Fix async gpu

app.py

@@ -21,8 +21,8 @@ async def handle(system_input: str, user_input: str):
             buffers[i] += str(outputs[i])
 
         yield list(buffers) + ["", ""]
-    yield list(buffers) + [await generate_openllama(system_input, user_input),
-                           await generate_bloom(system_input, user_input)]
+    yield list(buffers) + [generate_openllama(system_input, user_input),
+                           generate_bloom(system_input, user_input)]
 
 
 with gr.Blocks() as demo:

generators.py

@@ -61,7 +61,7 @@ async def generate_llama2(system_input, user_input):
 
 
 @spaces.GPU(duration=120)
-async def generate_openllama(system_input, user_input):
+def generate_openllama(system_input, user_input):
     model_path = 'openlm-research/open_llama_3b_v2'
     tokenizer = LlamaTokenizer.from_pretrained(model_path)
     model = LlamaForCausalLM.from_pretrained(
@@ -75,7 +75,7 @@ async def generate_openllama(system_input, user_input):
 
 
 @spaces.GPU(duration=120)
-async def generate_bloom(system_input, user_input):
+def generate_bloom(system_input, user_input):
     model_path = 'bigscience/bloom-7b1'
     tokenizer = AutoTokenizer.from_pretrained(model_path)
     model = LlamaForCausalLM.from_pretrained(