lemonade-sdk 8.0.4__py3-none-any.whl → 8.0.6__py3-none-any.whl

lemonade_server/cli.py

@@ -27,43 +27,75 @@ class DeleteError(Exception):
     """
 
 
+class ServerTimeoutError(Exception):
+    """
+    The server failed to start within the timeout period
+    """
+
+
+class ModelNotAvailableError(Exception):
+    """
+    The specified model is not available on the server
+    """
+
+
 def serve(
-    port: int,
+    port: int = None,
     log_level: str = None,
     tray: bool = False,
+    use_thread: bool = False,
 ):
     """
     Execute the serve command
     """
 
-    # Check if Lemonade Server is already running
-    _, running_port = get_server_info()
-    if running_port is not None:
-        print(
-            (
-                f"Lemonade Server is already running on port {running_port}\n"
-                "Please stop the existing server before starting a new instance."
-            ),
-        )
-        sys.exit(ExitCodes.SERVER_ALREADY_RUNNING)
-
     # Otherwise, start the server
     print("Starting Lemonade Server...")
     from lemonade.tools.server.serve import Server, DEFAULT_PORT, DEFAULT_LOG_LEVEL
 
-    server = Server()
     port = port if port is not None else DEFAULT_PORT
     log_level = log_level if log_level is not None else DEFAULT_LOG_LEVEL
 
     # Hidden environment variable to enable input truncation (experimental feature)
     truncate_inputs = "LEMONADE_TRUNCATE_INPUTS" in os.environ
 
-    server.run(
-        port=port,
-        log_level=log_level,
-        truncate_inputs=truncate_inputs,
-        tray=tray,
-    )
+    # Start the server
+    serve_kwargs = {
+        "log_level": log_level,
+        "truncate_inputs": truncate_inputs,
+        "tray": tray,
+    }
+    server = Server()
+    if not use_thread:
+        server.run(
+            port=port,
+            **serve_kwargs,
+        )
+    else:
+        from threading import Thread
+        import time
+
+        # Start a background thread to run the server
+        server_thread = Thread(
+            target=server.run,
+            args=(port,),
+            kwargs=serve_kwargs,
+            daemon=True,
+        )
+        server_thread.start()
+
+        # Wait for the server to be ready
+        max_wait_time = 30
+        wait_interval = 0.5
+        waited = 0
+        while waited < max_wait_time:
+            time.sleep(wait_interval)
+            _, running_port = get_server_info()
+            if running_port is not None:
+                break
+            waited += wait_interval
+
+        return port, server_thread
 
 
 def stop():
@@ -161,9 +193,8 @@ def pull(
             if pull_response.status_code != 200:
                 raise PullError(
                     f"Failed to install {model_name}. Check the "
-                    "Lemonade Server log for more information. A list of supported models "
-                    "is provided at "
-                    "https://github.com/lemonade-sdk/lemonade/blob/main/docs/server/server_models.md"
+                    "Lemonade Server log for more information. You can list "
+                    "supported models with `lemonade-server list`"
                 )
     else:
         from lemonade_server.model_manager import ModelManager
@@ -212,6 +243,53 @@ def delete(model_names: List[str]):
             ModelManager().delete_model(model_name)
 
 
+def run(model_name: str):
+    """
+    Start the server if not running and open the webapp with the specified model
+    """
+    import webbrowser
+    import time
+
+    # Start the server if not running
+    _, port = get_server_info()
+    server_previously_running = port is not None
+    if not server_previously_running:
+        port, server_thread = serve(use_thread=True, tray=True, log_level="info")
+
+    # Pull model
+    pull([model_name])
+
+    # Load model
+    load(model_name, port)
+
+    # Open the webapp with the specified model
+    url = f"http://localhost:{port}/?model={model_name}#llm-chat"
+    print(f"You can now chat with {model_name} at {url}")
+    webbrowser.open(url)
+
+    # Keep the server running if we started it
+    if not server_previously_running:
+        while server_thread.is_alive():
+            time.sleep(0.5)
+
+
+def load(model_name: str, port: int):
+    """
+    Load a model using the endpoint
+    """
+    import requests
+
+    base_url = f"http://localhost:{port}/api/v1"
+
+    # Load the model
+    load_response = requests.post(f"{base_url}/load", json={"model_name": model_name})
+    if load_response.status_code != 200:
+        raise ModelLoadError(
+            f"Failed to load {model_name}. Check the "
+            "Lemonade Server log for more information."
+        )
+
+
 def version():
     """
     Print the version number
@@ -294,6 +372,46 @@ def get_server_info() -> Tuple[int | None, int | None]:
     return None, None
 
 
+def list_models():
+    """
+    List recommended models and their download status
+    """
+    from tabulate import tabulate
+    from lemonade_server.model_manager import ModelManager
+
+    model_manager = ModelManager()
+
+    # Get all supported models and downloaded models
+    supported_models = model_manager.supported_models
+    downloaded_models = model_manager.downloaded_models
+
+    # Filter to only show recommended models
+    recommended_models = {
+        model_name: model_info
+        for model_name, model_info in supported_models.items()
+        if model_info.get("suggested", False)
+    }
+
+    # Create table data
+    table_data = []
+    for model_name, model_info in recommended_models.items():
+        downloaded_status = "Yes" if model_name in downloaded_models else "No"
+
+        # Get model labels/type
+        labels = model_info.get("labels", [])
+        model_type = ", ".join(labels) if labels else "-"
+
+        table_data.append([model_name, downloaded_status, model_type])
+
+    # Sort by model name for consistent display
+    # Show downloaded models first
+    table_data.sort(key=lambda x: (x[1] == "No", x[0].lower()))
+
+    # Display table
+    headers = ["Model Name", "Downloaded", "Details"]
+    print(tabulate(table_data, headers=headers, tablefmt="simple"))
+
+
 def main():
     parser = argparse.ArgumentParser(
         description="Serve LLMs on CPU, GPU, and NPU.",
@@ -333,6 +451,11 @@ def main():
     # Stop command
     stop_parser = subparsers.add_parser("stop", help="Stop the server")
 
+    # List command
+    list_parser = subparsers.add_parser(
+        "list", help="List recommended models and their download status"
+    )
+
     # Pull command
     pull_parser = subparsers.add_parser(
         "pull",
@@ -381,6 +504,16 @@ def main():
         nargs="+",
     )
 
+    # Run command
+    run_parser = subparsers.add_parser(
+        "run",
+        help="Chat with specified model (starts server if needed)",
+    )
+    run_parser.add_argument(
+        "model",
+        help="Lemonade Server model name to run",
+    )
+
     args = parser.parse_args()
 
     if os.name != "nt":
@@ -389,6 +522,15 @@ def main():
     if args.version:
         version()
     elif args.command == "serve":
+        _, running_port = get_server_info()
+        if running_port is not None:
+            print(
+                (
+                    f"Lemonade Server is already running on port {running_port}\n"
+                    "Please stop the existing server before starting a new instance."
+                ),
+            )
+            sys.exit(ExitCodes.SERVER_ALREADY_RUNNING)
         serve(
             port=args.port,
             log_level=args.log_level,
@@ -396,6 +538,8 @@ def main():
         )
     elif args.command == "status":
         status()
+    elif args.command == "list":
+        list_models()
     elif args.command == "pull":
         pull(
             args.model,
@@ -408,6 +552,8 @@ def main():
         delete(args.model)
     elif args.command == "stop":
         stop()
+    elif args.command == "run":
+        run(args.model)
     elif args.command == "help" or not args.command:
         parser.print_help()
 

lemonade_server/model_manager.py

@@ -6,31 +6,13 @@ import huggingface_hub
 from importlib.metadata import distributions
 from lemonade_server.pydantic_models import PullConfig
 from lemonade.cache import DEFAULT_CACHE_DIR
+from lemonade.tools.llamacpp.utils import parse_checkpoint, download_gguf
 
 USER_MODELS_FILE = os.path.join(DEFAULT_CACHE_DIR, "user_models.json")
 
 
 class ModelManager:
 
-    @staticmethod
-    def parse_checkpoint(checkpoint: str) -> tuple[str, str | None]:
-        """
-        Parse a checkpoint string that may contain a variant separated by a colon.
-
-        For GGUF models, the format is "repository:variant" (e.g., "unsloth/Qwen3-0.6B-GGUF:Q4_0").
-        For other models, there is no variant.
-
-        Args:
-            checkpoint: The checkpoint string, potentially with variant
-
-        Returns:
-            tuple: (base_checkpoint, variant) where variant is None if no colon is present
-        """
-        if ":" in checkpoint:
-            base_checkpoint, variant = checkpoint.split(":", 1)
-            return base_checkpoint, variant
-        return checkpoint, None
-
     @property
     def supported_models(self) -> dict:
         """
@@ -98,7 +80,7 @@ class ModelManager:
         downloaded_models = {}
         downloaded_checkpoints = self.downloaded_hf_checkpoints
         for model in self.supported_models:
-            base_checkpoint = self.parse_checkpoint(
+            base_checkpoint = parse_checkpoint(
                 self.supported_models[model]["checkpoint"]
             )[0]
             if base_checkpoint in downloaded_checkpoints:
@@ -113,132 +95,6 @@ class ModelManager:
         """
         return self.filter_models_by_backend(self.downloaded_models)
 
-    def identify_gguf_models(
-        self, checkpoint: str, variant: str, mmproj: str
-    ) -> tuple[dict, list[str]]:
-        """
-        Identifies the GGUF model files in the repository that match the variant.
-        """
-
-        hint = """
-        The CHECKPOINT:VARIANT scheme is used to specify model files in Hugging Face repositories.
-
-        The VARIANT format can be one of several types:
-        1. Full filename: exact file to download
-        2. None/empty: gets the first .gguf file in the repository (excludes mmproj files)
-        3. Quantization variant: find a single file ending with the variant name (case insensitive)
-        4. Folder name: downloads all .gguf files in the folder that matches the variant name (case insensitive)
-        
-        Examples:
-        - "unsloth/Qwen3-8B-GGUF:qwen3.gguf" -> downloads "qwen3.gguf"
-        - "unsloth/Qwen3-30B-A3B-GGUF" -> downloads "Qwen3-30B-A3B-GGUF.gguf"
-        - "unsloth/Qwen3-8B-GGUF:Q4_1" -> downloads "Qwen3-8B-GGUF-Q4_1.gguf"
-        - "unsloth/Qwen3-30B-A3B-GGUF:Q4_0" -> downloads all files in "Q4_0/" folder
-        """
-
-        repo_files = huggingface_hub.list_repo_files(checkpoint)
-        sharded_files = []
-
-        # (case 1) If variant ends in .gguf, use it directly
-        if variant and variant.endswith(".gguf"):
-            variant_name = variant
-            if variant_name not in repo_files:
-                raise ValueError(
-                    f"File {variant} not found in Hugging Face repository {checkpoint}. {hint}"
-                )
-        # (case 2) If no variant is provided, get the first .gguf file in the repository
-        elif variant is None:
-            all_variants = [
-                f for f in repo_files if f.endswith(".gguf") and "mmproj" not in f
-            ]
-            if len(all_variants) == 0:
-                raise ValueError(
-                    f"No .gguf files found in Hugging Face repository {checkpoint}. {hint}"
-                )
-            variant_name = all_variants[0]
-        else:
-            # (case 3) Find a single file ending with the variant name (case insensitive)
-            end_with_variant = [
-                f
-                for f in repo_files
-                if f.lower().endswith(f"{variant}.gguf".lower())
-                and "mmproj" not in f.lower()
-            ]
-            if len(end_with_variant) == 1:
-                variant_name = end_with_variant[0]
-            elif len(end_with_variant) > 1:
-                raise ValueError(
-                    f"Multiple .gguf files found for variant {variant}, but only one is allowed. {hint}"
-                )
-            # (case 4) Check whether the variant corresponds to a folder with sharded files (case insensitive)
-            else:
-                sharded_files = [
-                    f
-                    for f in repo_files
-                    if f.endswith(".gguf")
-                    and f.lower().startswith(f"{variant}/".lower())
-                ]
-
-                if not sharded_files:
-                    raise ValueError(
-                        f"No .gguf files found for variant {variant}. {hint}"
-                    )
-
-                # Sort to ensure consistent ordering
-                sharded_files.sort()
-
-                # Use first file as primary (this is how llamacpp handles it)
-                variant_name = sharded_files[0]
-
-        core_files = {"variant": variant_name}
-
-        # If there is a mmproj file, add it to the patterns
-        if mmproj:
-            if mmproj not in repo_files:
-                raise ValueError(
-                    f"The provided mmproj file {mmproj} was not found in {checkpoint}."
-                )
-            core_files["mmproj"] = mmproj
-
-        return core_files, sharded_files
-
-    def download_gguf(self, model_config: PullConfig) -> dict:
-        """
-        Downloads the GGUF file for the given model configuration.
-
-        For sharded models, if the variant points to a folder (e.g. Q4_0), all files in that folder
-        will be downloaded but only the first file will be returned for loading.
-        """
-
-        # This code handles all cases by constructing the appropriate filename or pattern
-        checkpoint, variant = self.parse_checkpoint(model_config.checkpoint)
-
-        # Identify the GGUF model files in the repository that match the variant
-        core_files, sharded_files = self.identify_gguf_models(
-            checkpoint, variant, model_config.mmproj
-        )
-
-        # Download the files
-        snapshot_folder = huggingface_hub.snapshot_download(
-            repo_id=checkpoint,
-            allow_patterns=list(core_files.values()) + sharded_files,
-        )
-
-        # Ensure we downloaded all expected files
-        for file in list(core_files.values()) + sharded_files:
-            expected_path = os.path.join(snapshot_folder, file)
-            if not os.path.exists(expected_path):
-                raise ValueError(
-                    f"Hugging Face snapshot download for {model_config.checkpoint} "
-                    f"expected file {file} not found at {expected_path}"
-                )
-
-        # Return a dict of the full path of the core GGUF files
-        return {
-            file_name: os.path.join(snapshot_folder, file_path)
-            for file_name, file_path in core_files.items()
-        }
-
     def download_models(
         self,
         models: list[str],
@@ -317,7 +173,7 @@ class ModelManager:
             print(f"Downloading {model} ({checkpoint_to_download})")
 
             if "gguf" in checkpoint_to_download.lower():
-                self.download_gguf(gguf_model_config)
+                download_gguf(gguf_model_config.checkpoint, gguf_model_config.mmproj)
             else:
                 huggingface_hub.snapshot_download(repo_id=checkpoint_to_download)
 
@@ -373,7 +229,7 @@ class ModelManager:
         print(f"Deleting {model_name} ({checkpoint})")
 
         # Handle GGUF models that have the format "checkpoint:variant"
-        base_checkpoint = self.parse_checkpoint(checkpoint)[0]
+        base_checkpoint = parse_checkpoint(checkpoint)[0]
 
         try:
             # Get the local path using snapshot_download with local_files_only=True

lemonade_server/server_models.json

@@ -213,5 +213,16 @@
         "recipe": "llamacpp",
         "suggested": false,
         "labels": ["reranking"]
+    },
+    "Devstral-Small-2507-GGUF":{
+        "checkpoint": "mistralai/Devstral-Small-2507_gguf:Q4_K_M",
+        "recipe": "llamacpp",
+        "suggested": true
+    },
+    "Qwen2.5-Coder-32B-Instruct-GGUF": {
+        "checkpoint": "Qwen/Qwen2.5-Coder-32B-Instruct-GGUF:Q4_K_M",
+        "recipe": "llamacpp",
+        "suggested": true,  
+        "labels": ["reasoning"]
     }
 }

lemonade_sdk-8.0.4.dist-info/METADATA

@@ -1,176 +0,0 @@
-Metadata-Version: 2.4
-Name: lemonade-sdk
-Version: 8.0.4
-Summary: Lemonade SDK: Your LLM Aide for Validation and Deployment
-Author-email: lemonade@amd.com
-Requires-Python: >=3.10, <3.12
-Description-Content-Type: text/markdown
-License-File: LICENSE
-License-File: NOTICE.md
-Requires-Dist: invoke>=2.0.0
-Requires-Dist: onnx<1.18.0,>=1.11.0
-Requires-Dist: pyyaml>=5.4
-Requires-Dist: typeguard>=2.3.13
-Requires-Dist: packaging>=20.9
-Requires-Dist: numpy<2.0.0
-Requires-Dist: fasteners
-Requires-Dist: GitPython>=3.1.40
-Requires-Dist: psutil>=6.1.1
-Requires-Dist: wmi
-Requires-Dist: py-cpuinfo
-Requires-Dist: pytz
-Requires-Dist: zstandard
-Requires-Dist: fastapi
-Requires-Dist: uvicorn[standard]
-Requires-Dist: openai>=1.81.0
-Requires-Dist: transformers<=4.51.3
-Requires-Dist: jinja2
-Requires-Dist: tabulate
-Requires-Dist: sentencepiece
-Requires-Dist: huggingface-hub==0.33.0
-Provides-Extra: oga-hybrid
-Requires-Dist: onnx==1.16.1; extra == "oga-hybrid"
-Requires-Dist: numpy==1.26.4; extra == "oga-hybrid"
-Requires-Dist: protobuf>=6.30.1; extra == "oga-hybrid"
-Provides-Extra: oga-cpu
-Requires-Dist: onnxruntime-genai==0.8.2; extra == "oga-cpu"
-Requires-Dist: onnxruntime>=1.22.0; extra == "oga-cpu"
-Provides-Extra: dev
-Requires-Dist: torch>=2.6.0; extra == "dev"
-Requires-Dist: accelerate; extra == "dev"
-Requires-Dist: datasets; extra == "dev"
-Requires-Dist: pandas>=1.5.3; extra == "dev"
-Requires-Dist: matplotlib; extra == "dev"
-Requires-Dist: human-eval-windows==1.0.4; extra == "dev"
-Requires-Dist: lm-eval[api]; extra == "dev"
-Provides-Extra: oga-hybrid-minimal
-Requires-Dist: lemonade-sdk[oga-hybrid]; extra == "oga-hybrid-minimal"
-Provides-Extra: oga-cpu-minimal
-Requires-Dist: lemonade-sdk[oga-cpu]; extra == "oga-cpu-minimal"
-Provides-Extra: llm
-Requires-Dist: lemonade-sdk[dev]; extra == "llm"
-Provides-Extra: llm-oga-cpu
-Requires-Dist: lemonade-sdk[dev,oga-cpu]; extra == "llm-oga-cpu"
-Provides-Extra: llm-oga-igpu
-Requires-Dist: onnxruntime-genai-directml==0.6.0; extra == "llm-oga-igpu"
-Requires-Dist: onnxruntime-directml<1.22.0,>=1.19.0; extra == "llm-oga-igpu"
-Requires-Dist: transformers<4.45.0; extra == "llm-oga-igpu"
-Requires-Dist: lemonade-sdk[dev]; extra == "llm-oga-igpu"
-Provides-Extra: llm-oga-cuda
-Requires-Dist: onnxruntime-genai-cuda==0.8.2; extra == "llm-oga-cuda"
-Requires-Dist: onnxruntime-gpu>=1.22.0; extra == "llm-oga-cuda"
-Requires-Dist: transformers<=4.51.3; extra == "llm-oga-cuda"
-Requires-Dist: lemonade-sdk[dev]; extra == "llm-oga-cuda"
-Provides-Extra: llm-oga-npu
-Requires-Dist: onnx==1.16.0; extra == "llm-oga-npu"
-Requires-Dist: onnxruntime==1.18.0; extra == "llm-oga-npu"
-Requires-Dist: numpy==1.26.4; extra == "llm-oga-npu"
-Requires-Dist: protobuf>=6.30.1; extra == "llm-oga-npu"
-Requires-Dist: lemonade-sdk[dev]; extra == "llm-oga-npu"
-Provides-Extra: llm-oga-hybrid
-Requires-Dist: lemonade-sdk[dev,oga-hybrid]; extra == "llm-oga-hybrid"
-Provides-Extra: llm-oga-unified
-Requires-Dist: lemonade-sdk[llm-oga-hybrid]; extra == "llm-oga-unified"
-Dynamic: author-email
-Dynamic: description
-Dynamic: description-content-type
-Dynamic: license-file
-Dynamic: provides-extra
-Dynamic: requires-dist
-Dynamic: requires-python
-Dynamic: summary
-
-[![Lemonade tests](https://github.com/lemonade-sdk/lemonade/actions/workflows/test_lemonade.yml/badge.svg)](https://github.com/lemonade-sdk/lemonade/tree/main/test "Check out our tests")
-[![OS - Windows | Linux](https://img.shields.io/badge/OS-windows%20%7C%20linux-blue)](docs/README.md#installation "Check out our instructions")
-[![Made with Python](https://img.shields.io/badge/Python-3.10-blue?logo=python&logoColor=white)](docs/README.md#installation "Check out our instructions")
-
-## 🍋 Lemonade SDK: Quickly serve, benchmark and deploy LLMs
-
-The [Lemonade SDK](./docs/README.md) makes it easy to run Large Language Models (LLMs) on your PC. Our focus is using the best tools, such as neural processing units (NPUs) and Vulkan GPU acceleration, to maximize LLM speed and responsiveness.
-
-<div align="center">
-  <img src="https://download.amd.com/images/lemonade_640x480_1.gif" alt="Lemonade Demo" title="Lemonade in Action">
-</div>
-
-### Features
-
-The [Lemonade SDK](./docs/README.md) is comprised of the following:
-
-- 🌐 **[Lemonade Server](https://lemonade-server.ai/docs)**: A local LLM server for running ONNX and GGUF models using the OpenAI API standard. Install and enable your applications with NPU and GPU acceleration in minutes.
-- 🐍 **[Lemonade API](./docs/lemonade_api.md)**: High-level Python API to directly integrate Lemonade LLMs into Python applications.
-- 🖥️ **[Lemonade CLI](./docs/dev_cli/README.md)**: The `lemonade` CLI lets you mix-and-match LLMs (ONNX, GGUF, SafeTensors) with measurement tools to characterize your models on your hardware. The available tools are:
-  - Prompting with templates.
-  - Measuring accuracy with a variety of tests.
-  - Benchmarking to get the time-to-first-token and tokens per second.
-  - Profiling the memory utilization.
-
-### [Click here to get started with Lemonade.](./docs/README.md)
-
-### Supported Configurations
-
-Maximum LLM performance requires the right hardware accelerator with the right inference engine for your scenario. Lemonade supports the following configurations, while also making it easy to switch between them at runtime.
-
-<table border="1" cellpadding="6" cellspacing="0">
-  <thead>
-    <tr>
-      <th rowspan="2">Hardware</th>
-      <th colspan="3" align="center">🛠️ Engine Support</th>
-      <th colspan="2" align="center">🖥️ OS (x86/x64)</th>
-    </tr>
-    <tr>
-      <th align="center">OGA</th>
-      <th align="center">llamacpp</th>
-      <th align="center">HF</th>
-      <th align="center">Windows</th>
-      <th align="center">Linux</th>
-    </tr>
-  </thead>
-  <tbody>
-    <tr>
-      <td>🧠 CPU</td>
-      <td align="center">All platforms</td>
-      <td align="center">All platforms</td>
-      <td align="center">All platforms</td>
-      <td align="center">✅</td>
-      <td align="center">✅</td>
-    </tr>
-    <tr>
-      <td>🎮 GPU</td>
-      <td align="center">—</td>
-      <td align="center">Vulkan: All platforms<br><small>Focus:<br/>Ryzen™ AI 7000/8000/300<br/>Radeon™ 7000/9000</small></td>
-      <td align="center">—</td>
-      <td align="center">✅</td>
-      <td align="center">✅</td>
-    </tr>
-    <tr>
-      <td>🤖 NPU</td>
-      <td align="center">AMD Ryzen™ AI 300 series</td>
-      <td align="center">—</td>
-      <td align="center">—</td>
-      <td align="center">✅</td>
-      <td align="center">—</td>
-    </tr>
-  </tbody>
-</table>
-
-To learn more about the supported hardware and software, visit the documentation [here](./docs/README.md#software-and-hardware-overview).
-
-## Integrate Lemonade Server with Your Application
-
-Lemonade Server enables languages including Python, C++, Java, C#, Node.js, Go, Ruby, Rust, and PHP. For the full list and integration details, see [docs/server/README.md](./docs/server/README.md).
-
-## Contributing
-
-We are actively seeking collaborators from across the industry. If you would like to contribute to this project, please check out our [contribution guide](./docs/contribute.md).
-
-## Maintainers
-
-This project is sponsored by AMD. It is maintained by @danielholanda @jeremyfowers @ramkrishna @vgodsoe in equal measure. You can reach us by filing an [issue](https://github.com/lemonade-sdk/lemonade/issues) or email [lemonade@amd.com](mailto:lemonade@amd.com).
-
-## License
-
-This project is licensed under the [Apache 2.0 License](https://github.com/lemonade-sdk/lemonade/blob/main/LICENSE). Portions of the project are licensed as described in [NOTICE.md](./NOTICE.md).
-
-<!--This file was originally licensed under Apache 2.0. It has been modified.
-Modifications Copyright (c) 2025 AMD-->
-