immich机器学习基础和配置使用

[11/26/24 07:47:31] INFO Downloading detection model 'buffalo_l'. This may take a while. [11/26/24 07:47:32] WARNING Failed to load detection model 'buffalo_l'. Clearing cache. [11/26/24 07:47:32] WARNING Attempted to clear cache for model 'buffalo_l', but cache directory does not exist [11/26/24 07:47:32] WARNING Failed to load detection model 'buffalo_l'. Clearing cache. [11/26/24 07:47:32] WARNING Attempted to clear cache for model 'buffalo_l', but cache directory does not exist [11/26/24 07:47:33] INFO Downloading visual model 'ViT-B-32__openai'. This may take a while. [11/26/24 07:49:42] INFO Downloading detection model 'buffalo_l'. This may take a while. [11/26/24 07:49:42] WARNING Failed to load visual model 'ViT-B-32__openai'. Clearing cache. [11/26/24 07:49:42] WARNING Attempted to clear cache for model 'ViT-B-32__openai', but cache directory does not exist [11/26/24 07:51:54] INFO Downloading visual model 'ViT-B-32__openai'. This may take a while.

from huggingface_hub import snapshot_download
...

 def _download(self) -> None:
        ignore_patterns = [] if self.model_format == ModelFormat.ARMNN else ["*.armnn"]
        snapshot_download(
            f"immich-app/{clean_name(self.model_name)}",
            cache_dir=self.cache_dir,
            local_dir=self.cache_dir,
            ignore_patterns=ignore_patterns,
        )

# 确保安装了git和lfs，用如下命令验证是否安装了
git lfs install
# 如果没有安装

# 根据系统选择更新软件源，参考这里https://github.com/git-lfs/git-lfs/blob/main/INSTALLING.md
curl -s https://packagecloud.io/install/repositories/github/git-lfs/script.deb.sh | sudo bash
apt-get install git-lfs

# 安装完成之后直接下载对应得模型文件，XLM-Roberta-Large-Vit-B-16Plus和buffalo_l
git clone https://hf-mirror.com/immich-app/XLM-Roberta-Large-Vit-B-16Plus
git clone https://hf-mirror.com/immich-app/buffalo_l

 immich-machine-learning:
    container_name: immich_machine_learning
    # For hardware acceleration, add one of -[armnn, cuda, openvino] to the image tag.
    # Example tag: ${IMMICH_VERSION:-release}-cuda
    image: ghcr.io/immich-app/immich-machine-learning:${IMMICH_VERSION:-release}
    # extends: # uncomment this section for hardware acceleration - see https://immich.app/docs/features/ml-hardware-acceleration
    #   file: hwaccel.ml.yml
    #   service: cpu # set to one of [armnn, cuda, openvino, openvino-wsl] for accelerated inference - use the `-wsl` version for WSL2 where applicable
    volumes:
      - /data/immich/modelcache:/cache   # 修改这里，根据个人情况，我这里是/data/immich/modelcache
    env_file:
      - immich.env
    restart: always
    healthcheck:
      disable: false

.
└── modelcache
    ├── buffalo_l
    └── XLM-Roberta-Large-Vit-B-16Plus

    def download(self) -> None:
        if not self.cached:
            log.info(self.model_path)  # 增加得日志打印
            log.info(
                f"Downloading {self.model_type.replace('-', ' ')} model '{self.model_name}'. This may take a while."
            )
            self._download()

[11/28/24 17:10:53] INFO     /cache/clip/XLM-Roberta-Large-Vit-B-16Plus/visual/model.onnx   
...
[11/28/24 20:29:14] INFO     /cache/facial-recognition/buffalo_l/detection/model.onnx  

├── modelcache
│   ├── clip
│   │ └── XLM-Roberta-Large-Vit-B-16Plus
│   └── facial-recognition
│       └── buffalo_l