|
|
|
@ -51,12 +51,12 @@ def repl(
|
|
|
|
|
n_threads: Annotated[
|
|
|
|
|
int,
|
|
|
|
|
typer.Option("--n-threads", "-t", help="Number of threads to use for chatbot"),
|
|
|
|
|
] = 4,
|
|
|
|
|
] = None,
|
|
|
|
|
):
|
|
|
|
|
gpt4all_instance = GPT4All(model)
|
|
|
|
|
|
|
|
|
|
# if threads are passed, set them
|
|
|
|
|
if n_threads != 4:
|
|
|
|
|
if n_threads is not None:
|
|
|
|
|
num_threads = gpt4all_instance.model.thread_count()
|
|
|
|
|
print(f"\nAdjusted: {num_threads} →", end="")
|
|
|
|
|
|
|
|
|
@ -65,7 +65,8 @@ def repl(
|
|
|
|
|
|
|
|
|
|
num_threads = gpt4all_instance.model.thread_count()
|
|
|
|
|
print(f" {num_threads} threads", end="", flush=True)
|
|
|
|
|
|
|
|
|
|
else:
|
|
|
|
|
print(f"\nUsing {gpt4all_instance.model.thread_count()} threads", end="")
|
|
|
|
|
|
|
|
|
|
# overwrite _response_callback on model
|
|
|
|
|
gpt4all_instance.model._response_callback = _cli_override_response_callback
|
|
|
|
|