Update app.py

app.py

@@ -88,7 +88,7 @@ def text_to_3d(
     slat_guidance_strength: float,
     slat_sampling_steps: int,
     req: gr.Request,
-) -> Tuple[dict, str]:
+) -> str:
     """
     Convert an text prompt to a 3D model.
 
@@ -109,7 +109,7 @@ def text_to_3d(
     outputs = pipeline.run(
         prompt,
         seed=seed,
-        formats=["gaussian", "mesh"],
+        formats=["gaussian"],
         sparse_structure_sampler_params={
             "steps": ss_sampling_steps,
             "cfg_strength": ss_guidance_strength,
@@ -119,13 +119,13 @@ def text_to_3d(
             "cfg_strength": slat_guidance_strength,
         },
     )
-    state = pack_state(outputs['gaussian'][0], outputs['mesh'][0])
     ply_path = os.path.join(user_dir, 'point_cloud.ply')
     gaussian_data = outputs['gaussian'][0]
     with open(ply_path, "wb") as f:
         gaussian_data.save_ply(f)
+    del outputs, gaussian_data
     torch.cuda.empty_cache()
-    return state, ply_path
+    return ply_path
 
 
 def extract_glb(
@@ -229,7 +229,7 @@ with gr.Blocks(delete_cache=(600, 600)) as demo:
     ).then(
         text_to_3d,
         inputs=[text_prompt, seed, ss_guidance_strength, ss_sampling_steps, slat_guidance_strength, slat_sampling_steps],
-        outputs=[output_buf, video_output],
+        outputs=video_output,
     ).then(
         lambda: tuple([gr.Button(interactive=True), gr.Button(interactive=True)]),
         outputs=[extract_glb_btn, extract_gs_btn],
@@ -267,7 +267,7 @@ with gr.Blocks(delete_cache=(600, 600)) as demo:
 # Launch the Gradio app
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(description="Gradio app with command-line port argument")
-    parser.add_argument("--port", type=int, default=10000, help="Port to run the Gradio app on")
+    parser.add_argument("--port", type=int, default=8000, help="Port to run the Gradio app on")
     args = parser.parse_args()
     port = args.port
     pipeline = TrellisTextTo3DPipeline.from_pretrained("JeffreyXiang/TRELLIS-text-xlarge")