#1126513 llama.cpp-tools: man pages for llama-cli and other tools are garbled

Package:
llama.cpp-tools
Source:
llama.cpp-tools
Description:
LLM inference in C/C++ - main utilities
Submitter:
Daniel Serpell
Date:
2026-01-27 20:19:13 UTC
Severity:
normal
#1126513#5
Date:
2026-01-27 19:59:09 UTC
From:
To:
Dear maintainer,

Current man pages for the tools are garbled, for example:
---------------------------------------------------------------------------
$ man llama-cli

LLAMA-CLI(1)                   User Commands                   LLAMA-CLI(1)

NAME
   llama-cli - llama-cli

DESCRIPTION
   load_backend:          loaded          CPU          backend         from
   /usr/lib/x86_64-linux-gnu/ggml/backends0/libggml-cpu-zen4.so -----  com‐
   mon params -----

   -h,     --help,  --usage                  print usage and exit --version
   show version and build info -cl,   --cache-list                     show
   list of models in  cache  --completion-bash                        print
   source-able   bash  completion  script  for  llama.cpp  --verbose-prompt
   print  a  verbose  prompt  before  generation   (default:   false)   -t,
   --threads  N                       number  of  CPU threads to use during
   generation (default: <num_cpus>)

          (env: LLAMA_ARG_THREADS)

---------------------------------------------------------------------------

It seems that the parsing of the 'llama-cli --help' command should redirect
standard error to /dev/null, and the parsing of the initial options should
be fixed.

Regards,
    Daniel.