Your message dated Tue, 14 Jul 2026 20:03:01 +0200
with message-id <20260714200301.29451869@crunch>
and subject line Re: Bug#1141599: llama.cpp-tools: seems to have no way to save
a transcript to a file
has caused the Debian Bug report #1141599,
regarding llama.cpp-tools: seems to have no way to save a transcript to a file
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
1141599: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1141599
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: llama.cpp-tools
Version: 9721+dfsg-1
Severity: normal
X-Debbugs-Cc: [email protected]
Hello and thanks a lot for packaging and maintaining llama.cpp in Debian!
I have tested llama-cli and llama-server. They allow me to interact with
a LLM obtained (for instance) from the Hugging Face model repository.
However, I haven't been able to find a way to save a transcript of the
conversation I had with the LLM.
I thought there could be a command (something like /save <file> in
llama-cli), but there seems to be nothing of the kind...
I also failed to find something in the web interface of llama-server.
I tried the good old UNIX way:
$ llama-cli -hf <GGUF-file> | tee chat.log
but the user prompts and the replies of the LLM are apparently not sent
to stdout, hence this approach also fails to work.
Manually copying and pasting chunks of the conversation works, but... how
tedious and error-prone!!!
If there's a better strategy, I think it should be documented (the man pages
do not seem to say anything about this topic).
If there's no better strategy, I really think this missing feature is a
bug that should be fixed: logging the conversations is important in
order to compare different LLMs and to keep track of a useful reply that
a LLM could (occasionally) give...
Please fix this bug and/or forward this bug report upstream, as appropriate.
Thanks for your time and dedication!
-- System Information:
Debian Release: forky/sid
APT prefers testing
APT policy: (800, 'testing'), (500, 'unstable')
Architecture: amd64 (x86_64)
Kernel: Linux 7.0.13+deb14-amd64 (SMP w/12 CPU threads; PREEMPT)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8),
LANGUAGE=en_US:en
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled
Versions of packages llama.cpp-tools depends on:
ii debconf [debconf-2.0] 1.5.92
ii libc6 2.42-17
ii libgcc-s1 16.1.0-2
ii libggml0 0.15.2-1
ii libllama0 9721+dfsg-1
ii libssl3t64 3.6.3-1
ii libstdc++6 16.1.0-2
ii systemd [systemd-tmpfiles] 261.1-2
ii wget 1.25.0-2
llama.cpp-tools recommends no packages.
Versions of packages llama.cpp-tools suggests:
pn llama.cpp-tools-webui-downloader <none>
-- debconf information:
llama.cpp-tools/path-error:
llama.cpp-tools/default-ui-custom-path:
llama.cpp-tools/default-ui: simplechat
--- End Message ---
--- Begin Message ---
Source: llama.cpp
Source-Version: 9951+dfsg-1
On Fri, 10 Jul 2026 16:59:50 +0200 Francesco Poli wrote:
[...]
> I brought it up with upstream.
>
> I obtained some suggestions to partially work around the missing
> feature and then the feature has also been apparently implemented!
>
> I am looking forward to testing the new option.
>
> Please keep an eye on the mentioned commit and close this Debian bug
> report, once the feature enters a version uploaded to Debian unstable.
[...]
Well, the commit appears to be included in llama.cpp/9951+dfsg-1 .
I tried the new `--output` command-line option and I can confirm that
it seems to meet my needs.
I think this Debian BTS bug report can be closed as fixed in
llama.cpp/9951+dfsg-1 .
I am doing so right now, in the hope that you agree.
Please feel free to reopen it, otherwise.
Bye!
--
http://www.inventati.org/frx/
There's not a second to spare! To the laboratory!
..................................................... Francesco Poli .
GnuPG key fpr == CA01 1147 9CD2 EFDF FB82 3925 3E1C 27E1 1F69 BFFE
pgpu0XKYJy7lX.pgp
Description: OpenPGP digital signature
--- End Message ---
