only gradio

app.py

@@ -1,4 +1,7 @@
 import os
+# Prevent libgomp crashes in Spaces
+os.environ["OMP_NUM_THREADS"] = "1"
+
 os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "0"
 os.environ["HF_HUB_DISABLE_PROGRESS_BARS"] = "1"
 os.environ["HTTP_PROXY"] = ""
@@ -283,23 +286,23 @@ def pose_transfer(source_image, test_pair_index):
         bsz = 1
 
         c_new = torch.cat([c_new[:bsz], c_new[:bsz], c_new[bsz:]])
-        down_block_additional_residuals = [torch.cat([torch.zeros_like(sample), sample, sample]).to(dtype=weight_dtype) \
-                                            for sample in down_block_additional_residuals]
-        up_block_additional_residuals = {k: torch.cat([torch.zeros_like(v), torch.zeros_like(v), v]).to(dtype=weight_dtype) \
-                                            for k, v in up_block_additional_residuals.items()}
+        down_block_additional_residuals = [torch.cat([torch.zeros_like(sample), sample, sample]).to(dtype=weight_dtype)
+                                           for sample in down_block_additional_residuals]
+        up_block_additional_residuals = {k: torch.cat([torch.zeros_like(v), torch.zeros_like(v), v]).to(dtype=weight_dtype)
+                                         for k, v in up_block_additional_residuals.items()}
 
         noise_scheduler.set_timesteps(cfg.TEST.NUM_INFERENCE_STEPS)
         for t in noise_scheduler.timesteps:
             inputs = torch.cat([noisy_latents, noisy_latents, noisy_latents], dim=0)
             inputs = noise_scheduler.scale_model_input(inputs, timestep=t)
             noise_pred = unet(sample=inputs, timestep=t, encoder_hidden_states=c_new,
-                down_block_additional_residuals=copy.deepcopy(down_block_additional_residuals),
-                up_block_additional_residuals=copy.deepcopy(up_block_additional_residuals))
+                              down_block_additional_residuals=copy.deepcopy(down_block_additional_residuals),
+                              up_block_additional_residuals=copy.deepcopy(up_block_additional_residuals))
 
             noise_pred_uc, noise_pred_down, noise_pred_full = noise_pred.chunk(3)
             noise_pred = noise_pred_uc + \
-                            cfg.TEST.DOWN_BLOCK_GUIDANCE_SCALE * (noise_pred_down - noise_pred_uc) + \
-                            cfg.TEST.FULL_GUIDANCE_SCALE * (noise_pred_full - noise_pred_down)
+                         cfg.TEST.DOWN_BLOCK_GUIDANCE_SCALE * (noise_pred_down - noise_pred_uc) + \
+                         cfg.TEST.FULL_GUIDANCE_SCALE * (noise_pred_full - noise_pred_down)
             noisy_latents = noise_scheduler.step(noise_pred, t, noisy_latents)[0]
 
         sampling_imgs = vae.decode(noisy_latents) * 0.5 + 0.5  # denormalize
@@ -342,52 +345,7 @@ def cancel_download_fn():
     return "Download cancelled."
 
 # ==============================
-# API Functions (using Gradio's built-in API)
-# ==============================
-def api_pose_transfer(source_image_base64: str, test_pair_index: int):
-    """API function for pose transfer."""
-    try:
-        if not model_ready:
-            return {
-                "output_image": "",
-                "success": False,
-                "message": "Models not ready. Please download models first."
-            }
-        
-        # Decode base64 image
-        image_data = base64.b64decode(source_image_base64)
-        source_image = Image.open(io.BytesIO(image_data)).convert("RGB")
-        
-        # Perform pose transfer
-        output_image = pose_transfer(source_image, test_pair_index)
-        
-        # Convert output to base64
-        buffered = io.BytesIO()
-        output_image.save(buffered, format="PNG")
-        output_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")
-        
-        return {
-            "output_image": output_base64,
-            "success": True,
-            "message": "Pose transfer successful"
-        }
-        
-    except Exception as e:
-        return {
-            "output_image": "",
-            "success": False,
-            "message": f"Error during pose transfer: {str(e)}"
-        }
-
-def api_status():
-    """API function to check model status."""
-    return {
-        "model_ready": model_ready,
-        "test_pairs_count": len(test_pairs) if test_pairs is not None else 0
-    }
-
-# ==============================
-# GRADIO UI
+# GRADIO UI ONLY (NO API ENDPOINTS)
 # ==============================
 def gradio_pose_transfer(source_image, test_pair_index):
     """Gradio interface for pose transfer."""
@@ -433,9 +391,6 @@ with gr.Blocks() as demo:
     with gr.Tab("Pose Transfer"):
         gr.Markdown("## Pose Transfer")
         
-        # Initialize test_pairs_count after models are loaded
-        test_pairs_count = gr.State(value=len(test_pairs) if test_pairs is not None else 0)
-        
         with gr.Row():
             with gr.Column():
                 source_image = gr.Image(label="Source Image", type="pil")
@@ -443,7 +398,7 @@ with gr.Blocks() as demo:
                     label="Test Pair Index", 
                     value=0, 
                     minimum=0, 
-                    maximum=4039  # Will be updated when models load
+                    maximum=4039
                 )
                 generate_btn = gr.Button("🚀 Generate Pose Transfer")
             
@@ -456,24 +411,6 @@ with gr.Blocks() as demo:
             inputs=[source_image, test_pair_index],
             outputs=[output_image, status_message]
         )
-    
-    # Add API endpoints using Gradio's built-in API
-    demo.add_api(
-        [
-            gr.Endpoint(
-                fn=api_pose_transfer,
-                inputs=[gr.Textbox(label="source_image"), gr.Number(label="test_pair_index")],
-                outputs=gr.JSON(label="response")
-            ),
-            gr.Endpoint(
-                fn=api_status,
-                inputs=[],
-                outputs=gr.JSON(label="status")
-            )
-        ],
-        title="Pose Transfer API",
-        description="API for pose transfer functionality"
-    )
 
 # ==============================
 # START APP
@@ -484,4 +421,4 @@ if __name__ == "__main__":
         model_ready = True
         initialize_models()
     
-    demo.launch(server_name="0.0.0.0", server_port=7860)
+    demo.launch(server_name="0.0.0.0", server_port=7860)