Merge pull request #19 from lombokai/feature/downloader

Feature/downloader

Author: nunenuh

predict.py

@@ -8,7 +8,7 @@
 
 def predict_image(
     image_path: str = typer.Argument(help="image path", show_default=True),
-    model_path: str = typer.Argument("checkpoint/checkpoint_notebook.pth", help="model path (pth)", show_default=True),
+    model_path: str = typer.Argument(None, help="path to your model (pth)", show_default=True),
     device: str = typer.Argument("cpu", help="use cuda if your device has cuda", show_default=True)
 ):
     predictor = ImageRecognition(model_path=model_path, device=device)

src/onepiece_classify/infer/recognition.py

@@ -1,3 +1,5 @@
+import os
+import sys
 from pathlib import Path
 from typing import Dict, Optional, Tuple
 
@@ -7,13 +9,23 @@
 
 from onepiece_classify.models import image_recog
 from onepiece_classify.transforms import get_test_transforms
+from onepiece_classify.utils import downloader
 
 from .base import BaseInference
 
 
 class ImageRecognition(BaseInference):
-    def __init__(self, model_path: str, device: str):
-        self.model_path = Path(model_path)
+    def __init__(self, device: str, model_path=None):
+
+        path_to_save = str(self._get_cache_dir()) + "/model.pth"
+        if model_path is None:
+            downloader(path_to_save)
+            self.model_path = path_to_save
+        else:
+            self.model_path = Path(model_path)
+            if not self.model_path.exists():
+                raise FileNotFoundError("Model does not exist, check your model location and read README for more information")
+
         self.device = device
         self.class_dict = {
             0: "Ace",
@@ -44,6 +56,15 @@ def _build_model(self):
         model_backbone = image_recog(self.nclass)
         model_backbone.load_state_dict(state_dict)
         return model_backbone
+    
+    def _get_cache_dir(self):
+        if sys.platform.startswith("win"):
+            cache_dir = Path(os.getenv("LOCALAPPDATA", Path.home() / "AppData" / "Local")) / "OnepieceClassifyCache"
+        else:
+            cache_dir = Path.home() / ".cache" / "OnepieceClassifyCache"
+
+        cache_dir.mkdir(parents=True, exist_ok=True)
+        return cache_dir
 
     def pre_process(
         self, image: Optional[str | np.ndarray | Image.Image]

src/onepiece_classify/utils/__init__.py

@@ -0,0 +1 @@
+from .downloader import *

src/onepiece_classify/utils/downloader.py

@@ -0,0 +1,14 @@
+import gdown
+from pathlib import Path
+
+
+def downloader(output_file):
+    file_id = "1M1-1Hs198XDD6Xx-kSWLThv1elZBzJ0j"
+    prefix = 'https://drive.google.com/uc?/export=download&id='
+
+    url_download = prefix+file_id
+    
+    if not Path(output_file).exists():
+        print("Downloading...")
+        gdown.download(url_download, output_file)
+        print("Download Finish...")