custom server we go... soon (#3)

- custom server we go... soon (f9a9db52e528a51f33d30ce35074ef4e4bfa1aca)

app.py

@@ -46,7 +46,7 @@ log_queue = collections.deque(maxlen=1000) # Store last 1000 log messages
 gradio_handler = GradioLogHandler(log_queue)
 
 # Set root logger level to DEBUG to capture all messages from agents
-logging.getLogger().setLevel(logging.DEBUG)
+logging.getLogger().setLevel(logging.INFO)
 logging.getLogger().addHandler(gradio_handler)
 # --- End Gradio Log Handler ---
 
@@ -110,14 +110,14 @@ def register_model_with_metadata(model_id, model, preprocess, postprocess, class
     MODEL_REGISTRY[model_id] = entry
 
 # Load and register models (copied from app_mcp.py)
-image_processor_1 = AutoImageProcessor.from_pretrained(MODEL_PATHS["model_1"], use_fast=True)
-model_1 = Swinv2ForImageClassification.from_pretrained(MODEL_PATHS["model_1"]).to(device)
-clf_1 = pipeline(model=model_1, task="image-classification", image_processor=image_processor_1, device=device)
-register_model_with_metadata(
-    "model_1", clf_1, preprocess_resize_256, postprocess_pipeline, CLASS_NAMES["model_1"],
-    display_name="SWIN1", contributor="haywoodsloan", model_path=MODEL_PATHS["model_1"],
-    architecture="SwinV2", dataset="TBA"
-)
+# image_processor_1 = AutoImageProcessor.from_pretrained(MODEL_PATHS["model_1"], use_fast=True)
+# model_1 = Swinv2ForImageClassification.from_pretrained(MODEL_PATHS["model_1"]).to(device)
+# clf_1 = pipeline(model=model_1, task="image-classification", image_processor=image_processor_1, device=device)
+# register_model_with_metadata(
+#     "model_1", clf_1, preprocess_resize_256, postprocess_pipeline, CLASS_NAMES["model_1"],
+#     display_name="SWIN1", contributor="haywoodsloan", model_path=MODEL_PATHS["model_1"],
+#     architecture="SwinV2", dataset="TBA"
+# )
 
 # --- ONNX Quantized Model Example ---
 ONNX_QUANTIZED_MODEL_PATH = "./models/model_1_quantized.onnx"
@@ -714,7 +714,6 @@ with gr.Blocks() as app:
     demo.render()
     footer.render()
     
-app.queue(max_size=3)
+    app.unload(demo)
 
-if __name__ == "__main__":
-    app.launch(mcp_server=True) 
+app.queue(max_size=10, default_concurrency_limit=2).launch(mcp_server=True)