From 610257b36e6455e28a316eb6d97c5c442403a5be Mon Sep 17 00:00:00 2001 From: Henk Date: Sat, 6 Aug 2022 16:47:04 +0200 Subject: [PATCH] Output Streaming on by Default --- aiserver.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/aiserver.py b/aiserver.py index c2a273b0..e841433f 100644 --- a/aiserver.py +++ b/aiserver.py @@ -364,7 +364,7 @@ class vars: lazy_load = True # Whether or not to use torch_lazy_loader.py for transformers models in order to reduce CPU memory usage use_colab_tpu = os.environ.get("COLAB_TPU_ADDR", "") != "" or os.environ.get("TPU_NAME", "") != "" # Whether or not we're in a Colab TPU instance or Kaggle TPU instance and are going to use the TPU rather than the CPU revision = None - output_streaming = False + output_streaming = True token_stream_queue = TokenStreamQueue() # Queue for the token streaming show_probs = False # Whether or not to show token probabilities