest-image: 'latest-cpu'
-            latest-image-aio: 'latest-aio-cpu'
-            makeflags: "--jobs=4 --output-sync=target"
+            runs-on: 'ubuntu-latest'
           - build-type: 'cublas'
             cuda-major-version: "11"
             cuda-minor-version: "7"
@@ -310,19 +224,17 @@ jobs:
             ffmpeg: ''
             image-type: 'core'
             base-image: "ubuntu:22.04"
-            runs-on: 'arc-runner-set'
-            makeflags: "--jobs=4 --output-sync=target"
+            runs-on: 'ubuntu-latest'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "0"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda12-core'
             ffmpeg: ''
             image-type: 'core'
             base-image: "ubuntu:22.04"
-            runs-on: 'arc-runner-set'
-            makeflags: "--jobs=4 --output-sync=target"
+            runs-on: 'ubuntu-latest'
           - build-type: 'cublas'
             cuda-major-version: "11"
             cuda-minor-version: "7"
@@ -331,27 +243,15 @@ jobs:
             tag-suffix: '-cublas-cuda11-ffmpeg-core'
             ffmpeg: 'true'
             image-type: 'core'
-            runs-on: 'arc-runner-set'
+            runs-on: 'ubuntu-latest'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=4 --output-sync=target"
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "0"
+            cuda-minor-version: "1"
             platforms: 'linux/amd64'
             tag-latest: 'false'
             tag-suffix: '-cublas-cuda12-ffmpeg-core'
             ffmpeg: 'true'
             image-type: 'core'
-            runs-on: 'arc-runner-set'
+            runs-on: 'ubuntu-latest'
             base-image: "ubuntu:22.04"
-            makeflags: "--jobs=4 --output-sync=target"
-          - build-type: 'vulkan'
-            platforms: 'linux/amd64'
-            tag-latest: 'false'
-            tag-suffix: '-vulkan-ffmpeg-core'
-            latest-image: 'latest-vulkan-ffmpeg-core'
-            ffmpeg: 'true'
-            image-type: 'core'
-            runs-on: 'arc-runner-set'
-            base-image: "ubuntu:22.04"
-            makeflags: "--jobs=4 --output-sync=target"

.github/workflows/image_build.yml

@@ -6,10 +6,6 @@ on:
     inputs:
       base-image:
         description: 'Base image'
-        required: true
-        type: string
-      grpc-base-image:
-        description: 'GRPC Base image, must be a compatible image with base-image'
         required: false
         default: ''
         type: string
@@ -19,11 +15,11 @@ on:
         type: string
       cuda-major-version:
         description: 'CUDA major version'
-        default: "12"
+        default: "11"
         type: string
       cuda-minor-version:
         description: 'CUDA minor version'
-        default: "4"
+        default: "7"
         type: string
       platforms:
         description: 'Platforms'
@@ -33,14 +29,6 @@ on:
         description: 'Tag latest'
         default: ''
         type: string
-      latest-image:
-          description: 'Tag latest'
-          default: ''
-          type: string
-      latest-image-aio:
-          description: 'Tag latest'
-          default: ''
-          type: string
       tag-suffix:
         description: 'Tag suffix'
         default: ''
@@ -58,16 +46,6 @@ on:
         required: true
         default: ''
         type: string
-      makeflags:
-        description: 'Make Flags'
-        required: false
-        default: '--jobs=4 --output-sync=target'
-        type: string
-      aio:
-        description: 'AIO Image Name'
-        required: false
-        default: ''
-        type: string
     secrets:
       dockerUsername:
         required: true
@@ -91,7 +69,6 @@ jobs:
           && sudo apt-get install -y git
       - name: Checkout
         uses: actions/checkout@v4
-
       - name: Release space from worker
         if: inputs.runs-on == 'ubuntu-latest'
         run: |
@@ -133,10 +110,8 @@ jobs:
           sudo rm -rf "/usr/local/share/boost" || true
           sudo rm -rf "$AGENT_TOOLSDIRECTORY" || true
           df -h
-
       - name: Docker meta
         id: meta
-        if: github.event_name != 'pull_request'
         uses: docker/metadata-action@v5
         with:
           images: |
@@ -149,46 +124,6 @@ jobs:
           flavor: |
             latest=${{ inputs.tag-latest }}
             suffix=${{ inputs.tag-suffix }}
-      - name: Docker meta for PR
-        id: meta_pull_request
-        if: github.event_name == 'pull_request'
-        uses: docker/metadata-action@v5
-        with:
-          images: |
-            ttl.sh/localai-ci-pr-${{ github.event.number }}
-          tags: |
-            type=ref,event=branch
-            type=semver,pattern={{raw}}
-            type=sha
-          flavor: |
-            latest=${{ inputs.tag-latest }}
-            suffix=${{ inputs.tag-suffix }}
-      - name: Docker meta AIO (quay.io)
-        if: inputs.aio != ''
-        id: meta_aio
-        uses: docker/metadata-action@v5
-        with:
-          images: |
-            quay.io/go-skynet/local-ai
-          tags: |
-            type=ref,event=branch
-            type=semver,pattern={{raw}}
-          flavor: |
-            latest=${{ inputs.tag-latest }}
-            suffix=${{ inputs.aio }}
-
-      - name: Docker meta AIO (dockerhub)
-        if: inputs.aio != ''
-        id: meta_aio_dockerhub
-        uses: docker/metadata-action@v5
-        with:
-          images: |
-            localai/localai
-          tags: |
-            type=ref,event=branch
-            type=semver,pattern={{raw}}
-          flavor: |
-            suffix=${{ inputs.aio }}
 
       - name: Set up QEMU
         uses: docker/setup-qemu-action@master
@@ -215,14 +150,9 @@ jobs:
           password: ${{ secrets.quayPassword }}
 
       - name: Build and push
-        uses: docker/build-push-action@v6
-        if: github.event_name != 'pull_request'
+        uses: docker/build-push-action@v5
         with:
           builder: ${{ steps.buildx.outputs.name }}
-          # The build-args MUST be an EXACT match between the image cache and other workflow steps that want to use that cache.
-          # This means that even the MAKEFLAGS have to be an EXACT match.
-          # If the build-args are not an EXACT match, it will result in a cache miss, which will require GRPC to be built from scratch.
-          # This is why some build args like GRPC_VERSION and MAKEFLAGS are hardcoded
           build-args: |
             BUILD_TYPE=${{ inputs.build-type }}
             CUDA_MAJOR_VERSION=${{ inputs.cuda-major-version }}
@@ -230,106 +160,12 @@ jobs:
             FFMPEG=${{ inputs.ffmpeg }}
             IMAGE_TYPE=${{ inputs.image-type }}
             BASE_IMAGE=${{ inputs.base-image }}
-            GRPC_BASE_IMAGE=${{ inputs.grpc-base-image || inputs.base-image }}
-            GRPC_MAKEFLAGS=--jobs=4 --output-sync=target
-            GRPC_VERSION=v1.65.0
-            MAKEFLAGS=${{ inputs.makeflags }}
           context: .
           file: ./Dockerfile
-          cache-from: type=gha
           platforms: ${{ inputs.platforms }}
           push: ${{ github.event_name != 'pull_request' }}
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
-### Start testing image
-      - name: Build and push
-        uses: docker/build-push-action@v6
-        if: github.event_name == 'pull_request'
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          # The build-args MUST be an EXACT match between the image cache and other workflow steps that want to use that cache.
-          # This means that even the MAKEFLAGS have to be an EXACT match.
-          # If the build-args are not an EXACT match, it will result in a cache miss, which will require GRPC to be built from scratch.
-          # This is why some build args like GRPC_VERSION and MAKEFLAGS are hardcoded
-          build-args: |
-            BUILD_TYPE=${{ 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 }}
-            BASE_IMAGE=${{ inputs.base-image }}
-            GRPC_BASE_IMAGE=${{ inputs.grpc-base-image || inputs.base-image }}
-            GRPC_MAKEFLAGS=--jobs=4 --output-sync=target
-            GRPC_VERSION=v1.65.0
-            MAKEFLAGS=${{ inputs.makeflags }}
-          context: .
-          file: ./Dockerfile
-          cache-from: type=gha
-          platforms: ${{ inputs.platforms }}
-          push: true
-          tags: ${{ steps.meta_pull_request.outputs.tags }}
-          labels: ${{ steps.meta_pull_request.outputs.labels }}
-      - name: Testing image
-        if: github.event_name == 'pull_request'
-        run: |
-          echo "Image is available at ttl.sh/localai-ci-pr-${{ github.event.number }}:${{ steps.meta_pull_request.outputs.version }}" >> $GITHUB_STEP_SUMMARY
-## End testing image
-      - name: Build and push AIO image
-        if: inputs.aio != ''
-        uses: docker/build-push-action@v6
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          build-args: |
-            BASE_IMAGE=quay.io/go-skynet/local-ai:${{ steps.meta.outputs.version }}
-            MAKEFLAGS=${{ inputs.makeflags }}
-          context: .
-          file: ./Dockerfile.aio
-          platforms: ${{ inputs.platforms }}
-          push: ${{ github.event_name != 'pull_request' }}
-          tags: ${{ steps.meta_aio.outputs.tags }}
-          labels: ${{ steps.meta_aio.outputs.labels }}
-
-      - name: Build and push AIO image (dockerhub)
-        if: inputs.aio != ''
-        uses: docker/build-push-action@v6
-        with:
-          builder: ${{ steps.buildx.outputs.name }}
-          build-args: |
-            BASE_IMAGE=localai/localai:${{ steps.meta.outputs.version }}
-            MAKEFLAGS=${{ inputs.makeflags }}
-          context: .
-          file: ./Dockerfile.aio
-          platforms: ${{ inputs.platforms }}
-          push: ${{ github.event_name != 'pull_request' }}
-          tags: ${{ steps.meta_aio_dockerhub.outputs.tags }}
-          labels: ${{ steps.meta_aio_dockerhub.outputs.labels }}
-
-      - name: Latest tag
-        # run this on branches, when it is a tag and there is a latest-image defined
-        if: github.event_name != 'pull_request' && inputs.latest-image != ''  && github.ref_type == 'tag'
-        run: |
-          docker pull localai/localai:${{ steps.meta.outputs.version }}
-          docker tag localai/localai:${{ steps.meta.outputs.version }} localai/localai:${{ inputs.latest-image }}
-          docker push localai/localai:${{ inputs.latest-image }}
-          docker pull quay.io/go-skynet/local-ai:${{ steps.meta.outputs.version }}
-          docker tag quay.io/go-skynet/local-ai:${{ steps.meta.outputs.version }} quay.io/go-skynet/local-ai:${{ inputs.latest-image }}
-          docker push quay.io/go-skynet/local-ai:${{ inputs.latest-image }}
-      - name: Latest AIO tag
-        # run this on branches, when it is a tag and there is a latest-image defined
-        if: github.event_name != 'pull_request' && inputs.latest-image-aio != ''  && github.ref_type == 'tag'
-        run: |
-          docker pull localai/localai:${{ steps.meta_aio_dockerhub.outputs.version }}
-          docker tag localai/localai:${{ steps.meta_aio_dockerhub.outputs.version }} localai/localai:${{ inputs.latest-image-aio }}
-          docker push localai/localai:${{ inputs.latest-image-aio }}
-          docker pull quay.io/go-skynet/local-ai:${{ steps.meta_aio.outputs.version }}
-          docker tag quay.io/go-skynet/local-ai:${{ steps.meta_aio.outputs.version }} quay.io/go-skynet/local-ai:${{ inputs.latest-image-aio }}
-          docker push quay.io/go-skynet/local-ai:${{ inputs.latest-image-aio }}
-
       - name: job summary
         run: |
           echo "Built image: ${{ steps.meta.outputs.labels }}" >> $GITHUB_STEP_SUMMARY
-
-      - name: job summary(AIO)
-        if: inputs.aio != ''
-        run: |
-          echo "Built image: ${{ steps.meta_aio.outputs.labels }}" >> $GITHUB_STEP_SUMMARY

.github/workflows/labeler.yml

@@ -1,12 +0,0 @@
-name: "Pull Request Labeler"
-on:
-- pull_request_target
-
-jobs:
-  labeler:
-    permissions:
-      contents: read
-      pull-requests: write
-    runs-on: ubuntu-latest
-    steps:
-    - uses: actions/labeler@v5

.github/workflows/localaibot_automerge.yml

@@ -1,35 +0,0 @@
-name: LocalAI-bot auto-merge
-on:
-- pull_request_target
-
-permissions:
-  contents: write
-  pull-requests: write
-  packages: read
-
-jobs:
-  dependabot:
-    runs-on: ubuntu-latest
-    if: ${{ github.actor == 'localai-bot' }}
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Approve a PR if not already approved
-        run: |
-          gh pr checkout "$PR_URL"
-            if [ "$(gh pr status --json reviewDecision -q .currentBranch.reviewDecision)" != "APPROVED" ];
-          then
-            gh pr review --approve "$PR_URL"
-          else
-            echo "PR already approved.";
-          fi
-        env:
-          PR_URL: ${{github.event.pull_request.html_url}}
-          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
-
-      - name: Enable auto-merge for LocalAIBot PRs
-        run: gh pr merge --auto --squash "$PR_URL"
-        env:
-          PR_URL: ${{github.event.pull_request.html_url}}
-          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}

.github/workflows/notify-models.yaml

@@ -1,168 +0,0 @@
-name: Notifications for new models
-on:
-  pull_request:
-     types:
-       - closed
-
-jobs:
-  notify-discord:
-    if: ${{ (github.event.pull_request.merged == true) && (contains(github.event.pull_request.labels.*.name, 'area/ai-model')) }}
-    env:
-        MODEL_NAME: hermes-2-theta-llama-3-8b
-    runs-on: ubuntu-latest
-    steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0 # needed to checkout all branches for this Action to work
-    - uses: mudler/localai-github-action@v1
-      with:
-        model: 'hermes-2-theta-llama-3-8b' # Any from models.localai.io, or from huggingface.com with: "huggingface://<repository>/file"
-        # Check the PR diff using the current branch and the base branch of the PR
-    - uses: GrantBirki/git-diff-action@v2.7.0
-      id: git-diff-action
-      with:
-            json_diff_file_output: diff.json
-            raw_diff_file_output: diff.txt
-            file_output_only: "true"
-    - name: Summarize
-      env:
-        DIFF: ${{ steps.git-diff-action.outputs.raw-diff-path }}
-      id: summarize
-      run: |
-            input="$(cat $DIFF)"
-
-            # Define the LocalAI API endpoint
-            API_URL="http://localhost:8080/chat/completions"
-
-            # Create a JSON payload using jq to handle special characters
-            json_payload=$(jq -n --arg input "$input" '{
-            model: "'$MODEL_NAME'",
-            messages: [
-                {
-                role: "system",
-                content: "You are LocalAI-bot. Write a discord message to notify everyone about the new model from the git diff. Make it informal. An example can include: the URL of the model, the name, and a brief description of the model if exists. Also add an hint on how to install it in LocalAI and that can be browsed over https://models.localai.io. For example: local-ai run model_name_here"
-                },
-                {
-                role: "user",
-                content: $input
-                }
-            ]
-            }')
-
-            # Send the request to LocalAI
-            response=$(curl -s -X POST $API_URL \
-            -H "Content-Type: application/json" \
-            -d "$json_payload")
-
-            # Extract the summary from the response
-            summary="$(echo $response | jq -r '.choices[0].message.content')"
-
-            # Print the summary
-            #  -H "Authorization: Bearer $API_KEY" \
-            echo "Summary:"
-            echo "$summary"
-            echo "payload sent"
-            echo "$json_payload"
-            {
-                echo 'message<<EOF'
-                echo "$summary"
-                echo EOF
-              } >> "$GITHUB_OUTPUT"
-            docker logs --tail 10 local-ai
-    - name: Discord notification
-      env:
-        DISCORD_WEBHOOK: ${{ secrets.DISCORD_WEBHOOK_URL }}
-        DISCORD_USERNAME: "LocalAI-Bot"
-        DISCORD_AVATAR: "https://avatars.githubusercontent.com/u/139863280?v=4"
-      uses: Ilshidur/action-discord@master
-      with:
-        args: ${{ steps.summarize.outputs.message }}
-    - name: Setup tmate session if fails
-      if: ${{ failure() }}
-      uses: mxschmitt/action-tmate@v3.19
-      with:
-        detached: true
-        connect-timeout-seconds: 180
-        limit-access-to-actor: true
-  notify-twitter:
-    if: ${{ (github.event.pull_request.merged == true) && (contains(github.event.pull_request.labels.*.name, 'area/ai-model')) }}
-    env:
-        MODEL_NAME: hermes-2-theta-llama-3-8b
-    runs-on: ubuntu-latest
-    steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0 # needed to checkout all branches for this Action to work
-    - name: Start LocalAI
-      run: |
-        echo "Starting LocalAI..."
-        docker run -e -ti -d --name local-ai -p 8080:8080 localai/localai:master-ffmpeg-core run --debug $MODEL_NAME
-        until [ "`docker inspect -f {{.State.Health.Status}} local-ai`" == "healthy" ]; do echo "Waiting for container to be ready";  docker logs --tail 10 local-ai; sleep 2; done
-      # Check the PR diff using the current branch and the base branch of the PR
-    - uses: GrantBirki/git-diff-action@v2.7.0
-      id: git-diff-action
-      with:
-            json_diff_file_output: diff.json
-            raw_diff_file_output: diff.txt
-            file_output_only: "true"
-    - name: Summarize
-      env:
-        DIFF: ${{ steps.git-diff-action.outputs.raw-diff-path }}
-      id: summarize
-      run: |
-            input="$(cat $DIFF)"
-
-            # Define the LocalAI API endpoint
-            API_URL="http://localhost:8080/chat/completions"
-
-            # Create a JSON payload using jq to handle special characters
-            json_payload=$(jq -n --arg input "$input" '{
-            model: "'$MODEL_NAME'",
-            messages: [
-                {
-                role: "system",
-                content: "You are LocalAI-bot. Write a twitter message to notify everyone about the new model from the git diff. Make it informal and really short. An example can include: the name, and a brief description of the model if exists. Also add an hint on how to install it in LocalAI. For example: local-ai run model_name_here"
-                },
-                {
-                role: "user",
-                content: $input
-                }
-            ]
-            }')
-
-            # Send the request to LocalAI
-            response=$(curl -s -X POST $API_URL \
-            -H "Content-Type: application/json" \
-            -d "$json_payload")
-
-            # Extract the summary from the response
-            summary="$(echo $response | jq -r '.choices[0].message.content')"
-
-            # Print the summary
-            #  -H "Authorization: Bearer $API_KEY" \
-            echo "Summary:"
-            echo "$summary"
-            echo "payload sent"
-            echo "$json_payload"
-            {
-                echo 'message<<EOF'
-                echo "$summary"
-                echo EOF
-              } >> "$GITHUB_OUTPUT"
-            docker logs --tail 10 local-ai
-    - uses: Eomm/why-don-t-you-tweet@v2
-      with:
-        tweet-message: ${{ steps.summarize.outputs.message }}
-      env:
-        # Get your tokens from https://developer.twitter.com/apps
-        TWITTER_CONSUMER_API_KEY: ${{ secrets.TWITTER_APP_KEY }}
-        TWITTER_CONSUMER_API_SECRET: ${{ secrets.TWITTER_APP_SECRET }}
-        TWITTER_ACCESS_TOKEN: ${{ secrets.TWITTER_ACCESS_TOKEN }}
-        TWITTER_ACCESS_TOKEN_SECRET: ${{ secrets.TWITTER_ACCESS_TOKEN_SECRET }}
-    - name: Setup tmate session if fails
-      if: ${{ failure() }}
-      uses: mxschmitt/action-tmate@v3.19
-      with:
-        detached: true
-        connect-timeout-seconds: 180
-        limit-access-to-actor: true

.github/workflows/notify-releases.yaml

@@ -1,63 +0,0 @@
-name: Release notifications
-on:
-  release:
-    types:
-      - published
-
-jobs:
-  notify-discord:
-    runs-on: ubuntu-latest
-    env:
-        RELEASE_BODY: ${{ github.event.release.body }}
-        RELEASE_TITLE: ${{ github.event.release.name }}
-        RELEASE_TAG_NAME: ${{ github.event.release.tag_name }}
-    steps:
-    - uses: mudler/localai-github-action@v1
-      with:
-        model: 'hermes-2-theta-llama-3-8b' # Any from models.localai.io, or from huggingface.com with: "huggingface://<repository>/file"
-    - name: Summarize
-      id: summarize
-      run: |
-            input="$RELEASE_TITLE\b$RELEASE_BODY"
-
-            # Define the LocalAI API endpoint
-            API_URL="http://localhost:8080/chat/completions"
-
-            # Create a JSON payload using jq to handle special characters
-            json_payload=$(jq -n --arg input "$input" '{
-            model: "'$MODEL_NAME'",
-            messages: [
-                {
-                role: "system",
-                content: "Write a discord message with a bullet point summary of the release notes."
-                },
-                {
-                role: "user",
-                content: $input
-                }
-            ]
-            }')
-
-            # Send the request to LocalAI API
-            response=$(curl -s -X POST $API_URL \
-            -H "Content-Type: application/json" \
-            -d "$json_payload")
-
-            # Extract the summary from the response
-            summary=$(echo $response | jq -r '.choices[0].message.content')
-
-            # Print the summary
-            #  -H "Authorization: Bearer $API_KEY" \
-            {
-                echo 'message<<EOF'
-                echo "$summary"
-                echo EOF
-              } >> "$GITHUB_OUTPUT"
-    - name: Discord notification
-      env:
-        DISCORD_WEBHOOK: ${{ secrets.DISCORD_WEBHOOK_URL_RELEASE }}
-        DISCORD_USERNAME: "LocalAI-Bot"
-        DISCORD_AVATAR: "https://avatars.githubusercontent.com/u/139863280?v=4"
-      uses: Ilshidur/action-discord@master
-      with:
-        args: ${{ steps.summarize.outputs.message }}

.github/workflows/prlint.yaml

@@ -1,28 +0,0 @@
-name: Check PR style
-
-on:
-  pull_request_target:
-    types:
-      - opened
-      - reopened
-      - edited
-      - synchronize
-
-jobs:
-  title-lint:
-    runs-on: ubuntu-latest
-    permissions:
-      statuses: write
-    steps:
-      - uses: aslafy-z/conventional-pr-title-action@v3
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-#  check-pr-description:
-#    runs-on: ubuntu-latest
-#    steps:
-#      - uses: actions/checkout@v2
-#      - uses: jadrol/pr-description-checker-action@v1.0.0
-#        id: description-checker
-#        with:
-#          repo-token: ${{ secrets.GITHUB_TOKEN }}
-#          exempt-labels: no qa

.github/workflows/release.yaml

@@ -1,15 +1,6 @@
 name: Build and Release
 
-on:
-  push:
-    branches:
-      - master
-    tags:
-      - 'v*'
-  pull_request:
-
-env:
-  GRPC_VERSION: v1.65.0
+on: push
 
 permissions:
   contents: write
@@ -19,224 +10,85 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-
-  build-linux-arm:
+  build-linux:
+    strategy:
+      matrix:
+        include:
+          - build: 'avx2'
+            defines: ''
+          - build: 'avx'
+            defines: '-DLLAMA_AVX2=OFF'
+          - build: 'avx512'
+            defines: '-DLLAMA_AVX512=ON'
+          - build: 'cuda12'
+            defines: ''
+          - build: 'cuda11'
+            defines: ''
     runs-on: ubuntu-latest
     steps:
       - name: Clone
         uses: actions/checkout@v4
         with:
           submodules: true
-      - uses: actions/setup-go@v5
+      - uses: actions/setup-go@v4
         with:
-          go-version: '1.21.x'
-          cache: false
+          go-version: '>=1.21.0'
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ffmpeg protobuf-compiler ccache upx-ucl gawk
-          sudo apt-get install -qy binutils-aarch64-linux-gnu gcc-aarch64-linux-gnu g++-aarch64-linux-gnu libgmock-dev
-      - name: Install CUDA Dependencies
-        run: |
-          curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/cross-linux-aarch64/cuda-keyring_1.1-1_all.deb
-          sudo dpkg -i cuda-keyring_1.1-1_all.deb
-          sudo apt-get update
-          sudo apt-get install -y cuda-cross-aarch64 cuda-nvcc-cross-aarch64-${CUDA_VERSION} libcublas-cross-aarch64-${CUDA_VERSION}
-        env:
-          CUDA_VERSION: 12-4
-      - name: Cache grpc
-        id: cache-grpc
-        uses: actions/cache@v4
-        with:
-          path: grpc
-          key: ${{ runner.os }}-arm-grpc-${{ env.GRPC_VERSION }}
-      - name: Build grpc
-        if: steps.cache-grpc.outputs.cache-hit != 'true'
-        run: |
-
-          git clone --recurse-submodules -b ${{ env.GRPC_VERSION }} --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
-          cd grpc && sed -i "216i\  TESTONLY" "third_party/abseil-cpp/absl/container/CMakeLists.txt" && mkdir -p cmake/build && \
-          cd cmake/build && cmake -DgRPC_INSTALL=ON \
-            -DgRPC_BUILD_TESTS=OFF \
-            ../.. && sudo make --jobs 5 --output-sync=target
-      - name: Install gRPC
-        run: |
-          GNU_HOST=aarch64-linux-gnu
-          C_COMPILER_ARM_LINUX=$GNU_HOST-gcc
-          CXX_COMPILER_ARM_LINUX=$GNU_HOST-g++
-
-          CROSS_TOOLCHAIN=/usr/$GNU_HOST
-          CROSS_STAGING_PREFIX=$CROSS_TOOLCHAIN/stage
-          CMAKE_CROSS_TOOLCHAIN=/tmp/arm.toolchain.cmake
-
-          # https://cmake.org/cmake/help/v3.13/manual/cmake-toolchains.7.html#cross-compiling-for-linux
-          echo "set(CMAKE_SYSTEM_NAME Linux)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_SYSTEM_PROCESSOR arm)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_STAGING_PREFIX $CROSS_STAGING_PREFIX)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_SYSROOT ${CROSS_TOOLCHAIN}/sysroot)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_C_COMPILER /usr/bin/$C_COMPILER_ARM_LINUX)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_CXX_COMPILER /usr/bin/$CXX_COMPILER_ARM_LINUX)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_FIND_ROOT_PATH_MODE_PROGRAM NEVER)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_FIND_ROOT_PATH_MODE_LIBRARY ONLY)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_FIND_ROOT_PATH_MODE_INCLUDE ONLY)" >> $CMAKE_CROSS_TOOLCHAIN && \
-            echo "set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE ONLY)" >> $CMAKE_CROSS_TOOLCHAIN
-          GRPC_DIR=$PWD/grpc
-          cd grpc && cd cmake/build && sudo make --jobs 5 --output-sync=target install && \
-          GRPC_CROSS_BUILD_DIR=$GRPC_DIR/cmake/cross_build && \
-          mkdir -p $GRPC_CROSS_BUILD_DIR && \
-          cd $GRPC_CROSS_BUILD_DIR && \
-          cmake -DCMAKE_TOOLCHAIN_FILE=$CMAKE_CROSS_TOOLCHAIN \
-            -DCMAKE_BUILD_TYPE=Release \
-            -DCMAKE_INSTALL_PREFIX=$CROSS_TOOLCHAIN/grpc_install \
-            ../.. && \
-          sudo make -j`nproc` install
-      - name: Build
-        id: build
-        run: |
-          GNU_HOST=aarch64-linux-gnu
-          C_COMPILER_ARM_LINUX=$GNU_HOST-gcc
-          CXX_COMPILER_ARM_LINUX=$GNU_HOST-g++
-
-          CROSS_TOOLCHAIN=/usr/$GNU_HOST
-          CROSS_STAGING_PREFIX=$CROSS_TOOLCHAIN/stage
-          CMAKE_CROSS_TOOLCHAIN=/tmp/arm.toolchain.cmake
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-          export PATH=$PATH:$GOPATH/bin
-          export PATH=/usr/local/cuda/bin:$PATH
-          sudo rm -rf /usr/aarch64-linux-gnu/lib/libstdc++.so.6
-          sudo cp -rf /usr/aarch64-linux-gnu/lib/libstdc++.so* /usr/aarch64-linux-gnu/lib/libstdc++.so.6
-          sudo cp /usr/aarch64-linux-gnu/lib/ld-linux-aarch64.so.1 ld.so
-          BACKEND_LIBS="./grpc/cmake/cross_build/third_party/re2/libre2.a ./grpc/cmake/cross_build/libgrpc.a ./grpc/cmake/cross_build/libgrpc++.a ./grpc/cmake/cross_build/third_party/protobuf/libprotobuf.a /usr/aarch64-linux-gnu/lib/libc.so.6 /usr/aarch64-linux-gnu/lib/libstdc++.so.6 /usr/aarch64-linux-gnu/lib/libgomp.so.1 /usr/aarch64-linux-gnu/lib/libm.so.6 /usr/aarch64-linux-gnu/lib/libgcc_s.so.1 /usr/aarch64-linux-gnu/lib/libdl.so.2 /usr/aarch64-linux-gnu/lib/libpthread.so.0 ./ld.so" \
-          GOOS=linux \
-          GOARCH=arm64 \
-          CMAKE_ARGS="-DProtobuf_INCLUDE_DIRS=$CROSS_STAGING_PREFIX/include -DProtobuf_DIR=$CROSS_STAGING_PREFIX/lib/cmake/protobuf -DgRPC_DIR=$CROSS_STAGING_PREFIX/lib/cmake/grpc -DCMAKE_TOOLCHAIN_FILE=$CMAKE_CROSS_TOOLCHAIN -DCMAKE_C_COMPILER=aarch64-linux-gnu-gcc -DCMAKE_CXX_COMPILER=aarch64-linux-gnu-g++" make dist-cross-linux-arm64
-      - uses: actions/upload-artifact@v4
-        with:
-          name: LocalAI-linux-arm64
-          path: release/
-      - name: Release
-        uses: softprops/action-gh-release@v2
-        if: startsWith(github.ref, 'refs/tags/')
-        with:
-          files: |
-            release/*
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
-  build-linux:
-    runs-on: arc-runner-set
-    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: Clone
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-      - uses: actions/setup-go@v5
-        with:
-          go-version: '1.21.x'
-          cache: false
-      - name: Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y wget curl build-essential ffmpeg protobuf-compiler ccache upx-ucl gawk cmake libgmock-dev
-      - name: Intel Dependencies
-        run: |
-          wget -O- https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB | gpg --dearmor | sudo tee /usr/share/keyrings/oneapi-archive-keyring.gpg > /dev/null
-          echo "deb [signed-by=/usr/share/keyrings/oneapi-archive-keyring.gpg] https://apt.repos.intel.com/oneapi all main" | sudo tee /etc/apt/sources.list.d/oneAPI.list
-          sudo apt update
-          sudo apt install -y intel-basekit
+          sudo apt-get install build-essential ffmpeg
       - name: Install CUDA Dependencies
+        if: ${{ matrix.build == 'cuda12' || matrix.build == 'cuda11' }}
         run: |
+          if [ "${{ matrix.build }}" == "cuda12" ]; then
+            export CUDA_VERSION=12-3
+          else
+            export CUDA_VERSION=11-7
+          fi
           curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
           sudo dpkg -i cuda-keyring_1.1-1_all.deb
           sudo apt-get update
           sudo apt-get install -y cuda-nvcc-${CUDA_VERSION} libcublas-dev-${CUDA_VERSION}
-        env:
-          CUDA_VERSION: 12-5
-      - name: "Install Hipblas"
-        env:
-          ROCM_VERSION: "6.1"
-          AMDGPU_VERSION: "6.1"
-        run: |
-            set -ex
-
-            sudo apt-get update
-            sudo DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends ca-certificates curl libnuma-dev gnupg
-
-            curl -sL https://repo.radeon.com/rocm/rocm.gpg.key | sudo apt-key add -
-
-            printf "deb [arch=amd64] https://repo.radeon.com/rocm/apt/$ROCM_VERSION/ jammy main" | sudo tee /etc/apt/sources.list.d/rocm.list
-
-            printf "deb [arch=amd64] https://repo.radeon.com/amdgpu/$AMDGPU_VERSION/ubuntu jammy main" | sudo tee /etc/apt/sources.list.d/amdgpu.list
-            printf 'Package: *\nPin: release o=repo.radeon.com\nPin-Priority: 600' | sudo tee /etc/apt/preferences.d/rocm-pin-600
-            sudo apt-get update
-
-            sudo DEBIAN_FRONTEND=noninteractive apt-get install -y \
-                hipblas-dev rocm-dev \
-                rocblas-dev
-
-            sudo apt-get clean
-            sudo rm -rf /var/lib/apt/lists/*
-            sudo ldconfig
       - name: Cache grpc
         id: cache-grpc
-        uses: actions/cache@v4
+        uses: actions/cache@v3
         with:
           path: grpc
-          key: ${{ runner.os }}-grpc-${{ env.GRPC_VERSION }}
+          key: ${{ runner.os }}-grpc
       - name: Build grpc
         if: steps.cache-grpc.outputs.cache-hit != 'true'
         run: |
-          git clone --recurse-submodules -b ${{ env.GRPC_VERSION }} --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
-          cd grpc && sed -i "216i\  TESTONLY" "third_party/abseil-cpp/absl/container/CMakeLists.txt" && mkdir -p cmake/build && \
-          cd cmake/build && cmake -DgRPC_INSTALL=ON \
+          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 --jobs 5 --output-sync=target
+            ../.. && sudo make -j12
       - name: Install gRPC
         run: |
-          cd grpc && cd cmake/build && sudo make --jobs 5 --output-sync=target install
-      # BACKEND_LIBS needed for gpu-workload: /opt/intel/oneapi/*/lib/libiomp5.so /opt/intel/oneapi/*/lib/libmkl_core.so /opt/intel/oneapi/*/lib/libmkl_core.so.2 /opt/intel/oneapi/*/lib/libmkl_intel_ilp64.so /opt/intel/oneapi/*/lib/libmkl_intel_ilp64.so.2 /opt/intel/oneapi/*/lib/libmkl_sycl_blas.so /opt/intel/oneapi/*/lib/libmkl_sycl_blas.so.4 /opt/intel/oneapi/*/lib/libmkl_tbb_thread.so /opt/intel/oneapi/*/lib/libmkl_tbb_thread.so.2 /opt/intel/oneapi/*/lib/libsycl.so /opt/intel/oneapi/*/lib/libsycl.so.7 /opt/intel/oneapi/*/lib/libsycl.so.7.1.0 /opt/rocm-*/lib/libamdhip64.so /opt/rocm-*/lib/libamdhip64.so.5 /opt/rocm-*/lib/libamdhip64.so.6 /opt/rocm-*/lib/libamdhip64.so.6.1.60100 /opt/rocm-*/lib/libhipblas.so /opt/rocm-*/lib/libhipblas.so.2 /opt/rocm-*/lib/libhipblas.so.2.1.60100 /opt/rocm-*/lib/librocblas.so /opt/rocm-*/lib/librocblas.so.4 /opt/rocm-*/lib/librocblas.so.4.1.60100 /usr/lib/x86_64-linux-gnu/libstdc++.so.6 /usr/lib/x86_64-linux-gnu/libOpenCL.so.1 /usr/lib/x86_64-linux-gnu/libOpenCL.so.1.0.0 /usr/lib/x86_64-linux-gnu/libm.so.6 /usr/lib/x86_64-linux-gnu/libgcc_s.so.1 /usr/lib/x86_64-linux-gnu/libc.so.6 /usr/lib/x86_64-linux-gnu/librt.so.1 /usr/local/cuda-*/targets/x86_64-linux/lib/libcublas.so /usr/local/cuda-*/targets/x86_64-linux/lib/libcublasLt.so /usr/local/cuda-*/targets/x86_64-linux/lib/libcudart.so /usr/local/cuda-*/targets/x86_64-linux/lib/stubs/libcuda.so
+          cd grpc && cd cmake/build && sudo make -j12 install
       - name: Build
         id: build
+        env:
+          CMAKE_ARGS: "${{ matrix.defines }}"
+          BUILD_ID: "${{ matrix.build }}"
         run: |
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-          export PATH=$PATH:$GOPATH/bin
-          export PATH=/usr/local/cuda/bin:$PATH
-          export PATH=/opt/rocm/bin:$PATH
-          source /opt/intel/oneapi/setvars.sh
-          sudo cp /lib64/ld-linux-x86-64.so.2 ld.so
-          BACKEND_LIBS="./ld.so ./sources/go-piper/piper/build/fi/lib/libfmt.a ./sources/go-piper/piper-phonemize/pi/lib/libonnxruntime.so.1.14.1 ./sources/go-piper/piper-phonemize/pi/src/libespeak-ng/libespeak-ng.so /usr/lib/x86_64-linux-gnu/libdl.so.2 /usr/lib/x86_64-linux-gnu/librt.so.1 /usr/lib/x86_64-linux-gnu/libpthread.so.0 ./sources/go-piper/piper-phonemize/pi/lib/libpiper_phonemize.so.1 ./sources/go-piper/piper/build/si/lib/libspdlog.a ./sources/go-piper/espeak/ei/lib/libucd.so" \
-          make -j4 dist
-      - uses: actions/upload-artifact@v4
+          if [ "${{ matrix.build }}" == "cuda12" ] || [ "${{ matrix.build }}" == "cuda11" ]; then
+            export BUILD_TYPE=cublas
+            export PATH=/usr/local/cuda/bin:$PATH
+            make dist
+          else
+            STATIC=true make dist
+          fi
+      - uses: actions/upload-artifact@v3
         with:
-          name: LocalAI-linux
+          name: ${{ matrix.build }}
           path: release/
       - name: Release
-        uses: softprops/action-gh-release@v2
+        uses: softprops/action-gh-release@v1
         if: startsWith(github.ref, 'refs/tags/')
         with:
           files: |
             release/*
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
+
   build-stablediffusion:
     runs-on: ubuntu-latest
     steps:
@@ -244,114 +96,66 @@ jobs:
         uses: actions/checkout@v4
         with:
           submodules: true
-      - uses: actions/setup-go@v5
+      - uses: actions/setup-go@v4
         with:
-          go-version: '1.21.x'
-          cache: false
+          go-version: '>=1.21.0'
       - name: Dependencies
         run: |
-          sudo apt-get update
-          sudo apt-get install -y --no-install-recommends libopencv-dev protobuf-compiler ccache upx-ucl
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
+          sudo apt-get install -y --no-install-recommends libopencv-dev
+          sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
       - name: Build stablediffusion
         run: |
-          export PATH=$PATH:$GOPATH/bin
           make backend-assets/grpc/stablediffusion
           mkdir -p release && cp backend-assets/grpc/stablediffusion release
-        env:
-          GO_TAGS: stablediffusion
-      - uses: actions/upload-artifact@v4
+      - uses: actions/upload-artifact@v3
         with:
           name: stablediffusion
           path: release/
       - name: Release
-        uses: softprops/action-gh-release@v2
+        uses: softprops/action-gh-release@v1
         if: startsWith(github.ref, 'refs/tags/')
         with:
           files: |
             release/*
 
-  build-macOS-x86_64:
-    runs-on: macos-13
+  build-macOS:
+    strategy:
+      matrix:
+        include:
+          - build: 'avx2'
+            defines: ''
+          - build: 'avx'
+            defines: '-DLLAMA_AVX2=OFF'
+          - build: 'avx512'
+            defines: '-DLLAMA_AVX512=ON'
+    runs-on: macOS-latest
     steps:
       - name: Clone
         uses: actions/checkout@v4
         with:
           submodules: true
-      - uses: actions/setup-go@v5
+      - uses: actions/setup-go@v4
         with:
-          go-version: '1.21.x'
-          cache: false
+          go-version: '>=1.21.0'
       - name: Dependencies
         run: |
           brew install protobuf grpc
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@8ba23be9613c672d40ae261d2a1335d639bdd59b
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.0
       - 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
-          export PATH=$PATH:$GOPATH/bin
-          export SKIP_GRPC_BACKEND=backend-assets/grpc/whisper
           make dist
-      - uses: actions/upload-artifact@v4
+      - uses: actions/upload-artifact@v3
         with:
-          name: LocalAI-MacOS-x86_64
+          name: ${{ matrix.build }}
           path: release/
       - name: Release
-        uses: softprops/action-gh-release@v2
+        uses: softprops/action-gh-release@v1
         if: startsWith(github.ref, 'refs/tags/')
         with:
           files: |
             release/*
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
-
-  build-macOS-arm64:
-    runs-on: macos-14
-    steps:
-      - name: Clone
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-      - uses: actions/setup-go@v5
-        with:
-          go-version: '1.21.x'
-          cache: false
-      - name: Dependencies
-        run: |
-          brew install protobuf grpc libomp llvm
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-      - name: Build
-        id: build
-        run: |
-          export C_INCLUDE_PATH=/usr/local/include
-          export CPLUS_INCLUDE_PATH=/usr/local/include
-          export PATH=$PATH:$GOPATH/bin
-          export CC=/opt/homebrew/opt/llvm/bin/clang
-          make dist
-      - uses: actions/upload-artifact@v4
-        with:
-          name: LocalAI-MacOS-arm64
-          path: release/
-      - name: Release
-        uses: softprops/action-gh-release@v2
-        if: startsWith(github.ref, 'refs/tags/')
-        with:
-          files: |
-            release/*
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true

.github/workflows/secscan.yaml

@@ -1,30 +0,0 @@
-name: "Security Scan"
-
-# Run workflow each time code is pushed to your repository and on a schedule.
-# The scheduled workflow runs every at 00:00 on Sunday UTC time.
-on:
-  push:
-  schedule:
-  - cron: '0 0 * * 0'
-
-jobs:
-  tests:
-    runs-on: ubuntu-latest
-    env:
-      GO111MODULE: on
-    steps:
-      - name: Checkout Source
-        uses: actions/checkout@v4
-        if: ${{ github.actor != 'dependabot[bot]' }}
-      - name: Run Gosec Security Scanner
-        if: ${{ github.actor != 'dependabot[bot]' }}
-        uses: securego/gosec@v2.21.4
-        with:
-          # we let the report trigger content trigger a failure using the GitHub Security features.
-          args: '-no-fail -fmt sarif -out results.sarif ./...'
-      - name: Upload SARIF file
-        if: ${{ github.actor != 'dependabot[bot]' }}
-        uses: github/codeql-action/upload-sarif@v3
-        with:
-          # Path to SARIF file relative to the root of the repository
-          sarif_file: results.sarif

.github/workflows/test-extra.yml

@@ -19,161 +19,150 @@ jobs:
     steps:
       - name: Clone
         uses: actions/checkout@v4
-        with:
+        with: 
           submodules: true
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
+          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: |
-           make --jobs=5 --output-sync=target -C backend/python/transformers
-           make --jobs=5 --output-sync=target -C backend/python/transformers test
+           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:
+        with: 
           submodules: true
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
+          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: |
-           make --jobs=5 --output-sync=target -C backend/python/sentencetransformers
-           make --jobs=5 --output-sync=target -C backend/python/sentencetransformers test
-
-
-  tests-rerankers:
-    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
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
-
-      - name: Test rerankers
-        run: |
-           make --jobs=5 --output-sync=target -C backend/python/rerankers
-           make --jobs=5 --output-sync=target -C backend/python/rerankers test
+           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:
+        with: 
           submodules: true
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install -y build-essential ffmpeg
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
+          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: |
-          make --jobs=5 --output-sync=target -C backend/python/diffusers
-          make --jobs=5 --output-sync=target -C backend/python/diffusers test
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/diffusers
+           make -C backend/python/diffusers test
 
-  tests-parler-tts:
-    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
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
-
-      - name: Test parler-tts
-        run: |
-           make --jobs=5 --output-sync=target -C backend/python/parler-tts
-           make --jobs=5 --output-sync=target -C backend/python/parler-tts test
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
-
-  tests-openvoice:
-    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
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
-
-      - name: Test openvoice
-        run: |
-           make --jobs=5 --output-sync=target -C backend/python/openvoice
-           make --jobs=5 --output-sync=target -C backend/python/openvoice test
 
   tests-transformers-musicgen:
     runs-on: ubuntu-latest
     steps:
       - name: Clone
         uses: actions/checkout@v4
-        with:
+        with: 
           submodules: true
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
+          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: |
-           make --jobs=5 --output-sync=target -C backend/python/transformers-musicgen
-           make --jobs=5 --output-sync=target -C backend/python/transformers-musicgen test
+           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
@@ -220,24 +209,31 @@ jobs:
   #           df -h
   #     - name: Clone
   #       uses: actions/checkout@v4
-  #       with:
+  #       with: 
   #         submodules: true
   #     - name: Dependencies
   #       run: |
   #         sudo apt-get update
   #         sudo apt-get install build-essential ffmpeg
-  #         # Install UV
-  #         curl -LsSf https://astral.sh/uv/install.sh | sh
-  #         sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-  #         sudo apt-get install -y libopencv-dev
-  #         pip install --user --no-cache-dir grpcio-tools==1.64.1
+  #         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: |
-  #          make --jobs=5 --output-sync=target -C backend/python/bark
-  #          make --jobs=5 --output-sync=target -C backend/python/bark test
-
+  #          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:
@@ -245,58 +241,77 @@ jobs:
   #   steps:
   #     - name: Clone
   #       uses: actions/checkout@v4
-  #       with:
+  #       with: 
   #         submodules: true
   #     - name: Dependencies
   #       run: |
   #         sudo apt-get update
   #         sudo apt-get install build-essential ffmpeg
-  #         # Install UV
-  #         curl -LsSf https://astral.sh/uv/install.sh | sh
-  #         sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-  #         sudo apt-get install -y libopencv-dev
-  #         pip install --user --no-cache-dir grpcio-tools==1.64.1
+  #         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: |
-  #          make --jobs=5 --output-sync=target -C backend/python/vllm
-  #          make --jobs=5 --output-sync=target -C backend/python/vllm test
+  #          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:
+        with: 
           submodules: true
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
-          sudo apt-get install -y libopencv-dev
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
+          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: |
-           make --jobs=5 --output-sync=target -C backend/python/vall-e-x
-           make --jobs=5 --output-sync=target -C backend/python/vall-e-x test
+           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:
+        with: 
           submodules: true
       - name: Dependencies
         run: |
           sudo apt-get update
           sudo apt-get install build-essential ffmpeg
-          sudo apt-get install -y ca-certificates cmake curl patch espeak espeak-ng python3-pip
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          pip install --user --no-cache-dir grpcio-tools==1.64.1
+          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: |
-          make --jobs=5 --output-sync=target -C backend/python/coqui
-          make --jobs=5 --output-sync=target -C backend/python/coqui test
+           export PATH=$PATH:/opt/conda/bin
+           make -C backend/python/coqui
+           make -C backend/python/coqui test

.github/workflows/test.yml

@@ -9,9 +9,6 @@ on:
     tags:
       - '*'
 
-env:
-  GRPC_VERSION: v1.65.0
-
 concurrency:
   group: ci-tests-${{ github.head_ref || github.ref }}-${{ github.repository }}
   cancel-in-progress: true
@@ -57,49 +54,29 @@ jobs:
           df -h
       - name: Clone
         uses: actions/checkout@v4
-        with:
+        with: 
           submodules: true
       - name: Setup Go ${{ matrix.go-version }}
-        uses: actions/setup-go@v5
+        uses: actions/setup-go@v4
         with:
           go-version: ${{ matrix.go-version }}
-          cache: false
       # 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 ccache upx-ucl curl ffmpeg
-          sudo apt-get install -y libgmock-dev
+          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 && \
+              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
-          # Install UV
-          curl -LsSf https://astral.sh/uv/install.sh | sh
-          sudo apt-get install -y ca-certificates cmake patch python3-pip unzip
-          sudo apt-get install -y libopencv-dev
-
-          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
-          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
-          rm protoc.zip
-
-          curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
-          sudo dpkg -i cuda-keyring_1.1-1_all.deb
-          sudo apt-get update
-          sudo apt-get install -y cuda-nvcc-${CUDA_VERSION} libcublas-dev-${CUDA_VERSION}
-          export CUDACXX=/usr/local/cuda/bin/nvcc
-
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-
-          # The python3-grpc-tools package in 22.04 is too old
-          pip install --user grpcio-tools
-
+          sudo apt-get install -y ca-certificates cmake curl patch
+          sudo apt-get install -y libopencv-dev && sudo ln -s /usr/include/opencv4/opencv2 /usr/include/opencv2
+          
           sudo rm -rfv /usr/bin/conda || true
           PATH=$PATH:/opt/conda/bin make -C backend/python/sentencetransformers
 
@@ -108,135 +85,49 @@ jobs:
           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)
-          PATH="$PATH:/root/go/bin" GO_TAGS="stablediffusion tts" GRPC_BACKENDS=backend-assets/grpc/stablediffusion make build
-        env:
-          CUDA_VERSION: 12-4
+          GO_TAGS="stablediffusion tts" GRPC_BACKENDS=backend-assets/grpc/stablediffusion make build
       - name: Cache grpc
         id: cache-grpc
-        uses: actions/cache@v4
+        uses: actions/cache@v3
         with:
           path: grpc
-          key: ${{ runner.os }}-grpc-${{ env.GRPC_VERSION }}
+          key: ${{ runner.os }}-grpc
       - name: Build grpc
         if: steps.cache-grpc.outputs.cache-hit != 'true'
         run: |
-          git clone --recurse-submodules -b ${{ env.GRPC_VERSION }} --depth 1 --jobs 5 --shallow-submodules https://github.com/grpc/grpc && \
-          cd grpc && sed -i "216i\  TESTONLY" "third_party/abseil-cpp/absl/container/CMakeLists.txt" && mkdir -p cmake/build && cd cmake/build && \
-          cmake -DgRPC_INSTALL=ON \
+          git clone --recurse-submodules -b v1.58.0 --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
+          cd grpc && mkdir -p cmake/build && cd cmake/build && cmake -DgRPC_INSTALL=ON \
             -DgRPC_BUILD_TESTS=OFF \
-            ../.. && sudo make --jobs 5
+            ../.. && sudo make -j12
       - name: Install gRPC
         run: |
-          cd grpc && cd cmake/build && sudo make --jobs 5 install
+          cd grpc && cd cmake/build && sudo make -j12 install
       - name: Test
         run: |
-          PATH="$PATH:/root/go/bin" GO_TAGS="stablediffusion tts" make --jobs 5 --output-sync=target test
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
-
-  tests-aio-container:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Release space from worker
-        run: |
-          echo "Listing top largest packages"
-          pkgs=$(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@v4
-        with:
-          submodules: true
-      - name: Dependencies
-        run: |
-          # Install protoc
-          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
-          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
-          rm protoc.zip
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          PATH="$PATH:$HOME/go/bin" make protogen-go
-      - name: Build images
-        run: |
-          docker build --build-arg FFMPEG=true --build-arg IMAGE_TYPE=extras --build-arg EXTRA_BACKENDS=rerankers --build-arg MAKEFLAGS="--jobs=5 --output-sync=target" -t local-ai:tests -f Dockerfile .
-          BASE_IMAGE=local-ai:tests DOCKER_AIO_IMAGE=local-ai-aio:test make docker-aio
-      - name: Test
-        run: |
-            PATH="$PATH:$HOME/go/bin" LOCALAI_MODELS_DIR=$PWD/models LOCALAI_IMAGE_TAG=test LOCALAI_IMAGE=local-ai-aio \
-            make run-e2e-aio
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
+          GO_TAGS="stablediffusion tts" make test
 
   tests-apple:
-    runs-on: macOS-14
+    runs-on: macOS-latest
     strategy:
       matrix:
         go-version: ['1.21.x']
     steps:
       - name: Clone
         uses: actions/checkout@v4
-        with:
+        with: 
           submodules: true
       - name: Setup Go ${{ matrix.go-version }}
-        uses: actions/setup-go@v5
+        uses: actions/setup-go@v4
         with:
           go-version: ${{ matrix.go-version }}
-          cache: false
       # You can test your matrix by printing the current Go version
       - name: Display Go version
         run: go version
       - name: Dependencies
         run: |
-          brew install protobuf grpc make protoc-gen-go protoc-gen-go-grpc libomp llvm
-          pip install --user --no-cache-dir grpcio-tools
+          brew install protobuf grpc
       - name: Test
         run: |
           export C_INCLUDE_PATH=/usr/local/include
           export CPLUS_INCLUDE_PATH=/usr/local/include
-          export CC=/opt/homebrew/opt/llvm/bin/clang
-          # Used to run the newer GNUMake version from brew that supports --output-sync
-          export PATH="/opt/homebrew/opt/make/libexec/gnubin:$PATH"
-          BUILD_TYPE="GITHUB_CI_HAS_BROKEN_METAL" CMAKE_ARGS="-DGGML_F16C=OFF -DGGML_AVX512=OFF -DGGML_AVX2=OFF -DGGML_FMA=OFF" make --jobs 4 --output-sync=target test
-      - name: Setup tmate session if tests fail
-        if: ${{ failure() }}
-        uses: mxschmitt/action-tmate@v3.19
-        with:
-          detached: true
-          connect-timeout-seconds: 180
-          limit-access-to-actor: true
+          CMAKE_ARGS="-DLLAMA_F16C=OFF -DLLAMA_AVX512=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF" make test

.github/workflows/update_swagger.yaml

@@ -1,37 +0,0 @@
-name: Update swagger
-on:
-  schedule:
-    - cron: 0 20 * * *
-  workflow_dispatch:
-jobs:
-  swagger:
-    strategy:
-      fail-fast: false
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - uses: actions/setup-go@v5
-        with:
-          go-version: 'stable'
-      - name: Dependencies
-        run: |
-          sudo apt-get update
-          sudo apt-get install protobuf-compiler
-      - run: |
-          go install github.com/swaggo/swag/cmd/swag@latest
-          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
-          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
-      - name: Bump swagger 🔧
-        run: |
-          make protogen-go swagger
-      - name: Create Pull Request
-        uses: peter-evans/create-pull-request@v7
-        with:
-          token: ${{ secrets.UPDATE_BOT_TOKEN }}
-          push-to-fork: ci-forks/LocalAI
-          commit-message: 'feat(swagger): update swagger'
-          title: 'feat(swagger): update swagger'
-          branch: "update/swagger"
-          body:  Update swagger
-          signoff: true
-

.github/workflows/yaml-check.yml

@@ -1,18 +0,0 @@
-name: 'Yamllint GitHub Actions'
-on:
-  - pull_request
-jobs:
-  yamllint:
-    name: 'Yamllint'
-    runs-on: ubuntu-latest
-    steps:
-      - name: 'Checkout'
-        uses: actions/checkout@master
-      - name: 'Yamllint'
-        uses: karancode/yamllint-github-action@master
-        with:
-          yamllint_file_or_dir: 'gallery'
-          yamllint_strict: false
-          yamllint_comment: true
-        env:
-          GITHUB_ACCESS_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -2,17 +2,14 @@
 /sources/
 __pycache__/
 *.a
-*.o
 get-sources
 prepare-sources
 /backend/cpp/llama/grpc-server
 /backend/cpp/llama/llama.cpp
-/backend/cpp/llama-*
-
-*.log
 
 go-ggml-transformers
 go-gpt2
+go-rwkv
 whisper.cpp
 /bloomz
 go-bert
@@ -42,18 +39,3 @@ backend-assets/*
 !backend-assets/.keep
 prepare
 /ggml-metal.metal
-docs/static/gallery.html
-
-# Protobuf generated files
-*.pb.go
-*pb2.py
-*pb2_grpc.py
-
-# SonarQube
-.scannerwork
-
-# backend virtual environments
-**/venv
-
-# per-developer customization files for the development container
-.devcontainer/customization/*

.vscode/extensions.json

@@ -1,5 +0,0 @@
-{
-    "recommendations": [
-        "golang.go"
-    ]
-}

.vscode/launch.json

@@ -3,12 +3,12 @@
     "configurations": [
         {
             "name": "Python: Current File",
-            "type": "debugpy",
+            "type": "python",
             "request": "launch",
             "program": "${file}",
             "console": "integratedTerminal",
             "justMyCode": false,
-            "cwd": "${fileDirname}",
+            "cwd": "${workspaceFolder}/examples/langchain-chroma",
             "env": {
                 "OPENAI_API_BASE": "http://localhost:8080/v1",
                 "OPENAI_API_KEY": "abc"
@@ -19,16 +19,15 @@
             "type": "go",
             "request": "launch",
             "mode": "debug",
-            "program": "${workspaceRoot}",
-            "args": [],
+            "program": "${workspaceFolder}/main.go",
+            "args": [
+                "api"
+            ],
             "env": {
-                "LOCALAI_LOG_LEVEL": "debug",
-                "LOCALAI_P2P": "true",
-                "LOCALAI_FEDERATED": "true"
-            },
-            "buildFlags": ["-tags", "stablediffusion p2p tts", "-v"],
-            "envFile": "${workspaceFolder}/.env",
-            "cwd": "${workspaceRoot}"
+                "C_INCLUDE_PATH": "${workspaceFolder}/go-llama:${workspaceFolder}/go-stable-diffusion/:${workspaceFolder}/gpt4all/gpt4all-bindings/golang/:${workspaceFolder}/go-gpt2:${workspaceFolder}/go-rwkv:${workspaceFolder}/whisper.cpp:${workspaceFolder}/go-bert:${workspaceFolder}/bloomz",
+                "LIBRARY_PATH": "${workspaceFolder}/go-llama:${workspaceFolder}/go-stable-diffusion/:${workspaceFolder}/gpt4all/gpt4all-bindings/golang/:${workspaceFolder}/go-gpt2:${workspaceFolder}/go-rwkv:${workspaceFolder}/whisper.cpp:${workspaceFolder}/go-bert:${workspaceFolder}/bloomz",
+                "DEBUG": "true"
+            }
         }
     ]
 }

.yamllint

@@ -1,4 +0,0 @@
-extends: default
-
-rules:
-    line-length: disable

CONTRIBUTING.md

@@ -1,4 +1,4 @@
-# Contributing to LocalAI
+# 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.
 
@@ -15,6 +15,8 @@ Thank you for your interest in contributing to LocalAI! We appreciate your time
 - [Documentation](#documentation)
 - [Community and Communication](#community-and-communication)
 
+
+
 ## Getting Started
 
 ### Prerequisites
@@ -27,9 +29,8 @@ Thank you for your interest in contributing to LocalAI! We appreciate your time
 
 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 ( see https://localai.io/basics/build/#build-localai-locally )
-4. Build LocalAI: `make build`
-5. Run LocalAI: `./local-ai`
+3. Install the required dependencies: `make prepare`
+4. Run LocalAI: `make run`
 
 ## Contributing
 
@@ -52,33 +53,20 @@ If you find a bug, have a feature request, or encounter any issues, please check
 
 ## 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.
+- 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.
 
-### Running AIO tests
-
-All-In-One images has a set of tests that automatically verifies that most of the endpoints works correctly, a flow can be :
-
-```bash
-# Build the LocalAI docker image
-make DOCKER_IMAGE=local-ai docker
-
-# Build the corresponding AIO image
-BASE_IMAGE=local-ai DOCKER_AIO_IMAGE=local-ai-aio:test make docker-aio
-
-# Run the AIO e2e tests
-LOCALAI_IMAGE_TAG=test LOCALAI_IMAGE=local-ai-aio make run-e2e-aio
-```
-
 ## Documentation
 
-We are welcome the contribution of the documents, please open new PR or create a new issue. The documentation is available under `docs/` https://github.com/mudler/LocalAI/tree/master/docs
- 
+- 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

@@ -1,410 +1,165 @@
 ARG IMAGE_TYPE=extras
 ARG BASE_IMAGE=ubuntu:22.04
-ARG GRPC_BASE_IMAGE=${BASE_IMAGE}
-ARG INTEL_BASE_IMAGE=${BASE_IMAGE}
 
-# The requirements-core target is common to all images.  It should not be placed in requirements-core unless every single build will use it.
-FROM ${BASE_IMAGE} AS requirements-core
+# extras or core
+FROM ${BASE_IMAGE} as requirements-core
 
 USER root
 
-ARG GO_VERSION=1.22.6
-ARG CMAKE_VERSION=3.26.4
-ARG CMAKE_FROM_SOURCE=false
+ARG GO_VERSION=1.21.7
+ARG BUILD_TYPE
+ARG CUDA_MAJOR_VERSION=11
+ARG CUDA_MINOR_VERSION=7
 ARG TARGETARCH
 ARG TARGETVARIANT
 
+ENV BUILD_TYPE=${BUILD_TYPE}
 ENV DEBIAN_FRONTEND=noninteractive
-ENV EXTERNAL_GRPC_BACKENDS="coqui:/build/backend/python/coqui/run.sh,huggingface-embeddings:/build/backend/python/sentencetransformers/run.sh,transformers:/build/backend/python/transformers/run.sh,sentencetransformers:/build/backend/python/sentencetransformers/run.sh,rerankers:/build/backend/python/rerankers/run.sh,autogptq:/build/backend/python/autogptq/run.sh,bark:/build/backend/python/bark/run.sh,diffusers:/build/backend/python/diffusers/run.sh,openvoice:/build/backend/python/openvoice/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,parler-tts:/build/backend/python/parler-tts/run.sh"
+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 --no-install-recommends \
-        build-essential \
-        ccache \
-        ca-certificates \
-        curl libssl-dev \
-        git \
-        unzip upx-ucl && \
-    apt-get clean && \
-    rm -rf /var/lib/apt/lists/*
-
-# Install CMake (the version in 22.04 is too old)
-RUN <<EOT bash
-    if [ "${CMAKE_FROM_SOURCE}}" = "true" ]; then
-        curl -L -s https://github.com/Kitware/CMake/releases/download/v${CMAKE_VERSION}/cmake-${CMAKE_VERSION}.tar.gz -o cmake.tar.gz && tar xvf cmake.tar.gz && cd cmake-${CMAKE_VERSION} && ./configure && make && make install
-    else
-        apt-get update && \
-        apt-get install -y \
-            cmake && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/*
-    fi
-EOT
+    apt-get install -y ca-certificates curl patch pip cmake git && apt-get clean
 
 # Install Go
-RUN curl -L -s https://go.dev/dl/go${GO_VERSION}.linux-${TARGETARCH}.tar.gz | tar -C /usr/local -xz
-ENV PATH=$PATH:/root/go/bin:/usr/local/go/bin
-
-# Install grpc compilers
-RUN go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2 && \
-    go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
+RUN curl -L -s https://go.dev/dl/go$GO_VERSION.linux-$TARGETARCH.tar.gz | tar -C /usr/local -xz
+ENV PATH $PATH:/usr/local/go/bin
 
 COPY --chmod=644 custom-ca-certs/* /usr/local/share/ca-certificates/
 RUN update-ca-certificates
 
-RUN test -n "$TARGETARCH" \
-    || (echo 'warn: missing $TARGETARCH, either set this `ARG` manually, or run using `docker buildkit`')
-
 # Use the variables in subsequent instructions
 RUN echo "Target Architecture: $TARGETARCH"
 RUN echo "Target Variant: $TARGETVARIANT"
 
+# CuBLAS requirements
+RUN if [ "${BUILD_TYPE}" = "cublas" ]; then \
+    apt-get install -y software-properties-common && \
+    curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb && \
+    dpkg -i cuda-keyring_1.1-1_all.deb && \
+    rm -f cuda-keyring_1.1-1_all.deb && \
+    apt-get update && \
+    apt-get install -y cuda-nvcc-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} libcurand-dev-${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}-${CUDA_MINOR_VERSION}  && apt-get clean \
+    ; fi
+
 # Cuda
-ENV PATH=/usr/local/cuda/bin:${PATH}
+ENV PATH /usr/local/cuda/bin:${PATH}
 
 # HipBLAS requirements
-ENV PATH=/opt/rocm/bin:${PATH}
+ENV PATH /opt/rocm/bin:${PATH}
 
 # OpenBLAS requirements and stable diffusion
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends \
-        libopenblas-dev \
-        libopencv-dev && \
-    apt-get clean && \
-    rm -rf /var/lib/apt/lists/*
+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
 
-###################################
-###################################
+RUN test -n "$TARGETARCH" \
+    || (echo 'warn: missing $TARGETARCH, either set this `ARG` manually, or run using `docker buildkit`')
 
-# The requirements-extras target is for any builds with IMAGE_TYPE=extras. It should not be placed in this target unless every IMAGE_TYPE=extras build will use it
-FROM requirements-core AS requirements-extras
+# 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
 
-# Install uv as a system package
-RUN curl -LsSf https://astral.sh/uv/install.sh | UV_INSTALL_DIR=/usr/bin sh
 ENV PATH="/root/.cargo/bin:${PATH}"
+RUN apt-get install -y python3-pip && apt-get clean
+RUN pip install --upgrade pip
 
 RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends \
-        espeak-ng \
-        espeak \
-        python3-pip \
-        python-is-python3 \
-        python3-dev llvm \
-        python3-venv && \
-    apt-get clean && \
-    rm -rf /var/lib/apt/lists/* && \
-    pip install --upgrade pip
+RUN apt-get install -y espeak-ng espeak && apt-get clean
 
-# Install grpcio-tools (the version in 22.04 is too old)
-RUN pip install --user grpcio-tools
-
-###################################
-###################################
-
-# The requirements-drivers target is for BUILD_TYPE specific items.  If you need to install something specific to CUDA, or specific to ROCM, it goes here.
-# This target will be built on top of requirements-core or requirements-extras as retermined by the IMAGE_TYPE build-arg
-FROM requirements-${IMAGE_TYPE} AS requirements-drivers
-
-ARG BUILD_TYPE
-ARG CUDA_MAJOR_VERSION=12
-ARG CUDA_MINOR_VERSION=0
-
-ENV BUILD_TYPE=${BUILD_TYPE}
-
-# Vulkan requirements
-RUN <<EOT bash
-    if [ "${BUILD_TYPE}" = "vulkan" ]; then
-        apt-get update && \
-        apt-get install -y  --no-install-recommends \
-            software-properties-common pciutils wget gpg-agent && \
-        wget -qO - https://packages.lunarg.com/lunarg-signing-key-pub.asc | apt-key add - && \
-        wget -qO /etc/apt/sources.list.d/lunarg-vulkan-jammy.list https://packages.lunarg.com/vulkan/lunarg-vulkan-jammy.list && \
-        apt-get update && \
-        apt-get install -y \
-            vulkan-sdk && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/*
-    fi
-EOT
-
-# CuBLAS requirements
-RUN <<EOT bash
-    if [ "${BUILD_TYPE}" = "cublas" ]; then
-        apt-get update && \
-        apt-get install -y  --no-install-recommends \
-            software-properties-common pciutils
-        if [ "amd64" = "$TARGETARCH" ]; then
-            curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
-        fi
-        if [ "arm64" = "$TARGETARCH" ]; then
-            curl -O https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/arm64/cuda-keyring_1.1-1_all.deb
-        fi
-        dpkg -i cuda-keyring_1.1-1_all.deb && \
-        rm -f cuda-keyring_1.1-1_all.deb && \
-        apt-get update && \
-        apt-get install -y --no-install-recommends \
-            cuda-nvcc-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} \
-            libcufft-dev-${CUDA_MAJOR_VERSION}-${CUDA_MINOR_VERSION} \
-            libcurand-dev-${CUDA_MAJOR_VERSION}-${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 && \
-        rm -rf /var/lib/apt/lists/*
-    fi
-EOT
-
-# If we are building with clblas support, we need the libraries for the builds
-RUN if [ "${BUILD_TYPE}" = "clblas" ]; then \
-        apt-get update && \
-        apt-get install -y --no-install-recommends \
-            libclblast-dev && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/* \
-    ; fi
-
-RUN if [ "${BUILD_TYPE}" = "hipblas" ]; then \
-        apt-get update && \
-        apt-get install -y --no-install-recommends \
-            hipblas-dev \
-            rocblas-dev && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/* && \
-        # I have no idea why, but the ROCM lib packages don't trigger ldconfig after they install, which results in local-ai and others not being able
-        # to locate the libraries. We run ldconfig ourselves to work around this packaging deficiency
-        ldconfig \
+RUN if [ ! -e /usr/bin/python ]; then \
+	  ln -s /usr/bin/python3 /usr/bin/python \
     ; fi
 
 ###################################
 ###################################
 
-# Temporary workaround for Intel's repository to work correctly
-# https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/APT-Repository-not-working-signatures-invalid/m-p/1599436/highlight/true#M36143
-# This is a temporary workaround until Intel fixes their repository
-FROM ${INTEL_BASE_IMAGE} AS intel
-RUN wget -qO - https://repositories.intel.com/gpu/intel-graphics.key | \
-gpg --yes --dearmor --output /usr/share/keyrings/intel-graphics.gpg
-RUN echo "deb [arch=amd64 signed-by=/usr/share/keyrings/intel-graphics.gpg] https://repositories.intel.com/gpu/ubuntu jammy/lts/2350 unified" > /etc/apt/sources.list.d/intel-graphics.list
+FROM requirements-${IMAGE_TYPE} as builder
 
-###################################
-###################################
-
-# The grpc target does one thing, it builds and installs GRPC.  This is in it's own layer so that it can be effectively cached by CI.
-# You probably don't need to change anything here, and if you do, make sure that CI is adjusted so that the cache continues to work.
-FROM ${GRPC_BASE_IMAGE} AS grpc
-
-# This is a bit of a hack, but it's required in order to be able to effectively cache this layer in CI
-ARG GRPC_MAKEFLAGS="-j4 -Otarget"
-ARG GRPC_VERSION=v1.65.0
-ARG CMAKE_FROM_SOURCE=false
-ARG CMAKE_VERSION=3.26.4
-
-ENV MAKEFLAGS=${GRPC_MAKEFLAGS}
-
-WORKDIR /build
-
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends \
-        ca-certificates \
-        build-essential curl libssl-dev \
-        git && \
-    apt-get clean && \
-    rm -rf /var/lib/apt/lists/*
-
-# Install CMake (the version in 22.04 is too old)
-RUN <<EOT bash
-    if [ "${CMAKE_FROM_SOURCE}}" = "true" ]; then
-        curl -L -s https://github.com/Kitware/CMake/releases/download/v${CMAKE_VERSION}/cmake-${CMAKE_VERSION}.tar.gz -o cmake.tar.gz && tar xvf cmake.tar.gz && cd cmake-${CMAKE_VERSION} && ./configure && make && make install
-    else
-        apt-get update && \
-        apt-get install -y \
-            cmake && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/*
-    fi
-EOT
-
-# We install GRPC to a different prefix here so that we can copy in only the build artifacts later
-# saves several hundred MB on the final docker image size vs copying in the entire GRPC source tree
-# and running make install in the target container
-RUN git clone --recurse-submodules --jobs 4 -b ${GRPC_VERSION} --depth 1 --shallow-submodules https://github.com/grpc/grpc && \
-    mkdir -p /build/grpc/cmake/build && \
-    cd /build/grpc/cmake/build && \
-    sed -i "216i\  TESTONLY" "../../third_party/abseil-cpp/absl/container/CMakeLists.txt" && \
-    cmake -DgRPC_INSTALL=ON -DgRPC_BUILD_TESTS=OFF -DCMAKE_INSTALL_PREFIX:PATH=/opt/grpc ../.. && \
-    make && \
-    make install && \
-    rm -rf /build
-
-###################################
-###################################
-
-# The builder-base target has the arguments, variables, and copies shared between full builder images and the uncompiled devcontainer
-
-FROM requirements-drivers AS builder-base
-
-ARG GO_TAGS="stablediffusion tts p2p"
+ARG GO_TAGS="stablediffusion tts"
 ARG GRPC_BACKENDS
-ARG MAKEFLAGS
-ARG LD_FLAGS="-s -w"
-
+ARG BUILD_GRPC=true
 ENV GRPC_BACKENDS=${GRPC_BACKENDS}
 ENV GO_TAGS=${GO_TAGS}
-ENV MAKEFLAGS=${MAKEFLAGS}
 ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
 ENV NVIDIA_REQUIRE_CUDA="cuda>=${CUDA_MAJOR_VERSION}.0"
 ENV NVIDIA_VISIBLE_DEVICES=all
-ENV LD_FLAGS=${LD_FLAGS}
 
-RUN echo "GO_TAGS: $GO_TAGS" && echo "TARGETARCH: $TARGETARCH"
-
-WORKDIR /build
-
-
-# We need protoc installed, and the version in 22.04 is too old.  We will create one as part installing the GRPC build below
-# but that will also being in a newer version of absl which stablediffusion cannot compile with.  This version of protoc is only
-# here so that we can generate the grpc code for the stablediffusion build
-RUN <<EOT bash
-    if [ "amd64" = "$TARGETARCH" ]; then
-        curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v27.1/protoc-27.1-linux-x86_64.zip -o protoc.zip && \
-        unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
-        rm protoc.zip
-    fi
-    if [ "arm64" = "$TARGETARCH" ]; then
-        curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v27.1/protoc-27.1-linux-aarch_64.zip -o protoc.zip && \
-        unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
-        rm protoc.zip
-    fi
-EOT
-
-
-###################################
-###################################
-
-# This first portion of builder holds the layers specifically used to build backend-assets/grpc/stablediffusion
-# In most cases, builder is the image you should be using - however, this can save build time if one just needs to copy backend-assets/grpc/stablediffusion and nothing else.
-FROM builder-base AS builder-sd
-
-# stablediffusion does not tolerate a newer version of abseil, copy only over enough elements to build it
-COPY Makefile .
-COPY go.mod .
-COPY go.sum .
-COPY backend/backend.proto ./backend/backend.proto
-COPY backend/go/image/stablediffusion ./backend/go/image/stablediffusion
-COPY pkg/grpc ./pkg/grpc
-COPY pkg/stablediffusion ./pkg/stablediffusion
-RUN git init
-RUN make sources/go-stable-diffusion
-RUN touch prepare-sources
-
-# Actually build the backend
-RUN GRPC_BACKENDS=backend-assets/grpc/stablediffusion make backend-assets/grpc/stablediffusion
-
-###################################
-###################################
-
-# The builder target compiles LocalAI. This target is not the target that will be uploaded to the registry.
-# Adjustments to the build process should likely be made here.
-FROM builder-sd AS builder
-
-# Install the pre-built GRPC
-COPY --from=grpc /opt/grpc /usr/local
-
-# Rebuild with defaults backends
 WORKDIR /build
 
 COPY . .
 COPY .git .
-
 RUN make prepare
 
-## Build the binary
-## If it's CUDA or hipblas, we want to skip some of the llama-compat backends to save space
-## We only leave the most CPU-optimized variant and the fallback for the cublas/hipblas build
-## (both will use CUDA or hipblas for the actual computation)
-RUN if [ "${BUILD_TYPE}" = "cublas" ] || [ "${BUILD_TYPE}" = "hipblas" ]; then \
-        SKIP_GRPC_BACKEND="backend-assets/grpc/llama-cpp-avx backend-assets/grpc/llama-cpp-avx2" make build; \
-    else \
-        make build; \
-    fi
+# If we are building with clblas support, we need the libraries for the builds
+RUN if [ "${BUILD_TYPE}" = "clblas" ]; then \
+    apt-get update && \
+    apt-get install -y libclblast-dev && \
+    apt-get clean \
+    ; fi
+
+# stablediffusion does not tolerate a newer version of abseil, build it first
+RUN GRPC_BACKENDS=backend-assets/grpc/stablediffusion make build
+
+RUN if [ "${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 \
+    mkdir -p /build/sources/go-piper/piper-phonemize/pi/lib/ \
+    touch /build/sources/go-piper/piper-phonemize/pi/lib/keep \
     ; fi
 
 ###################################
 ###################################
 
-# The devcontainer target is not used on CI. It is a target for developers to use locally -
-# rather than copying files it mounts them locally and leaves building to the developer
-
-FROM builder-base AS devcontainer
-
-ARG FFMPEG
-
-COPY --from=grpc /opt/grpc /usr/local
-
-COPY --from=builder-sd /build/backend-assets/grpc/stablediffusion /build/backend-assets/grpc/stablediffusion
-
-COPY .devcontainer-scripts /.devcontainer-scripts
-
-# Add FFmpeg
-RUN if [ "${FFMPEG}" = "true" ]; then \
-        apt-get update && \
-        apt-get install -y --no-install-recommends \
-            ffmpeg && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/* \
-    ; fi
-
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends \
-        ssh less wget
-# For the devcontainer, leave apt functional in case additional devtools are needed at runtime.
-
-RUN go install github.com/go-delve/delve/cmd/dlv@latest
-
-RUN go install github.com/mikefarah/yq/v4@latest
-
-###################################
-###################################
-
-# This is the final target. The result of this target will be the image uploaded to the registry.
-# If you cannot find a more suitable place for an addition, this layer is a suitable place for it.
-FROM requirements-drivers
+FROM requirements-${IMAGE_TYPE}
 
 ARG FFMPEG
 ARG BUILD_TYPE
 ARG TARGETARCH
 ARG IMAGE_TYPE=extras
-ARG EXTRA_BACKENDS
-ARG MAKEFLAGS
 
 ENV BUILD_TYPE=${BUILD_TYPE}
 ENV REBUILD=false
 ENV HEALTHCHECK_ENDPOINT=http://localhost:8080/readyz
-ENV MAKEFLAGS=${MAKEFLAGS}
 
-ARG CUDA_MAJOR_VERSION=12
+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 [ "${FFMPEG}" = "true" ]; then \
-        apt-get update && \
-        apt-get install -y --no-install-recommends \
-            ffmpeg && \
-        apt-get clean && \
-        rm -rf /var/lib/apt/lists/* \
+    apt-get install -y ffmpeg && apt-get clean \
+    ; fi
+
+# Add OpenCL
+RUN if [ "${BUILD_TYPE}" = "clblas" ]; then \
+    apt-get update && \
+    apt-get install -y libclblast1 && \
+    apt-get clean \
     ; fi
 
 WORKDIR /build
@@ -416,9 +171,9 @@ WORKDIR /build
 COPY . .
 
 COPY --from=builder /build/sources ./sources/
-COPY --from=grpc /opt/grpc /usr/local
+COPY --from=builder /build/grpc ./grpc/
 
-RUN make prepare-sources
+RUN make prepare-sources && cd /build/grpc/cmake/build && make install && rm -rf grpc
 
 # Copy the binary
 COPY --from=builder /build/local-ai ./
@@ -427,57 +182,47 @@ COPY --from=builder /build/local-ai ./
 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-sd /build/backend-assets/grpc/stablediffusion ./backend-assets/grpc/stablediffusion
+COPY --from=builder /build/backend-assets/grpc/stablediffusion ./backend-assets/grpc/stablediffusion
 
-# Change the shell to bash so we can use [[ tests below
-SHELL ["/bin/bash", "-c"]
-# We try to strike a balance between individual layer size (as that affects total push time) and total image size
-# Splitting the backends into more groups with fewer items results in a larger image, but a smaller size for the largest layer
-# Splitting the backends into fewer groups with more items results in a smaller image, but a larger size for the largest layer
-
-RUN if [[ ( "${EXTRA_BACKENDS}" =~ "coqui" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/coqui \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "parler-tts" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/parler-tts \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "diffusers" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/diffusers \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "transformers-musicgen" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/transformers-musicgen \
+## Duplicated from Makefile to avoid having a big layer that's hard to push
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/autogptq \
     ; fi
-
-RUN if [[ ( "${EXTRA_BACKENDS}" =~ "vall-e-x" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/vall-e-x \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "openvoice" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/openvoice \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "sentencetransformers" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/sentencetransformers \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "exllama2" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/exllama2 \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "transformers" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/transformers \
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/bark \
     ; fi
-
-RUN if [[ ( "${EXTRA_BACKENDS}" =~ "vllm" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/vllm \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "autogptq" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/autogptq \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "bark" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/bark \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "rerankers" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/rerankers \
-    ; fi && \
-    if [[ ( "${EXTRA_BACKENDS}" =~ "mamba" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \
-        make -C backend/python/mamba \
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/diffusers \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/vllm \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/mamba \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/sentencetransformers \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/transformers \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/vall-e-x \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/exllama \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+     make -C backend/python/exllama2 \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/petals \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/transformers-musicgen \
+    ; fi
+RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
+	 make -C backend/python/coqui \
     ; fi
 
 # Make sure the models directory exists
@@ -485,8 +230,7 @@ RUN mkdir -p /build/models
 
 # Define the health check command
 HEALTHCHECK --interval=1m --timeout=10m --retries=10 \
-  CMD curl -f ${HEALTHCHECK_ENDPOINT} || exit 1
+  CMD curl -f $HEALTHCHECK_ENDPOINT || exit 1
 
-VOLUME /build/models
 EXPOSE 8080
 ENTRYPOINT [ "/build/entrypoint.sh" ]

Dockerfile.aio

@@ -1,8 +0,0 @@
-ARG BASE_IMAGE=ubuntu:22.04
-
-FROM ${BASE_IMAGE} 
-
-RUN apt-get update && apt-get install -y pciutils && apt-get clean
-
-COPY aio/ /aio
-ENTRYPOINT [ "/aio/entrypoint.sh" ]

README.md

@@ -20,14 +20,14 @@
 </a>
 </p>
 
-<p align="center">
-<a href="https://hub.docker.com/r/localai/localai" target="blank">
-<img src="https://img.shields.io/badge/dockerhub-images-important.svg?logo=Docker" alt="LocalAI Docker hub"/>
-</a>
-<a href="https://quay.io/repository/go-skynet/local-ai?tab=tags&tag=latest" target="blank">
-<img src="https://img.shields.io/badge/quay.io-images-important.svg?" alt="LocalAI Quay.io"/>
-</a>
-</p>
+[<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)
+
+> :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)
+
+[![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)
 
 <p align="center">
 <a href="https://twitter.com/LocalAI_API" target="blank">
@@ -36,108 +36,53 @@
 <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>
-</p>
 
-<p align="center">
-<a href="https://trendshift.io/repositories/1484" target="_blank"><img src="https://trendshift.io/api/badge/repositories/1484" alt="go-skynet%2FLocalAI | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
-</p>
+**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.
 
-> :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/) [🖼️ Models](https://models.localai.io/) [🚀 Roadmap](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap) [🥽 Demo](https://demo.localai.io) [🌍 Explorer](https://explorer.localai.io) [🛫 Examples](https://github.com/mudler/LocalAI-examples) 
+## 🔥🔥 Hot topics / Roadmap
 
-[![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)
+[Roadmap](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap)
 
-**LocalAI** is the free, Open Source OpenAI alternative. LocalAI act as a drop-in replacement REST API that’s compatible with OpenAI (Elevenlabs, Anthropic... ) API specifications for local AI 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. It is created and maintained by [Ettore Di Giacinto](https://github.com/mudler).
+- Parallel function calling: https://github.com/mudler/LocalAI/pull/1726
+- Upload file API: https://github.com/mudler/LocalAI/pull/1703
+- Tools API support: https://github.com/mudler/LocalAI/pull/1715
+- LLaVa 1.6: https://github.com/mudler/LocalAI/pull/1714
+- ROCm container images: https://github.com/mudler/LocalAI/pull/1595
+- Intel GPU support (sycl, transformers, diffusers): https://github.com/mudler/LocalAI/issues/1653
+- Deprecation of old backends: https://github.com/mudler/LocalAI/issues/1651
+- 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
+- Img2vid https://github.com/mudler/LocalAI/pull/1442
 
-![screen](https://github.com/mudler/LocalAI/assets/2420543/20b5ccd2-8393-44f0-aaf6-87a23806381e)
-
-Run the installer script:
-
-```bash
-curl https://localai.io/install.sh | sh
-```
-
-Or run with docker:
-```bash
-# CPU only image:
-docker run -ti --name local-ai -p 8080:8080 localai/localai:latest-cpu
-
-# Nvidia GPU:
-docker run -ti --name local-ai -p 8080:8080 --gpus all localai/localai:latest-gpu-nvidia-cuda-12
-
-# CPU and GPU image (bigger size):
-docker run -ti --name local-ai -p 8080:8080 localai/localai:latest
-
-# AIO images (it will pre-download a set of models ready for use, see https://localai.io/basics/container/)
-docker run -ti --name local-ai -p 8080:8080 localai/localai:latest-aio-cpu
-```
-
-To load models:
-
-```bash
-# From the model gallery (see available models with `local-ai models list`, in the WebUI from the model tab, or visiting https://models.localai.io)
-local-ai run llama-3.2-1b-instruct:q4_k_m
-# Start LocalAI with the phi-2 model directly from huggingface
-local-ai run huggingface://TheBloke/phi-2-GGUF/phi-2.Q8_0.gguf
-# Install and run a model from the Ollama OCI registry
-local-ai run ollama://gemma:2b
-# Run a model from a configuration file
-local-ai run https://gist.githubusercontent.com/.../phi-2.yaml
-# Install and run a model from a standard OCI registry (e.g., Docker Hub)
-local-ai run oci://localai/phi-2:latest
-```
-
-[💻 Getting started](https://localai.io/basics/getting_started/index.html)
-
-## 📰 Latest project news
-
-- Dec 2024: stablediffusion.cpp backend (ggml) added ( https://github.com/mudler/LocalAI/pull/4289 )
-- Nov 2024: Bark.cpp backend added ( https://github.com/mudler/LocalAI/pull/4287 )
-- Nov 2024: Voice activity detection models (**VAD**) added to the API: https://github.com/mudler/LocalAI/pull/4204
-- Oct 2024: examples moved to [LocalAI-examples](https://github.com/mudler/LocalAI-examples)
-- Aug 2024:  🆕 FLUX-1, [P2P Explorer](https://explorer.localai.io)
-- July 2024: 🔥🔥 🆕 P2P Dashboard, LocalAI Federated mode and AI Swarms: https://github.com/mudler/LocalAI/pull/2723
-- June 2024: 🆕 You can browse now the model gallery without LocalAI! Check out https://models.localai.io
-- June 2024: Support for models from OCI registries: https://github.com/mudler/LocalAI/pull/2628
-- May 2024: 🔥🔥 Decentralized P2P llama.cpp:  https://github.com/mudler/LocalAI/pull/2343 (peer2peer llama.cpp!) 👉 Docs  https://localai.io/features/distribute/
-- May 2024: 🔥🔥 Openvoice: https://github.com/mudler/LocalAI/pull/2334
-- May 2024: 🆕 Function calls without grammars and mixed mode: https://github.com/mudler/LocalAI/pull/2328
-- May 2024: 🔥🔥 Distributed inferencing: https://github.com/mudler/LocalAI/pull/2324
-- May 2024: Chat, TTS, and Image generation in the WebUI: https://github.com/mudler/LocalAI/pull/2222
-- April 2024: Reranker API: https://github.com/mudler/LocalAI/pull/2121
-
-Roadmap items: [List of issues](https://github.com/mudler/LocalAI/issues?q=is%3Aissue+is%3Aopen+label%3Aroadmap)
-
-## 🔥🔥 Hot topics (looking for help):
-
-- Multimodal with vLLM and Video understanding: https://github.com/mudler/LocalAI/pull/3729
-- Realtime API https://github.com/mudler/LocalAI/issues/3714
-- 🔥🔥 Distributed, P2P Global community pools: https://github.com/mudler/LocalAI/issues/3113
-- WebUI improvements: https://github.com/mudler/LocalAI/issues/2156
+Hot topics (looking for contributors):
 - Backends v2: https://github.com/mudler/LocalAI/issues/1126
 - Improving UX v2: https://github.com/mudler/LocalAI/issues/1373
 - Assistant API: https://github.com/mudler/LocalAI/issues/1273
 - Moderation endpoint: https://github.com/mudler/LocalAI/issues/999
 - Vulkan: https://github.com/mudler/LocalAI/issues/1647
-- Anthropic API: https://github.com/mudler/LocalAI/issues/1808
 
 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
 
+## 💻 [Getting started](https://localai.io/basics/getting_started/index.html)
+
+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`:
+
+```
+docker run -ti -p 8080:8080 localai/localai:v2.9.0-ffmpeg-core phi-2
+```
+
 ## 🚀 [Features](https://localai.io/features/)
 
 - 📖 [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-alike tools API](https://localai.io/features/openai-functions/) 
+- 🔥 [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/)
-- 📈 [Reranker API](https://localai.io/features/reranker/)
-- 🆕🖧 [P2P Inferencing](https://localai.io/features/distribute/)
-- 🌍 Integrated WebUI!
+- 🆕 [Vision API](https://localai.io/features/gpt-vision/)
 
 ## 💻 Usage
 
@@ -151,7 +96,6 @@ Build and deploy custom containers:
 WebUIs:
 - https://github.com/Jirubizu/localai-admin
 - https://github.com/go-skynet/LocalAI-frontend
-- QA-Pilot(An interactive chat project that leverages LocalAI LLMs for rapid understanding and navigation of GitHub code repository) https://github.com/reid41/QA-Pilot
 
 Model galleries
 - https://github.com/go-skynet/model-gallery
@@ -159,23 +103,17 @@ Model galleries
 Other:
 - Helm chart https://github.com/go-skynet/helm-charts
 - VSCode extension https://github.com/badgooooor/localai-vscode-plugin
-- Terminal utility https://github.com/djcopley/ShellOracle
 - Local Smart assistant https://github.com/mudler/LocalAGI
-- Home Assistant https://github.com/sammcj/homeassistant-localai / https://github.com/drndos/hass-openai-custom-conversation / https://github.com/valentinfrlch/ha-gpt4vision
+- 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
-- Shell-Pilot(Interact with LLM using LocalAI models via pure shell scripts on your Linux or MacOS system) https://github.com/reid41/shell-pilot
 - Telegram bot https://github.com/mudler/LocalAI/tree/master/examples/telegram-bot
-- Another Telegram Bot https://github.com/JackBekket/Hellper
-- Auto-documentation https://github.com/JackBekket/Reflexia
-- Github bot which answer on issues, with code and documentation as context https://github.com/JackBekket/GitHelper
-- Github Actions: https://github.com/marketplace/actions/start-localai
 - Examples: https://github.com/mudler/LocalAI/tree/master/examples/
   
 
 ### 🔗 Resources
 
-- [LLM finetuning guide](https://localai.io/docs/advanced/fine-tuning/)
+- 🆕 New! [LLM finetuning guide](https://localai.io/docs/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/docs/integrations/)
@@ -183,9 +121,7 @@ Other:
 
 ## :book: 🎥 [Media, Blogs, Social](https://localai.io/basics/news/#media-blogs-social)
 
-- [Run Visual studio code with LocalAI (SUSE)](https://www.suse.com/c/running-ai-locally/)
-- 🆕 [Run LocalAI on Jetson Nano Devkit](https://mudler.pm/posts/local-ai-jetson-nano-devkit/)
-- [Run LocalAI on AWS EKS with Pulumi](https://www.pulumi.com/blog/low-code-llm-apps-with-local-ai-flowise-and-pulumi/)
+- [Run LocalAI on AWS EKS with Pulumi](https://www.pulumi.com/ai/answers/tiZMDoZzZV6TLxgDXNBnFE/deploying-helm-charts-on-aws-eks)
 - [Run LocalAI on AWS](https://staleks.hashnode.dev/installing-localai-on-aws-ec2-instance)
 - [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)
@@ -212,16 +148,17 @@ If you utilize this repository, data in a downstream project, please consider ci
 
 Support the project by becoming [a backer or sponsor](https://github.com/sponsors/mudler). Your logo will show up here with a link to your website.
 
-A huge thank you to our generous sponsors who support this project covering CI expenses, and our [Sponsor list](https://github.com/sponsors/mudler):
+A huge thank you to our generous sponsors who support this project:
 
-<p align="center">
-  <a href="https://www.spectrocloud.com/" target="blank">
-    <img height="200" src="https://github.com/go-skynet/LocalAI/assets/2420543/68a6f3cb-8a65-4a4d-99b5-6417a8905512">
-  </a>
-  <a href="https://www.premai.io/" target="blank">
-    <img height="200" src="https://github.com/mudler/LocalAI/assets/2420543/42e4ca83-661e-4f79-8e46-ae43689683d6"> <br>
-  </a>
-</p>
+| ![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 kindly supports LocalAI by providing GPU and computing resources to run tests on lamdalabs!  |
+
+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
 
@@ -231,7 +168,7 @@ A huge thank you to our generous sponsors who support this project covering CI e
 
 LocalAI is a community-driven project created by [Ettore Di Giacinto](https://github.com/mudler/).
 
-MIT - Author Ettore Di Giacinto <mudler@localai.io>
+MIT - Author Ettore Di Giacinto
 
 ## 🙇 Acknowledgements
 
@@ -243,6 +180,7 @@ LocalAI couldn't have been built without the help of great software already avai
 - https://github.com/antimatter15/alpaca.cpp
 - https://github.com/EdVince/Stable-Diffusion-NCNN
 - https://github.com/ggerganov/whisper.cpp
+- https://github.com/saharNooby/rwkv.cpp
 - https://github.com/rhasspy/piper
 
 ## 🤗 Contributors

aio/cpu/README.md

@@ -1,5 +0,0 @@
-## AIO CPU size
-
-Use this image with CPU-only.
-
-Please keep using only C++ backends so the base image is as small as possible (without CUDA, cuDNN, python, etc).

aio/cpu/embeddings.yaml

@@ -1,12 +0,0 @@
-name: text-embedding-ada-002
-embeddings: true
-parameters:
-  model: huggingface://hugging-quants/Llama-3.2-1B-Instruct-Q4_K_M-GGUF/llama-3.2-1b-instruct-q4_k_m.gguf
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/embeddings -X POST -H "Content-Type: application/json" -d '{
-      "input": "Your text string goes here",
-      "model": "text-embedding-ada-002"
-    }'

aio/cpu/image-gen.yaml

@@ -1,62 +0,0 @@
-name: stablediffusion
-backend: stablediffusion
-parameters:
-  model: stablediffusion_assets
-
-license: "BSD-3"
-urls:
-- https://github.com/EdVince/Stable-Diffusion-NCNN
-- https://github.com/EdVince/Stable-Diffusion-NCNN/blob/main/LICENSE
-
-description: |
-     Stable Diffusion in NCNN with c++, supported txt2img and img2img
-
-download_files:
-- filename: "stablediffusion_assets/AutoencoderKL-256-256-fp16-opt.param"
-  sha256: "18ca4b66685e21406bcf64c484b3b680b4949900415536d599cc876579c85c82"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/AutoencoderKL-256-256-fp16-opt.param"
-- filename: "stablediffusion_assets/AutoencoderKL-512-512-fp16-opt.param"
-  sha256: "cf45f63aacf3dbbab0f59ed92a6f2c14d9a1801314631cd3abe91e3c85639a20"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/AutoencoderKL-512-512-fp16-opt.param"
-- filename: "stablediffusion_assets/AutoencoderKL-base-fp16.param"
-  sha256: "0254a056dce61b0c27dc9ec1b78b53bcf55315c540f55f051eb841aa992701ba"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/AutoencoderKL-base-fp16.param"
-- filename: "stablediffusion_assets/AutoencoderKL-encoder-512-512-fp16.bin"
-  sha256: "ddcb79a9951b9f91e05e087739ed69da2c1c4ae30ba4168cce350b49d617c9fa"
-  uri: "https://github.com/EdVince/Stable-Diffusion-NCNN/releases/download/naifu/AutoencoderKL-encoder-512-512-fp16.bin"
-- filename: "stablediffusion_assets/AutoencoderKL-fp16.bin"
-  sha256: "f02e71f80e70252734724bbfaed5c4ddd3a8ed7e61bb2175ff5f53099f0e35dd"
-  uri: "https://github.com/EdVince/Stable-Diffusion-NCNN/releases/download/naifu/AutoencoderKL-fp16.bin"
-- filename: "stablediffusion_assets/FrozenCLIPEmbedder-fp16.bin"
-  sha256: "1c9a12f4e1dd1b295a388045f7f28a2352a4d70c3dc96a542189a3dd7051fdd6"
-  uri: "https://github.com/EdVince/Stable-Diffusion-NCNN/releases/download/naifu/FrozenCLIPEmbedder-fp16.bin"
-- filename: "stablediffusion_assets/FrozenCLIPEmbedder-fp16.param"
-  sha256: "471afbe678dd1fd3fe764ef9c6eccaccb0a7d7e601f27b462aa926b20eb368c9"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/FrozenCLIPEmbedder-fp16.param"
-- filename: "stablediffusion_assets/log_sigmas.bin"
-  sha256: "a2089f8aa4c61f9c200feaec541ab3f5c94233b28deb6d5e8bcd974fa79b68ac"
-  uri: "https://github.com/EdVince/Stable-Diffusion-NCNN/raw/main/x86/linux/assets/log_sigmas.bin"
-- filename: "stablediffusion_assets/UNetModel-256-256-MHA-fp16-opt.param"
-  sha256: "a58c380229f09491776df837b7aa7adffc0a87821dc4708b34535da2e36e3da1"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/UNetModel-256-256-MHA-fp16-opt.param"
-- filename: "stablediffusion_assets/UNetModel-512-512-MHA-fp16-opt.param"
-  sha256: "f12034067062827bd7f43d1d21888d1f03905401acf6c6eea22be23c259636fa"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/UNetModel-512-512-MHA-fp16-opt.param"
-- filename: "stablediffusion_assets/UNetModel-base-MHA-fp16.param"
-  sha256: "696f6975de49f4325b53ce32aff81861a6d6c07cd9ce3f0aae2cc405350af38d"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/UNetModel-base-MHA-fp16.param"
-- filename: "stablediffusion_assets/UNetModel-MHA-fp16.bin"
-  sha256: "d618918d011bfc1f644c0f2a33bf84931bd53b28a98492b0a8ed6f3a818852c3"
-  uri: "https://github.com/EdVince/Stable-Diffusion-NCNN/releases/download/naifu/UNetModel-MHA-fp16.bin"
-- filename: "stablediffusion_assets/vocab.txt"
-  sha256: "e30e57b6f1e47616982ef898d8922be24e535b4fa3d0110477b3a6f02ebbae7d"
-  uri: "https://raw.githubusercontent.com/EdVince/Stable-Diffusion-NCNN/main/x86/linux/assets/vocab.txt"
-
-usage: |
-        curl http://localhost:8080/v1/images/generations \
-          -H "Content-Type: application/json" \
-          -d '{
-            "prompt": "<positive prompt>|<negative prompt>",
-            "step": 25,
-            "size": "512x512"
-          }'

aio/cpu/rerank.yaml

@@ -1,27 +0,0 @@
-name: jina-reranker-v1-base-en
-backend: rerankers
-parameters:
-  model: cross-encoder
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/v1/rerank \
-      -H "Content-Type: application/json" \
-      -d '{
-      "model": "jina-reranker-v1-base-en",
-      "query": "Organic skincare products for sensitive skin",
-      "documents": [
-        "Eco-friendly kitchenware for modern homes",
-        "Biodegradable cleaning supplies for eco-conscious consumers",
-        "Organic cotton baby clothes for sensitive skin",
-        "Natural organic skincare range for sensitive skin",
-        "Tech gadgets for smart homes: 2024 edition",
-        "Sustainable gardening tools and compost solutions",
-        "Sensitive skin-friendly facial cleansers and toners",
-        "Organic food wraps and storage solutions",
-        "All-natural pet food for dogs with allergies",
-        "Yoga mats made from recycled materials"
-      ],
-      "top_n": 3
-    }'

aio/cpu/speech-to-text.yaml

@@ -1,18 +0,0 @@
-name: whisper-1
-backend: whisper
-parameters:
-  model: ggml-whisper-base.bin
-
-usage: |
-    ## example audio file
-    wget --quiet --show-progress -O gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
-
-    ## Send the example audio file to the transcriptions endpoint
-    curl http://localhost:8080/v1/audio/transcriptions \
-         -H "Content-Type: multipart/form-data" \
-         -F file="@$PWD/gb1.ogg" -F model="whisper-1"
-
-download_files:
-- filename: "ggml-whisper-base.bin"
-  sha256: "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
-  uri: "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin"

aio/cpu/text-to-speech.yaml

@@ -1,15 +0,0 @@
-name: tts-1
-download_files:
-  - filename: voice-en-us-amy-low.tar.gz
-    uri: https://github.com/rhasspy/piper/releases/download/v0.0.2/voice-en-us-amy-low.tar.gz
-
-parameters:
-  model: en-us-amy-low.onnx
-
-usage: |
-    To test if this model works as expected, you can use the following curl command:
-
-    curl http://localhost:8080/tts -H "Content-Type: application/json" -d '{
-      "model":"voice-en-us-amy-low",
-      "input": "Hi, this is a test."
-    }'

aio/cpu/text-to-text.yaml

@@ -1,101 +0,0 @@
-name: gpt-4
-mmap: true
-parameters:
-  model: huggingface://NousResearch/Hermes-2-Pro-Llama-3-8B-GGUF/Hermes-2-Pro-Llama-3-8B-Q4_K_M.gguf
-context_size: 8192
-
-stopwords:
-- "<|im_end|>"
-- "<dummy32000>"
-- "</tool_call>"
-- "<|eot_id|>"
-- "<|end_of_text|>"
-
-function:
-  # disable injecting the "answer" tool
-  disable_no_action: true
-
-  grammar:
-    # This allows the grammar to also return messages
-    mixed_mode: true
-    # Suffix to add to the grammar
-    #prefix: '<tool_call>\n'
-    # Force parallel calls in the grammar
-    # parallel_calls: true
-
-  return_name_in_function_response: true
-  # Without grammar uncomment the lines below
-  # Warning: this is relying only on the capability of the
-  # LLM model to generate the correct function call.
-  json_regex_match: 
-   - "(?s)<tool_call>(.*?)</tool_call>"
-   - "(?s)<tool_call>(.*?)"
-  replace_llm_results:
-  # Drop the scratchpad content from responses
-  - key: "(?s)<scratchpad>.*</scratchpad>"
-    value: ""
-  replace_function_results: 
-  # Replace everything that is not JSON array or object
-  # 
-  - key: '(?s)^[^{\[]*'
-    value: ""
-  - key: '(?s)[^}\]]*$'
-    value: ""
-  - key: "'([^']*?)'"
-    value: "_DQUOTE_${1}_DQUOTE_"
-  - key: '\\"'
-    value: "__TEMP_QUOTE__"
-  - key: "\'"
-    value: "'"
-  - key: "_DQUOTE_"
-    value: '"'
-  - key: "__TEMP_QUOTE__"
-    value: '"'
-  # Drop the scratchpad content from responses
-  - key: "(?s)<scratchpad>.*</scratchpad>"
-    value: ""
-
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{if eq .RoleName "assistant"}}assistant{{else if eq .RoleName "system"}}system{{else if eq .RoleName "tool"}}tool{{else if eq .RoleName "user"}}user{{end}}
-    {{- if .FunctionCall }}
-    <tool_call>
-    {{- else if eq .RoleName "tool" }}
-    <tool_response>
-    {{- end }}
-    {{- if .Content}}
-    {{.Content }}
-    {{- end }}
-    {{- if .FunctionCall}}
-    {{toJson .FunctionCall}}
-    {{- end }}
-    {{- if .FunctionCall }}
-    </tool_call>
-    {{- else if eq .RoleName "tool" }}
-    </tool_response>
-    {{- end }}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |-
-    <|im_start|>system
-    You are a function calling AI model.
-    Here are the available tools:
-    <tools>
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    </tools>
-    You should call the tools provided to you sequentially
-    Please use <scratchpad> XML tags to record your reasoning and planning before you call the functions as follows:
-    <scratchpad>
-    {step-by-step reasoning and plan in bullet points}
-    </scratchpad>
-    For each function call return a json object with function name and arguments within <tool_call> XML tags as follows:
-    <tool_call>
-    {"arguments": <args-dict>, "name": <function-name>}
-    </tool_call><|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant

aio/cpu/vision.yaml

@@ -1,31 +0,0 @@
-backend: llama-cpp
-context_size: 4096
-f16: true
-mmap: true
-name: gpt-4o
-
-roles:
-  user: "USER:"
-  assistant: "ASSISTANT:"
-  system: "SYSTEM:"
-
-mmproj: bakllava-mmproj.gguf
-parameters:
-  model: bakllava.gguf
-
-template:
-  chat: |
-    A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.
-    {{.Input}}
-    ASSISTANT:
-
-download_files:
-- filename: bakllava.gguf
-  uri: huggingface://mys/ggml_bakllava-1/ggml-model-q4_k.gguf
-- filename: bakllava-mmproj.gguf
-  uri: huggingface://mys/ggml_bakllava-1/mmproj-model-f16.gguf
-
-usage: |
-    curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
-        "model": "gpt-4-vision-preview",
-        "messages": [{"role": "user", "content": [{"type":"text", "text": "What is in the image?"}, {"type": "image_url", "image_url": {"url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg" }}], "temperature": 0.9}]}'

aio/entrypoint.sh

@@ -1,138 +0,0 @@
-#!/bin/bash
-
-echo "===> LocalAI All-in-One (AIO) container starting..."
-
-GPU_ACCELERATION=false
-GPU_VENDOR=""
-
-function check_intel() {
-    if lspci | grep -E 'VGA|3D' | grep -iq intel; then
-        echo "Intel GPU detected"
-        if [ -d /opt/intel ]; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=intel
-        else
-            echo "Intel GPU detected, but Intel GPU drivers are not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_nvidia_wsl() {
-    if lspci | grep -E 'VGA|3D' | grep -iq "Microsoft Corporation Device 008e"; then
-        # We make the assumption this WSL2 cars is NVIDIA, then check for nvidia-smi
-        # Make sure the container was run with `--gpus all` as the only required parameter
-        echo "NVIDIA GPU detected via WSL2"
-        # nvidia-smi should be installed in the container
-        if nvidia-smi; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=nvidia
-        else
-            echo "NVIDIA GPU detected via WSL2, but nvidia-smi is not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_amd() {
-    if lspci | grep -E 'VGA|3D' | grep -iq amd; then
-        echo "AMD GPU detected"
-        # Check if ROCm is installed
-        if [ -d /opt/rocm ]; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=amd
-        else
-            echo "AMD GPU detected, but ROCm is not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_nvidia() {
-    if lspci | grep -E 'VGA|3D' | grep -iq nvidia; then
-        echo "NVIDIA GPU detected"
-        # nvidia-smi should be installed in the container
-        if nvidia-smi; then
-            GPU_ACCELERATION=true
-            GPU_VENDOR=nvidia
-        else
-            echo "NVIDIA GPU detected, but nvidia-smi is not installed. GPU acceleration will not be available."
-        fi
-    fi
-}
-
-function check_metal() {
-    if system_profiler SPDisplaysDataType | grep -iq 'Metal'; then
-        echo "Apple Metal supported GPU detected"
-        GPU_ACCELERATION=true
-        GPU_VENDOR=apple
-    fi
-}
-
-function detect_gpu() {
-    case "$(uname -s)" in
-        Linux)
-            check_nvidia
-            check_amd
-            check_intel
-            check_nvidia_wsl
-            ;;
-        Darwin)
-            check_metal
-            ;;
-    esac
-}
-
-function detect_gpu_size() {
-    # Attempting to find GPU memory size for NVIDIA GPUs
-    if [ "$GPU_ACCELERATION" = true ] && [ "$GPU_VENDOR" = "nvidia" ]; then
-        echo "NVIDIA GPU detected. Attempting to find memory size..."
-        # Using head -n 1 to get the total memory of the 1st NVIDIA GPU detected.
-        # If handling multiple GPUs is required in the future, this is the place to do it
-        nvidia_sm=$(nvidia-smi --query-gpu=memory.total --format=csv,noheader,nounits | head -n 1)
-        if [ ! -z "$nvidia_sm" ]; then
-            echo "Total GPU Memory: $nvidia_sm MiB"
-            # if bigger than 8GB, use 16GB
-            #if [ "$nvidia_sm" -gt 8192 ]; then
-            #    GPU_SIZE=gpu-16g
-            #else
-            GPU_SIZE=gpu-8g
-            #fi
-        else
-            echo "Unable to determine NVIDIA GPU memory size. Falling back to CPU."
-            GPU_SIZE=gpu-8g
-        fi
-    elif [ "$GPU_ACCELERATION" = true ] && [ "$GPU_VENDOR" = "intel" ]; then
-        GPU_SIZE=intel
-    # Default to a generic GPU size until we implement GPU size detection for non NVIDIA GPUs
-    elif [ "$GPU_ACCELERATION" = true ]; then
-        echo "Non-NVIDIA GPU detected. Specific GPU memory size detection is not implemented."
-        GPU_SIZE=gpu-8g
-
-    # default to cpu if GPU_SIZE is not set
-    else
-        echo "GPU acceleration is not enabled or supported. Defaulting to CPU."
-        GPU_SIZE=cpu
-    fi
-}
-
-function check_vars() {
-    if [ -z "$MODELS" ]; then
-        echo "MODELS environment variable is not set. Please set it to a comma-separated list of model YAML files to load."
-        exit 1
-    fi
-
-    if [ -z "$PROFILE" ]; then
-        echo "PROFILE environment variable is not set. Please set it to one of the following: cpu, gpu-8g, gpu-16g, apple"
-        exit 1
-    fi
-}
-
-detect_gpu
-detect_gpu_size
-
-PROFILE="${PROFILE:-$GPU_SIZE}" # default to cpu
-export MODELS="${MODELS:-/aio/${PROFILE}/embeddings.yaml,/aio/${PROFILE}/rerank.yaml,/aio/${PROFILE}/text-to-speech.yaml,/aio/${PROFILE}/image-gen.yaml,/aio/${PROFILE}/text-to-text.yaml,/aio/${PROFILE}/speech-to-text.yaml,/aio/${PROFILE}/vision.yaml}"
-
-check_vars
-
-echo "===> Starting LocalAI[$PROFILE] with the following models: $MODELS"
-
-exec /build/entrypoint.sh "$@"

aio/gpu-8g/embeddings.yaml

@@ -1,12 +0,0 @@
-name: text-embedding-ada-002
-backend: sentencetransformers
-parameters:
-  model: all-MiniLM-L6-v2
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/embeddings -X POST -H "Content-Type: application/json" -d '{
-      "input": "Your text string goes here",
-      "model": "text-embedding-ada-002"
-    }'

aio/gpu-8g/image-gen.yaml

@@ -1,25 +0,0 @@
-name: stablediffusion
-parameters:
-  model: DreamShaper_8_pruned.safetensors
-backend: diffusers
-step: 25
-f16: true
-
-diffusers:
-  pipeline_type: StableDiffusionPipeline
-  cuda: true
-  enable_parameters: "negative_prompt,num_inference_steps"
-  scheduler_type: "k_dpmpp_2m"
-
-download_files:
-- filename: DreamShaper_8_pruned.safetensors
-  uri: huggingface://Lykon/DreamShaper/DreamShaper_8_pruned.safetensors
-
-usage: |
-        curl http://localhost:8080/v1/images/generations \
-          -H "Content-Type: application/json" \
-          -d '{
-            "prompt": "<positive prompt>|<negative prompt>",
-            "step": 25,
-            "size": "512x512"
-          }'

aio/gpu-8g/rerank.yaml

@@ -1,27 +0,0 @@
-name: jina-reranker-v1-base-en
-backend: rerankers
-parameters:
-  model: cross-encoder
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/v1/rerank \
-      -H "Content-Type: application/json" \
-      -d '{
-      "model": "jina-reranker-v1-base-en",
-      "query": "Organic skincare products for sensitive skin",
-      "documents": [
-        "Eco-friendly kitchenware for modern homes",
-        "Biodegradable cleaning supplies for eco-conscious consumers",
-        "Organic cotton baby clothes for sensitive skin",
-        "Natural organic skincare range for sensitive skin",
-        "Tech gadgets for smart homes: 2024 edition",
-        "Sustainable gardening tools and compost solutions",
-        "Sensitive skin-friendly facial cleansers and toners",
-        "Organic food wraps and storage solutions",
-        "All-natural pet food for dogs with allergies",
-        "Yoga mats made from recycled materials"
-      ],
-      "top_n": 3
-    }'

aio/gpu-8g/speech-to-text.yaml

@@ -1,18 +0,0 @@
-name: whisper-1
-backend: whisper
-parameters:
-  model: ggml-whisper-base.bin
-
-usage: |
-    ## example audio file
-    wget --quiet --show-progress -O gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
-
-    ## Send the example audio file to the transcriptions endpoint
-    curl http://localhost:8080/v1/audio/transcriptions \
-         -H "Content-Type: multipart/form-data" \
-         -F file="@$PWD/gb1.ogg" -F model="whisper-1"
-
-download_files:
-- filename: "ggml-whisper-base.bin"
-  sha256: "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
-  uri: "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin"

aio/gpu-8g/text-to-speech.yaml

@@ -1,15 +0,0 @@
-name: tts-1
-download_files:
-  - filename: voice-en-us-amy-low.tar.gz
-    uri: https://github.com/rhasspy/piper/releases/download/v0.0.2/voice-en-us-amy-low.tar.gz
-
-parameters:
-  model: en-us-amy-low.onnx
-
-usage: |
-    To test if this model works as expected, you can use the following curl command:
-
-    curl http://localhost:8080/tts -H "Content-Type: application/json" -d '{
-      "model":"tts-1",
-      "input": "Hi, this is a test."
-    }'

aio/gpu-8g/text-to-text.yaml

@@ -1,101 +0,0 @@
-name: gpt-4
-mmap: true
-parameters:
-  model: huggingface://NousResearch/Hermes-2-Pro-Llama-3-8B-GGUF/Hermes-2-Pro-Llama-3-8B-Q4_K_M.gguf
-context_size: 8192
-
-stopwords:
-- "<|im_end|>"
-- "<dummy32000>"
-- "</tool_call>"
-- "<|eot_id|>"
-- "<|end_of_text|>"
-
-function:
-  # disable injecting the "answer" tool
-  disable_no_action: true
-
-  grammar:
-    # This allows the grammar to also return messages
-    mixed_mode: true
-    # Suffix to add to the grammar
-    #prefix: '<tool_call>\n'
-    # Force parallel calls in the grammar
-    # parallel_calls: true
-
-  return_name_in_function_response: true
-  # Without grammar uncomment the lines below
-  # Warning: this is relying only on the capability of the
-  # LLM model to generate the correct function call.
-  json_regex_match: 
-   - "(?s)<tool_call>(.*?)</tool_call>"
-   - "(?s)<tool_call>(.*?)"
-  replace_llm_results:
-  # Drop the scratchpad content from responses
-  - key: "(?s)<scratchpad>.*</scratchpad>"
-    value: ""
-  replace_function_results: 
-  # Replace everything that is not JSON array or object
-  # 
-  - key: '(?s)^[^{\[]*'
-    value: ""
-  - key: '(?s)[^}\]]*$'
-    value: ""
-  - key: "'([^']*?)'"
-    value: "_DQUOTE_${1}_DQUOTE_"
-  - key: '\\"'
-    value: "__TEMP_QUOTE__"
-  - key: "\'"
-    value: "'"
-  - key: "_DQUOTE_"
-    value: '"'
-  - key: "__TEMP_QUOTE__"
-    value: '"'
-  # Drop the scratchpad content from responses
-  - key: "(?s)<scratchpad>.*</scratchpad>"
-    value: ""
-
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{if eq .RoleName "assistant"}}assistant{{else if eq .RoleName "system"}}system{{else if eq .RoleName "tool"}}tool{{else if eq .RoleName "user"}}user{{end}}
-    {{- if .FunctionCall }}
-    <tool_call>
-    {{- else if eq .RoleName "tool" }}
-    <tool_response>
-    {{- end }}
-    {{- if .Content}}
-    {{.Content }}
-    {{- end }}
-    {{- if .FunctionCall}}
-    {{toJson .FunctionCall}}
-    {{- end }}
-    {{- if .FunctionCall }}
-    </tool_call>
-    {{- else if eq .RoleName "tool" }}
-    </tool_response>
-    {{- end }}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |-
-    <|im_start|>system
-    You are a function calling AI model.
-    Here are the available tools:
-    <tools>
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    </tools>
-    You should call the tools provided to you sequentially
-    Please use <scratchpad> XML tags to record your reasoning and planning before you call the functions as follows:
-    <scratchpad>
-    {step-by-step reasoning and plan in bullet points}
-    </scratchpad>
-    For each function call return a json object with function name and arguments within <tool_call> XML tags as follows:
-    <tool_call>
-    {"arguments": <args-dict>, "name": <function-name>}
-    </tool_call><|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant

aio/gpu-8g/vision.yaml

@@ -1,35 +0,0 @@
-backend: llama-cpp
-context_size: 4096
-f16: true
-mmap: true
-name: gpt-4o
-
-roles:
-  user: "USER:"
-  assistant: "ASSISTANT:"
-  system: "SYSTEM:"
-
-mmproj: llava-v1.6-7b-mmproj-f16.gguf
-parameters:
-  model: llava-v1.6-mistral-7b.Q5_K_M.gguf
-  temperature: 0.2
-  top_k: 40
-  top_p: 0.95
-  seed: -1
-
-template:
-  chat: |
-    A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.
-    {{.Input}}
-    ASSISTANT:
-
-download_files:
-- filename: llava-v1.6-mistral-7b.Q5_K_M.gguf
-  uri: huggingface://cjpais/llava-1.6-mistral-7b-gguf/llava-v1.6-mistral-7b.Q5_K_M.gguf
-- filename: llava-v1.6-7b-mmproj-f16.gguf
-  uri: huggingface://cjpais/llava-1.6-mistral-7b-gguf/mmproj-model-f16.gguf
-
-usage: |
-    curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
-        "model": "gpt-4-vision-preview",
-        "messages": [{"role": "user", "content": [{"type":"text", "text": "What is in the image?"}, {"type": "image_url", "image_url": {"url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg" }}], "temperature": 0.9}]}'

aio/intel/embeddings.yaml

@@ -1,12 +0,0 @@
-name: text-embedding-ada-002
-backend: sentencetransformers
-parameters:
-  model: all-MiniLM-L6-v2
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/embeddings -X POST -H "Content-Type: application/json" -d '{
-      "input": "Your text string goes here",
-      "model": "text-embedding-ada-002"
-    }'

aio/intel/image-gen.yaml

@@ -1,20 +0,0 @@
-name: stablediffusion
-parameters:
-  model: Lykon/dreamshaper-8
-backend: diffusers
-step: 25
-f16: true
-diffusers:
-  pipeline_type: StableDiffusionPipeline
-  cuda: true
-  enable_parameters: "negative_prompt,num_inference_steps"
-  scheduler_type: "k_dpmpp_2m"
-
-usage: |
-        curl http://localhost:8080/v1/images/generations \
-          -H "Content-Type: application/json" \
-          -d '{
-            "prompt": "<positive prompt>|<negative prompt>",
-            "step": 25,
-            "size": "512x512"
-          }'

aio/intel/rerank.yaml

@@ -1,27 +0,0 @@
-name: jina-reranker-v1-base-en
-backend: rerankers
-parameters:
-  model: cross-encoder
-
-usage: |
-    You can test this model with curl like this:
-
-    curl http://localhost:8080/v1/rerank \
-      -H "Content-Type: application/json" \
-      -d '{
-      "model": "jina-reranker-v1-base-en",
-      "query": "Organic skincare products for sensitive skin",
-      "documents": [
-        "Eco-friendly kitchenware for modern homes",
-        "Biodegradable cleaning supplies for eco-conscious consumers",
-        "Organic cotton baby clothes for sensitive skin",
-        "Natural organic skincare range for sensitive skin",
-        "Tech gadgets for smart homes: 2024 edition",
-        "Sustainable gardening tools and compost solutions",
-        "Sensitive skin-friendly facial cleansers and toners",
-        "Organic food wraps and storage solutions",
-        "All-natural pet food for dogs with allergies",
-        "Yoga mats made from recycled materials"
-      ],
-      "top_n": 3
-    }'

aio/intel/speech-to-text.yaml

@@ -1,18 +0,0 @@
-name: whisper-1
-backend: whisper
-parameters:
-  model: ggml-whisper-base.bin
-
-usage: |
-    ## example audio file
-    wget --quiet --show-progress -O gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
-
-    ## Send the example audio file to the transcriptions endpoint
-    curl http://localhost:8080/v1/audio/transcriptions \
-         -H "Content-Type: multipart/form-data" \
-         -F file="@$PWD/gb1.ogg" -F model="whisper-1"
-
-download_files:
-- filename: "ggml-whisper-base.bin"
-  sha256: "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
-  uri: "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin"

aio/intel/text-to-speech.yaml

@@ -1,15 +0,0 @@
-name: tts-1
-download_files:
-  - filename: voice-en-us-amy-low.tar.gz
-    uri: https://github.com/rhasspy/piper/releases/download/v0.0.2/voice-en-us-amy-low.tar.gz
-
-parameters:
-  model: en-us-amy-low.onnx
-
-usage: |
-    To test if this model works as expected, you can use the following curl command:
-
-    curl http://localhost:8080/tts -H "Content-Type: application/json" -d '{
-      "model":"tts-1",
-      "input": "Hi, this is a test."
-    }'

aio/intel/text-to-text.yaml

@@ -1,103 +0,0 @@
-name: gpt-4
-mmap: false
-context_size: 8192
-
-f16: false
-parameters:
-  model: huggingface://NousResearch/Hermes-2-Pro-Llama-3-8B-GGUF/Hermes-2-Pro-Llama-3-8B-Q4_K_M.gguf
-
-stopwords:
-- "<|im_end|>"
-- "<dummy32000>"
-- "</tool_call>"
-- "<|eot_id|>"
-- "<|end_of_text|>"
-
-function:
-  # disable injecting the "answer" tool
-  disable_no_action: true
-
-  grammar:
-    # This allows the grammar to also return messages
-    mixed_mode: true
-    # Suffix to add to the grammar
-    #prefix: '<tool_call>\n'
-    # Force parallel calls in the grammar
-    # parallel_calls: true
-
-  return_name_in_function_response: true
-  # Without grammar uncomment the lines below
-  # Warning: this is relying only on the capability of the
-  # LLM model to generate the correct function call.
-  json_regex_match: 
-   - "(?s)<tool_call>(.*?)</tool_call>"
-   - "(?s)<tool_call>(.*?)"
-  replace_llm_results:
-  # Drop the scratchpad content from responses
-  - key: "(?s)<scratchpad>.*</scratchpad>"
-    value: ""
-  replace_function_results: 
-  # Replace everything that is not JSON array or object
-  # 
-  - key: '(?s)^[^{\[]*'
-    value: ""
-  - key: '(?s)[^}\]]*$'
-    value: ""
-  - key: "'([^']*?)'"
-    value: "_DQUOTE_${1}_DQUOTE_"
-  - key: '\\"'
-    value: "__TEMP_QUOTE__"
-  - key: "\'"
-    value: "'"
-  - key: "_DQUOTE_"
-    value: '"'
-  - key: "__TEMP_QUOTE__"
-    value: '"'
-  # Drop the scratchpad content from responses
-  - key: "(?s)<scratchpad>.*</scratchpad>"
-    value: ""
-
-template:
-  chat: |
-    {{.Input -}}
-    <|im_start|>assistant
-  chat_message: |
-    <|im_start|>{{if eq .RoleName "assistant"}}assistant{{else if eq .RoleName "system"}}system{{else if eq .RoleName "tool"}}tool{{else if eq .RoleName "user"}}user{{end}}
-    {{- if .FunctionCall }}
-    <tool_call>
-    {{- else if eq .RoleName "tool" }}
-    <tool_response>
-    {{- end }}
-    {{- if .Content}}
-    {{.Content }}
-    {{- end }}
-    {{- if .FunctionCall}}
-    {{toJson .FunctionCall}}
-    {{- end }}
-    {{- if .FunctionCall }}
-    </tool_call>
-    {{- else if eq .RoleName "tool" }}
-    </tool_response>
-    {{- end }}<|im_end|>
-  completion: |
-    {{.Input}}
-  function: |-
-    <|im_start|>system
-    You are a function calling AI model.
-    Here are the available tools:
-    <tools>
-    {{range .Functions}}
-    {'type': 'function', 'function': {'name': '{{.Name}}', 'description': '{{.Description}}', 'parameters': {{toJson .Parameters}} }}
-    {{end}}
-    </tools>
-    You should call the tools provided to you sequentially
-    Please use <scratchpad> XML tags to record your reasoning and planning before you call the functions as follows:
-    <scratchpad>
-    {step-by-step reasoning and plan in bullet points}
-    </scratchpad>
-    For each function call return a json object with function name and arguments within <tool_call> XML tags as follows:
-    <tool_call>
-    {"arguments": <args-dict>, "name": <function-name>}
-    </tool_call><|im_end|>
-    {{.Input -}}
-    <|im_start|>assistant

aio/intel/vision.yaml

@@ -1,35 +0,0 @@
-backend: llama-cpp
-context_size: 4096
-mmap: false
-f16: false
-name: gpt-4o
-
-roles:
-  user: "USER:"
-  assistant: "ASSISTANT:"
-  system: "SYSTEM:"
-
-mmproj: llava-v1.6-7b-mmproj-f16.gguf
-parameters:
-  model: llava-v1.6-mistral-7b.Q5_K_M.gguf
-  temperature: 0.2
-  top_k: 40
-  top_p: 0.95
-  seed: -1
-
-template:
-  chat: |
-    A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.
-    {{.Input}}
-    ASSISTANT:
-
-download_files:
-- filename: llava-v1.6-mistral-7b.Q5_K_M.gguf
-  uri: huggingface://cjpais/llava-1.6-mistral-7b-gguf/llava-v1.6-mistral-7b.Q5_K_M.gguf
-- filename: llava-v1.6-7b-mmproj-f16.gguf
-  uri: huggingface://cjpais/llava-1.6-mistral-7b-gguf/mmproj-model-f16.gguf
-
-usage: |
-    curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
-        "model": "gpt-4-vision-preview",
-        "messages": [{"role": "user", "content": [{"type":"text", "text": "What is in the image?"}, {"type": "image_url", "image_url": {"url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg" }}], "temperature": 0.9}]}'

backend/backend.proto

@@ -16,90 +16,8 @@ service Backend {
   rpc GenerateImage(GenerateImageRequest) returns (Result) {}
   rpc AudioTranscription(TranscriptRequest) returns (TranscriptResult) {}
   rpc TTS(TTSRequest) returns (Result) {}
-  rpc SoundGeneration(SoundGenerationRequest) returns (Result) {}
   rpc TokenizeString(PredictOptions) returns (TokenizationResponse) {}
   rpc Status(HealthMessage) returns (StatusResponse) {}
-
-  rpc StoresSet(StoresSetOptions) returns (Result) {}
-  rpc StoresDelete(StoresDeleteOptions) returns (Result) {}
-  rpc StoresGet(StoresGetOptions) returns (StoresGetResult) {}
-  rpc StoresFind(StoresFindOptions) returns (StoresFindResult) {}
-
-  rpc Rerank(RerankRequest) returns (RerankResult) {}
-
-  rpc GetMetrics(MetricsRequest) returns (MetricsResponse);
-
-  rpc VAD(VADRequest) returns (VADResponse) {}
-}
-
-// Define the empty request
-message MetricsRequest {}
-
-message MetricsResponse {
-  int32 slot_id = 1;
-  string prompt_json_for_slot = 2;  // Stores the prompt as a JSON string.
-  float tokens_per_second = 3;
-  int32 tokens_generated = 4;
-  int32 prompt_tokens_processed = 5;
-}
-
-message RerankRequest {
-  string query = 1;
-  repeated string documents = 2;
-  int32 top_n = 3;
-}
-
-message RerankResult {
-  Usage usage = 1;
-  repeated DocumentResult results = 2;
-}
-
-message Usage {
-  int32 total_tokens = 1;
-  int32 prompt_tokens = 2;
-}
-
-message DocumentResult {
-  int32 index = 1;
-  string text = 2;
-  float relevance_score = 3;
-}
-
-message StoresKey {
-  repeated float Floats = 1;
-}
-
-message StoresValue {
-  bytes Bytes = 1;
-}
-
-message StoresSetOptions {
-  repeated StoresKey Keys = 1;
-  repeated StoresValue Values = 2;
-}
-
-message StoresDeleteOptions {
-  repeated StoresKey Keys = 1;
-}
-
-message StoresGetOptions {
-  repeated StoresKey Keys = 1;
-}
-
-message StoresGetResult {
-  repeated StoresKey Keys = 1;
-  repeated StoresValue Values = 2;
-}
-
-message StoresFindOptions {
-  StoresKey Key = 1;
-  int32 TopK = 2;
-}
-
-message StoresFindResult {
-  repeated StoresKey Keys = 1;
-  repeated StoresValue Values = 2;
-  repeated float Similarities = 3;
 }
 
 message HealthMessage {}
@@ -147,18 +65,11 @@ message PredictOptions {
   string NegativePrompt = 40;
   int32 NDraft = 41;
   repeated string Images = 42;
-  bool UseTokenizerTemplate = 43;
-  repeated Message Messages = 44;
-  repeated string Videos = 45;
-  repeated string Audios = 46;
-  string CorrelationId = 47;
 }
 
 // The response message containing the result
 message Reply {
   bytes message = 1;
-  int32 tokens = 2;
-  int32 prompt_tokens = 3;
 }
 
 message ModelOptions {
@@ -210,7 +121,7 @@ message ModelOptions {
 
   bool NoMulMatQ = 37;
   string DraftModel = 39;
-
+  
   string AudioPath = 38;
 
   // vllm
@@ -220,8 +131,6 @@ message ModelOptions {
   bool   EnforceEager = 52;
   int32  SwapSpace = 53;
   int32  MaxModelLen = 54;
-  int32  TensorParallelSize = 55;
-  string LoadFormat = 58;
 
   string MMProj = 41;
 
@@ -232,16 +141,6 @@ message ModelOptions {
   float YarnBetaSlow = 47;
 
   string Type = 49;
-
-  bool FlashAttention = 56;
-  bool NoKVOffload = 57;
-
-  string ModelPath = 59;
-
-  repeated string LoraAdapters = 60;
-  repeated float LoraScales = 61;
-
-  repeated string Options = 62;
 }
 
 message Result {
@@ -257,7 +156,6 @@ message TranscriptRequest {
   string dst = 2;
   string language = 3;
   uint32 threads = 4;
-  bool translate = 5;
 }
 
 message TranscriptResult {
@@ -294,31 +192,6 @@ message TTSRequest {
   string model = 2;
   string dst = 3;
   string voice = 4;
-  optional string language = 5;
-}
-
-message VADRequest {
-  repeated float audio = 1;
-}
-
-message VADSegment {
-  float start = 1;
-  float end = 2;
-}
-
-message VADResponse {
-  repeated VADSegment segments = 1;
-}
-
-message SoundGenerationRequest {
-  string text = 1;
-  string model = 2;
-  string dst = 3;
-  optional float duration = 4;
-  optional float temperature = 5;
-  optional bool sample = 6;
-  optional string src = 7;
-  optional int32 src_divisor = 8;
 }
 
 message TokenizationResponse {
@@ -340,9 +213,4 @@ message StatusResponse {
   }
   State state = 1;
   MemoryUsageData memory = 2;
-}
-
-message Message {
-  string role = 1;
-  string content = 2;
 }

backend/backend_grpc.pb.go

@@ -0,0 +1,457 @@
+// Code generated by protoc-gen-go-grpc. DO NOT EDIT.
+// versions:
+// - protoc-gen-go-grpc v1.2.0
+// - protoc             v4.23.4
+// source: backend/backend.proto
+
+package proto
+
+import (
+	context "context"
+	grpc "google.golang.org/grpc"
+	codes "google.golang.org/grpc/codes"
+	status "google.golang.org/grpc/status"
+)
+
+// This is a compile-time assertion to ensure that this generated file
+// is compatible with the grpc package it is being compiled against.
+// Requires gRPC-Go v1.32.0 or later.
+const _ = grpc.SupportPackageIsVersion7
+
+// BackendClient is the client API for Backend service.
+//
+// For semantics around ctx use and closing/ending streaming RPCs, please refer to https://pkg.go.dev/google.golang.org/grpc/?tab=doc#ClientConn.NewStream.
+type BackendClient interface {
+	Health(ctx context.Context, in *HealthMessage, opts ...grpc.CallOption) (*Reply, error)
+	Predict(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (*Reply, error)
+	LoadModel(ctx context.Context, in *ModelOptions, opts ...grpc.CallOption) (*Result, error)
+	PredictStream(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (Backend_PredictStreamClient, error)
+	Embedding(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (*EmbeddingResult, error)
+	GenerateImage(ctx context.Context, in *GenerateImageRequest, opts ...grpc.CallOption) (*Result, error)
+	AudioTranscription(ctx context.Context, in *TranscriptRequest, opts ...grpc.CallOption) (*TranscriptResult, error)
+	TTS(ctx context.Context, in *TTSRequest, opts ...grpc.CallOption) (*Result, error)
+	TokenizeString(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (*TokenizationResponse, error)
+	Status(ctx context.Context, in *HealthMessage, opts ...grpc.CallOption) (*StatusResponse, error)
+}
+
+type backendClient struct {
+	cc grpc.ClientConnInterface
+}
+
+func NewBackendClient(cc grpc.ClientConnInterface) BackendClient {
+	return &backendClient{cc}
+}
+
+func (c *backendClient) Health(ctx context.Context, in *HealthMessage, opts ...grpc.CallOption) (*Reply, error) {
+	out := new(Reply)
+	err := c.cc.Invoke(ctx, "/backend.Backend/Health", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) Predict(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (*Reply, error) {
+	out := new(Reply)
+	err := c.cc.Invoke(ctx, "/backend.Backend/Predict", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) LoadModel(ctx context.Context, in *ModelOptions, opts ...grpc.CallOption) (*Result, error) {
+	out := new(Result)
+	err := c.cc.Invoke(ctx, "/backend.Backend/LoadModel", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) PredictStream(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (Backend_PredictStreamClient, error) {
+	stream, err := c.cc.NewStream(ctx, &Backend_ServiceDesc.Streams[0], "/backend.Backend/PredictStream", opts...)
+	if err != nil {
+		return nil, err
+	}
+	x := &backendPredictStreamClient{stream}
+	if err := x.ClientStream.SendMsg(in); err != nil {
+		return nil, err
+	}
+	if err := x.ClientStream.CloseSend(); err != nil {
+		return nil, err
+	}
+	return x, nil
+}
+
+type Backend_PredictStreamClient interface {
+	Recv() (*Reply, error)
+	grpc.ClientStream
+}
+
+type backendPredictStreamClient struct {
+	grpc.ClientStream
+}
+
+func (x *backendPredictStreamClient) Recv() (*Reply, error) {
+	m := new(Reply)
+	if err := x.ClientStream.RecvMsg(m); err != nil {
+		return nil, err
+	}
+	return m, nil
+}
+
+func (c *backendClient) Embedding(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (*EmbeddingResult, error) {
+	out := new(EmbeddingResult)
+	err := c.cc.Invoke(ctx, "/backend.Backend/Embedding", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) GenerateImage(ctx context.Context, in *GenerateImageRequest, opts ...grpc.CallOption) (*Result, error) {
+	out := new(Result)
+	err := c.cc.Invoke(ctx, "/backend.Backend/GenerateImage", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) AudioTranscription(ctx context.Context, in *TranscriptRequest, opts ...grpc.CallOption) (*TranscriptResult, error) {
+	out := new(TranscriptResult)
+	err := c.cc.Invoke(ctx, "/backend.Backend/AudioTranscription", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) TTS(ctx context.Context, in *TTSRequest, opts ...grpc.CallOption) (*Result, error) {
+	out := new(Result)
+	err := c.cc.Invoke(ctx, "/backend.Backend/TTS", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) TokenizeString(ctx context.Context, in *PredictOptions, opts ...grpc.CallOption) (*TokenizationResponse, error) {
+	out := new(TokenizationResponse)
+	err := c.cc.Invoke(ctx, "/backend.Backend/TokenizeString", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+func (c *backendClient) Status(ctx context.Context, in *HealthMessage, opts ...grpc.CallOption) (*StatusResponse, error) {
+	out := new(StatusResponse)
+	err := c.cc.Invoke(ctx, "/backend.Backend/Status", in, out, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+// BackendServer is the server API for Backend service.
+// All implementations must embed UnimplementedBackendServer
+// for forward compatibility
+type BackendServer interface {
+	Health(context.Context, *HealthMessage) (*Reply, error)
+	Predict(context.Context, *PredictOptions) (*Reply, error)
+	LoadModel(context.Context, *ModelOptions) (*Result, error)
+	PredictStream(*PredictOptions, Backend_PredictStreamServer) error
+	Embedding(context.Context, *PredictOptions) (*EmbeddingResult, error)
+	GenerateImage(context.Context, *GenerateImageRequest) (*Result, error)
+	AudioTranscription(context.Context, *TranscriptRequest) (*TranscriptResult, error)
+	TTS(context.Context, *TTSRequest) (*Result, error)
+	TokenizeString(context.Context, *PredictOptions) (*TokenizationResponse, error)
+	Status(context.Context, *HealthMessage) (*StatusResponse, error)
+	mustEmbedUnimplementedBackendServer()
+}
+
+// UnimplementedBackendServer must be embedded to have forward compatible implementations.
+type UnimplementedBackendServer struct {
+}
+
+func (UnimplementedBackendServer) Health(context.Context, *HealthMessage) (*Reply, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method Health not implemented")
+}
+func (UnimplementedBackendServer) Predict(context.Context, *PredictOptions) (*Reply, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method Predict not implemented")
+}
+func (UnimplementedBackendServer) LoadModel(context.Context, *ModelOptions) (*Result, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method LoadModel not implemented")
+}
+func (UnimplementedBackendServer) PredictStream(*PredictOptions, Backend_PredictStreamServer) error {
+	return status.Errorf(codes.Unimplemented, "method PredictStream not implemented")
+}
+func (UnimplementedBackendServer) Embedding(context.Context, *PredictOptions) (*EmbeddingResult, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method Embedding not implemented")
+}
+func (UnimplementedBackendServer) GenerateImage(context.Context, *GenerateImageRequest) (*Result, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method GenerateImage not implemented")
+}
+func (UnimplementedBackendServer) AudioTranscription(context.Context, *TranscriptRequest) (*TranscriptResult, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method AudioTranscription not implemented")
+}
+func (UnimplementedBackendServer) TTS(context.Context, *TTSRequest) (*Result, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method TTS not implemented")
+}
+func (UnimplementedBackendServer) TokenizeString(context.Context, *PredictOptions) (*TokenizationResponse, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method TokenizeString not implemented")
+}
+func (UnimplementedBackendServer) Status(context.Context, *HealthMessage) (*StatusResponse, error) {
+	return nil, status.Errorf(codes.Unimplemented, "method Status not implemented")
+}
+func (UnimplementedBackendServer) mustEmbedUnimplementedBackendServer() {}
+
+// UnsafeBackendServer may be embedded to opt out of forward compatibility for this service.
+// Use of this interface is not recommended, as added methods to BackendServer will
+// result in compilation errors.
+type UnsafeBackendServer interface {
+	mustEmbedUnimplementedBackendServer()
+}
+
+func RegisterBackendServer(s grpc.ServiceRegistrar, srv BackendServer) {
+	s.RegisterService(&Backend_ServiceDesc, srv)
+}
+
+func _Backend_Health_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(HealthMessage)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).Health(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/Health",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).Health(ctx, req.(*HealthMessage))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_Predict_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(PredictOptions)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).Predict(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/Predict",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).Predict(ctx, req.(*PredictOptions))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_LoadModel_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(ModelOptions)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).LoadModel(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/LoadModel",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).LoadModel(ctx, req.(*ModelOptions))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_PredictStream_Handler(srv interface{}, stream grpc.ServerStream) error {
+	m := new(PredictOptions)
+	if err := stream.RecvMsg(m); err != nil {
+		return err
+	}
+	return srv.(BackendServer).PredictStream(m, &backendPredictStreamServer{stream})
+}
+
+type Backend_PredictStreamServer interface {
+	Send(*Reply) error
+	grpc.ServerStream
+}
+
+type backendPredictStreamServer struct {
+	grpc.ServerStream
+}
+
+func (x *backendPredictStreamServer) Send(m *Reply) error {
+	return x.ServerStream.SendMsg(m)
+}
+
+func _Backend_Embedding_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(PredictOptions)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).Embedding(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/Embedding",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).Embedding(ctx, req.(*PredictOptions))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_GenerateImage_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(GenerateImageRequest)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).GenerateImage(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/GenerateImage",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).GenerateImage(ctx, req.(*GenerateImageRequest))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_AudioTranscription_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(TranscriptRequest)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).AudioTranscription(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/AudioTranscription",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).AudioTranscription(ctx, req.(*TranscriptRequest))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_TTS_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(TTSRequest)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).TTS(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/TTS",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).TTS(ctx, req.(*TTSRequest))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_TokenizeString_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(PredictOptions)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).TokenizeString(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/TokenizeString",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).TokenizeString(ctx, req.(*PredictOptions))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+func _Backend_Status_Handler(srv interface{}, ctx context.Context, dec func(interface{}) error, interceptor grpc.UnaryServerInterceptor) (interface{}, error) {
+	in := new(HealthMessage)
+	if err := dec(in); err != nil {
+		return nil, err
+	}
+	if interceptor == nil {
+		return srv.(BackendServer).Status(ctx, in)
+	}
+	info := &grpc.UnaryServerInfo{
+		Server:     srv,
+		FullMethod: "/backend.Backend/Status",
+	}
+	handler := func(ctx context.Context, req interface{}) (interface{}, error) {
+		return srv.(BackendServer).Status(ctx, req.(*HealthMessage))
+	}
+	return interceptor(ctx, in, info, handler)
+}
+
+// Backend_ServiceDesc is the grpc.ServiceDesc for Backend service.
+// It's only intended for direct use with grpc.RegisterService,
+// and not to be introspected or modified (even as a copy)
+var Backend_ServiceDesc = grpc.ServiceDesc{
+	ServiceName: "backend.Backend",
+	HandlerType: (*BackendServer)(nil),
+	Methods: []grpc.MethodDesc{
+		{
+			MethodName: "Health",
+			Handler:    _Backend_Health_Handler,
+		},
+		{
+			MethodName: "Predict",
+			Handler:    _Backend_Predict_Handler,
+		},
+		{
+			MethodName: "LoadModel",
+			Handler:    _Backend_LoadModel_Handler,
+		},
+		{
+			MethodName: "Embedding",
+			Handler:    _Backend_Embedding_Handler,
+		},
+		{
+			MethodName: "GenerateImage",
+			Handler:    _Backend_GenerateImage_Handler,
+		},
+		{
+			MethodName: "AudioTranscription",
+			Handler:    _Backend_AudioTranscription_Handler,
+		},
+		{
+			MethodName: "TTS",
+			Handler:    _Backend_TTS_Handler,
+		},
+		{
+			MethodName: "TokenizeString",
+			Handler:    _Backend_TokenizeString_Handler,
+		},
+		{
+			MethodName: "Status",
+			Handler:    _Backend_Status_Handler,
+		},
+	},
+	Streams: []grpc.StreamDesc{
+		{
+			StreamName:    "PredictStream",
+			Handler:       _Backend_PredictStream_Handler,
+			ServerStreams: true,
+		},
+	},
+	Metadata: "backend/backend.proto",
+}

backend/cpp/grpc/Makefile

@@ -5,6 +5,7 @@ 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
@@ -46,17 +47,12 @@ endif
 $(INSTALLED_PACKAGES): grpc_build
 
 $(GRPC_REPO):
-	mkdir -p $(GRPC_REPO)/grpc
-	cd $(GRPC_REPO)/grpc && \
-	git init && \
-	git remote add origin $(GIT_REPO_LIB_GRPC)  && \
-	git fetch origin && \
-	git checkout $(TAG_LIB_GRPC) && \
-	git submodule update --init --recursive --depth 1 --single-branch
-	
+	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 . && cmake --build . --target install
+	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)
 

backend/cpp/llama/Makefile

@@ -1,80 +1,72 @@
 
 LLAMA_VERSION?=
-LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
 
 CMAKE_ARGS?=
 BUILD_TYPE?=
 ONEAPI_VARS?=/opt/intel/oneapi/setvars.sh
-TARGET?=--target grpc-server
 
-# Disable Shared libs as we are linking on static gRPC and we can't mix shared and static
-CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF
-
-# If build type is cublas, then we set -DGGML_CUDA=ON to CMAKE_ARGS automatically
+# If build type is cublas, then we set -DLLAMA_CUBLAS=ON to CMAKE_ARGS automatically
 ifeq ($(BUILD_TYPE),cublas)
-	CMAKE_ARGS+=-DGGML_CUDA=ON
-# If build type is openblas then we set -DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS
+	CMAKE_ARGS+=-DLLAMA_CUBLAS=ON
+# If build type is openblas then we set -DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS
 # to CMAKE_ARGS automatically
 else ifeq ($(BUILD_TYPE),openblas)
-	CMAKE_ARGS+=-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS
-# If build type is clblas (openCL) we set -DGGML_CLBLAST=ON -DCLBlast_DIR=/some/path
+	CMAKE_ARGS+=-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS
+# If build type is clblas (openCL) we set -DLLAMA_CLBLAST=ON -DCLBlast_DIR=/some/path
 else ifeq ($(BUILD_TYPE),clblas)
-	CMAKE_ARGS+=-DGGML_CLBLAST=ON -DCLBlast_DIR=/some/path
+	CMAKE_ARGS+=-DLLAMA_CLBLAST=ON -DCLBlast_DIR=/some/path
 # If it's hipblas we do have also to set CC=/opt/rocm/llvm/bin/clang CXX=/opt/rocm/llvm/bin/clang++ 
 else ifeq ($(BUILD_TYPE),hipblas)
-	CMAKE_ARGS+=-DGGML_HIP=ON
-# If it's OSX, DO NOT embed the metal library - -DGGML_METAL_EMBED_LIBRARY=ON requires further investigation
-# But if it's OSX without metal, disable it here
-else ifeq ($(OS),Darwin)
-	ifneq ($(BUILD_TYPE),metal)
-		CMAKE_ARGS+=-DGGML_METAL=OFF
-	else
-		CMAKE_ARGS+=-DGGML_METAL=ON
-		CMAKE_ARGS+=-DGGML_METAL_EMBED_LIBRARY=ON
-		TARGET+=--target ggml-metal
-	endif
+	CMAKE_ARGS+=-DLLAMA_HIPBLAS=ON
+# If it's OSX, DO NOT embed the metal library - -DLLAMA_METAL_EMBED_LIBRARY=ON requires further investigation
 endif
 
 ifeq ($(BUILD_TYPE),sycl_f16)
-	CMAKE_ARGS+=-DGGML_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DGGML_SYCL_F16=ON
+	CMAKE_ARGS+=-DLLAMA_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DLLAMA_SYCL_F16=ON
 endif
 
 ifeq ($(BUILD_TYPE),sycl_f32)
-	CMAKE_ARGS+=-DGGML_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx
+	CMAKE_ARGS+=-DLLAMA_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx
 endif
 
 llama.cpp:
-	mkdir -p llama.cpp
-	cd llama.cpp && \
-	git init && \
-	git remote add origin $(LLAMA_REPO)  && \
-	git fetch origin && \
-	git checkout -b build $(LLAMA_VERSION) && \
-	git submodule update --init --recursive --depth 1 --single-branch
+	git clone --recurse-submodules https://github.com/ggerganov/llama.cpp llama.cpp
+	if [ -z "$(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: llama.cpp
+llama.cpp/examples/grpc-server:
 	mkdir -p llama.cpp/examples/grpc-server
-	bash prepare.sh
+	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/
+	cp -rfv $(abspath ./)/utils.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/llava.cpp llama.cpp/examples/grpc-server/llava.cpp
+	echo '#include "llama.h"' > llama.cpp/examples/grpc-server/llava.h
+	cat llama.cpp/examples/llava/llava.h >> llama.cpp/examples/grpc-server/llava.h
+	cp -rfv llama.cpp/examples/llava/clip.cpp llama.cpp/examples/grpc-server/clip.cpp
 
 rebuild:
-	bash prepare.sh
+	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
 
-purge:
-	rm -rf llama.cpp/build
-	rm -rf llama.cpp/examples/grpc-server
+clean:
+	rm -rf llama.cpp
 	rm -rf grpc-server
 
-clean: purge
-	rm -rf llama.cpp
-
 grpc-server: llama.cpp llama.cpp/examples/grpc-server
-	@echo "Building grpc-server with $(BUILD_TYPE) build type and $(CMAKE_ARGS)"
 ifneq (,$(findstring sycl,$(BUILD_TYPE)))
-	+bash -c "source $(ONEAPI_VARS); \
-	cd llama.cpp && mkdir -p build && cd build && cmake .. $(CMAKE_ARGS) && cmake --build . --config Release $(TARGET)"
+	bash -c "source $(ONEAPI_VARS); \
+	cd llama.cpp && mkdir -p build && cd build && cmake .. $(CMAKE_ARGS) && cmake --build . --config Release"	
 else
-	+cd llama.cpp && mkdir -p build && cd build && cmake .. $(CMAKE_ARGS) && cmake --build . --config Release $(TARGET)
+	cd llama.cpp && mkdir -p build && cd build && cmake .. $(CMAKE_ARGS) && cmake --build . --config Release
 endif
 	cp llama.cpp/build/bin/grpc-server .

backend/cpp/llama/grpc-server.cpp

@@ -13,15 +13,15 @@
 #include <getopt.h>
 #include "clip.h"
 #include "llava.h"
-#include "log.h"
 #include "stb_image.h"
 #include "common.h"
 #include "json.hpp"
 #include "llama.h"
+#include "grammar-parser.h"
 #include "backend.pb.h"
 #include "backend.grpc.pb.h"
 #include "utils.hpp"
-#include "sampling.h"
+
 // include std::regex
 #include <cstddef>
 #include <thread>
@@ -113,7 +113,7 @@ static std::string tokens_to_str(llama_context *ctx, Iter begin, Iter end)
     std::string ret;
     for (; begin != end; ++begin)
     {
-        ret += common_token_to_piece(ctx, *begin);
+        ret += llama_token_to_piece(ctx, *begin);
     }
     return ret;
 }
@@ -121,7 +121,7 @@ static std::string tokens_to_str(llama_context *ctx, Iter begin, Iter end)
 // format incomplete utf-8 multibyte character for output
 static std::string tokens_to_output_formatted_string(const llama_context *ctx, const llama_token token)
 {
-    std::string out = token == -1 ? "" : common_token_to_piece(ctx, token);
+    std::string out = token == -1 ? "" : llama_token_to_piece(ctx, token);
     // if the size is 1 and first bit is 1, meaning it's a partial character
     //   (size > 1 meaning it's already a known token)
     if (out.size() == 1 && (out[0] & 0x80) == 0x80)
@@ -203,8 +203,8 @@ struct llama_client_slot
     std::string stopping_word;
 
     // sampling
-    struct common_params_sampling sparams;
-    common_sampler *ctx_sampling = nullptr;
+    struct llama_sampling_params sparams;
+    llama_sampling_context *ctx_sampling = nullptr;
 
     int32_t ga_i = 0;   // group-attention state
     int32_t ga_n = 1;   // group-attention factor
@@ -257,7 +257,7 @@ struct llama_client_slot
         images.clear();
     }
 
-    bool has_budget(common_params &global_params) {
+    bool has_budget(gpt_params &global_params) {
         if (params.n_predict == -1 && global_params.n_predict == -1)
         {
             return true; // limitless
@@ -391,39 +391,6 @@ struct llama_metrics {
     }
 };
 
-struct llava_embd_batch {
-    std::vector<llama_pos>      pos;
-    std::vector<int32_t>        n_seq_id;
-    std::vector<llama_seq_id>   seq_id_0;
-    std::vector<llama_seq_id *> seq_ids;
-    std::vector<int8_t>         logits;
-    llama_batch batch;
-    llava_embd_batch(float * embd, int32_t n_tokens, llama_pos pos_0, llama_seq_id seq_id) {
-        pos     .resize(n_tokens);
-        n_seq_id.resize(n_tokens);
-        seq_ids .resize(n_tokens + 1);
-        logits  .resize(n_tokens);
-        seq_id_0.resize(1);
-        seq_id_0[0] = seq_id;
-        seq_ids [n_tokens] = nullptr;
-        batch = {
-            /*n_tokens       =*/ n_tokens,
-            /*tokens         =*/ nullptr,
-            /*embd           =*/ embd,
-            /*pos            =*/ pos.data(),
-            /*n_seq_id       =*/ n_seq_id.data(),
-            /*seq_id         =*/ seq_ids.data(),
-            /*logits         =*/ logits.data(),
-        };
-        for (int i = 0; i < n_tokens; i++) {
-            batch.pos     [i] = pos_0 + i;
-            batch.n_seq_id[i] = 1;
-            batch.seq_id  [i] = seq_id_0.data();
-            batch.logits  [i] = false;
-        }
-    }
-};
-
 struct llama_server_context
 {
     llama_model *model = nullptr;
@@ -431,7 +398,7 @@ struct llama_server_context
 
     clip_ctx *clp_ctx = nullptr;
 
-    common_params params;
+    gpt_params params;
 
     llama_batch batch;
 
@@ -474,7 +441,7 @@ struct llama_server_context
         }
     }
 
-    bool load_model(const common_params &params_)
+    bool load_model(const gpt_params &params_)
     {
         params = params_;
         if (!params.mmproj.empty()) {
@@ -482,7 +449,7 @@ struct llama_server_context
             LOG_INFO("Multi Modal Mode Enabled", {});
             clp_ctx = clip_model_load(params.mmproj.c_str(), /*verbosity=*/ 1);
             if(clp_ctx == nullptr) {
-                LOG_ERR("unable to load clip model: %s", params.mmproj.c_str());
+                LOG_ERROR("unable to load clip model", {{"model", params.mmproj}});
                 return false;
             }
 
@@ -491,12 +458,10 @@ struct llama_server_context
             }
         }
 
-        common_init_result common_init = common_init_from_params(params);
-        model = common_init.model;
-        ctx = common_init.context;
+        std::tie(model, ctx) = llama_init_from_gpt_params(params);
         if (model == nullptr)
         {
-            LOG_ERR("unable to load model: %s", params.model.c_str());
+            LOG_ERROR("unable to load model", {{"model", params.model}});
             return false;
         }
 
@@ -504,7 +469,7 @@ struct llama_server_context
             const int n_embd_clip = clip_n_mmproj_embd(clp_ctx);
             const int n_embd_llm  = llama_n_embd(model);
             if (n_embd_clip != n_embd_llm) {
-                LOG("%s: embedding dim of the multimodal projector (%d) is not equal to that of LLaMA (%d). Make sure that you use the correct mmproj file.\n", __func__, n_embd_clip, n_embd_llm);
+                LOG_TEE("%s: embedding dim of the multimodal projector (%d) is not equal to that of LLaMA (%d). Make sure that you use the correct mmproj file.\n", __func__, n_embd_clip, n_embd_llm);
                 llama_free(ctx);
                 llama_free_model(model);
                 return false;
@@ -513,7 +478,7 @@ struct llama_server_context
 
         n_ctx = llama_n_ctx(ctx);
 
-        add_bos_token = llama_add_bos_token(model);
+        add_bos_token = llama_should_add_bos_token(model);
 
         return true;
     }
@@ -523,21 +488,11 @@ struct llama_server_context
         std::vector<char> buf(1);
         int res = llama_chat_apply_template(model, nullptr, chat, 1, true, buf.data(), buf.size());
         if (res < 0) {
-            LOG_ERR("The chat template comes with this model is not yet supported, falling back to chatml. This may cause the model to output suboptimal responses", __func__);
+            LOG_ERROR("The chat template comes with this model is not yet supported, falling back to chatml. This may cause the model to output suboptimal responses", {});
             sparams.chat_template = "<|im_start|>"; // llama_chat_apply_template only checks if <|im_start|> exist in the template
         }
     }
 
-    llama_client_slot* get_active_slot() {
-        for (llama_client_slot& slot : slots) {
-            // Check if the slot is currently processing
-            if (slot.is_processing()) {
-                return &slot;  // Return the active slot
-            }
-        }
-        return nullptr;  // No active slot found
-    }
-
     void initialize() {
         // create slots
         all_slots_are_idle = true;
@@ -611,12 +566,12 @@ struct llama_server_context
                     std::vector<llama_token> p;
                     if (first)
                     {
-                        p = common_tokenize(ctx, s, add_bos, TMP_FORCE_SPECIAL);
+                        p = ::llama_tokenize(ctx, s, add_bos, TMP_FORCE_SPECIAL);
                         first = false;
                     }
                     else
                     {
-                        p = common_tokenize(ctx, s, false, TMP_FORCE_SPECIAL);
+                        p = ::llama_tokenize(ctx, s, false, TMP_FORCE_SPECIAL);
                     }
                     prompt_tokens.insert(prompt_tokens.end(), p.begin(), p.end());
                 }
@@ -633,7 +588,7 @@ struct llama_server_context
         else
         {
             auto s = json_prompt.template get<std::string>();
-            prompt_tokens = common_tokenize(ctx, s, add_bos, TMP_FORCE_SPECIAL);
+            prompt_tokens = ::llama_tokenize(ctx, s, add_bos, TMP_FORCE_SPECIAL);
         }
 
         return prompt_tokens;
@@ -662,7 +617,7 @@ struct llama_server_context
 
     bool launch_slot_with_data(llama_client_slot* &slot, json data) {
         slot_params default_params;
-        common_params_sampling default_sparams;
+        llama_sampling_params default_sparams;
  
         slot->params.stream             = json_value(data, "stream",            false);
         slot->params.cache_prompt       = json_value(data, "cache_prompt",      false);
@@ -670,7 +625,8 @@ struct llama_server_context
         slot->sparams.top_k             = json_value(data, "top_k",             default_sparams.top_k);
         slot->sparams.top_p             = json_value(data, "top_p",             default_sparams.top_p);
         slot->sparams.min_p             = json_value(data, "min_p",             default_sparams.min_p);
-        slot->sparams.typ_p             = json_value(data, "typical_p",         default_sparams.typ_p);
+        slot->sparams.tfs_z             = json_value(data, "tfs_z",             default_sparams.tfs_z);
+        slot->sparams.typical_p         = json_value(data, "typical_p",         default_sparams.typical_p);
         slot->sparams.temp              = json_value(data, "temperature",       default_sparams.temp);
         slot->sparams.dynatemp_range    = json_value(data, "dynatemp_range",    default_sparams.dynatemp_range);
         slot->sparams.dynatemp_exponent = json_value(data, "dynatemp_exponent", default_sparams.dynatemp_exponent);
@@ -683,7 +639,7 @@ struct llama_server_context
         slot->sparams.mirostat_eta      = json_value(data, "mirostat_eta",      default_sparams.mirostat_eta);
         slot->sparams.penalize_nl       = json_value(data, "penalize_nl",       default_sparams.penalize_nl);
         slot->params.n_keep             = json_value(data, "n_keep",            slot->params.n_keep);
-        slot->sparams.seed               = json_value(data, "seed",              default_sparams.seed);
+        slot->params.seed               = json_value(data, "seed",              default_params.seed);
         slot->sparams.grammar           = json_value(data, "grammar",           default_sparams.grammar);
         slot->sparams.n_probs           = json_value(data, "n_probs",           default_sparams.n_probs);
         slot->sparams.min_keep          = json_value(data, "min_keep",          default_sparams.min_keep);
@@ -707,7 +663,6 @@ struct llama_server_context
             slot->params.input_prefix = "";
         }
 
-
         if (data.count("input_suffix") != 0)
         {
             slot->params.input_suffix = data["input_suffix"];
@@ -726,10 +681,6 @@ struct llama_server_context
             slot->prompt = "";
         }
 
-        if (json_value(data, "ignore_eos", false)) {
-                slot->sparams.logit_bias.push_back({llama_token_eos(model), -INFINITY});
-        }
-        /*
         slot->sparams.penalty_prompt_tokens.clear();
         slot->sparams.use_penalty_prompt_tokens = false;
         const auto &penalty_prompt = data.find("penalty_prompt");
@@ -765,10 +716,14 @@ struct llama_server_context
                 slot->sparams.use_penalty_prompt_tokens = true;
             }
         }
-      */
 
         slot->sparams.logit_bias.clear();
 
+        if (json_value(data, "ignore_eos", false))
+        {
+            slot->sparams.logit_bias[llama_token_eos(model)] = -INFINITY;
+        }
+
         const auto &logit_bias = data.find("logit_bias");
         if (logit_bias != data.end() && logit_bias->is_array())
         {
@@ -796,21 +751,21 @@ struct llama_server_context
                         llama_token tok = el[0].get<llama_token>();
                         if (tok >= 0 && tok < n_vocab)
                         {
-                            slot->sparams.logit_bias.push_back({tok, bias});
+                            slot->sparams.logit_bias[tok] = bias;
                         }
                     }
                     else if (el[0].is_string())
                     {
-                        auto toks = common_tokenize(model, el[0].get<std::string>(), false);
+                        auto toks = llama_tokenize(model, el[0].get<std::string>(), false);
                         for (auto tok : toks)
                         {
-                            slot->sparams.logit_bias.push_back({tok, bias});
+                            slot->sparams.logit_bias[tok] = bias;
                         }
                     }
                 }
             }
         }
-        
+
         slot->params.antiprompt.clear();
 
         const auto &stop = data.find("stop");
@@ -824,22 +779,24 @@ struct llama_server_context
                 }
             }
         }
-        
-        const auto & samplers = data.find("samplers");
-        if (samplers != data.end() && samplers->is_array()) {
+
+        const auto &samplers_sequence = data.find("samplers");
+        if (samplers_sequence != data.end() && samplers_sequence->is_array())
+        {
             std::vector<std::string> sampler_names;
-                for (const auto & name : *samplers) {
-                    if (name.is_string()) {
-                        sampler_names.emplace_back(name);
-                    }
+            for (const auto &sampler_name : *samplers_sequence)
+            {
+                if (sampler_name.is_string())
+                {
+                    sampler_names.emplace_back(sampler_name);
                 }
-                slot->sparams.samplers = common_sampler_types_from_names(sampler_names, false);
+            }
+            slot->sparams.samplers_sequence = sampler_types_from_names(sampler_names, false);
         }
         else
         {
-                slot->sparams.samplers = default_sparams.samplers;
+            slot->sparams.samplers_sequence = default_sparams.samplers_sequence;
         }
-        
 
         if (multimodal)
         {
@@ -855,11 +812,10 @@ struct llama_server_context
                     img_sl.img_data = clip_image_u8_init();
                     if (!clip_image_load_from_bytes(image_buffer.data(), image_buffer.size(), img_sl.img_data))
                     {
-                        LOG_ERR("%s: failed to load image, slot_id: %d, img_sl_id: %d", 
-                             __func__,
-                             slot->id,
-                             img_sl.id
-                        );
+                        LOG_ERROR("failed to load image", {
+                            {"slot_id",   slot->id},
+                            {"img_sl_id", img_sl.id}
+                        });
                         return false;
                     }
                     LOG_VERBOSE("image loaded", {
@@ -897,12 +853,12 @@ struct llama_server_context
                                     }
                                 }
                                 if (!found) {
-                                    LOG("ERROR: Image with id: %i, not found.\n", img_id);
+                                    LOG_TEE("ERROR: Image with id: %i, not found.\n", img_id);
                                     slot->images.clear();
                                     return false;
                                 }
                             } catch (const std::invalid_argument& e) {
-                                LOG("Invalid image number id in prompt\n");
+                                LOG_TEE("Invalid image number id in prompt\n");
                                 slot->images.clear();
                                 return false;
                             }
@@ -917,10 +873,10 @@ struct llama_server_context
 
         if (slot->ctx_sampling != nullptr)
         {
-            common_sampler_free(slot->ctx_sampling);
+            llama_sampling_free(slot->ctx_sampling);
         }
-        slot->ctx_sampling = common_sampler_init(model, slot->sparams);
-        //llama_set_rng_seed(ctx, slot->params.seed);
+        slot->ctx_sampling = llama_sampling_init(slot->sparams);
+        llama_set_rng_seed(ctx, slot->params.seed);
         slot->command = LOAD_PROMPT;
 
         all_slots_are_idle = false;
@@ -930,8 +886,6 @@ struct llama_server_context
             {"task_id", slot->task_id},
         });
 
-      //  LOG("sampling: \n%s\n", llama_sampling_print(slot->sparams).c_str());
-
         return true;
     }
 
@@ -946,13 +900,13 @@ struct llama_server_context
         system_tokens.clear();
 
         if (!system_prompt.empty()) {
-            system_tokens = common_tokenize(ctx, system_prompt, add_bos_token);
+            system_tokens = ::llama_tokenize(ctx, system_prompt, add_bos_token);
 
-            common_batch_clear(batch);
+            llama_batch_clear(batch);
 
             for (int i = 0; i < (int)system_tokens.size(); ++i)
             {
-                common_batch_add(batch, system_tokens[i], i, { 0 }, false);
+                llama_batch_add(batch, system_tokens[i], i, { 0 }, false);
             }
 
             for (int32_t i = 0; i < (int32_t) batch.n_tokens; i += params.n_batch)
@@ -966,10 +920,11 @@ struct llama_server_context
                     batch.n_seq_id + i,
                     batch.seq_id   + i,
                     batch.logits   + i,
+                    0, 0, 0, // unused
                 };
                 if (llama_decode(ctx, batch_view) != 0)
                 {
-                    LOG("%s: llama_decode() failed\n", __func__);
+                    LOG_TEE("%s: llama_decode() failed\n", __func__);
                     return;
                 }
             }
@@ -981,7 +936,7 @@ struct llama_server_context
             }
         }
 
-        LOG("system prompt updated\n");
+        LOG_TEE("system prompt updated\n");
         system_need_update = false;
     }
 
@@ -1040,20 +995,18 @@ struct llama_server_context
 
     bool process_token(completion_token_output &result, llama_client_slot &slot) {
         // remember which tokens were sampled - used for repetition penalties during sampling
-        const std::string token_str = common_token_to_piece(ctx, result.tok);
+        const std::string token_str = llama_token_to_piece(ctx, result.tok);
         slot.sampled = result.tok;
 
         // search stop word and delete it
         slot.generated_text += token_str;
         slot.has_next_token = true;
 
-/*
         if (slot.ctx_sampling->params.use_penalty_prompt_tokens && result.tok != -1)
         {
             // we can change penalty_prompt_tokens because it is always created from scratch each request
             slot.ctx_sampling->params.penalty_prompt_tokens.push_back(result.tok);
         }
-        */
 
         // check if there is incomplete UTF-8 character at the end
         bool incomplete = false;
@@ -1162,8 +1115,8 @@ struct llama_server_context
                 continue;
             }
 
-            if (!llava_image_embed_make_with_clip_img(clp_ctx, params.cpuparams.n_threads, img.img_data, &img.image_embedding, &img.image_tokens)) {
-                LOG("Error processing the given image");
+            if (!llava_image_embed_make_with_clip_img(clp_ctx, params.n_threads, img.img_data, &img.image_embedding, &img.image_tokens)) {
+                LOG_TEE("Error processing the given image");
                 return false;
             }
 
@@ -1175,7 +1128,7 @@ struct llama_server_context
 
     void send_error(task_server& task, const std::string &error)
     {
-        LOG("task %i - error: %s\n", task.id, error.c_str());
+        LOG_TEE("task %i - error: %s\n", task.id, error.c_str());
         task_result res;
         res.id = task.id;
         res.multitask_id = task.multitask_id;
@@ -1187,11 +1140,13 @@ struct llama_server_context
 
     json get_formated_generation(llama_client_slot &slot)
     {
-        std::vector<std::string> samplers;
-        samplers.reserve(slot.sparams.samplers.size());
-        for (const auto & sampler : slot.sparams.samplers)
+        const auto eos_bias = slot.sparams.logit_bias.find(llama_token_eos(model));
+        const bool ignore_eos = eos_bias != slot.sparams.logit_bias.end() &&
+                                eos_bias->second < 0.0f && std::isinf(eos_bias->second);
+        std::vector<std::string> samplers_sequence;
+        for (const auto &sampler_type : slot.sparams.samplers_sequence)
         {
-            samplers.emplace_back(common_sampler_type_to_str(sampler));
+            samplers_sequence.emplace_back(sampler_type_to_name_string(sampler_type));
         }
 
         return json {
@@ -1205,11 +1160,14 @@ struct llama_server_context
             {"top_k",             slot.sparams.top_k},
             {"top_p",             slot.sparams.top_p},
             {"min_p",             slot.sparams.min_p},
-            {"typical_p",         slot.sparams.typ_p},
+            {"tfs_z",             slot.sparams.tfs_z},
+            {"typical_p",         slot.sparams.typical_p},
             {"repeat_last_n",     slot.sparams.penalty_last_n},
             {"repeat_penalty",    slot.sparams.penalty_repeat},
             {"presence_penalty",  slot.sparams.penalty_present},
             {"frequency_penalty", slot.sparams.penalty_freq},
+            {"penalty_prompt_tokens", slot.sparams.penalty_prompt_tokens},
+            {"use_penalty_prompt_tokens", slot.sparams.use_penalty_prompt_tokens},
             {"mirostat",          slot.sparams.mirostat},
             {"mirostat_tau",      slot.sparams.mirostat_tau},
             {"mirostat_eta",      slot.sparams.mirostat_eta},
@@ -1217,13 +1175,13 @@ struct llama_server_context
             {"stop",              slot.params.antiprompt},
             {"n_predict",         slot.params.n_predict},
             {"n_keep",            params.n_keep},
-            {"ignore_eos",        slot.sparams.ignore_eos},
+            {"ignore_eos",        ignore_eos},
             {"stream",            slot.params.stream},
-      //      {"logit_bias",        slot.sparams.logit_bias},
+            {"logit_bias",        slot.sparams.logit_bias},
             {"n_probs",           slot.sparams.n_probs},
             {"min_keep",          slot.sparams.min_keep},
             {"grammar",           slot.sparams.grammar},
-            {"samplers",          samplers}
+            {"samplers",          samplers_sequence}
         };
     }
 
@@ -1246,7 +1204,7 @@ struct llama_server_context
         if (slot.sparams.n_probs > 0)
         {
             std::vector<completion_token_output> probs_output = {};
-            const std::vector<llama_token> to_send_toks = common_tokenize(ctx, tkn.text_to_send, false);
+            const std::vector<llama_token> to_send_toks = llama_tokenize(ctx, tkn.text_to_send, false);
             size_t probs_pos      = std::min(slot.sent_token_probs_index,                       slot.generated_token_probs.size());
             size_t probs_stop_pos = std::min(slot.sent_token_probs_index + to_send_toks.size(), slot.generated_token_probs.size());
             if (probs_pos < probs_stop_pos)
@@ -1298,7 +1256,7 @@ struct llama_server_context
             std::vector<completion_token_output> probs = {};
             if (!slot.params.stream && slot.stopped_word)
             {
-                const std::vector<llama_token> stop_word_toks = common_tokenize(ctx, slot.stopping_word, false);
+                const std::vector<llama_token> stop_word_toks = llama_tokenize(ctx, slot.stopping_word, false);
                 probs = std::vector<completion_token_output>(slot.generated_token_probs.begin(), slot.generated_token_probs.end() - stop_word_toks.size());
             }
             else
@@ -1409,10 +1367,11 @@ struct llama_server_context
                     batch.n_seq_id + i,
                     batch.seq_id   + i,
                     batch.logits   + i,
+                    0, 0, 0, // unused
                 };
                 if (llama_decode(ctx, batch_view))
                 {
-                    LOG("%s : failed to eval\n", __func__);
+                    LOG_TEE("%s : failed to eval\n", __func__);
                     return false;
                 }
             }
@@ -1427,18 +1386,17 @@ struct llama_server_context
                 }
 
                 const int n_embd = llama_n_embd(model);
-                float * embd = img.image_embedding + i * n_embd;
-                llava_embd_batch llava_batch = llava_embd_batch(embd, n_eval, slot.n_past, 0);
-                if (llama_decode(ctx, llava_batch.batch))
+                llama_batch batch_img = { n_eval, nullptr, (img.image_embedding + i * n_embd), nullptr, nullptr, nullptr, nullptr, slot.n_past, 1, 0, };
+                if (llama_decode(ctx, batch_img))
                 {
-                    LOG("%s : failed to eval image\n", __func__);
+                    LOG_TEE("%s : failed to eval image\n", __func__);
                     return false;
                 }
                 slot.n_past += n_eval;
             }
             image_idx++;
 
-            common_batch_clear(batch);
+            llama_batch_clear(batch);
 
             // append prefix of next image
             const auto json_prompt = (image_idx >= (int) slot.images.size()) ?
@@ -1448,7 +1406,7 @@ struct llama_server_context
             std::vector<llama_token> append_tokens = tokenize(json_prompt, false); // has next image
             for (int i = 0; i < (int) append_tokens.size(); ++i)
             {
-                common_batch_add(batch, append_tokens[i], system_tokens.size() + slot.n_past, { slot.id }, true);
+                llama_batch_add(batch, append_tokens[i], system_tokens.size() + slot.n_past, { slot.id }, true);
                 slot.n_past += 1;
             }
         }
@@ -1580,7 +1538,7 @@ struct llama_server_context
             update_system_prompt();
         }
 
-        common_batch_clear(batch);
+        llama_batch_clear(batch);
 
         if (all_slots_are_idle)
         {
@@ -1614,7 +1572,7 @@ struct llama_server_context
                     slot.n_past = 0;
                     slot.truncated = false;
                     slot.has_next_token = true;
-                    LOG("Context exhausted. Slot %d released (%d tokens in cache)\n", slot.id, (int) slot.cache_tokens.size());
+                    LOG_TEE("Context exhausted. Slot %d released (%d tokens in cache)\n", slot.id, (int) slot.cache_tokens.size());
 
                     continue;
                     // END LOCALAI changes
@@ -1658,7 +1616,7 @@ struct llama_server_context
 
             // TODO: we always have to take into account the "system_tokens"
             //       this is not great and needs to be improved somehow
-            common_batch_add(batch, slot.sampled, system_tokens.size() + slot_npast, { slot.id }, true);
+            llama_batch_add(batch, slot.sampled, system_tokens.size() + slot_npast, { slot.id }, true);
             slot.n_past += 1;
         }
 
@@ -1752,7 +1710,7 @@ struct llama_server_context
 
                     if (!slot.params.cache_prompt)
                     {
-                        common_sampler_reset(slot.ctx_sampling);
+                        llama_sampling_reset(slot.ctx_sampling);
 
                         slot.n_past = 0;
                         slot.n_past_se = 0;
@@ -1764,7 +1722,7 @@ struct llama_server_context
                         // push the prompt into the sampling context (do not apply grammar)
                         for (auto &token : prompt_tokens)
                         {
-                            common_sampler_accept(slot.ctx_sampling, token, false);
+                            llama_sampling_accept(slot.ctx_sampling, ctx, token, false);
                         }
 
                         slot.n_past = common_part(slot.cache_tokens, prompt_tokens);
@@ -1856,17 +1814,16 @@ struct llama_server_context
                                 ga_i += ga_w/ga_n;
                             }
                         }
-                        common_batch_add(batch, prefix_tokens[slot.n_past], system_tokens.size() + slot_npast, {slot.id }, false);
+                        llama_batch_add(batch, prefix_tokens[slot.n_past], system_tokens.size() + slot_npast, {slot.id }, false);
                         slot_npast++;
                     }
 
                     if (has_images && !ingest_images(slot, n_batch))
                     {
-                        LOG_ERR("%s: failed processing images Slot id : %d, Task id: %d", 
-                            __func__,
-                            slot.id,
-                            slot.task_id
-                        );
+                        LOG_ERROR("failed processing images", {
+                            "slot_id", slot.id,
+                            "task_id", slot.task_id,
+                        });
                         // FIXME @phymbert: to be properly tested
                         //  early returning without changing the slot state will block the slot for ever
                         // no one at the moment is checking the return value
@@ -1906,10 +1863,10 @@ struct llama_server_context
                         const int bd = (slot.ga_w / slot.ga_n) * (slot.ga_n - 1);
                         const int dd = (slot.ga_w / slot.ga_n) - ib * bd - slot.ga_w;
 
-                        LOG("\n");
-                        LOG("shift: [%6d, %6d] + %6d -> [%6d, %6d]\n", slot.ga_i, slot.n_past_se, ib * bd, slot.ga_i + ib * bd, slot.n_past_se + ib * bd);
-                        LOG("div:   [%6d, %6d] / %6d -> [%6d, %6d]\n", slot.ga_i + ib * bd, slot.ga_i + ib * bd + slot.ga_w, slot.ga_n, (slot.ga_i + ib * bd) / slot.ga_n, (slot.ga_i + ib * bd + slot.ga_w) / slot.ga_n);
-                        LOG("shift: [%6d, %6d] + %6d -> [%6d, %6d]\n", slot.ga_i + ib * bd + slot.ga_w, slot.n_past_se + ib * bd, dd, slot.ga_i + ib * bd + slot.ga_w + dd, slot.n_past_se + ib * bd + dd);
+                        LOG_TEE("\n");
+                        LOG_TEE("shift: [%6d, %6d] + %6d -> [%6d, %6d]\n", slot.ga_i, slot.n_past_se, ib * bd, slot.ga_i + ib * bd, slot.n_past_se + ib * bd);
+                        LOG_TEE("div:   [%6d, %6d] / %6d -> [%6d, %6d]\n", slot.ga_i + ib * bd, slot.ga_i + ib * bd + slot.ga_w, slot.ga_n, (slot.ga_i + ib * bd) / slot.ga_n, (slot.ga_i + ib * bd + slot.ga_w) / slot.ga_n);
+                        LOG_TEE("shift: [%6d, %6d] + %6d -> [%6d, %6d]\n", slot.ga_i + ib * bd + slot.ga_w, slot.n_past_se + ib * bd, dd, slot.ga_i + ib * bd + slot.ga_w + dd, slot.n_past_se + ib * bd + dd);
 
                         llama_kv_cache_seq_add(ctx, slot.id, slot.ga_i, slot.n_past_se, ib * bd);
                         llama_kv_cache_seq_div(ctx, slot.id, slot.ga_i + ib * bd, slot.ga_i + ib * bd + slot.ga_w,slot.ga_n);
@@ -1919,7 +1876,7 @@ struct llama_server_context
 
                         slot.ga_i += slot.ga_w / slot.ga_n;
 
-                        LOG("\nn_past_old = %d, n_past = %d, ga_i = %d\n\n", slot.n_past_se + bd, slot.n_past_se, slot.ga_i);
+                        LOG_TEE("\nn_past_old = %d, n_past = %d, ga_i = %d\n\n", slot.n_past_se + bd, slot.n_past_se, slot.ga_i);
                     }
                     slot.n_past_se += n_tokens;
                 }
@@ -1934,6 +1891,7 @@ struct llama_server_context
                 batch.n_seq_id + i,
                 batch.seq_id   + i,
                 batch.logits   + i,
+                0, 0, 0, // unused
             };
 
             const int ret = llama_decode(ctx, batch_view);
@@ -1943,11 +1901,11 @@ struct llama_server_context
                 if (n_batch == 1 || ret < 0)
                 {
                     // if you get here, it means the KV cache is full - try increasing it via the context size
-                    LOG("%s : failed to decode the batch, n_batch = %d, ret = %d\n", __func__, n_batch, ret);
+                    LOG_TEE("%s : failed to decode the batch, n_batch = %d, ret = %d\n", __func__, n_batch, ret);
                     return false;
                 }
 
-                LOG("%s : failed to find free space in the KV cache, retrying with smaller n_batch = %d\n", __func__, n_batch / 2);
+                LOG_TEE("%s : failed to find free space in the KV cache, retrying with smaller n_batch = %d\n", __func__, n_batch / 2);
 
                 // retry with half the batch size to try to find a free slot in the KV cache
                 n_batch /= 2;
@@ -1972,9 +1930,9 @@ struct llama_server_context
                 }
 
                 completion_token_output result;
-                const llama_token id = common_sampler_sample(slot.ctx_sampling, ctx, slot.i_batch - i);
+                const llama_token id = llama_sampling_sample(slot.ctx_sampling, ctx, NULL, slot.i_batch - i);
 
-                common_sampler_accept(slot.ctx_sampling, id, true);
+                llama_sampling_accept(slot.ctx_sampling, ctx, id, true);
 
                 slot.n_decoded += 1;
                 if (slot.n_decoded == 1)
@@ -1984,14 +1942,19 @@ struct llama_server_context
                     metrics.on_prompt_eval(slot);
                 }
 
+                llama_token_data_array cur_p = { slot.ctx_sampling->cur.data(), slot.ctx_sampling->cur.size(), false };
                 result.tok = id;
-                const auto * cur_p = common_sampler_get_candidates(slot.ctx_sampling);
 
-                for (size_t i = 0; i < (size_t) slot.sparams.n_probs; ++i) {
-                    result.probs.push_back({
-                        cur_p->data[i].id,
-                        i >= cur_p->size ? 0.0f : cur_p->data[i].p,
-                    });
+                const int32_t n_probs = slot.sparams.n_probs;
+                if (slot.sparams.temp <= 0 && n_probs > 0)
+                {
+                    // for llama_sample_token_greedy we need to sort candidates
+                    llama_sample_softmax(ctx, &cur_p);
+                }
+
+                for (size_t i = 0; i < std::min(cur_p.size, (size_t)n_probs); ++i)
+                {
+                    result.probs.push_back({cur_p.data[i].id, cur_p.data[i].p});
                 }
 
                 if (!process_token(result, slot))
@@ -2038,7 +2001,7 @@ static json format_partial_response(
 struct token_translator
 {
     llama_context * ctx;
-    std::string operator()(llama_token tok)                    const { return common_token_to_piece(ctx, tok); }
+    std::string operator()(llama_token tok)                    const { return llama_token_to_piece(ctx, tok); }
     std::string operator()(const completion_token_output &cto) const { return (*this)(cto.tok); }
 };
 
@@ -2103,6 +2066,7 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, llama
     //     slot->params.n_predict        = json_value(data, "n_predict",         default_params.n_predict);
     //     slot->sparams.top_k           = json_value(data, "top_k",             default_sparams.top_k);
     //     slot->sparams.top_p           = json_value(data, "top_p",             default_sparams.top_p);
+    //     slot->sparams.tfs_z           = json_value(data, "tfs_z",             default_sparams.tfs_z);
     //     slot->sparams.typical_p       = json_value(data, "typical_p",         default_sparams.typical_p);
     //     slot->sparams.temp            = json_value(data, "temperature",       default_sparams.temp);
     //     slot->sparams.penalty_last_n  = json_value(data, "repeat_last_n",     default_sparams.penalty_last_n);
@@ -2126,6 +2090,7 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, llama
     data["n_predict"] = predict->tokens() == 0 ? -1 : predict->tokens();
     data["top_k"] = predict->topk();
     data["top_p"] = predict->topp();
+    data["tfs_z"] = predict->tailfreesamplingz();
     data["typical_p"] = predict->typicalp();
     data["temperature"] = predict->temperature();
     data["repeat_last_n"] = predict->repeat();
@@ -2141,10 +2106,6 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, llama
     data["grammar"] = predict->grammar();
     data["prompt"] = predict->prompt();
     data["ignore_eos"] = predict->ignoreeos();
-    data["embeddings"] = predict->embeddings();
-
-    // Add the correlationid to json data
-    data["correlation_id"] = predict->correlationid();
 
     // for each image in the request, add the image data
     //
@@ -2172,6 +2133,7 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, llama
 //     llama.params.n_predict = predict->tokens() == 0 ? -1 : predict->tokens();
 //     llama.params.sparams.top_k = predict->topk();
 //     llama.params.sparams.top_p = predict->topp();
+//     llama.params.sparams.tfs_z = predict->tailfreesamplingz();
 //     llama.params.sparams.typical_p = predict->typicalp();
 //     llama.params.sparams.penalty_last_n = predict->repeat();
 //     llama.params.sparams.temp = predict->temperature();
@@ -2229,7 +2191,7 @@ json parse_options(bool streaming, const backend::PredictOptions* predict, llama
 // }
 
 static void params_parse(const backend::ModelOptions* request,
-                                common_params & params) {
+                                gpt_params & params) {
    
     // this is comparable to: https://github.com/ggerganov/llama.cpp/blob/d9b33fe95bd257b36c84ee5769cc048230067d6f/examples/server/server.cpp#L1809
 
@@ -2243,7 +2205,7 @@ static void params_parse(const backend::ModelOptions* request,
     params.model_alias =  request->modelfile();
     params.n_ctx = request->contextsize();
     //params.memory_f16 = request->f16memory();
-    params.cpuparams.n_threads = request->threads();
+    params.n_threads = request->threads();
     params.n_gpu_layers = request->ngpulayers();
     params.n_batch = request->nbatch();
     // Set params.n_parallel by environment variable (LLAMA_PARALLEL), defaults to 1
@@ -2255,12 +2217,6 @@ static void params_parse(const backend::ModelOptions* request,
     } else {
         params.n_parallel = 1;
     }
-
-    const char *llama_grpc_servers = std::getenv("LLAMACPP_GRPC_SERVERS");
-    if (llama_grpc_servers != NULL) {
-        params.rpc_servers = std::string(llama_grpc_servers);
-    }
-    
     // TODO: Add yarn
 
     if (!request->tensorsplit().empty()) {
@@ -2293,14 +2249,11 @@ static void params_parse(const backend::ModelOptions* request,
      }
      // get the directory of modelfile
      std::string model_dir = params.model.substr(0, params.model.find_last_of("/\\"));
-     params.lora_adapters.push_back({ model_dir + "/"+request->loraadapter(), scale_factor });
+     params.lora_adapter.push_back(std::make_tuple(model_dir + "/"+request->loraadapter(), scale_factor));
+     params.lora_base  =  model_dir + "/"+request->lorabase();
     }
     params.use_mlock = request->mlock();
     params.use_mmap = request->mmap();
-    params.flash_attn = request->flashattention();
-    params.no_kv_offload = request->nokvoffload();
-    params.ctx_shift = false; // We control context-shifting in any case (and we disable it as it could just lead to infinite loops)
-
     params.embedding = request->embeddings();
 
     if (request->ropescaling() == "none")   { params.rope_scaling_type = LLAMA_ROPE_SCALING_TYPE_NONE; }
@@ -2338,7 +2291,7 @@ public:
 
   grpc::Status LoadModel(ServerContext* context, const backend::ModelOptions* request, backend::Result* result) {
     // Implement LoadModel RPC
-    common_params params;
+    gpt_params params;
     params_parse(request, params);
 
     llama_backend_init();
@@ -2379,15 +2332,6 @@ public:
                 std::string completion_text = result.result_json.value("content", "");
 
                 reply.set_message(completion_text);
-                int32_t tokens_predicted = result.result_json.value("tokens_predicted", 0);
-                reply.set_tokens(tokens_predicted);
-                int32_t tokens_evaluated = result.result_json.value("tokens_evaluated", 0);
-                reply.set_prompt_tokens(tokens_evaluated);
-
-                // Log Request Correlation Id
-                LOG_VERBOSE("correlation:", {
-                    { "id", data["correlation_id"] }
-                });
 
                 // Send the reply
                 writer->Write(reply);
@@ -2412,17 +2356,7 @@ public:
         std::string completion_text;
         task_result result = llama.queue_results.recv(task_id);
         if (!result.error && result.stop) {
-            
-            // Log Request Correlation Id
-            LOG_VERBOSE("correlation:", {
-                { "id", data["correlation_id"] }
-            });
-
             completion_text = result.result_json.value("content", "");
-            int32_t tokens_predicted = result.result_json.value("tokens_predicted", 0);
-            int32_t tokens_evaluated = result.result_json.value("tokens_evaluated", 0);
-            reply->set_prompt_tokens(tokens_evaluated);
-            reply->set_tokens(tokens_predicted);
             reply->set_message(completion_text);
         }
         else
@@ -2432,56 +2366,6 @@ public:
 
         return grpc::Status::OK;
     }
-
-    /// https://github.com/ggerganov/llama.cpp/blob/aa2341298924ac89778252015efcb792f2df1e20/examples/server/server.cpp#L2969
-    grpc::Status Embedding(ServerContext* context, const backend::PredictOptions* request, backend::EmbeddingResult* embeddingResult) {
-        json data = parse_options(false, request, llama);
-        const int task_id = llama.queue_tasks.get_new_id();
-        llama.queue_results.add_waiting_task_id(task_id);
-        llama.request_completion(task_id, { {"prompt", data["embeddings"]}, { "n_predict", 0}, {"image_data", ""} }, false, true, -1);
-        // get the result
-        task_result result = llama.queue_results.recv(task_id);
-        //std::cout << "Embedding result JSON" << result.result_json.dump() << std::endl;
-        llama.queue_results.remove_waiting_task_id(task_id);
-        if (!result.error && result.stop) {
-            std::vector<float> embeddings = result.result_json.value("embedding", std::vector<float>());
-            // loop the vector and set the embeddings results
-            for (int i = 0; i < embeddings.size(); i++) {
-                embeddingResult->add_embeddings(embeddings[i]);
-            }
-        }
-        else
-        {
-            return grpc::Status::OK;
-        }
-
-        return grpc::Status::OK;
-    }
-
-    grpc::Status GetMetrics(ServerContext* context, const backend::MetricsRequest* request, backend::MetricsResponse* response) {
-        llama_client_slot* active_slot = llama.get_active_slot();
-
-        if (active_slot != nullptr) {
-            // Calculate the tokens per second using existing logic
-            double tokens_per_second = 1e3 / active_slot->t_token_generation * active_slot->n_decoded;
-
-            // Populate the response with metrics
-            response->set_slot_id(active_slot->id);
-            response->set_prompt_json_for_slot(active_slot->prompt.dump());
-            response->set_tokens_per_second(tokens_per_second);
-            response->set_tokens_generated(active_slot->n_decoded);
-            response->set_prompt_tokens_processed(active_slot->num_prompt_tokens_processed);
-        } else {
-            // Handle case when no active slot exists
-            response->set_slot_id(0);
-            response->set_prompt_json_for_slot("");
-            response->set_tokens_per_second(0);
-            response->set_tokens_generated(0);
-            response->set_prompt_tokens_processed(0);
-        }
-
-        return grpc::Status::OK;
-    } 
 };
 
 void RunServer(const std::string& server_address) {

backend/cpp/llama/patches/01-llava.patch

@@ -1,13 +0,0 @@
-diff --git a/examples/llava/clip.cpp b/examples/llava/clip.cpp
-index 342042ff..224db9b5 100644
---- a/examples/llava/clip.cpp
-+++ b/examples/llava/clip.cpp
-@@ -2419,7 +2419,7 @@ bool clip_image_batch_encode(clip_ctx * ctx, const int n_threads, const clip_ima
-             struct ggml_tensor * patches = ggml_graph_get_tensor(gf, "patches");
-             int* patches_data = (int*)malloc(ggml_nbytes(patches));
-             for (int i = 0; i < num_patches; i++) {
--                patches_data[i] = i + 1;
-+                patches_data[i] = i;
-             }
-             ggml_backend_tensor_set(patches, patches_data, 0, ggml_nbytes(patches));
-             free(patches_data);

backend/cpp/llama/prepare.sh

@@ -1,27 +0,0 @@
-#!/bin/bash
-
-## Patches
-## Apply patches from the `patches` directory
-for patch in $(ls patches); do
-    echo "Applying patch $patch"
-    patch -d llama.cpp/ -p1 < patches/$patch
-done 
-
-cp -r CMakeLists.txt llama.cpp/examples/grpc-server/
-cp -r grpc-server.cpp llama.cpp/examples/grpc-server/
-cp -rfv json.hpp llama.cpp/examples/grpc-server/
-cp -rfv utils.hpp llama.cpp/examples/grpc-server/
-    
-if grep -q "grpc-server" llama.cpp/examples/CMakeLists.txt; then
-    echo "grpc-server already added"
-else
-    echo "add_subdirectory(grpc-server)" >> llama.cpp/examples/CMakeLists.txt
-fi
-
-## 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/llava.cpp llama.cpp/examples/grpc-server/llava.cpp
-echo '#include "llama.h"' > llama.cpp/examples/grpc-server/llava.h
-cat llama.cpp/examples/llava/llava.h >> llama.cpp/examples/grpc-server/llava.h
-cp -rfv llama.cpp/examples/llava/clip.cpp llama.cpp/examples/grpc-server/clip.cpp

backend/cpp/llama/utils.hpp

@@ -480,4 +480,31 @@ static inline std::vector<uint8_t> base64_decode(const std::string & encoded_str
     }
 
     return ret;
+}
+
+//
+// random string / id
+//
+
+static std::string random_string()
+{
+    static const std::string str("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz");
+
+    std::random_device rd;
+    std::mt19937 generator(rd());
+
+    std::string result(32, ' ');
+
+    for (int i = 0; i < 32; ++i) {
+        result[i] = str[generator() % str.size()];
+    }
+
+    return result;
+}
+
+static std::string gen_chatcmplid()
+{
+    std::stringstream chatcmplid;
+    chatcmplid << "chatcmpl-" << random_string();
+    return chatcmplid.str();
 }

backend/go/bark/Makefile

@@ -1,25 +0,0 @@
-INCLUDE_PATH := $(abspath ./)
-LIBRARY_PATH := $(abspath ./)
-
-AR?=ar
-
-BUILD_TYPE?=
-# keep standard at C11 and C++11
-CXXFLAGS = -I. -I$(INCLUDE_PATH)/../../../sources/bark.cpp/examples -I$(INCLUDE_PATH)/../../../sources/bark.cpp/spm-headers -I$(INCLUDE_PATH)/../../../sources/bark.cpp -O3 -DNDEBUG -std=c++17 -fPIC
-LDFLAGS  = -L$(LIBRARY_PATH) -L$(LIBRARY_PATH)/../../../sources/bark.cpp/build/examples -lbark -lstdc++ -lm
-
-# warnings
-CXXFLAGS += -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function
-
-gobark.o:
-	$(CXX) $(CXXFLAGS) gobark.cpp -o gobark.o -c $(LDFLAGS)
-
-libbark.a: gobark.o
-	cp $(INCLUDE_PATH)/../../../sources/bark.cpp/build/libbark.a ./
-	$(AR) rcs libbark.a gobark.o
-	$(AR) rcs libbark.a $(LIBRARY_PATH)/../../../sources/bark.cpp/build/encodec.cpp/ggml/src/CMakeFiles/ggml.dir/ggml.c.o
-	$(AR) rcs libbark.a $(LIBRARY_PATH)/../../../sources/bark.cpp/build/encodec.cpp/ggml/src/CMakeFiles/ggml.dir/ggml-alloc.c.o
-	$(AR) rcs libbark.a $(LIBRARY_PATH)/../../../sources/bark.cpp/build/encodec.cpp/ggml/src/CMakeFiles/ggml.dir/ggml-backend.c.o
-
-clean:
-	rm -f gobark.o libbark.a

backend/go/bark/gobark.cpp

@@ -1,85 +0,0 @@
-#include <iostream>
-#include <tuple>
-
-#include "bark.h"
-#include "gobark.h"
-#include "common.h"
-#include "ggml.h"
-
-struct bark_context *c;
-
-void bark_print_progress_callback(struct bark_context *bctx, enum bark_encoding_step step, int progress, void *user_data) {
-    if (step == bark_encoding_step::SEMANTIC) {
-        printf("\rGenerating semantic tokens... %d%%", progress);
-    } else if (step == bark_encoding_step::COARSE) {
-        printf("\rGenerating coarse tokens... %d%%", progress);
-    } else if (step == bark_encoding_step::FINE) {
-        printf("\rGenerating fine tokens... %d%%", progress);
-    }
-    fflush(stdout);
-}
-
-int load_model(char *model) {
-    // initialize bark context
-    struct bark_context_params ctx_params = bark_context_default_params();
-    bark_params params;
-
-    params.model_path = model;
-
-   // ctx_params.verbosity = verbosity;
-    ctx_params.progress_callback = bark_print_progress_callback;
-    ctx_params.progress_callback_user_data = nullptr;
-
-    struct bark_context *bctx = bark_load_model(params.model_path.c_str(), ctx_params, params.seed);
-    if (!bctx) {
-        fprintf(stderr, "%s: Could not load model\n", __func__);
-        return 1;
-    }
-
-    c = bctx;
-
-    return 0;
-}
-
-int tts(char *text,int  threads, char *dst ) {
-
-    ggml_time_init();
-    const int64_t t_main_start_us = ggml_time_us();
-
-    // generate audio
-    if (!bark_generate_audio(c, text, threads)) {
-        fprintf(stderr, "%s: An error occured. If the problem persists, feel free to open an issue to report it.\n", __func__);
-        return 1;
-    }
-
-    const float *audio_data = bark_get_audio_data(c);
-    if (audio_data == NULL) {
-        fprintf(stderr, "%s: Could not get audio data\n", __func__);
-        return 1;
-    }
-
-    const int audio_arr_size = bark_get_audio_data_size(c);
-
-    std::vector<float> audio_arr(audio_data, audio_data + audio_arr_size);
-
-    write_wav_on_disk(audio_arr, dst);
-
-    // report timing
-    {
-        const int64_t t_main_end_us = ggml_time_us();
-        const int64_t t_load_us = bark_get_load_time(c);
-        const int64_t t_eval_us = bark_get_eval_time(c);
-
-        printf("\n\n");
-        printf("%s:     load time = %8.2f ms\n", __func__, t_load_us / 1000.0f);
-        printf("%s:     eval time = %8.2f ms\n", __func__, t_eval_us / 1000.0f);
-        printf("%s:    total time = %8.2f ms\n", __func__, (t_main_end_us - t_main_start_us) / 1000.0f);
-    }
-    
-    return 0;
-}
-
-int unload() {
-    bark_free(c);
-}
-

backend/go/bark/gobark.go

@@ -1,52 +0,0 @@
-package main
-
-// #cgo CXXFLAGS: -I${SRCDIR}/../../../sources/bark.cpp/ -I${SRCDIR}/../../../sources/bark.cpp/encodec.cpp -I${SRCDIR}/../../../sources/bark.cpp/examples -I${SRCDIR}/../../../sources/bark.cpp/spm-headers
-// #cgo LDFLAGS: -L${SRCDIR}/ -L${SRCDIR}/../../../sources/bark.cpp/build/examples -L${SRCDIR}/../../../sources/bark.cpp/build/encodec.cpp/ -lbark -lencodec -lcommon
-// #include <gobark.h>
-// #include <stdlib.h>
-import "C"
-
-import (
-	"fmt"
-	"unsafe"
-
-	"github.com/mudler/LocalAI/pkg/grpc/base"
-	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
-)
-
-type Bark struct {
-	base.SingleThread
-	threads int
-}
-
-func (sd *Bark) Load(opts *pb.ModelOptions) error {
-
-	sd.threads = int(opts.Threads)
-
-	modelFile := C.CString(opts.ModelFile)
-	defer C.free(unsafe.Pointer(modelFile))
-
-	ret := C.load_model(modelFile)
-	if ret != 0 {
-		return fmt.Errorf("inference failed")
-	}
-
-	return nil
-}
-
-func (sd *Bark) TTS(opts *pb.TTSRequest) error {
-	t := C.CString(opts.Text)
-	defer C.free(unsafe.Pointer(t))
-
-	dst := C.CString(opts.Dst)
-	defer C.free(unsafe.Pointer(dst))
-
-	threads := C.int(sd.threads)
-
-	ret := C.tts(t, threads, dst)
-	if ret != 0 {
-		return fmt.Errorf("inference failed")
-	}
-
-	return nil
-}

backend/go/bark/gobark.h

@@ -1,8 +0,0 @@
-#ifdef __cplusplus
-extern "C" {
-#endif
-int load_model(char *model);
-int tts(char *text,int  threads, char *dst );
-#ifdef __cplusplus
-}
-#endif

backend/go/image/stablediffusion-ggml/Makefile

@@ -1,21 +0,0 @@
-INCLUDE_PATH := $(abspath ./)
-LIBRARY_PATH := $(abspath ./)
-
-AR?=ar
-
-BUILD_TYPE?=
-# keep standard at C11 and C++11
-CXXFLAGS = -I. -I$(INCLUDE_PATH)/../../../../sources/stablediffusion-ggml.cpp/thirdparty -I$(INCLUDE_PATH)/../../../../sources/stablediffusion-ggml.cpp/ggml/include -I$(INCLUDE_PATH)/../../../../sources/stablediffusion-ggml.cpp -O3 -DNDEBUG -std=c++17 -fPIC
-
-# warnings
-CXXFLAGS += -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function
-
-gosd.o:
-	$(CXX) $(CXXFLAGS) gosd.cpp -o gosd.o -c
-
-libsd.a: gosd.o
-	cp $(INCLUDE_PATH)/../../../../sources/stablediffusion-ggml.cpp/build/libstable-diffusion.a ./libsd.a
-	$(AR) rcs libsd.a gosd.o
-
-clean:
-	rm -f gosd.o libsd.a

backend/go/image/stablediffusion-ggml/gosd.cpp

@@ -1,228 +0,0 @@
-#include <stdio.h>
-#include <string.h>
-#include <time.h>
-#include <iostream>
-#include <random>
-#include <string>
-#include <vector>
-#include "gosd.h"
-
-// #include "preprocessing.hpp"
-#include "flux.hpp"
-#include "stable-diffusion.h"
-
-#define STB_IMAGE_IMPLEMENTATION
-#define STB_IMAGE_STATIC
-#include "stb_image.h"
-
-#define STB_IMAGE_WRITE_IMPLEMENTATION
-#define STB_IMAGE_WRITE_STATIC
-#include "stb_image_write.h"
-
-#define STB_IMAGE_RESIZE_IMPLEMENTATION
-#define STB_IMAGE_RESIZE_STATIC
-#include "stb_image_resize.h"
-
-// Names of the sampler method, same order as enum sample_method in stable-diffusion.h
-const char* sample_method_str[] = {
-    "euler_a",
-    "euler",
-    "heun",
-    "dpm2",
-    "dpm++2s_a",
-    "dpm++2m",
-    "dpm++2mv2",
-    "ipndm",
-    "ipndm_v",
-    "lcm",
-};
-
-// Names of the sigma schedule overrides, same order as sample_schedule in stable-diffusion.h
-const char* schedule_str[] = {
-    "default",
-    "discrete",
-    "karras",
-    "exponential",
-    "ays",
-    "gits",
-};
-
-sd_ctx_t* sd_c;
-
-sample_method_t sample_method;
-
-int load_model(char *model, char* options[], int threads, int diff) {
-    fprintf (stderr, "Loading model!\n");
-
-    char *stableDiffusionModel = "";
-    if (diff == 1 ) {
-        stableDiffusionModel = model;
-        model = "";
-    }
-
-    // decode options. Options are in form optname:optvale, or if booleans only optname.
-    char *clip_l_path  = "";
-    char *clip_g_path  = "";
-    char *t5xxl_path  = "";
-    char *vae_path  = "";
-    char *scheduler = "";
-    char *sampler = "";
-
-    // If options is not NULL, parse options
-    for (int i = 0; options[i] != NULL; i++) {
-        char *optname = strtok(options[i], ":");
-        char *optval = strtok(NULL, ":");
-        if (optval == NULL) {
-            optval = "true";
-        }
-
-        if (!strcmp(optname, "clip_l_path")) {
-            clip_l_path = optval;
-        }
-        if (!strcmp(optname, "clip_g_path")) {
-            clip_g_path = optval;
-        }
-        if (!strcmp(optname, "t5xxl_path")) {
-            t5xxl_path = optval;
-        }
-        if (!strcmp(optname, "vae_path")) {
-            vae_path = optval;
-        }
-        if (!strcmp(optname, "scheduler")) {
-            scheduler = optval;
-        }
-        if (!strcmp(optname, "sampler")) {
-            sampler = optval;
-        }
-    }
-
-    int sample_method_found = -1;
-    for (int m = 0; m < N_SAMPLE_METHODS; m++) {
-        if (!strcmp(sampler, sample_method_str[m])) {
-            sample_method_found = m;
-        }
-    }
-    if (sample_method_found == -1) {
-        fprintf(stderr, "Invalid sample method, default to EULER_A!\n");
-        sample_method_found = EULER_A;
-    }
-    sample_method = (sample_method_t)sample_method_found;
-
-    int schedule_found            = -1;
-    for (int d = 0; d < N_SCHEDULES; d++) {
-        if (!strcmp(scheduler, schedule_str[d])) {
-            schedule_found = d;
-                fprintf (stderr, "Found scheduler: %s\n", scheduler);
-
-        }
-    }
-
-    if (schedule_found == -1) {
-        fprintf (stderr, "Invalid scheduler! using DEFAULT\n");
-        schedule_found = DEFAULT;
-    }
-
-    schedule_t schedule = (schedule_t)schedule_found;
-    
-    fprintf (stderr, "Creating context\n");
-    sd_ctx_t* sd_ctx = new_sd_ctx(model,
-                                  clip_l_path,
-                                  clip_g_path,
-                                  t5xxl_path,
-                                  stableDiffusionModel,
-                                  vae_path,
-                                  "",
-                                  "",
-                                  "",
-                                  "",
-                                  "",
-                                  false,
-                                  false,
-                                  false,
-                                  threads,
-                                  SD_TYPE_COUNT,
-                                  STD_DEFAULT_RNG,
-                                  schedule,
-                                  false,
-                                  false,
-                                  false,
-                                  false);
-
-    if (sd_ctx == NULL) {
-        fprintf (stderr, "failed loading model (generic error)\n");
-        return 1;
-    }
-    fprintf (stderr, "Created context: OK\n");
-
-    sd_c = sd_ctx;
-
-    return 0;
-}
-
-int gen_image(char *text, char *negativeText, int width, int height, int steps, int seed , char *dst, float cfg_scale) {
-
-    sd_image_t* results;
-
-    std::vector<int> skip_layers = {7, 8, 9};
-
-    fprintf (stderr, "Generating image\n");
-
-    results = txt2img(sd_c,
-                            text,
-                            negativeText,
-                            -1, //clip_skip
-                            cfg_scale, // sfg_scale
-                            3.5f,
-                            width,
-                            height,
-                            sample_method, 
-                            steps,
-                            seed,
-                            1,
-                            NULL,
-                            0.9f,
-                            20.f,
-                            false,
-                            "",
-                            skip_layers.data(),
-                            skip_layers.size(),
-                            0,
-                            0.01,
-                            0.2);
-
-    if (results == NULL) {
-        fprintf (stderr, "NO results\n");
-        return 1;
-    }
-
-    if (results[0].data == NULL) {
-        fprintf (stderr, "Results with no data\n");
-        return 1;
-    }
-
-    fprintf (stderr, "Writing PNG\n");
-
-    fprintf (stderr, "DST: %s\n", dst);
-    fprintf (stderr, "Width: %d\n", results[0].width);
-    fprintf (stderr, "Height: %d\n", results[0].height);
-    fprintf (stderr, "Channel: %d\n", results[0].channel);
-    fprintf (stderr, "Data: %p\n", results[0].data);
-
-    stbi_write_png(dst, results[0].width, results[0].height, results[0].channel,
-                       results[0].data, 0, NULL);
-    fprintf (stderr, "Saved resulting image to '%s'\n", dst);
-
-    // TODO: free results. Why does it crash?
-
-    free(results[0].data);
-    results[0].data = NULL;
-    free(results);
-    fprintf (stderr, "gen_image is done", dst);
-
-    return 0;
-}
-
-int unload() {
-    free_sd_ctx(sd_c);
-}
-

backend/go/image/stablediffusion-ggml/gosd.go

@@ -1,96 +0,0 @@
-package main
-
-// #cgo CXXFLAGS: -I${SRCDIR}/../../../../sources/stablediffusion-ggml.cpp/thirdparty -I${SRCDIR}/../../../../sources/stablediffusion-ggml.cpp -I${SRCDIR}/../../../../sources/stablediffusion-ggml.cpp/ggml/include
-// #cgo LDFLAGS: -L${SRCDIR}/ -L${SRCDIR}/../../../../sources/stablediffusion-ggml.cpp/build/ggml/src/ggml-cpu -L${SRCDIR}/../../../../sources/stablediffusion-ggml.cpp/build/ggml/src -lsd -lstdc++ -lm -lggml -lggml-base -lggml-cpu -lgomp
-// #include <gosd.h>
-// #include <stdlib.h>
-import "C"
-
-import (
-	"fmt"
-	"os"
-	"path/filepath"
-	"strings"
-	"unsafe"
-
-	"github.com/mudler/LocalAI/pkg/grpc/base"
-	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
-	"github.com/mudler/LocalAI/pkg/utils"
-)
-
-type SDGGML struct {
-	base.SingleThread
-	threads      int
-	sampleMethod string
-	cfgScale     float32
-}
-
-func (sd *SDGGML) Load(opts *pb.ModelOptions) error {
-
-	sd.threads = int(opts.Threads)
-
-	modelFile := C.CString(opts.ModelFile)
-	defer C.free(unsafe.Pointer(modelFile))
-
-	var options **C.char
-	// prepare the options array to pass to C
-
-	size := C.size_t(unsafe.Sizeof((*C.char)(nil)))
-	length := C.size_t(len(opts.Options))
-	options = (**C.char)(C.malloc(length * size))
-	view := (*[1 << 30]*C.char)(unsafe.Pointer(options))[0:len(opts.Options):len(opts.Options)]
-
-	var diffusionModel int
-
-	var oo []string
-	for _, op := range opts.Options {
-		if op == "diffusion_model" {
-			diffusionModel = 1
-			continue
-		}
-
-		// If it's an option path, we resolve absolute path from the model path
-		if strings.Contains(op, ":") && strings.Contains(op, "path") {
-			data := strings.Split(op, ":")
-			data[1] = filepath.Join(opts.ModelPath, data[1])
-			if err := utils.VerifyPath(data[1], opts.ModelPath); err == nil {
-				oo = append(oo, strings.Join(data, ":"))
-			}
-		} else {
-			oo = append(oo, op)
-		}
-	}
-
-	fmt.Fprintf(os.Stderr, "Options: %+v\n", oo)
-
-	for i, x := range oo {
-		view[i] = C.CString(x)
-	}
-
-	sd.cfgScale = opts.CFGScale
-
-	ret := C.load_model(modelFile, options, C.int(opts.Threads), C.int(diffusionModel))
-	if ret != 0 {
-		return fmt.Errorf("could not load model")
-	}
-
-	return nil
-}
-
-func (sd *SDGGML) GenerateImage(opts *pb.GenerateImageRequest) error {
-	t := C.CString(opts.PositivePrompt)
-	defer C.free(unsafe.Pointer(t))
-
-	dst := C.CString(opts.Dst)
-	defer C.free(unsafe.Pointer(dst))
-
-	negative := C.CString(opts.NegativePrompt)
-	defer C.free(unsafe.Pointer(negative))
-
-	ret := C.gen_image(t, negative, C.int(opts.Width), C.int(opts.Height), C.int(opts.Step), C.int(opts.Seed), dst, C.float(sd.cfgScale))
-	if ret != 0 {
-		return fmt.Errorf("inference failed")
-	}
-
-	return nil
-}

backend/go/image/stablediffusion-ggml/gosd.h

@@ -1,8 +0,0 @@
-#ifdef __cplusplus
-extern "C" {
-#endif
-int load_model(char *model, char* options[], int threads, int diffusionModel);
-int gen_image(char *text, char *negativeText, int width, int height, int steps, int seed, char *dst, float cfg_scale);
-#ifdef __cplusplus
-}
-#endif

backend/go/image/stablediffusion/main.go

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

backend/go/image/stablediffusion/stablediffusion.go

@@ -3,9 +3,9 @@ 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/mudler/LocalAI/pkg/grpc/base"
-	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
-	"github.com/mudler/LocalAI/pkg/stablediffusion"
+	"github.com/go-skynet/LocalAI/pkg/grpc/base"
+	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
+	"github.com/go-skynet/LocalAI/pkg/stablediffusion"
 )
 
 type Image struct {

backend/go/image/tinydream/main.go

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

backend/go/image/tinydream/tinydream.go

@@ -3,9 +3,9 @@ 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/mudler/LocalAI/pkg/grpc/base"
-	pb "github.com/mudler/LocalAI/pkg/grpc/proto"
-	"github.com/mudler/LocalAI/pkg/tinydream"
+	"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 {

backend/go/llm/bert/bert.go

@@ -0,0 +1,34 @@
+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 (
+	bert "github.com/go-skynet/go-bert.cpp"
+
+	"github.com/go-skynet/LocalAI/pkg/grpc/base"
+	pb "github.com/go-skynet/LocalAI/pkg/grpc/proto"
+)
+
+type Embeddings struct {
+	base.SingleThread
+	bert *bert.Bert
+}
+
+func (llm *Embeddings) Load(opts *pb.ModelOptions) error {
+	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 {
+			tokens = append(tokens, int(t))
+		}
+		return llm.bert.TokenEmbeddings(tokens, bert.SetThreads(int(opts.Threads)))
+	}
+
+	return llm.bert.Embeddings(opts.Embeddings, bert.SetThreads(int(opts.Threads)))
+}

backend/go/llm/bert/main.go

@@ -1,10 +1,11 @@
 package main
 
 // Note: this is started internally by LocalAI and a server is allocated for each model
+
 import (
 	"flag"
 
-	grpc "github.com/mudler/LocalAI/pkg/grpc"
+	grpc "github.com/go-skynet/LocalAI/pkg/grpc"
 )
 
 var (
@@ -14,7 +15,7 @@ var (
 func main() {
 	flag.Parse()
 
-	if err := grpc.StartServer(*addr, &Bark{}); err != nil {
+	if err := grpc.StartServer(*addr, &Embeddings{}); err != nil {
 		panic(err)
 	}
 }