Update app.py

app.py

@@ -16,9 +16,9 @@ FILEPATH_PATTERN = "structured_data_doc.parquet"
 NUM_PROC = os.cpu_count()
 parent_path  = dirname(getcwd())
 
-cache_path = join(parent_path,'.cache')
-if not exists(cache_path):
-    makedirs(cache_path)
+temp_path  = join(parent_path,'temp')
+if not exists(temp_path ):
+    makedirs(temp_path )
 
 # Determine device based on GPU availability
 device = "cuda" if torch.cuda.is_available() else "cpu"
@@ -75,7 +75,7 @@ async def create_upload_file(file: UploadFile = File(...)):
     full_dataset = load_dataset("parquet",
                 data_files=FILEPATH_PATTERN,
                 split="train",
-                cache_path=cache_path,
+                cache_path=temp_path,
                 keep_in_memory=True,
                 num_proc=NUM_PROC*2)
     # Generate and append embeddings to the train split