Merge pull request #299 from makllama/xd/musa

Add MUSA Support

Author: ericcurtin

.github/workflows/release.yml

@@ -148,3 +148,18 @@ jobs:
           sbom: true
           provenance: mode=max
           tags: ${{ steps.tags.outputs.rocm }}
+
+      - name: Build MUSA image
+        uses: docker/build-push-action@ca052bb54ab0790a636c9b5f226502c73d547a25
+        with:
+          file: Dockerfile
+          target: final-llamacpp
+          platforms: linux/amd64
+          build-args: |
+            "LLAMA_SERVER_VERSION=${{ inputs.llamaServerVersion }}"
+            "LLAMA_SERVER_VARIANT=musa"
+            "BASE_IMAGE=mthreads/musa:rc4.3.0-runtime-ubuntu22.04-amd64"
+          push: true
+          sbom: true
+          provenance: mode=max
+          tags: ${{ steps.tags.outputs.musa }}

README.md

@@ -57,17 +57,17 @@ If you encounter errors like `Package 'docker-model-plugin' has no installation
    # Check Docker Model Runner version
    docker model version
    ```
-   
+
    Look for the source in the output. If it shows a package from your distro, you'll need to reinstall from Docker's official repositories.
 
 2. **Remove the distro version and install from Docker's official repository:**
    ```bash
    # Remove distro version (Ubuntu/Debian example)
    sudo apt-get purge docker docker.io containerd runc
-   
+
    # Install from Docker's official repository
    curl -fsSL https://get.docker.com | sudo bash
-   
+
    # Verify Docker Model Runner is available
    docker model --help
    ```
@@ -146,7 +146,7 @@ cd cmd/cli
 MODEL_RUNNER_HOST=http://localhost:13434 ./model-cli list
 
 # Pull and run a model
-MODEL_RUNNER_HOST=http://localhost:13434 ./model-cli run ai/smollm2 "Hello, how are you?" 
+MODEL_RUNNER_HOST=http://localhost:13434 ./model-cli run ai/smollm2 "Hello, how are you?"
 ```
 
 #### Option 2: Using Docker
@@ -173,7 +173,7 @@ MODEL_RUNNER_HOST=http://localhost:13434 ./model-cli list
 
 ## Using the Makefile
 
-This project includes a Makefile to simplify common development tasks. It requires Docker Desktop >= 4.41.0 
+This project includes a Makefile to simplify common development tasks. It requires Docker Desktop >= 4.41.0
 The Makefile provides the following targets:
 
 - `build` - Build the Go application
@@ -224,6 +224,7 @@ Available variants:
 - `cpu`: CPU-optimized version
 - `cuda`: CUDA-accelerated version for NVIDIA GPUs
 - `rocm`: ROCm-accelerated version for AMD GPUs
+- `musa`: MUSA-accelerated version for MTHREADS GPUs
 
 The binary path in the image follows this pattern: `/com.docker.llama-server.native.linux.${LLAMA_SERVER_VARIANT}.${TARGETARCH}`
 

cmd/cli/README.md

@@ -24,37 +24,37 @@ A powerful command-line interface for managing, running, packaging, and deployin
    ```
 3. **Install Model Runner:**
    ```bash
-   ./model install-runner
+   ./model-cli install-runner
    ```
    Use `--gpu cuda` for GPU support, or `--gpu auto` for automatic detection.
 
 ## Usage
-Run `./model --help` to see all commands and options.
+Run `./model-cli --help` to see all commands and options.
 
 ### Common Commands
-- `model install-runner` — Install the Docker Model Runner
-- `model start-runner` — Start the Docker Model Runner
-- `model stop-runner` — Stop the Docker Model Runner
-- `model restart-runner` — Restart the Docker Model Runner
-- `model run MODEL [PROMPT]` — Run a model with a prompt or enter chat mode
-- `model list` — List available models
-- `model package --gguf <path> --push <target>` — Package and push a model
-- `model logs` — View logs
-- `model status` — Check runner status
-- `model configure MODEL [flags]` — Configure model runtime
-- `model unload MODEL` — Unload a model
-- `model tag SOURCE TARGET` — Tag a model
-- `model pull MODEL` — Pull a model
-- `model push MODEL` — Push a model
-- `model rm MODEL` — Remove a model
+- `model-cli install-runner` — Install the Docker Model Runner
+- `model-cli start-runner` — Start the Docker Model Runner
+- `model-cli stop-runner` — Stop the Docker Model Runner
+- `model-cli restart-runner` — Restart the Docker Model Runner
+- `model-cli run MODEL [PROMPT]` — Run a model with a prompt or enter chat mode
+- `model-cli list` — List available models
+- `model-cli package --gguf <path> --push <target>` — Package and push a model
+- `model-cli logs` — View logs
+- `model-cli status` — Check runner status
+- `model-cli configure MODEL [flags]` — Configure model runtime
+- `model-cli unload MODEL` — Unload a model
+- `model-cli tag SOURCE TARGET` — Tag a model
+- `model-cli pull MODEL` — Pull a model
+- `model-cli push MODEL` — Push a model
+- `model-cli rm MODEL` — Remove a model
 
 ## Example: Interactive Chat
 ```bash
-./model run llama.cpp "What is the capital of France?"
+./model-cli run llama.cpp "What is the capital of France?"
 ```
 Or enter chat mode:
 ```bash
-./model run llama.cpp
+./model-cli run llama.cpp
 > """
 Tell me a joke.
 """

cmd/cli/commands/install-runner.go

@@ -4,10 +4,11 @@ import (
 	"context"
 	"errors"
 	"fmt"
-	"github.com/docker/model-runner/cmd/cli/pkg/types"
 	"os"
 	"time"
 
+	"github.com/docker/model-runner/cmd/cli/pkg/types"
+
 	"github.com/docker/docker/api/types/container"
 	"github.com/docker/model-runner/cmd/cli/commands/completion"
 	"github.com/docker/model-runner/cmd/cli/desktop"
@@ -247,6 +248,8 @@ func runInstallOrStart(cmd *cobra.Command, opts runnerOptions) error {
 		gpu = gpupkg.GPUSupportCUDA
 	} else if opts.gpuMode == "rocm" {
 		gpu = gpupkg.GPUSupportROCm
+	} else if opts.gpuMode == "musa" {
+		gpu = gpupkg.GPUSupportMUSA
 	} else if opts.gpuMode != "none" {
 		return fmt.Errorf("unknown GPU specification: %q", opts.gpuMode)
 	}
@@ -295,7 +298,7 @@ func newInstallRunner() *cobra.Command {
 	c.Flags().Uint16Var(&port, "port", 0,
 		"Docker container port for Docker Model Runner (default: 12434 for Docker Engine, 12435 for Cloud mode)")
 	c.Flags().StringVar(&host, "host", "127.0.0.1", "Host address to bind Docker Model Runner")
-	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda|rocm)")
+	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda|rocm|musa)")
 	c.Flags().BoolVar(&doNotTrack, "do-not-track", false, "Do not track models usage in Docker Model Runner")
 	return c
 }

cmd/cli/commands/reinstall-runner.go

@@ -28,7 +28,7 @@ func newReinstallRunner() *cobra.Command {
 	c.Flags().Uint16Var(&port, "port", 0,
 		"Docker container port for Docker Model Runner (default: 12434 for Docker Engine, 12435 for Cloud mode)")
 	c.Flags().StringVar(&host, "host", "127.0.0.1", "Host address to bind Docker Model Runner")
-	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda)")
+	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda|musa)")
 	c.Flags().BoolVar(&doNotTrack, "do-not-track", false, "Do not track models usage in Docker Model Runner")
 	return c
 }

cmd/cli/commands/restart-runner.go

@@ -36,7 +36,7 @@ func newRestartRunner() *cobra.Command {
 	c.Flags().Uint16Var(&port, "port", 0,
 		"Docker container port for Docker Model Runner (default: 12434 for Docker Engine, 12435 for Cloud mode)")
 	c.Flags().StringVar(&host, "host", "127.0.0.1", "Host address to bind Docker Model Runner")
-	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda)")
+	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda|musa)")
 	c.Flags().BoolVar(&doNotTrack, "do-not-track", false, "Do not track models usage in Docker Model Runner")
 	return c
 }

cmd/cli/commands/start-runner.go

@@ -24,7 +24,7 @@ func newStartRunner() *cobra.Command {
 	}
 	c.Flags().Uint16Var(&port, "port", 0,
 		"Docker container port for Docker Model Runner (default: 12434 for Docker Engine, 12435 for Cloud mode)")
-	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda)")
+	c.Flags().StringVar(&gpuMode, "gpu", "auto", "Specify GPU support (none|auto|cuda|musa)")
 	c.Flags().BoolVar(&doNotTrack, "do-not-track", false, "Do not track models usage in Docker Model Runner")
 	return c
 }

cmd/cli/docs/reference/docker_model_install-runner.yaml

@@ -19,7 +19,7 @@ options:
     - option: gpu
       value_type: string
       default_value: auto
-      description: Specify GPU support (none|auto|cuda|rocm)
+      description: Specify GPU support (none|auto|cuda|rocm|musa)
       deprecated: false
       hidden: false
       experimental: false

cmd/cli/docs/reference/docker_model_reinstall-runner.yaml

@@ -19,7 +19,7 @@ options:
     - option: gpu
       value_type: string
       default_value: auto
-      description: Specify GPU support (none|auto|cuda)
+      description: Specify GPU support (none|auto|cuda|musa)
       deprecated: false
       hidden: false
       experimental: false

cmd/cli/docs/reference/docker_model_restart-runner.yaml

@@ -21,7 +21,7 @@ options:
     - option: gpu
       value_type: string
       default_value: auto
-      description: Specify GPU support (none|auto|cuda)
+      description: Specify GPU support (none|auto|cuda|musa)
       deprecated: false
       hidden: false
       experimental: false