From 18bd78f1f038b2b69178d464184b937f7d2b15d6 Mon Sep 17 00:00:00 2001 From: oobabooga <112222186+oobabooga@users.noreply.github.com> Date: Tue, 10 Jun 2025 14:03:25 -0700 Subject: [PATCH] Make the llama.cpp prompt processing messages shorter --- modules/llama_cpp_server.py | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/modules/llama_cpp_server.py b/modules/llama_cpp_server.py index f0a72de8..a79e24e4 100644 --- a/modules/llama_cpp_server.py +++ b/modules/llama_cpp_server.py @@ -422,9 +422,17 @@ def filter_stderr_with_progress(process_stderr): if match: progress = float(match.group(1)) + + # Extract just the part from "prompt processing" onwards + prompt_processing_idx = line.find('prompt processing') + if prompt_processing_idx != -1: + display_line = line[prompt_processing_idx:] + else: + display_line = line # fallback to full line + # choose carriage return for in-progress or newline at completion end_char = '\r' if progress < 1.0 else '\n' - print(line, end=end_char, file=sys.stderr, flush=True) + print(display_line, end=end_char, file=sys.stderr, flush=True) last_was_progress = (progress < 1.0) # skip noise lines