Update app.py

app.py

@@ -12,7 +12,6 @@ os.environ['CUDA_LAUNCH_BLOCKING'] = '1'
 
 args = None
 lm_generation = None
-print(torch.__version__)
 
 def init_args():
     global args
@@ -22,7 +21,7 @@ def init_args():
     args.config_path = './config/llama_7b.json'
     args.spm_model_path = './model_file/tokenizer.model'
     args.batch_size = 1
-    args.seq_length = 512
+    args.seq_length = 1024
     args.world_size = 1
     args.use_int8 = False
     args.top_p = 0