feat(mlx): Add support for CUDA12, CUDA13, L4T, SBSA and CPU (#8380)

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

.github/workflows/backend.yml

@@ -91,10 +91,62 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-whisperx'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "whisperx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-mlx'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "mlx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-mlx-vlm'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "mlx-vlm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-mlx-audio'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "mlx-audio"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           # CUDA 12 builds
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-vibevoice'
@@ -107,7 +159,20 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-qwen-asr'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-qwen-tts'
@@ -120,7 +185,20 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-voxcpm'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "voxcpm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-pocket-tts'
@@ -146,7 +224,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-llama-cpp'
@@ -159,7 +237,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-vllm'
@@ -172,7 +250,20 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-vllm-omni'
+            runs-on: 'arc-runner-set'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "vllm-omni"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-transformers'
@@ -185,7 +276,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-diffusers'
@@ -198,7 +289,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-kokoro'
@@ -211,7 +302,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-faster-whisper'
@@ -222,6 +313,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-whisperx'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "whisperx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
             cuda-minor-version: "9"
@@ -237,20 +341,20 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
-            tag-suffix: '-gpu-nvidia-cuda-12-bark'
+            tag-suffix: '-gpu-nvidia-cuda-12-outetts'
             runs-on: 'ubuntu-latest'
             base-image: "ubuntu:24.04"
             skip-drivers: 'false'
-            backend: "bark"
+            backend: "outetts"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-chatterbox'
@@ -263,7 +367,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-moonshine'
@@ -276,7 +380,46 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-mlx'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-mlx-vlm'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx-vlm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-mlx-audio'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx-audio"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-stablediffusion-ggml'
@@ -289,7 +432,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-whisper'
@@ -302,7 +445,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-rfdetr'
@@ -315,20 +458,7 @@ jobs:
             ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
-            cuda-minor-version: "9"
-            platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            tag-suffix: '-gpu-nvidia-cuda-12-exllama2'
-            runs-on: 'ubuntu-latest'
-            base-image: "ubuntu:24.04"
-            skip-drivers: 'false'
-            backend: "exllama2"
-            dockerfile: "./backend/Dockerfile.python"
-            context: "./"
-            ubuntu-version: '2404'
-          - build-type: 'cublas'
-            cuda-major-version: "12"
-            cuda-minor-version: "9"
+            cuda-minor-version: "8"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
             tag-suffix: '-gpu-nvidia-cuda-12-neutts'
@@ -366,6 +496,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-qwen-asr'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -379,6 +522,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-voxcpm'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "voxcpm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -457,6 +613,19 @@ jobs:
             backend: "vibevoice"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-qwen-asr'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
           - build-type: 'l4t'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -496,6 +665,45 @@ jobs:
             backend: "diffusers"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-mlx'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "mlx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-mlx-vlm'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "mlx-vlm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+          - build-type: 'l4t'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-cuda-13-arm64-mlx-audio'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            ubuntu-version: '2404'
+            backend: "mlx-audio"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -527,11 +735,11 @@ jobs:
             cuda-minor-version: "0"
             platforms: 'linux/amd64'
             tag-latest: 'auto'
-            tag-suffix: '-gpu-nvidia-cuda-13-bark'
+            tag-suffix: '-gpu-nvidia-cuda-13-whisperx'
             runs-on: 'ubuntu-latest'
             base-image: "ubuntu:24.04"
             skip-drivers: 'false'
-            backend: "bark"
+            backend: "whisperx"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
@@ -561,6 +769,45 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-mlx'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-mlx-vlm'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx-vlm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-mlx-audio'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "mlx-audio"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"
@@ -666,6 +913,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'hipblas'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-rocm-hipblas-vllm-omni'
+            runs-on: 'arc-runner-set'
+            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
+            skip-drivers: 'false'
+            backend: "vllm-omni"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'hipblas'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -719,6 +979,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'hipblas'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-rocm-hipblas-qwen-asr'
+            runs-on: 'arc-runner-set'
+            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
+            skip-drivers: 'false'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'hipblas'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -732,6 +1005,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'hipblas'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-rocm-hipblas-voxcpm'
+            runs-on: 'arc-runner-set'
+            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
+            skip-drivers: 'false'
+            backend: "voxcpm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'hipblas'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -763,11 +1049,11 @@ jobs:
             cuda-minor-version: ""
             platforms: 'linux/amd64'
             tag-latest: 'auto'
-            tag-suffix: '-gpu-rocm-hipblas-coqui'
+            tag-suffix: '-gpu-rocm-hipblas-whisperx'
             runs-on: 'bigger-runner'
             base-image: "rocm/dev-ubuntu-24.04:6.4.4"
             skip-drivers: 'false'
-            backend: "coqui"
+            backend: "whisperx"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
@@ -776,11 +1062,11 @@ jobs:
             cuda-minor-version: ""
             platforms: 'linux/amd64'
             tag-latest: 'auto'
-            tag-suffix: '-gpu-rocm-hipblas-bark'
-            runs-on: 'arc-runner-set'
+            tag-suffix: '-gpu-rocm-hipblas-coqui'
+            runs-on: 'bigger-runner'
             base-image: "rocm/dev-ubuntu-24.04:6.4.4"
             skip-drivers: 'false'
-            backend: "bark"
+            backend: "coqui"
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
@@ -876,6 +1162,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-qwen-asr'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
           - build-type: 'l4t'
             cuda-major-version: "12"
             cuda-minor-version: "0"
@@ -915,6 +1214,45 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-mlx'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "mlx"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-mlx-vlm'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "mlx-vlm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
+          - build-type: 'l4t'
+            cuda-major-version: "12"
+            cuda-minor-version: "0"
+            platforms: 'linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-nvidia-l4t-mlx-audio'
+            runs-on: 'ubuntu-24.04-arm'
+            base-image: "nvcr.io/nvidia/l4t-jetpack:r36.4.0"
+            skip-drivers: 'true'
+            backend: "mlx-audio"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2204'
           # SYCL additional backends
           - build-type: 'intel'
             cuda-major-version: ""
@@ -955,6 +1293,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'intel'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-intel-qwen-asr'
+            runs-on: 'arc-runner-set'
+            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
+            skip-drivers: 'false'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'intel'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -968,6 +1319,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'intel'
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-intel-voxcpm'
+            runs-on: 'arc-runner-set'
+            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
+            skip-drivers: 'false'
+            backend: "voxcpm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'intel'
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -994,19 +1358,6 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
-          - build-type: 'intel'
-            cuda-major-version: ""
-            cuda-minor-version: ""
-            platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            tag-suffix: '-gpu-intel-bark'
-            runs-on: 'ubuntu-latest'
-            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
-            skip-drivers: 'false'
-            backend: "bark"
-            dockerfile: "./backend/Dockerfile.python"
-            context: "./"
-            ubuntu-version: '2404'
           # piper
           - build-type: ''
             cuda-major-version: ""
@@ -1021,20 +1372,6 @@ jobs:
             dockerfile: "./backend/Dockerfile.golang"
             context: "./"
             ubuntu-version: '2404'
-          # bark-cpp
-          - build-type: ''
-            cuda-major-version: ""
-            cuda-minor-version: ""
-            platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            tag-suffix: '-bark-cpp'
-            runs-on: 'ubuntu-latest'
-            base-image: "ubuntu:24.04"
-            skip-drivers: 'false'
-            backend: "bark-cpp"
-            dockerfile: "./backend/Dockerfile.golang"
-            context: "./"
-            ubuntu-version: '2404'
           - build-type: ''
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -1301,46 +1638,6 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2204'
-          # exllama2
-          - build-type: ''
-            cuda-major-version: ""
-            cuda-minor-version: ""
-            platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            tag-suffix: '-cpu-exllama2'
-            runs-on: 'ubuntu-latest'
-            base-image: "ubuntu:24.04"
-            skip-drivers: 'false'
-            backend: "exllama2"
-            dockerfile: "./backend/Dockerfile.python"
-            context: "./"
-            ubuntu-version: '2404'
-          - build-type: 'intel'
-            cuda-major-version: ""
-            cuda-minor-version: ""
-            platforms: 'linux/amd64'
-            tag-latest: 'auto'
-            tag-suffix: '-gpu-intel-exllama2'
-            runs-on: 'ubuntu-latest'
-            base-image: "intel/oneapi-basekit:2025.3.0-0-devel-ubuntu24.04"
-            skip-drivers: 'false'
-            backend: "exllama2"
-            dockerfile: "./backend/Dockerfile.python"
-            context: "./"
-            ubuntu-version: '2404'
-          - build-type: 'hipblas'
-            cuda-major-version: ""
-            cuda-minor-version: ""
-            platforms: 'linux/amd64'
-            skip-drivers: 'true'
-            tag-latest: 'auto'
-            tag-suffix: '-gpu-hipblas-exllama2'
-            base-image: "rocm/dev-ubuntu-24.04:6.4.4"
-            runs-on: 'ubuntu-latest'
-            backend: "exllama2"
-            dockerfile: "./backend/Dockerfile.python"
-            context: "./"
-            ubuntu-version: '2404'
           - build-type: 'l4t'
             cuda-major-version: "12"
             cuda-minor-version: "0"
@@ -1421,6 +1718,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64,linux/arm64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-qwen-asr'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "qwen-asr"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: ''
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -1434,6 +1744,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-voxcpm'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "voxcpm"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: ''
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -1447,6 +1770,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-outetts'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "outetts"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
   backend-jobs-darwin:
     uses: ./.github/workflows/backend_build_darwin.yml
     strategy:
@@ -1475,6 +1811,64 @@ jobs:
             tag-suffix: "-metal-darwin-arm64-whisper"
             build-type: "metal"
             lang: "go"
+          - backend: "vibevoice"
+            tag-suffix: "-metal-darwin-arm64-vibevoice"
+            build-type: "mps"
+          - backend: "qwen-asr"
+            tag-suffix: "-metal-darwin-arm64-qwen-asr"
+            build-type: "mps"
+          - backend: "qwen-tts"
+            tag-suffix: "-metal-darwin-arm64-qwen-tts"
+            build-type: "mps"
+          - backend: "voxcpm"
+            tag-suffix: "-metal-darwin-arm64-voxcpm"
+            build-type: "mps"
+          - backend: "pocket-tts"
+            tag-suffix: "-metal-darwin-arm64-pocket-tts"
+            build-type: "mps"
+          - backend: "moonshine"
+            tag-suffix: "-metal-darwin-arm64-moonshine"
+            build-type: "mps"
+          - backend: "whisperx"
+            tag-suffix: "-metal-darwin-arm64-whisperx"
+            build-type: "mps"
+          - backend: "rerankers"
+            tag-suffix: "-metal-darwin-arm64-rerankers"
+            build-type: "mps"
+          - backend: "transformers"
+            tag-suffix: "-metal-darwin-arm64-transformers"
+            build-type: "mps"
+          - backend: "kokoro"
+            tag-suffix: "-metal-darwin-arm64-kokoro"
+            build-type: "mps"
+          - backend: "faster-whisper"
+            tag-suffix: "-metal-darwin-arm64-faster-whisper"
+            build-type: "mps"
+          - backend: "coqui"
+            tag-suffix: "-metal-darwin-arm64-coqui"
+            build-type: "mps"
+          - backend: "rfdetr"
+            tag-suffix: "-metal-darwin-arm64-rfdetr"
+            build-type: "mps"
+          - backend: "kitten-tts"
+            tag-suffix: "-metal-darwin-arm64-kitten-tts"
+            build-type: "mps"
+          - backend: "piper"
+            tag-suffix: "-metal-darwin-arm64-piper"
+            build-type: "metal"
+            lang: "go"
+          - backend: "silero-vad"
+            tag-suffix: "-metal-darwin-arm64-silero-vad"
+            build-type: "metal"
+            lang: "go"
+          - backend: "local-store"
+            tag-suffix: "-metal-darwin-arm64-local-store"
+            build-type: "metal"
+            lang: "go"
+          - backend: "huggingface"
+            tag-suffix: "-metal-darwin-arm64-huggingface"
+            build-type: "metal"
+            lang: "go"
     with:
       backend: ${{ matrix.backend }}
       build-type: ${{ matrix.build-type }}

.github/workflows/deploy-explorer.yaml

@@ -33,7 +33,7 @@ jobs:
         run: |
           CGO_ENABLED=0 make build
       - name: rm
-        uses: appleboy/ssh-action@v1.2.4
+        uses: appleboy/ssh-action@v1.2.5
         with:
             host: ${{ secrets.EXPLORER_SSH_HOST }}
             username: ${{ secrets.EXPLORER_SSH_USERNAME }}
@@ -53,7 +53,7 @@ jobs:
             rm: true
             target: ./local-ai
       - name: restarting
-        uses: appleboy/ssh-action@v1.2.4
+        uses: appleboy/ssh-action@v1.2.5
         with:
             host: ${{ secrets.EXPLORER_SSH_HOST }}
             username: ${{ secrets.EXPLORER_SSH_USERNAME }}

.github/workflows/image-pr.yml

@@ -37,7 +37,7 @@
           include:
             - build-type: 'cublas'
               cuda-major-version: "12"
-              cuda-minor-version: "9"
+              cuda-minor-version: "8"
               platforms: 'linux/amd64'
               tag-latest: 'false'
               tag-suffix: '-gpu-nvidia-cuda-12'

.github/workflows/image.yml

@@ -88,7 +88,7 @@
               ubuntu-codename: 'noble'
             - build-type: 'cublas'
               cuda-major-version: "12"
-              cuda-minor-version: "9"
+              cuda-minor-version: "8"
               platforms: 'linux/amd64'
               tag-latest: 'auto'
               tag-suffix: '-gpu-nvidia-cuda-12'

.github/workflows/test-extra.yml

@@ -238,7 +238,7 @@ jobs:
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ffmpeg
+          sudo apt-get install -y 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
@@ -257,7 +257,7 @@ jobs:
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ffmpeg
+          sudo apt-get install -y build-essential ffmpeg
           sudo apt-get install -y ca-certificates cmake curl patch python3-pip
           # Install UV
           curl -LsSf https://astral.sh/uv/install.sh | sh
@@ -276,7 +276,7 @@ jobs:
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ffmpeg
+          sudo apt-get install -y build-essential ffmpeg
           sudo apt-get install -y ca-certificates cmake curl patch python3-pip
           # Install UV
           curl -LsSf https://astral.sh/uv/install.sh | sh
@@ -295,7 +295,7 @@ jobs:
       - name: Dependencies
         run: |
           sudo apt-get update
-          sudo apt-get install build-essential ffmpeg
+          sudo apt-get install -y build-essential ffmpeg
           sudo apt-get install -y ca-certificates cmake curl patch python3-pip
           # Install UV
           curl -LsSf https://astral.sh/uv/install.sh | sh
@@ -303,4 +303,42 @@ jobs:
       - name: Test qwen-tts
         run: |
           make --jobs=5 --output-sync=target -C backend/python/qwen-tts
-          make --jobs=5 --output-sync=target -C backend/python/qwen-tts test
+          make --jobs=5 --output-sync=target -C backend/python/qwen-tts test
+  tests-qwen-asr:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        with:
+          submodules: true
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential ffmpeg sox
+          sudo apt-get install -y ca-certificates cmake curl patch python3-pip
+          # Install UV
+          curl -LsSf https://astral.sh/uv/install.sh | sh
+          pip install --user --no-cache-dir grpcio-tools==1.64.1
+      - name: Test qwen-asr
+        run: |
+          make --jobs=5 --output-sync=target -C backend/python/qwen-asr
+          make --jobs=5 --output-sync=target -C backend/python/qwen-asr test
+  tests-voxcpm:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        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 python3-pip
+          # Install UV
+          curl -LsSf https://astral.sh/uv/install.sh | sh
+          pip install --user --no-cache-dir grpcio-tools==1.64.1
+      - name: Test voxcpm
+        run: |
+          make --jobs=5 --output-sync=target -C backend/python/voxcpm
+          make --jobs=5 --output-sync=target -C backend/python/voxcpm test

.github/workflows/tests-e2e.yml

@@ -0,0 +1,56 @@
+---
+name: 'E2E Backend Tests'
+
+on:
+  pull_request:
+  push:
+    branches:
+      - master
+    tags:
+      - '*'
+
+concurrency:
+  group: ci-tests-e2e-backend-${{ github.head_ref || github.ref }}-${{ github.repository }}
+  cancel-in-progress: true
+
+jobs:
+  tests-e2e-backend:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        go-version: ['1.25.x']
+    steps:
+      - name: Clone
+        uses: actions/checkout@v6
+        with:
+          submodules: true
+      - name: Setup Go ${{ matrix.go-version }}
+        uses: actions/setup-go@v5
+        with:
+          go-version: ${{ matrix.go-version }}
+          cache: false
+      - name: Display Go version
+        run: go version
+      - name: Proto Dependencies
+        run: |
+          # Install protoc
+          curl -L -s https://github.com/protocolbuffers/protobuf/releases/download/v26.1/protoc-26.1-linux-x86_64.zip -o protoc.zip && \
+          unzip -j -d /usr/local/bin protoc.zip bin/protoc && \
+          rm protoc.zip
+          go install google.golang.org/protobuf/cmd/protoc-gen-go@v1.34.2
+          go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@1958fcbe2ca8bd93af633f11e97d44e567e945af
+          PATH="$PATH:$HOME/go/bin" make protogen-go
+      - name: Dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y build-essential
+      - name: Test Backend E2E
+        run: |
+          PATH="$PATH:$HOME/go/bin" make build-mock-backend test-e2e
+      - name: Setup tmate session if tests fail
+        if: ${{ failure() }}
+        uses: mxschmitt/action-tmate@v3.23
+        with:
+          detached: true
+          connect-timeout-seconds: 180
+          limit-access-to-actor: true

.gitignore

@@ -36,6 +36,8 @@ LocalAI
 models/*
 test-models/
 test-dir/
+tests/e2e-aio/backends
+tests/e2e-aio/models
 
 release/
 

AGENTS.md

@@ -4,13 +4,13 @@ Building and testing the project depends on the components involved and the plat
 
 ## Building a specified backend
 
-Let's say the user wants to build a particular backend for a given platform. For example let's say they want to build bark for ROCM/hipblas
+Let's say the user wants to build a particular backend for a given platform. For example let's say they want to build coqui for ROCM/hipblas
 
-- The Makefile has targets like `docker-build-bark` created with `generate-docker-build-target` at the time of writing. Recently added backends may require a new target.
+- The Makefile has targets like `docker-build-coqui` created with `generate-docker-build-target` at the time of writing. Recently added backends may require a new target.
 - At a minimum we need to set the BUILD_TYPE, BASE_IMAGE build-args
   - Use .github/workflows/backend.yml as a reference it lists the needed args in the `include` job strategy matrix
   - l4t and cublas also requires the CUDA major and minor version
-- You can pretty print a command like `DOCKER_MAKEFLAGS=-j$(nproc --ignore=1) BUILD_TYPE=hipblas BASE_IMAGE=rocm/dev-ubuntu-24.04:6.4.4 make docker-build-bark`
+- You can pretty print a command like `DOCKER_MAKEFLAGS=-j$(nproc --ignore=1) BUILD_TYPE=hipblas BASE_IMAGE=rocm/dev-ubuntu-24.04:6.4.4 make docker-build-coqui`
 - Unless the user specifies that they want you to run the command, then just print it because not all agent frontends handle long running jobs well and the output may overflow your context
 - The user may say they want to build AMD or ROCM instead of hipblas, or Intel instead of SYCL or NVIDIA insted of l4t or cublas. Ask for confirmation if there is ambiguity.
 - Sometimes the user may need extra parameters to be added to `docker build` (e.g. `--platform` for cross-platform builds or `--progress` to view the full logs), in which case you can generate the `docker build` command directly.
@@ -95,7 +95,7 @@ test-extra: prepare-test-extra
 
 Add a backend definition variable in the backend definitions section (around line 428-457). The format depends on the backend type:
 
-**For Python backends with root context** (like `faster-whisper`, `bark`):
+**For Python backends with root context** (like `faster-whisper`, `coqui`):
 ```makefile
 BACKEND_<BACKEND_NAME> = <backend-name>|python|.|false|true
 ```
@@ -280,3 +280,11 @@ Always check `llama.cpp` for new model configuration options that should be supp
 - `llama.cpp/common/chat-parser.cpp` - Format presets and model-specific handlers
 - `llama.cpp/common/chat.h` - Format enums and parameter structures
 - `llama.cpp/tools/server/server-context.cpp` - Server configuration options
+
+# Documentation
+
+The project documentation is located in `docs/content`. When adding new features or changing existing functionality, it is crucial to update the documentation to reflect these changes. This helps users understand how to use the new capabilities and ensures the documentation stays relevant.
+
+- **Feature Documentation**: If you add a new feature (like a new backend or API endpoint), create a new markdown file in `docs/content/features/` explaining what it is, how to configure it, and how to use it.
+- **Configuration**: If you modify configuration options, update the relevant sections in `docs/content/`.
+- **Examples**: providing concrete examples (like YAML configuration blocks) is highly encouraged to help users get started quickly.

Makefile

@@ -1,5 +1,5 @@
 # Disable parallel execution for backend builds
-.NOTPARALLEL: backends/diffusers backends/llama-cpp backends/piper backends/stablediffusion-ggml backends/whisper backends/faster-whisper backends/silero-vad backends/local-store backends/huggingface backends/rfdetr backends/kitten-tts backends/kokoro backends/chatterbox backends/llama-cpp-darwin backends/neutts build-darwin-python-backend build-darwin-go-backend backends/mlx backends/diffuser-darwin backends/mlx-vlm backends/mlx-audio backends/stablediffusion-ggml-darwin backends/vllm backends/moonshine backends/pocket-tts backends/qwen-tts
+.NOTPARALLEL: backends/diffusers backends/llama-cpp backends/outetts backends/piper backends/stablediffusion-ggml backends/whisper backends/faster-whisper backends/silero-vad backends/local-store backends/huggingface backends/rfdetr backends/kitten-tts backends/kokoro backends/chatterbox backends/llama-cpp-darwin backends/neutts build-darwin-python-backend build-darwin-go-backend backends/mlx backends/diffuser-darwin backends/mlx-vlm backends/mlx-audio backends/stablediffusion-ggml-darwin backends/vllm backends/vllm-omni backends/moonshine backends/pocket-tts backends/qwen-tts backends/qwen-asr backends/voxcpm backends/whisperx
 
 GOCMD=go
 GOTEST=$(GOCMD) test
@@ -7,16 +7,14 @@ GOVET=$(GOCMD) vet
 BINARY_NAME=local-ai
 LAUNCHER_BINARY_NAME=local-ai-launcher
 
-CUDA_MAJOR_VERSION?=13
-CUDA_MINOR_VERSION?=0
 UBUNTU_VERSION?=2404
 UBUNTU_CODENAME?=noble
 
 GORELEASER?=
 
 export BUILD_TYPE?=
-export CUDA_MAJOR_VERSION?=12
-export CUDA_MINOR_VERSION?=9
+export CUDA_MAJOR_VERSION?=13
+export CUDA_MINOR_VERSION?=0
 
 GO_TAGS?=
 BUILD_ID?=
@@ -191,9 +189,6 @@ run-e2e-aio: protogen-go
 ########################################################
 
 prepare-e2e:
-	mkdir -p $(TEST_DIR)
-	cp -rfv $(abspath ./tests/e2e-fixtures)/gpu.yaml $(TEST_DIR)/gpu.yaml
-	test -e $(TEST_DIR)/ggllm-test-model.bin || wget -q https://huggingface.co/TheBloke/CodeLlama-7B-Instruct-GGUF/resolve/main/codellama-7b-instruct.Q2_K.gguf -O $(TEST_DIR)/ggllm-test-model.bin
 	docker build \
 		--build-arg IMAGE_TYPE=core \
 		--build-arg BUILD_TYPE=$(BUILD_TYPE) \
@@ -207,14 +202,16 @@ prepare-e2e:
 		-t localai-tests .
 
 run-e2e-image:
-	ls -liah $(abspath ./tests/e2e-fixtures)
-	docker run -p 5390:8080 -e MODELS_PATH=/models -e THREADS=1 -e DEBUG=true -d --rm -v $(TEST_DIR):/models --gpus all --name e2e-tests-$(RANDOM) localai-tests
+	docker run -p 5390:8080 -e MODELS_PATH=/models -e THREADS=1 -e DEBUG=true -d --rm -v $(TEST_DIR):/models --name e2e-tests-$(RANDOM) localai-tests
 
-test-e2e:
+test-e2e: build-mock-backend prepare-e2e run-e2e-image
 	@echo 'Running e2e tests'
 	BUILD_TYPE=$(BUILD_TYPE) \
-	LOCALAI_API=http://$(E2E_BRIDGE_IP):5390/v1 \
+	LOCALAI_API=http://$(E2E_BRIDGE_IP):5390 \
 	$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --flake-attempts $(TEST_FLAKES) -v -r ./tests/e2e
+	$(MAKE) clean-mock-backend
+	$(MAKE) teardown-e2e
+	docker rmi localai-tests
 
 teardown-e2e:
 	rm -rf $(TEST_DIR) || true
@@ -311,23 +308,33 @@ protogen-go-clean:
 
 prepare-test-extra: protogen-python
 	$(MAKE) -C backend/python/transformers
+	$(MAKE) -C backend/python/outetts
 	$(MAKE) -C backend/python/diffusers
 	$(MAKE) -C backend/python/chatterbox
 	$(MAKE) -C backend/python/vllm
+	$(MAKE) -C backend/python/vllm-omni
 	$(MAKE) -C backend/python/vibevoice
 	$(MAKE) -C backend/python/moonshine
 	$(MAKE) -C backend/python/pocket-tts
 	$(MAKE) -C backend/python/qwen-tts
+	$(MAKE) -C backend/python/qwen-asr
+	$(MAKE) -C backend/python/voxcpm
+	$(MAKE) -C backend/python/whisperx
 
 test-extra: prepare-test-extra
 	$(MAKE) -C backend/python/transformers test
+	$(MAKE) -C backend/python/outetts test
 	$(MAKE) -C backend/python/diffusers test
 	$(MAKE) -C backend/python/chatterbox test
 	$(MAKE) -C backend/python/vllm test
+	$(MAKE) -C backend/python/vllm-omni test
 	$(MAKE) -C backend/python/vibevoice test
 	$(MAKE) -C backend/python/moonshine test
 	$(MAKE) -C backend/python/pocket-tts test
 	$(MAKE) -C backend/python/qwen-tts test
+	$(MAKE) -C backend/python/qwen-asr test
+	$(MAKE) -C backend/python/voxcpm test
+	$(MAKE) -C backend/python/whisperx test
 
 DOCKER_IMAGE?=local-ai
 DOCKER_AIO_IMAGE?=local-ai-aio
@@ -436,7 +443,6 @@ backend-images:
 BACKEND_LLAMA_CPP = llama-cpp|llama-cpp|.|false|false
 
 # Golang backends
-BACKEND_BARK_CPP = bark-cpp|golang|.|false|true
 BACKEND_PIPER = piper|golang|.|false|true
 BACKEND_LOCAL_STORE = local-store|golang|.|false|true
 BACKEND_HUGGINGFACE = huggingface|golang|.|false|true
@@ -447,21 +453,24 @@ BACKEND_WHISPER = whisper|golang|.|false|true
 # Python backends with root context
 BACKEND_RERANKERS = rerankers|python|.|false|true
 BACKEND_TRANSFORMERS = transformers|python|.|false|true
+BACKEND_OUTETTS = outetts|python|.|false|true
 BACKEND_FASTER_WHISPER = faster-whisper|python|.|false|true
 BACKEND_COQUI = coqui|python|.|false|true
-BACKEND_BARK = bark|python|.|false|true
-BACKEND_EXLLAMA2 = exllama2|python|.|false|true
 BACKEND_RFDETR = rfdetr|python|.|false|true
 BACKEND_KITTEN_TTS = kitten-tts|python|.|false|true
 BACKEND_NEUTTS = neutts|python|.|false|true
 BACKEND_KOKORO = kokoro|python|.|false|true
 BACKEND_VLLM = vllm|python|.|false|true
+BACKEND_VLLM_OMNI = vllm-omni|python|.|false|true
 BACKEND_DIFFUSERS = diffusers|python|.|--progress=plain|true
 BACKEND_CHATTERBOX = chatterbox|python|.|false|true
 BACKEND_VIBEVOICE = vibevoice|python|.|--progress=plain|true
 BACKEND_MOONSHINE = moonshine|python|.|false|true
 BACKEND_POCKET_TTS = pocket-tts|python|.|false|true
 BACKEND_QWEN_TTS = qwen-tts|python|.|false|true
+BACKEND_QWEN_ASR = qwen-asr|python|.|false|true
+BACKEND_VOXCPM = voxcpm|python|.|false|true
+BACKEND_WHISPERX = whisperx|python|.|false|true
 
 # Helper function to build docker image for a backend
 # Usage: $(call docker-build-backend,BACKEND_NAME,DOCKERFILE_TYPE,BUILD_CONTEXT,PROGRESS_FLAG,NEEDS_BACKEND_ARG)
@@ -485,7 +494,6 @@ endef
 
 # Generate all docker-build targets
 $(eval $(call generate-docker-build-target,$(BACKEND_LLAMA_CPP)))
-$(eval $(call generate-docker-build-target,$(BACKEND_BARK_CPP)))
 $(eval $(call generate-docker-build-target,$(BACKEND_PIPER)))
 $(eval $(call generate-docker-build-target,$(BACKEND_LOCAL_STORE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_HUGGINGFACE)))
@@ -494,27 +502,40 @@ $(eval $(call generate-docker-build-target,$(BACKEND_STABLEDIFFUSION_GGML)))
 $(eval $(call generate-docker-build-target,$(BACKEND_WHISPER)))
 $(eval $(call generate-docker-build-target,$(BACKEND_RERANKERS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_TRANSFORMERS)))
+$(eval $(call generate-docker-build-target,$(BACKEND_OUTETTS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_FASTER_WHISPER)))
 $(eval $(call generate-docker-build-target,$(BACKEND_COQUI)))
-$(eval $(call generate-docker-build-target,$(BACKEND_BARK)))
-$(eval $(call generate-docker-build-target,$(BACKEND_EXLLAMA2)))
 $(eval $(call generate-docker-build-target,$(BACKEND_RFDETR)))
 $(eval $(call generate-docker-build-target,$(BACKEND_KITTEN_TTS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_NEUTTS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_KOKORO)))
 $(eval $(call generate-docker-build-target,$(BACKEND_VLLM)))
+$(eval $(call generate-docker-build-target,$(BACKEND_VLLM_OMNI)))
 $(eval $(call generate-docker-build-target,$(BACKEND_DIFFUSERS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_CHATTERBOX)))
 $(eval $(call generate-docker-build-target,$(BACKEND_VIBEVOICE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_MOONSHINE)))
 $(eval $(call generate-docker-build-target,$(BACKEND_POCKET_TTS)))
 $(eval $(call generate-docker-build-target,$(BACKEND_QWEN_TTS)))
+$(eval $(call generate-docker-build-target,$(BACKEND_QWEN_ASR)))
+$(eval $(call generate-docker-build-target,$(BACKEND_VOXCPM)))
+$(eval $(call generate-docker-build-target,$(BACKEND_WHISPERX)))
 
 # Pattern rule for docker-save targets
 docker-save-%: backend-images
 	docker save local-ai-backend:$* -o backend-images/$*.tar
 
-docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-transformers docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-bark docker-build-chatterbox docker-build-vibevoice docker-build-exllama2 docker-build-moonshine docker-build-pocket-tts docker-build-qwen-tts
+docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-vllm-omni docker-build-transformers docker-build-outetts docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-chatterbox docker-build-vibevoice docker-build-moonshine docker-build-pocket-tts docker-build-qwen-tts docker-build-qwen-asr docker-build-voxcpm docker-build-whisperx
+
+########################################################
+### Mock Backend for E2E Tests
+########################################################
+
+build-mock-backend: protogen-go
+	$(GOCMD) build -o tests/e2e/mock-backend/mock-backend ./tests/e2e/mock-backend
+
+clean-mock-backend:
+	rm -f tests/e2e/mock-backend/mock-backend
 
 ########################################################
 ### END Backends

README.md

@@ -51,34 +51,16 @@
 **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).
 
 
-## 📚🆕 Local Stack Family
+## Local Stack Family
 
-🆕 LocalAI is now part of a comprehensive suite of AI tools designed to work together:
+Liking LocalAI? LocalAI is part of an integrated suite of AI infrastructure tools, you might also like:
+
+- **[LocalAGI](https://github.com/mudler/LocalAGI)** - AI agent orchestration platform with OpenAI Responses API compatibility and advanced agentic capabilities
+- **[LocalRecall](https://github.com/mudler/LocalRecall)** - MCP/REST API knowledge base system providing persistent memory and storage for AI agents
+- 🆕 **[Cogito](https://github.com/mudler/cogito)** - Go library for building intelligent, co-operative agentic software and LLM-powered workflows, focusing on improving results for small, open source language models that scales to any LLM. Powers LocalAGI and LocalAI MCP/Agentic capabilities
+- 🆕 **[Wiz](https://github.com/mudler/wiz)** - Terminal-based AI agent accessible via Ctrl+Space keybinding. Portable, local-LLM friendly shell assistant with TUI/CLI modes, tool execution with approval, MCP protocol support, and multi-shell compatibility (zsh, bash, fish)
+- 🆕 **[SkillServer](https://github.com/mudler/skillserver)** - Simple, centralized skills database for AI agents via MCP. Manages skills as Markdown files with MCP server integration, web UI for editing, Git synchronization, and full-text search capabilities
 
-<table>
-  <tr>
-    <td width="50%" valign="top">
-      <a href="https://github.com/mudler/LocalAGI">
-        <img src="https://raw.githubusercontent.com/mudler/LocalAGI/refs/heads/main/webui/react-ui/public/logo_2.png" width="300" alt="LocalAGI Logo">
-      </a>
-    </td>
-    <td width="50%" valign="top">
-      <h3><a href="https://github.com/mudler/LocalAGI">LocalAGI</a></h3>
-      <p>A powerful Local AI agent management platform that serves as a drop-in replacement for OpenAI's Responses API, enhanced with advanced agentic capabilities.</p>
-    </td>
-  </tr>
-  <tr>
-    <td width="50%" valign="top">
-      <a href="https://github.com/mudler/LocalRecall">
-        <img src="https://raw.githubusercontent.com/mudler/LocalRecall/refs/heads/main/static/localrecall_horizontal.png" width="300" alt="LocalRecall Logo">
-      </a>
-    </td>
-    <td width="50%" valign="top">
-      <h3><a href="https://github.com/mudler/LocalRecall">LocalRecall</a></h3>
-      <p>A REST-ful API and knowledge base management system that provides persistent memory and storage capabilities for AI agents.</p>
-    </td>
-  </tr>
-</table>
 
 ## Screenshots / Video
 
@@ -257,6 +239,7 @@ Roadmap items: [List of issues](https://github.com/mudler/LocalAI/issues?q=is%3A
 - 🔈 [Audio to Text](https://localai.io/features/audio-to-text/) (Audio transcription with `whisper.cpp`)
 - 🎨 [Image generation](https://localai.io/features/image-generation)
 - 🔥 [OpenAI-alike tools API](https://localai.io/features/openai-functions/) 
+- ⚡ [Realtime API](https://localai.io/features/openai-realtime/) (Speech-to-speech) 
 - 🧠 [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/)
@@ -278,7 +261,6 @@ LocalAI supports a comprehensive range of AI backends with multiple acceleration
 | **llama.cpp** | LLM inference in C/C++ | CUDA 12/13, ROCm, Intel SYCL, Vulkan, Metal, CPU |
 | **vLLM** | Fast LLM inference with PagedAttention | CUDA 12/13, ROCm, Intel |
 | **transformers** | HuggingFace transformers framework | CUDA 12/13, ROCm, Intel, CPU |
-| **exllama2** | GPTQ inference library | CUDA 12/13 |
 | **MLX** | Apple Silicon LLM inference | Metal (M1/M2/M3+) |
 | **MLX-VLM** | Apple Silicon Vision-Language Models | Metal (M1/M2/M3+) |
 
@@ -287,8 +269,6 @@ LocalAI supports a comprehensive range of AI backends with multiple acceleration
 |---------|-------------|---------------------|
 | **whisper.cpp** | OpenAI Whisper in C/C++ | CUDA 12/13, ROCm, Intel SYCL, Vulkan, CPU |
 | **faster-whisper** | Fast Whisper with CTranslate2 | CUDA 12/13, ROCm, Intel, CPU |
-| **bark** | Text-to-audio generation | CUDA 12/13, ROCm, Intel |
-| **bark-cpp** | C++ implementation of Bark | CUDA, Metal, CPU |
 | **coqui** | Advanced TTS with 1100+ languages | CUDA 12/13, ROCm, Intel, CPU |
 | **kokoro** | Lightweight TTS model | CUDA 12/13, ROCm, Intel, CPU |
 | **chatterbox** | Production-grade TTS | CUDA 12/13, CPU |
@@ -320,9 +300,9 @@ LocalAI supports a comprehensive range of AI backends with multiple acceleration
 |-------------------|-------------------|------------------|
 | **NVIDIA CUDA 12** | All CUDA-compatible backends | Nvidia hardware |
 | **NVIDIA CUDA 13** | All CUDA-compatible backends | Nvidia hardware |
-| **AMD ROCm** | llama.cpp, whisper, vllm, transformers, diffusers, rerankers, coqui, kokoro, bark, neutts, vibevoice, pocket-tts, qwen-tts | AMD Graphics |
-| **Intel oneAPI** | llama.cpp, whisper, stablediffusion, vllm, transformers, diffusers, rfdetr, rerankers, exllama2, coqui, kokoro, bark, vibevoice, pocket-tts, qwen-tts | Intel Arc, Intel iGPUs |
-| **Apple Metal** | llama.cpp, whisper, diffusers, MLX, MLX-VLM, bark-cpp | Apple M1/M2/M3+ |
+| **AMD ROCm** | llama.cpp, whisper, vllm, transformers, diffusers, rerankers, coqui, kokoro, neutts, vibevoice, pocket-tts, qwen-tts | AMD Graphics |
+| **Intel oneAPI** | llama.cpp, whisper, stablediffusion, vllm, transformers, diffusers, rfdetr, rerankers, coqui, kokoro, vibevoice, pocket-tts, qwen-tts | Intel Arc, Intel iGPUs |
+| **Apple Metal** | llama.cpp, whisper, diffusers, MLX, MLX-VLM | Apple M1/M2/M3+ |
 | **Vulkan** | llama.cpp, whisper, stablediffusion | Cross-platform GPUs |
 | **NVIDIA Jetson (CUDA 12)** | llama.cpp, whisper, stablediffusion, diffusers, rfdetr | ARM64 embedded AI (AGX Orin, etc.) |
 | **NVIDIA Jetson (CUDA 13)** | llama.cpp, whisper, stablediffusion, diffusers, rfdetr | ARM64 embedded AI (DGX Spark) |
@@ -344,6 +324,10 @@ Agentic Libraries:
 MCPs:
 - https://github.com/mudler/MCPs
 
+OS Assistant:
+
+- https://github.com/mudler/Keygeist - Keygeist is an AI-powered keyboard operator that listens for key combinations and responds with AI-generated text typed directly into your Linux box.
+
 Model galleries
 - https://github.com/go-skynet/model-gallery
 

backend/README.md

@@ -46,7 +46,7 @@ The backend system provides language-specific Dockerfiles that handle the build
 - **vllm**: High-performance LLM inference
 - **mlx**: Apple Silicon optimization
 - **diffusers**: Stable Diffusion models
-- **Audio**: bark, coqui, faster-whisper, kitten-tts
+- **Audio**: coqui, faster-whisper, kitten-tts
 - **Vision**: mlx-vlm, rfdetr
 - **Specialized**: rerankers, chatterbox, kokoro
 
@@ -55,7 +55,6 @@ The backend system provides language-specific Dockerfiles that handle the build
 - **stablediffusion-ggml**: Stable Diffusion in Go with GGML Cpp backend
 - **huggingface**: Hugging Face model integration
 - **piper**: Text-to-speech synthesis Golang with C bindings using rhaspy/piper
-- **bark-cpp**: Bark TTS models Golang with Cpp bindings
 - **local-store**: Vector storage backend
 
 #### C++ Backends (`cpp/`)

backend/backend.proto

@@ -17,6 +17,7 @@ service Backend {
   rpc GenerateVideo(GenerateVideoRequest) returns (Result) {}
   rpc AudioTranscription(TranscriptRequest) returns (TranscriptResult) {}
   rpc TTS(TTSRequest) returns (Result) {}
+  rpc TTSStream(TTSRequest) returns (stream Reply) {}
   rpc SoundGeneration(SoundGenerationRequest) returns (Result) {}
   rpc TokenizeString(PredictOptions) returns (TokenizationResponse) {}
   rpc Status(HealthMessage) returns (StatusResponse) {}
@@ -298,6 +299,7 @@ message TranscriptSegment {
   int64 end = 3;
   string text = 4;
   repeated int32 tokens = 5;
+  string speaker = 6;
 }
 
 message GenerateImageRequest {

backend/cpp/llama-cpp/Makefile

@@ -1,5 +1,5 @@
 
-LLAMA_VERSION?=a5eaa1d6a3732bc0f460b02b61c95680bba5a012
+LLAMA_VERSION?=2634ed207a17db1a54bd8df0555bd8499a6ab691
 LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
 
 CMAKE_ARGS?=

backend/cpp/llama-cpp/grpc-server.cpp

@@ -778,8 +778,8 @@ public:
             if (!params.mmproj.path.empty()) {
                 error_msg += " (with mmproj: " + params.mmproj.path + ")";
             }
-            if (params.has_speculative() && !params.speculative.model.path.empty()) {
-                error_msg += " (with draft model: " + params.speculative.model.path + ")";
+            if (params.speculative.has_dft() && !params.speculative.mparams_dft.path.empty()) {
+                error_msg += " (with draft model: " + params.speculative.mparams_dft.path + ")";
             }
             
             // Add captured error details if available

backend/go/bark-cpp/Makefile

@@ -1,51 +0,0 @@
-INCLUDE_PATH := $(abspath ./)
-LIBRARY_PATH := $(abspath ./)
-
-AR?=ar
-
-CMAKE_ARGS?=-DGGML_NATIVE=OFF
-BUILD_TYPE?=
-GOCMD=go
-# keep standard at C11 and C++11
-CXXFLAGS = -I. -I$(INCLUDE_PATH)/sources/bark.cpp/examples -I$(INCLUDE_PATH)/sources/bark.cpp/encodec.cpp/ggml/include -I$(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
-
-# bark.cpp
-BARKCPP_REPO?=https://github.com/PABannier/bark.cpp.git
-BARKCPP_VERSION?=5d5be84f089ab9ea53b7a793f088d3fbf7247495
-
-# warnings
-CXXFLAGS += -Wall -Wextra -Wpedantic -Wcast-qual -Wno-unused-function
-
-## bark.cpp
-sources/bark.cpp:
-	git clone --recursive $(BARKCPP_REPO) sources/bark.cpp && \
-	cd sources/bark.cpp && \
-	git checkout $(BARKCPP_VERSION) && \
-	git submodule update --init --recursive --depth 1 --single-branch
-
-sources/bark.cpp/build/libbark.a: sources/bark.cpp
-	cd sources/bark.cpp && \
-	mkdir -p build && \
-	cd build && \
-	cmake $(CMAKE_ARGS) .. && \
-	cmake --build . --config Release
-
-gobark.o:
-	$(CXX) $(CXXFLAGS) gobark.cpp -o gobark.o -c $(LDFLAGS)
-
-libbark.a: sources/bark.cpp/build/libbark.a gobark.o
-	cp $(INCLUDE_PATH)/sources/bark.cpp/build/libbark.a ./
-	$(AR) rcs libbark.a gobark.o
-
-bark-cpp: libbark.a
-	CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH="$(CURDIR)" LIBRARY_PATH=$(CURDIR) \
-	$(GOCMD) build -v -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o bark-cpp ./
-
-package:
-	bash package.sh
-
-build: bark-cpp package
-
-clean:
-	rm -f gobark.o libbark.a

backend/go/bark-cpp/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 occurred. 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-cpp/gobark.go

@@ -1,52 +0,0 @@
-package main
-
-// #cgo CXXFLAGS: -I${SRCDIR}/sources/bark.cpp/ -I${SRCDIR}/sources/bark.cpp/encodec.cpp -I${SRCDIR}/sources/bark.cpp/encodec.cpp/ggml/include -I${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/ggml/src/ -L${SRCDIR}/sources/bark.cpp/build/encodec.cpp/ -lbark -lencodec -lcommon -lggml -lgomp
-// #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-cpp/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/bark-cpp/main.go

@@ -1,20 +0,0 @@
-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"
-)
-
-var (
-	addr = flag.String("addr", "localhost:50051", "the address to connect to")
-)
-
-func main() {
-	flag.Parse()
-
-	if err := grpc.StartServer(*addr, &Bark{}); err != nil {
-		panic(err)
-	}
-}

backend/go/bark-cpp/package.sh

@@ -1,41 +0,0 @@
-#!/bin/bash
-
-# Script to copy the appropriate libraries based on architecture
-# This script is used in the final stage of the Dockerfile
-
-set -e
-
-CURDIR=$(dirname "$(realpath $0)")
-
-# Create lib directory
-mkdir -p $CURDIR/package/lib
-cp -avrf $CURDIR/bark-cpp $CURDIR/package/
-cp -rfv $CURDIR/run.sh $CURDIR/package/
-
-# Detect architecture and copy appropriate libraries
-if [ -f "/lib64/ld-linux-x86-64.so.2" ]; then
-    # x86_64 architecture
-    echo "Detected x86_64 architecture, copying x86_64 libraries..."
-    cp -arfLv /lib64/ld-linux-x86-64.so.2 $CURDIR/package/lib/ld.so
-    cp -arfLv /lib/x86_64-linux-gnu/libc.so.6 $CURDIR/package/lib/libc.so.6
-    cp -arfLv /lib/x86_64-linux-gnu/libgcc_s.so.1 $CURDIR/package/lib/libgcc_s.so.1
-    cp -arfLv /lib/x86_64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
-    cp -arfLv /lib/x86_64-linux-gnu/libm.so.6 $CURDIR/package/lib/libm.so.6
-    cp -arfLv /lib/x86_64-linux-gnu/libgomp.so.1 $CURDIR/package/lib/libgomp.so.1
-elif [ -f "/lib/ld-linux-aarch64.so.1" ]; then
-    # ARM64 architecture
-    echo "Detected ARM64 architecture, copying ARM64 libraries..."
-    cp -arfLv /lib/ld-linux-aarch64.so.1 $CURDIR/package/lib/ld.so
-    cp -arfLv /lib/aarch64-linux-gnu/libc.so.6 $CURDIR/package/lib/libc.so.6
-    cp -arfLv /lib/aarch64-linux-gnu/libgcc_s.so.1 $CURDIR/package/lib/libgcc_s.so.1
-    cp -arfLv /lib/aarch64-linux-gnu/libstdc++.so.6 $CURDIR/package/lib/libstdc++.so.6
-    cp -arfLv /lib/aarch64-linux-gnu/libm.so.6 $CURDIR/package/lib/libm.so.6
-    cp -arfLv /lib/aarch64-linux-gnu/libgomp.so.1 $CURDIR/package/lib/libgomp.so.1
-else
-    echo "Error: Could not detect architecture"
-    exit 1
-fi
-
-echo "Packaging completed successfully" 
-ls -liah $CURDIR/package/
-ls -liah $CURDIR/package/lib/

backend/go/bark-cpp/run.sh

@@ -1,13 +0,0 @@
-#!/bin/bash
-set -ex
-
-CURDIR=$(dirname "$(realpath $0)")
-export LD_LIBRARY_PATH=$CURDIR/lib:$LD_LIBRARY_PATH
-
-# If there is a lib/ld.so, use it
-if [ -f $CURDIR/lib/ld.so ]; then
-	echo "Using lib/ld.so"
-	exec $CURDIR/lib/ld.so $CURDIR/bark-cpp "$@"
-fi
-
-exec $CURDIR/bark-cpp "$@"

backend/go/stablediffusion-ggml/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # stablediffusion.cpp (ggml)
 STABLEDIFFUSION_GGML_REPO?=https://github.com/leejet/stable-diffusion.cpp
-STABLEDIFFUSION_GGML_VERSION?=5e4579c11d0678f9765463582d024e58270faa9c
+STABLEDIFFUSION_GGML_VERSION?=e411520407663e1ddf8ff2e5ed4ff3a116fbbc97
 
 CMAKE_ARGS+=-DGGML_MAX_NAME=128
 

backend/go/whisper/Makefile

@@ -8,7 +8,7 @@ JOBS?=$(shell nproc --ignore=1)
 
 # whisper.cpp version
 WHISPER_REPO?=https://github.com/ggml-org/whisper.cpp
-WHISPER_CPP_VERSION?=7aa8818647303b567c3a21fe4220b2681988e220
+WHISPER_CPP_VERSION?=aa1bc0d1a6dfd70dbb9f60c11df12441e03a9075
 SO_TARGET?=libgowhisper.so
 
 CMAKE_ARGS+=-DBUILD_SHARED_LIBS=OFF

backend/go/whisper/gowhisper.go

@@ -130,8 +130,9 @@ func (w *Whisper) AudioTranscription(opts *pb.TranscriptRequest) (pb.TranscriptR
 	segments := []*pb.TranscriptSegment{}
 	text := ""
 	for i := range int(segsLen) {
-		s := CppGetSegmentStart(i)
-		t := CppGetSegmentEnd(i)
+		// segment start/end conversion factor taken from https://github.com/ggml-org/whisper.cpp/blob/master/examples/cli/cli.cpp#L895
+		s := CppGetSegmentStart(i) * (10000000)
+		t := CppGetSegmentEnd(i) * (10000000)
 		txt := strings.Clone(CppGetSegmentText(i))
 		tokens := make([]int32, CppNTokens(i))
 

backend/python/README.md

@@ -16,10 +16,8 @@ The Python backends use a unified build system based on `libbackend.sh` that pro
 - **transformers** - Hugging Face Transformers framework (PyTorch-based)
 - **vllm** - High-performance LLM inference engine
 - **mlx** - Apple Silicon optimized ML framework
-- **exllama2** - ExLlama2 quantized models
 
 ### Audio & Speech
-- **bark** - Text-to-speech synthesis
 - **coqui** - Coqui TTS models
 - **faster-whisper** - Fast Whisper speech recognition
 - **kitten-tts** - Lightweight TTS

backend/python/bark/README.md

@@ -1,16 +0,0 @@
-# Creating a separate environment for ttsbark project
-
-```
-make ttsbark
-```
-
-# Testing the gRPC server
-
-```
-<The path of your python interpreter> -m unittest test_ttsbark.py
-```
-
-For example
-```
-/opt/conda/envs/bark/bin/python -m unittest extra/grpc/bark/test_ttsbark.py
-``````

backend/python/bark/backend.py

@@ -1,98 +0,0 @@
-#!/usr/bin/env python3
-"""
-This is an extra gRPC server of LocalAI for Bark TTS
-"""
-from concurrent import futures
-import time
-import argparse
-import signal
-import sys
-import os
-from scipy.io.wavfile import write as write_wav
-
-import backend_pb2
-import backend_pb2_grpc
-from bark import SAMPLE_RATE, generate_audio, preload_models
-
-import grpc
-
-
-_ONE_DAY_IN_SECONDS = 60 * 60 * 24
-
-# If MAX_WORKERS are specified in the environment use it, otherwise default to 1
-MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1'))
-
-# Implement the BackendServicer class with the service methods
-class BackendServicer(backend_pb2_grpc.BackendServicer):
-    """
-    BackendServicer is the class that implements the gRPC service
-    """
-    def Health(self, request, context):
-        return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
-    def LoadModel(self, request, context):
-        model_name = request.Model
-        try:
-            print("Preparing models, please wait", file=sys.stderr)
-            # download and load all models
-            preload_models()
-        except Exception as err:
-            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
-        # Implement your logic here for the LoadModel service
-        # Replace this with your desired response
-        return backend_pb2.Result(message="Model loaded successfully", success=True)
-
-    def TTS(self, request, context):
-        model = request.model
-        print(request, file=sys.stderr)
-        try:
-            audio_array = None
-            if model != "":
-                audio_array = generate_audio(request.text, history_prompt=model)
-            else:
-                audio_array = generate_audio(request.text)
-            print("saving to", request.dst, file=sys.stderr)
-            # save audio to disk
-            write_wav(request.dst, SAMPLE_RATE, audio_array)
-            print("saved to", request.dst, file=sys.stderr)
-            print("tts for", file=sys.stderr)
-            print(request, file=sys.stderr)
-        except Exception as err:
-            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
-        return backend_pb2.Result(success=True)
-
-def serve(address):
-    server = grpc.server(futures.ThreadPoolExecutor(max_workers=MAX_WORKERS),
-        options=[
-            ('grpc.max_message_length', 50 * 1024 * 1024),  # 50MB
-            ('grpc.max_send_message_length', 50 * 1024 * 1024),  # 50MB
-            ('grpc.max_receive_message_length', 50 * 1024 * 1024),  # 50MB
-        ])
-    backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server)
-    server.add_insecure_port(address)
-    server.start()
-    print("Server started. Listening on: " + address, file=sys.stderr)
-
-    # Define the signal handler function
-    def signal_handler(sig, frame):
-        print("Received termination signal. Shutting down...")
-        server.stop(0)
-        sys.exit(0)
-
-    # Set the signal handlers for SIGINT and SIGTERM
-    signal.signal(signal.SIGINT, signal_handler)
-    signal.signal(signal.SIGTERM, signal_handler)
-
-    try:
-        while True:
-            time.sleep(_ONE_DAY_IN_SECONDS)
-    except KeyboardInterrupt:
-        server.stop(0)
-
-if __name__ == "__main__":
-    parser = argparse.ArgumentParser(description="Run the gRPC server.")
-    parser.add_argument(
-        "--addr", default="localhost:50051", help="The address to bind the server to."
-    )
-    args = parser.parse_args()
-
-    serve(args.addr)

backend/python/bark/install.sh

@@ -1,19 +0,0 @@
-#!/bin/bash
-set -e
-
-backend_dir=$(dirname $0)
-if [ -d $backend_dir/common ]; then
-    source $backend_dir/common/libbackend.sh
-else
-    source $backend_dir/../common/libbackend.sh
-fi
-
-# This is here because the Intel pip index is broken and returns 200 status codes for every package name, it just doesn't return any package links.
-# This makes uv think that the package exists in the Intel pip index, and by default it stops looking at other pip indexes once it finds a match.
-# We need uv to continue falling through to the pypi default index to find optimum[openvino] in the pypi index
-# the --upgrade actually allows us to *downgrade* torch to the version provided in the Intel pip index
-if [ "x${BUILD_PROFILE}" == "xintel" ]; then
-    EXTRA_PIP_INSTALL_FLAGS+=" --upgrade --index-strategy=unsafe-first-match"
-fi
-
-installRequirements

backend/python/bark/requirements-cpu.txt

@@ -1,4 +0,0 @@
-transformers
-accelerate
-torch==2.4.1
-torchaudio==2.4.1

backend/python/bark/requirements-cublas12.txt

@@ -1,4 +0,0 @@
-torch==2.4.1
-torchaudio==2.4.1
-transformers
-accelerate

backend/python/bark/requirements.txt

@@ -1,4 +0,0 @@
-bark==0.1.5
-grpcio==1.76.0
-protobuf
-certifi

backend/python/chatterbox/requirements-mps.txt

@@ -0,0 +1,7 @@
+torch
+torchaudio
+accelerate
+numpy>=1.24.0,<1.26.0
+transformers
+# https://github.com/mudler/LocalAI/pull/6240#issuecomment-3329518289
+chatterbox-tts@git+https://git@github.com/mudler/chatterbox.git@faster

backend/python/coqui/README.md

@@ -1,4 +1,4 @@
-# Creating a separate environment for ttsbark project
+# Creating a separate environment for coqui project
 
 ```
 make coqui

backend/python/coqui/backend.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
 """
-This is an extra gRPC server of LocalAI for Bark TTS
+This is an extra gRPC server of LocalAI for Coqui TTS
 """
 from concurrent import futures
 import time

backend/python/coqui/requirements-mps.txt

@@ -0,0 +1,4 @@
+torch==2.7.1
+transformers==4.48.3
+accelerate
+coqui-tts

backend/python/exllama2/.gitignore

@@ -1 +0,0 @@
-source

backend/python/exllama2/Makefile

@@ -1,17 +0,0 @@
-.PHONY: exllama2
-exllama2:
-	bash install.sh
-
-.PHONY: run
-run: exllama2
-	@echo "Running exllama2..."
-	bash run.sh
-	@echo "exllama2 run."
-
-.PHONY: protogen-clean
-protogen-clean:
-	$(RM) backend_pb2_grpc.py backend_pb2.py
-
-.PHONY: clean
-clean: protogen-clean
-	$(RM) -r venv source __pycache__

backend/python/exllama2/backend.py

@@ -1,143 +0,0 @@
-#!/usr/bin/env python3
-import grpc
-from concurrent import futures
-import time
-import backend_pb2
-import backend_pb2_grpc
-import argparse
-import signal
-import sys
-import os
-import glob
-
-from pathlib import Path
-import torch
-import torch.nn.functional as F
-from torch import version as torch_version
-
-
-from exllamav2.generator import (
-    ExLlamaV2BaseGenerator,
-    ExLlamaV2Sampler
-)
-
-
-from exllamav2 import (
-    ExLlamaV2,
-    ExLlamaV2Config,
-    ExLlamaV2Cache,
-    ExLlamaV2Cache_8bit,
-    ExLlamaV2Tokenizer,
-    model_init,
-)
-
-
-_ONE_DAY_IN_SECONDS = 60 * 60 * 24
-
-# If MAX_WORKERS are specified in the environment use it, otherwise default to 1
-MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1'))
-
-# Implement the BackendServicer class with the service methods
-class BackendServicer(backend_pb2_grpc.BackendServicer):
-    def Health(self, request, context):
-        return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
-
-    def LoadModel(self, request, context):
-        try:
-            model_directory = request.ModelFile
-
-            config = ExLlamaV2Config()
-            config.model_dir = model_directory
-            config.prepare()
-
-            model = ExLlamaV2(config)
-
-            cache = ExLlamaV2Cache(model, lazy=True)
-            model.load_autosplit(cache)
-
-            tokenizer = ExLlamaV2Tokenizer(config)
-
-            # Initialize generator
-
-            generator = ExLlamaV2BaseGenerator(model, cache, tokenizer)
-
-            self.generator = generator
-
-            generator.warmup()
-            self.model = model
-            self.tokenizer = tokenizer
-            self.cache = cache
-        except Exception as err:
-            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
-        return backend_pb2.Result(message="Model loaded successfully", success=True)
-
-    def Predict(self, request, context):
-
-        penalty = 1.15
-        if request.Penalty != 0.0:
-            penalty = request.Penalty
-
-        settings = ExLlamaV2Sampler.Settings()
-        settings.temperature = request.Temperature
-        settings.top_k = request.TopK
-        settings.top_p = request.TopP
-        settings.token_repetition_penalty = penalty
-        settings.disallow_tokens(self.tokenizer, [self.tokenizer.eos_token_id])
-        tokens = 512
-
-        if request.Tokens != 0:
-            tokens = request.Tokens
-        output = self.generator.generate_simple(
-            request.Prompt, settings, tokens)
-
-        # Remove prompt from response if present
-        if request.Prompt in output:
-            output = output.replace(request.Prompt, "")
-
-        return backend_pb2.Result(message=bytes(output, encoding='utf-8'))
-
-    def PredictStream(self, request, context):
-        # Implement PredictStream RPC
-        # for reply in some_data_generator():
-        #    yield reply
-        # Not implemented yet
-        return self.Predict(request, context)
-
-
-def serve(address):
-    server = grpc.server(futures.ThreadPoolExecutor(max_workers=MAX_WORKERS),
-        options=[
-            ('grpc.max_message_length', 50 * 1024 * 1024),  # 50MB
-            ('grpc.max_send_message_length', 50 * 1024 * 1024),  # 50MB
-            ('grpc.max_receive_message_length', 50 * 1024 * 1024),  # 50MB
-        ])
-    backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server)
-    server.add_insecure_port(address)
-    server.start()
-    print("Server started. Listening on: " + address, file=sys.stderr)
-
-    # Define the signal handler function
-    def signal_handler(sig, frame):
-        print("Received termination signal. Shutting down...")
-        server.stop(0)
-        sys.exit(0)
-
-    # Set the signal handlers for SIGINT and SIGTERM
-    signal.signal(signal.SIGINT, signal_handler)
-    signal.signal(signal.SIGTERM, signal_handler)
-
-    try:
-        while True:
-            time.sleep(_ONE_DAY_IN_SECONDS)
-    except KeyboardInterrupt:
-        server.stop(0)
-
-
-if __name__ == "__main__":
-    parser = argparse.ArgumentParser(description="Run the gRPC server.")
-    parser.add_argument(
-        "--addr", default="localhost:50051", help="The address to bind the server to."
-    )
-    args = parser.parse_args()
-
-    serve(args.addr)

backend/python/exllama2/install.sh

@@ -1,21 +0,0 @@
-#!/bin/bash
-set -e
-
-LIMIT_TARGETS="cublas"
-EXTRA_PIP_INSTALL_FLAGS="--no-build-isolation"
-EXLLAMA2_VERSION=c0ddebaaaf8ffd1b3529c2bb654e650bce2f790f
-
-backend_dir=$(dirname $0)
-if [ -d $backend_dir/common ]; then
-    source $backend_dir/common/libbackend.sh
-else
-    source $backend_dir/../common/libbackend.sh
-fi
-
-installRequirements
-
-git clone https://github.com/turboderp/exllamav2 $MY_DIR/source
-pushd ${MY_DIR}/source && git checkout -b build ${EXLLAMA2_VERSION} && popd
-
-# This installs exllamav2 in JIT mode so it will compile the appropriate torch extension at runtime
-EXLLAMA_NOCOMPILE= uv pip install ${EXTRA_PIP_INSTALL_FLAGS} ${MY_DIR}/source/

backend/python/exllama2/requirements-cpu.txt

@@ -1,3 +0,0 @@
-transformers
-accelerate
-torch==2.4.1

backend/python/exllama2/requirements-cublas12.txt

@@ -1,3 +0,0 @@
-torch==2.4.1
-transformers
-accelerate

backend/python/exllama2/requirements-install.txt

@@ -1,4 +0,0 @@
-# This is here to trigger the install script to add --no-build-isolation to the uv pip install commands
-# exllama2 does not specify it's build requirements per PEP517, so we need to provide some things ourselves
-wheel
-setuptools

backend/python/exllama2/requirements.txt

@@ -1,5 +0,0 @@
-grpcio==1.76.0
-protobuf
-certifi
-wheel
-setuptools

backend/python/faster-whisper/backend.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
 """
-This is an extra gRPC server of LocalAI for Bark TTS
+This is an extra gRPC server of LocalAI for Faster Whisper TTS
 """
 from concurrent import futures
 import time
@@ -40,7 +40,7 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
             device = "mps"
         try:
             print("Preparing models, please wait", file=sys.stderr)
-            self.model = WhisperModel(request.Model, device=device, compute_type="float16")
+            self.model = WhisperModel(request.Model, device=device, compute_type="default")
         except Exception as err:
             return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
         # Implement your logic here for the LoadModel service
@@ -55,11 +55,12 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
             id = 0
             for segment in segments:
                 print("[%.2fs -> %.2fs] %s" % (segment.start, segment.end, segment.text))
-                resultSegments.append(backend_pb2.TranscriptSegment(id=id, start=segment.start, end=segment.end, text=segment.text))
+                resultSegments.append(backend_pb2.TranscriptSegment(id=id, start=int(segment.start)*1e9, end=int(segment.end)*1e9, text=segment.text))
                 text += segment.text
-                id += 1            
+                id += 1
         except Exception as err:
             print(f"Unexpected {err=}, {type(err)=}", file=sys.stderr)
+            raise err
 
         return backend_pb2.TranscriptResult(segments=resultSegments, text=text)
 

backend/python/faster-whisper/requirements-mps.txt

@@ -0,0 +1,8 @@
+torch==2.7.1
+faster-whisper
+opencv-python
+accelerate
+compel
+peft
+sentencepiece
+optimum-quanto

backend/python/kitten-tts/requirements-mps.txt

@@ -0,0 +1,5 @@
+grpcio==1.71.0
+protobuf
+certifi
+packaging==24.1
+https://github.com/KittenML/KittenTTS/releases/download/0.1/kittentts-0.1.0-py3-none-any.whl

backend/python/kokoro/requirements-mps.txt

@@ -0,0 +1,5 @@
+torch==2.7.1
+transformers
+accelerate
+kokoro
+soundfile

backend/python/mlx-audio/requirements-cpu.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-audio
+mlx[cpu]

backend/python/mlx-audio/requirements-cublas12.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-audio
+mlx[cuda12]

backend/python/mlx-audio/requirements-cublas13.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-audio
+mlx[cuda13]

backend/python/mlx-audio/requirements-l4t12.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-audio
+mlx[cuda12]

backend/python/mlx-audio/requirements-l4t13.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-audio
+mlx[cuda13]

backend/python/mlx-vlm/requirements-cpu.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-vlm
+mlx[cpu]

backend/python/mlx-vlm/requirements-cublas12.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-vlm
+mlx[cuda12]

backend/python/mlx-vlm/requirements-cublas13.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-vlm
+mlx[cuda13]

backend/python/mlx-vlm/requirements-l4t12.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-vlm
+mlx[cuda12]

backend/python/mlx-vlm/requirements-l4t13.txt

@@ -0,0 +1,2 @@
+git+https://github.com/Blaizzy/mlx-vlm
+mlx[cuda13]

backend/python/mlx/requirements-cpu.txt

@@ -0,0 +1,2 @@
+mlx-lm
+mlx[cpu]

backend/python/mlx/requirements-cublas12.txt

@@ -0,0 +1,2 @@
+mlx-lm
+mlx[cuda12]

backend/python/mlx/requirements-cublas13.txt

@@ -0,0 +1,2 @@
+mlx-lm
+mlx[cuda13]

backend/python/mlx/requirements-l4t12.txt

@@ -0,0 +1,2 @@
+mlx-lm
+mlx[cuda12]

backend/python/mlx/requirements-l4t13.txt

@@ -0,0 +1,2 @@
+mlx-lm
+mlx[cuda13]

backend/python/moonshine/requirements-mps.txt

@@ -0,0 +1,4 @@
+grpcio==1.71.0
+protobuf
+grpcio-tools
+useful-moonshine-onnx@git+https://git@github.com/moonshine-ai/moonshine.git#subdirectory=moonshine-onnx

backend/python/outetts/Makefile

@@ -0,0 +1,23 @@
+.PHONY: outetts
+outetts:
+	bash install.sh
+
+.PHONY: run
+run: outetts
+	@echo "Running outetts..."
+	bash run.sh
+	@echo "outetts run."
+
+.PHONY: test
+test: outetts
+	@echo "Testing outetts..."
+	bash test.sh
+	@echo "outetts tested."
+
+.PHONY: protogen-clean
+protogen-clean:
+	$(RM) backend_pb2_grpc.py backend_pb2.py
+
+.PHONY: clean
+clean: protogen-clean
+	rm -rf venv __pycache__

backend/python/outetts/backend.py

@@ -0,0 +1,138 @@
+#!/usr/bin/env python3
+"""
+gRPC server for OuteTTS (OuteAI TTS) models.
+"""
+from concurrent import futures
+
+import argparse
+import signal
+import sys
+import os
+import asyncio
+
+import backend_pb2
+import backend_pb2_grpc
+
+import grpc
+import outetts
+
+_ONE_DAY_IN_SECONDS = 60 * 60 * 24
+
+MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1'))
+
+
+class BackendServicer(backend_pb2_grpc.BackendServicer):
+    def Health(self, request, context):
+        return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
+
+    def LoadModel(self, request, context):
+        model_name = request.Model
+        if os.path.exists(request.ModelFile):
+            model_name = request.ModelFile
+
+        self.options = {}
+        for opt in request.Options:
+            if ":" not in opt:
+                continue
+            key, value = opt.split(":", 1)
+            try:
+                if "." in value:
+                    value = float(value)
+                else:
+                    value = int(value)
+            except ValueError:
+                pass
+            self.options[key] = value
+
+        MODELNAME = "OuteAI/OuteTTS-0.3-1B"
+        TOKENIZER = "OuteAI/OuteTTS-0.3-1B"
+        VERSION = "0.3"
+        SPEAKER = "en_male_1"
+        for opt in request.Options:
+            if opt.startswith("tokenizer:"):
+                TOKENIZER = opt.split(":")[1]
+                break
+            if opt.startswith("version:"):
+                VERSION = opt.split(":")[1]
+                break
+            if opt.startswith("speaker:"):
+                SPEAKER = opt.split(":")[1]
+                break
+
+        if model_name != "":
+            MODELNAME = model_name
+
+        try:
+            model_config = outetts.HFModelConfig_v2(
+                model_path=MODELNAME,
+                tokenizer_path=TOKENIZER
+            )
+            self.interface = outetts.InterfaceHF(model_version=VERSION, cfg=model_config)
+
+            self.interface.print_default_speakers()
+            if request.AudioPath:
+                if os.path.isabs(request.AudioPath):
+                    self.AudioPath = request.AudioPath
+                else:
+                    self.AudioPath = os.path.join(request.ModelPath, request.AudioPath)
+                self.speaker = self.interface.create_speaker(audio_path=self.AudioPath)
+            else:
+                self.speaker = self.interface.load_default_speaker(name=SPEAKER)
+
+            if request.ContextSize > 0:
+                self.max_tokens = request.ContextSize
+            else:
+                self.max_tokens = self.options.get("max_new_tokens", 512)
+
+        except Exception as err:
+            print("Error:", err, file=sys.stderr)
+            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
+        return backend_pb2.Result(message="Model loaded successfully", success=True)
+
+    def TTS(self, request, context):
+        try:
+            text = request.text if request.text else "Speech synthesis is the artificial production of human speech."
+            print("[OuteTTS] generating TTS", file=sys.stderr)
+            gen_cfg = outetts.GenerationConfig(
+                text=text,
+                temperature=self.options.get("temperature", 0.1),
+                repetition_penalty=self.options.get("repetition_penalty", 1.1),
+                max_length=self.max_tokens,
+                speaker=self.speaker,
+            )
+            output = self.interface.generate(config=gen_cfg)
+            print("[OuteTTS] Generated TTS", file=sys.stderr)
+            output.save(request.dst)
+            print("[OuteTTS] TTS done", file=sys.stderr)
+        except Exception as err:
+            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
+        return backend_pb2.Result(success=True)
+
+
+async def serve(address):
+    server = grpc.aio.server(
+        migration_thread_pool=futures.ThreadPoolExecutor(max_workers=MAX_WORKERS),
+        options=[
+            ('grpc.max_message_length', 50 * 1024 * 1024),
+            ('grpc.max_send_message_length', 50 * 1024 * 1024),
+            ('grpc.max_receive_message_length', 50 * 1024 * 1024),
+        ])
+    backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server)
+    server.add_insecure_port(address)
+
+    loop = asyncio.get_event_loop()
+    for sig in (signal.SIGINT, signal.SIGTERM):
+        loop.add_signal_handler(
+            sig, lambda: asyncio.ensure_future(server.stop(5))
+        )
+
+    await server.start()
+    print("Server started. Listening on: " + address, file=sys.stderr)
+    await server.wait_for_termination()
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Run the OuteTTS gRPC server.")
+    parser.add_argument("--addr", default="localhost:50051", help="The address to bind the server to.")
+    args = parser.parse_args()
+    asyncio.run(serve(args.addr))

backend/python/outetts/install.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+set -e
+
+backend_dir=$(dirname $0)
+if [ -d $backend_dir/common ]; then
+    source $backend_dir/common/libbackend.sh
+else
+    source $backend_dir/../common/libbackend.sh
+fi
+
+installRequirements

backend/python/outetts/requirements-cpu.txt

@@ -0,0 +1,7 @@
+torch==2.7.1
+llvmlite==0.43.0
+numba==0.60.0
+accelerate
+bitsandbytes
+outetts
+protobuf==6.33.5

backend/python/outetts/requirements-cublas12.txt

@@ -0,0 +1,7 @@
+torch==2.7.1
+accelerate
+llvmlite==0.43.0
+numba==0.60.0
+bitsandbytes
+protobuf==6.33.5
+outetts

backend/python/outetts/requirements-cublas13.txt

@@ -0,0 +1,7 @@
+--extra-index-url https://download.pytorch.org/whl/cu130
+torch==2.9.0
+llvmlite==0.43.0
+numba==0.60.0
+bitsandbytes
+outetts
+protobuf==6.33.5

backend/python/outetts/requirements-hipblas.txt

@@ -1,5 +1,8 @@
 --extra-index-url https://download.pytorch.org/whl/rocm6.4
 torch==2.8.0+rocm6.4
-torchaudio==2.8.0+rocm6.4
-transformers
-accelerate
+accelerate
+llvmlite==0.43.0
+numba==0.60.0
+bitsandbytes
+outetts
+protobuf==6.33.5

backend/python/outetts/requirements-intel.txt

@@ -0,0 +1,8 @@
+--extra-index-url https://download.pytorch.org/whl/xpu
+torch
+optimum[openvino]
+llvmlite==0.43.0
+numba==0.60.0
+bitsandbytes
+outetts
+protobuf==6.33.5

backend/python/outetts/requirements.txt

@@ -0,0 +1,6 @@
+grpcio==1.76.0
+protobuf==6.33.5
+certifi
+setuptools
+scipy==1.15.1
+numpy>=2.0.0

backend/python/outetts/run.sh

@@ -6,4 +6,4 @@ else
     source $backend_dir/../common/libbackend.sh
 fi
 
-startBackend $@
+startBackend $@

backend/python/outetts/test.py

@@ -0,0 +1,35 @@
+"""
+Test script for the OuteTTS gRPC service.
+"""
+import unittest
+import subprocess
+import time
+import backend_pb2
+import backend_pb2_grpc
+
+import grpc
+
+
+class TestBackendServicer(unittest.TestCase):
+    def setUp(self):
+        self.service = subprocess.Popen(["python3", "backend.py", "--addr", "localhost:50051"])
+        time.sleep(5)
+
+    def tearDown(self):
+        self.service.terminate()
+        self.service.wait()
+
+    def test_health(self):
+        try:
+            with grpc.insecure_channel("localhost:50051") as channel:
+                stub = backend_pb2_grpc.BackendStub(channel)
+                response = stub.Health(backend_pb2.HealthMessage())
+                self.assertEqual(response.message, b'OK')
+        except Exception as err:
+            self.fail(f"Health check failed: {err}")
+        finally:
+            self.tearDown()
+
+
+if __name__ == "__main__":
+    unittest.main()

backend/python/qwen-asr/Makefile

@@ -0,0 +1,25 @@
+.DEFAULT_GOAL := install
+
+.PHONY: qwen-asr
+qwen-asr:
+	bash install.sh
+
+.PHONY: run
+run: qwen-asr
+	@echo "Running qwen-asr..."
+	bash run.sh
+	@echo "qwen-asr run."
+
+.PHONY: test
+test: qwen-asr
+	@echo "Testing qwen-asr..."
+	bash test.sh
+	@echo "qwen-asr tested."
+
+.PHONY: protogen-clean
+protogen-clean:
+	$(RM) backend_pb2_grpc.py backend_pb2.py
+
+.PHONY: clean
+clean: protogen-clean
+	rm -rf venv __pycache__

backend/python/qwen-asr/backend.py

@@ -0,0 +1,212 @@
+#!/usr/bin/env python3
+"""
+gRPC server of LocalAI for Qwen3-ASR (transformers backend, non-vLLM).
+"""
+from concurrent import futures
+import time
+import argparse
+import signal
+import sys
+import os
+import backend_pb2
+import backend_pb2_grpc
+import torch
+from qwen_asr import Qwen3ASRModel
+
+import grpc
+
+
+def is_float(s):
+    try:
+        float(s)
+        return True
+    except ValueError:
+        return False
+
+
+def is_int(s):
+    try:
+        int(s)
+        return True
+    except ValueError:
+        return False
+
+
+_ONE_DAY_IN_SECONDS = 60 * 60 * 24
+MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1'))
+
+
+class BackendServicer(backend_pb2_grpc.BackendServicer):
+    def Health(self, request, context):
+        return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
+
+    def LoadModel(self, request, context):
+        if torch.cuda.is_available():
+            device = "cuda"
+        else:
+            device = "cpu"
+        mps_available = hasattr(torch.backends, "mps") and torch.backends.mps.is_available()
+        if mps_available:
+            device = "mps"
+        if not torch.cuda.is_available() and request.CUDA:
+            return backend_pb2.Result(success=False, message="CUDA is not available")
+
+        self.device = device
+        self.options = {}
+
+        for opt in request.Options:
+            if ":" not in opt:
+                continue
+            key, value = opt.split(":", 1)
+            if is_float(value):
+                value = float(value)
+            elif is_int(value):
+                value = int(value)
+            elif value.lower() in ["true", "false"]:
+                value = value.lower() == "true"
+            self.options[key] = value
+
+        model_path = request.Model or "Qwen/Qwen3-ASR-1.7B"
+        default_dtype = torch.bfloat16 if self.device == "cuda" else torch.float32
+        load_dtype = default_dtype
+        if "torch_dtype" in self.options:
+            d = str(self.options["torch_dtype"]).lower()
+            if d == "fp16":
+                load_dtype = torch.float16
+            elif d == "bf16":
+                load_dtype = torch.bfloat16
+            elif d == "fp32":
+                load_dtype = torch.float32
+            del self.options["torch_dtype"]
+
+        self.max_inference_batch_size = self.options.get("max_inference_batch_size", 32)
+        self.max_new_tokens = self.options.get("max_new_tokens", 256)
+
+        forced_aligner = self.options.get("forced_aligner")
+        if forced_aligner is not None and isinstance(forced_aligner, str):
+            forced_aligner = forced_aligner.strip() or None
+        attn_implementation = self.options.get("attn_implementation")
+        if attn_implementation is not None and isinstance(attn_implementation, str):
+            attn_implementation = attn_implementation.strip() or None
+
+        if self.device == "mps":
+            device_map = None
+        elif self.device == "cuda":
+            device_map = "cuda:0"
+        else:
+            device_map = "cpu"
+
+        load_kwargs = dict(
+            dtype=load_dtype,
+            device_map=device_map,
+            max_inference_batch_size=self.max_inference_batch_size,
+            max_new_tokens=self.max_new_tokens,
+        )
+        if attn_implementation:
+            load_kwargs["attn_implementation"] = attn_implementation
+        if forced_aligner:
+            load_kwargs["forced_aligner"] = forced_aligner
+            forced_aligner_kwargs = dict(
+                dtype=load_dtype,
+                device_map=device_map,
+            )
+            if attn_implementation:
+                forced_aligner_kwargs["attn_implementation"] = attn_implementation
+            load_kwargs["forced_aligner_kwargs"] = forced_aligner_kwargs
+
+        try:
+            print(f"Loading Qwen3-ASR from {model_path}", file=sys.stderr)
+            if attn_implementation:
+                print(f"Using attn_implementation: {attn_implementation}", file=sys.stderr)
+            if forced_aligner:
+                print(f"Loading with forced_aligner: {forced_aligner}", file=sys.stderr)
+            self.model = Qwen3ASRModel.from_pretrained(model_path, **load_kwargs)
+            print("Qwen3-ASR model loaded successfully", file=sys.stderr)
+        except Exception as err:
+            print(f"[ERROR] LoadModel failed: {err}", file=sys.stderr)
+            import traceback
+            traceback.print_exc(file=sys.stderr)
+            return backend_pb2.Result(success=False, message=str(err))
+
+        return backend_pb2.Result(message="Model loaded successfully", success=True)
+
+    def AudioTranscription(self, request, context):
+        result_segments = []
+        text = ""
+        try:
+            audio_path = request.dst
+            if not audio_path or not os.path.exists(audio_path):
+                print(f"Error: Audio file not found: {audio_path}", file=sys.stderr)
+                return backend_pb2.TranscriptResult(segments=[], text="")
+
+            language = None
+            if request.language and request.language.strip():
+                language = request.language.strip()
+
+            results = self.model.transcribe(audio=audio_path, language=language)
+
+            if not results:
+                return backend_pb2.TranscriptResult(segments=[], text="")
+
+            r = results[0]
+            text = r.text or ""
+
+            if getattr(r, 'time_stamps', None) and len(r.time_stamps) > 0:
+                for idx, ts in enumerate(r.time_stamps):
+                    start_ms = 0
+                    end_ms = 0
+                    seg_text = text
+                    if isinstance(ts, (list, tuple)) and len(ts) >= 3:
+                        start_ms = int(float(ts[0]) * 1000) if ts[0] is not None else 0
+                        end_ms = int(float(ts[1]) * 1000) if ts[1] is not None else 0
+                        seg_text = ts[2] if len(ts) > 2 and ts[2] is not None else ""
+                    result_segments.append(backend_pb2.TranscriptSegment(
+                        id=idx, start=start_ms, end=end_ms, text=seg_text
+                    ))
+            else:
+                if text:
+                    result_segments.append(backend_pb2.TranscriptSegment(
+                        id=0, start=0, end=0, text=text
+                    ))
+        except Exception as err:
+            print(f"Error in AudioTranscription: {err}", file=sys.stderr)
+            import traceback
+            traceback.print_exc(file=sys.stderr)
+            return backend_pb2.TranscriptResult(segments=[], text="")
+
+        return backend_pb2.TranscriptResult(segments=result_segments, text=text)
+
+
+def serve(address):
+    server = grpc.server(
+        futures.ThreadPoolExecutor(max_workers=MAX_WORKERS),
+        options=[
+            ('grpc.max_message_length', 50 * 1024 * 1024),
+            ('grpc.max_send_message_length', 50 * 1024 * 1024),
+            ('grpc.max_receive_message_length', 50 * 1024 * 1024),
+        ])
+    backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server)
+    server.add_insecure_port(address)
+    server.start()
+    print("Server started. Listening on: " + address, file=sys.stderr)
+
+    def signal_handler(sig, frame):
+        print("Received termination signal. Shutting down...")
+        server.stop(0)
+        sys.exit(0)
+
+    signal.signal(signal.SIGINT, signal_handler)
+    signal.signal(signal.SIGTERM, signal_handler)
+
+    try:
+        while True:
+            time.sleep(_ONE_DAY_IN_SECONDS)
+    except KeyboardInterrupt:
+        server.stop(0)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Run the gRPC server.")
+    parser.add_argument("--addr", default="localhost:50051", help="The address to bind the server to.")
+    args = parser.parse_args()
+    serve(args.addr)

backend/python/qwen-asr/install.sh

@@ -0,0 +1,21 @@
+#!/bin/bash
+set -e
+
+EXTRA_PIP_INSTALL_FLAGS="--no-build-isolation"
+
+backend_dir=$(dirname $0)
+if [ -d $backend_dir/common ]; then
+    source $backend_dir/common/libbackend.sh
+else
+    source $backend_dir/../common/libbackend.sh
+fi
+
+if [ "x${BUILD_PROFILE}" == "xintel" ]; then
+    EXTRA_PIP_INSTALL_FLAGS+=" --upgrade --index-strategy=unsafe-first-match"
+fi
+
+PYTHON_VERSION="3.12"
+PYTHON_PATCH="12"
+PY_STANDALONE_TAG="20251120"
+
+installRequirements

backend/python/qwen-asr/requirements-cpu.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://download.pytorch.org/whl/cpu
+torch
+qwen-asr

backend/python/qwen-asr/requirements-cublas12-after.txt

@@ -0,0 +1 @@
+flash-attn

backend/python/qwen-asr/requirements-cublas12.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://download.pytorch.org/whl/cu121
+torch
+qwen-asr

backend/python/qwen-asr/requirements-cublas13.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://download.pytorch.org/whl/cu130
+torch
+qwen-asr

backend/python/qwen-asr/requirements-hipblas.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://download.pytorch.org/whl/rocm6.3
+torch==2.7.1+rocm6.3
+qwen-asr

backend/python/qwen-asr/requirements-intel-after.txt

@@ -0,0 +1 @@
+flash-attn

backend/python/qwen-asr/requirements-intel.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://download.pytorch.org/whl/xpu
+torch
+qwen-asr

backend/python/qwen-asr/requirements-l4t12.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://pypi.jetson-ai-lab.io/jp6/cu129/
+torch
+qwen-asr

backend/python/qwen-asr/requirements-l4t13.txt

@@ -0,0 +1,3 @@
+--extra-index-url https://download.pytorch.org/whl/cu130
+torch
+qwen-asr

backend/python/qwen-asr/requirements-mps.txt

@@ -0,0 +1,2 @@
+torch==2.7.1
+qwen-asr

backend/python/qwen-asr/requirements.txt

@@ -0,0 +1,5 @@
+grpcio==1.71.0
+protobuf
+certifi
+packaging==24.1
+setuptools

backend/python/qwen-asr/run.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+backend_dir=$(dirname $0)
+if [ -d $backend_dir/common ]; then
+    source $backend_dir/common/libbackend.sh
+else
+    source $backend_dir/../common/libbackend.sh
+fi
+
+startBackend $@

backend/python/qwen-asr/test.py

@@ -0,0 +1,94 @@
+"""
+Tests for the Qwen3-ASR gRPC backend.
+"""
+import unittest
+import subprocess
+import time
+import os
+import tempfile
+import shutil
+import backend_pb2
+import backend_pb2_grpc
+
+import grpc
+
+# Skip heavy transcription test in CI (model download + inference)
+SKIP_ASR_TESTS = os.environ.get("SKIP_ASR_TESTS", "false").lower() == "true"
+
+
+class TestBackendServicer(unittest.TestCase):
+    def setUp(self):
+        self.service = subprocess.Popen(["python3", "backend.py", "--addr", "localhost:50051"])
+        time.sleep(15)
+
+    def tearDown(self):
+        self.service.terminate()
+        self.service.wait()
+
+    def test_server_startup(self):
+        try:
+            self.setUp()
+            with grpc.insecure_channel("localhost:50051") as channel:
+                stub = backend_pb2_grpc.BackendStub(channel)
+                response = stub.Health(backend_pb2.HealthMessage())
+                self.assertEqual(response.message, b'OK')
+        except Exception as err:
+            print(err)
+            self.fail("Server failed to start")
+        finally:
+            self.tearDown()
+
+    def test_load_model(self):
+        try:
+            self.setUp()
+            with grpc.insecure_channel("localhost:50051") as channel:
+                stub = backend_pb2_grpc.BackendStub(channel)
+                response = stub.LoadModel(backend_pb2.ModelOptions(Model="Qwen/Qwen3-ASR-1.7B"))
+                self.assertTrue(response.success, response.message)
+                self.assertEqual(response.message, "Model loaded successfully")
+        except Exception as err:
+            print(err)
+            self.fail("LoadModel service failed")
+        finally:
+            self.tearDown()
+
+    @unittest.skipIf(SKIP_ASR_TESTS, "ASR transcription test skipped (SKIP_ASR_TESTS=true)")
+    def test_audio_transcription(self):
+        temp_dir = tempfile.mkdtemp()
+        audio_file = os.path.join(temp_dir, 'audio.wav')
+        try:
+            url = "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen3-ASR-Repo/asr_en.wav"
+            result = subprocess.run(
+                ["wget", "-q", url, "-O", audio_file],
+                capture_output=True,
+                text=True,
+                timeout=30,
+            )
+            if result.returncode != 0:
+                self.skipTest(f"Could not download sample audio: {result.stderr}")
+            if not os.path.exists(audio_file):
+                self.skipTest("Sample audio file not found after download")
+
+            self.setUp()
+            with grpc.insecure_channel("localhost:50051") as channel:
+                stub = backend_pb2_grpc.BackendStub(channel)
+                load_response = stub.LoadModel(backend_pb2.ModelOptions(Model="Qwen/Qwen3-ASR-0.6B"))
+                self.assertTrue(load_response.success, load_response.message)
+
+                transcript_response = stub.AudioTranscription(
+                    backend_pb2.TranscriptRequest(dst=audio_file)
+                )
+                self.assertIsNotNone(transcript_response)
+                self.assertIsNotNone(transcript_response.text)
+                self.assertGreaterEqual(len(transcript_response.segments), 0)
+                all_text = ""
+                for segment in transcript_response.segments:
+                    all_text += segment.text
+                print(f"All text: {all_text}")
+                self.assertIn("big", all_text)
+                if transcript_response.segments:
+                    self.assertIsNotNone(transcript_response.segments[0].text)
+        finally:
+            self.tearDown()
+            if os.path.exists(temp_dir):
+                shutil.rmtree(temp_dir)

backend/python/qwen-tts/test.py

@@ -28,7 +28,7 @@ class TestBackendServicer(unittest.TestCase):
             stderr=subprocess.PIPE,
             text=True
         )
-        time.sleep(5)
+        time.sleep(30)
 
     def tearDown(self) -> None:
         """

backend/python/rerankers/requirements-mps.txt

@@ -0,0 +1,4 @@
+torch==2.7.1
+transformers
+accelerate
+rerankers[transformers]

backend/python/rfdetr/requirements-mps.txt

@@ -0,0 +1,7 @@
+torch==2.7.1
+rfdetr
+opencv-python
+accelerate
+peft
+inference
+optimum-quanto

backend/python/transformers/backend.py

@@ -24,7 +24,6 @@ XPU=os.environ.get("XPU", "0") == "1"
 from transformers import AutoTokenizer, AutoModel, set_seed, TextIteratorStreamer, StoppingCriteriaList, StopStringCriteria, MambaConfig, MambaForCausalLM
 from transformers import AutoProcessor, MusicgenForConditionalGeneration, DiaForConditionalGeneration
 from scipy.io import wavfile
-import outetts
 from sentence_transformers import SentenceTransformer
 
 
@@ -89,7 +88,6 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
 
         self.CUDA = torch.cuda.is_available()
         self.OV=False
-        self.OuteTTS=False
         self.DiaTTS=False
         self.SentenceTransformer = False
 
@@ -239,45 +237,6 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
                     self.processor = self.processor.to("cuda")
                 print("DiaForConditionalGeneration loaded", file=sys.stderr)
                 self.DiaTTS = True
-            elif request.Type == "OuteTTS":
-                autoTokenizer = False
-                options = request.Options
-                MODELNAME = "OuteAI/OuteTTS-0.3-1B"
-                TOKENIZER = "OuteAI/OuteTTS-0.3-1B"
-                VERSION = "0.3"
-                SPEAKER = "en_male_1"
-                for opt in options:
-                    if opt.startswith("tokenizer:"):
-                        TOKENIZER = opt.split(":")[1]
-                        break
-                    if opt.startswith("version:"):
-                        VERSION = opt.split(":")[1]
-                        break
-                    if opt.startswith("speaker:"):
-                        SPEAKER = opt.split(":")[1]
-                        break
-                
-                if model_name != "":
-                    MODELNAME = model_name
-
-                # Configure the model
-                model_config = outetts.HFModelConfig_v2(
-                    model_path=MODELNAME,
-                    tokenizer_path=TOKENIZER
-                )
-                # Initialize the interface
-                self.interface = outetts.InterfaceHF(model_version=VERSION, cfg=model_config)
-                self.OuteTTS = True
-
-                self.interface.print_default_speakers()
-                if request.AudioPath:
-                    if os.path.isabs(request.AudioPath):
-                        self.AudioPath = request.AudioPath
-                    else:
-                        self.AudioPath = os.path.join(request.ModelPath, request.AudioPath)
-                    self.speaker = self.interface.create_speaker(audio_path=self.AudioPath)
-                else:
-                    self.speaker = self.interface.load_default_speaker(name=SPEAKER)               
             elif request.Type == "SentenceTransformer":
                 autoTokenizer = False
                 self.model = SentenceTransformer(model_name, trust_remote_code=request.TrustRemoteCode)
@@ -588,30 +547,8 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
         return backend_pb2.Result(success=True)
 
 
-    def CallOuteTTS(self, request, context):
-        try:
-            print("[OuteTTS] generating TTS", file=sys.stderr)
-            gen_cfg = outetts.GenerationConfig(
-                text="Speech synthesis is the artificial production of human speech.",
-                temperature=self.options.get("temperature", 0.1),
-                repetition_penalty=self.options.get("repetition_penalty", 1.1),
-                max_length=self.max_tokens,
-                speaker=self.speaker,
-                # voice_characteristics="upbeat enthusiasm, friendliness, clarity, professionalism, and trustworthiness"
-            )
-            output = self.interface.generate(config=gen_cfg)
-            print("[OuteTTS] Generated TTS", file=sys.stderr)
-            output.save(request.dst)
-            print("[OuteTTS] TTS done", file=sys.stderr)
-        except Exception as err:
-            return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}")
-        return backend_pb2.Result(success=True)
-
 # The TTS endpoint is older, and provides fewer features, but exists for compatibility reasons
     def TTS(self, request, context):
-        if self.OuteTTS:
-            return self.CallOuteTTS(request, context)
-        
         if self.DiaTTS:
             print("DiaTTS", file=sys.stderr)
             return self.CallDiaTTS(request, context)

backend/python/transformers/requirements-cpu.txt

@@ -4,6 +4,5 @@ numba==0.60.0
 accelerate
 transformers
 bitsandbytes
-outetts
-sentence-transformers==5.2.0
-protobuf==6.33.4
+sentence-transformers==5.2.2
+protobuf==6.33.5

backend/python/transformers/requirements-cublas12.txt

@@ -4,6 +4,5 @@ llvmlite==0.43.0
 numba==0.60.0
 transformers
 bitsandbytes
-outetts
-sentence-transformers==5.2.0
-protobuf==6.33.4
+sentence-transformers==5.2.2
+protobuf==6.33.5

backend/python/transformers/requirements-cublas13.txt

@@ -4,6 +4,5 @@ llvmlite==0.43.0
 numba==0.60.0
 transformers
 bitsandbytes
-outetts
-sentence-transformers==5.2.0
-protobuf==6.33.4
+sentence-transformers==5.2.2
+protobuf==6.33.5