Update app.py

app.py

@@ -4,11 +4,7 @@ import os
 os.putenv('PYTORCH_NVML_BASED_CUDA_CHECK','1')
 os.putenv('TORCH_LINALG_PREFER_CUSOLVER','1')
 alloc_conf_parts = [
-    'max_split_size_mb:128',
-    #'backend:cudaMallocAsync',
     'expandable_segments:True',
-    #'pinned_use_cuda_host_register:True', # May require careful testing; can sometimes be problematic if not fully supported or needed.
-    'pinned_num_register_threads:8',      # Specific to pinned memory.
     'pinned_use_background_threads:True'  # Specific to pinned memory.
 ]
 os.environ['PYTORCH_CUDA_ALLOC_CONF'] = ','.join(alloc_conf_parts)