Currently we are defaulting to /tmp and in some cases this means pulling
huge models to the /tmp directory which on a lot of linux platforms is a
tmpfs. /var/tmp tends to have much more space, so setting this as a
default is better.
Fixes: https://github.com/containers/ramalama/issues/1778
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
This was added when we didn't have good installation techniques
for mac. We have pipx which was not intuitive and a hacked
together shell script as an alternative. Now that we have brew and
uv integrated we don't need this code.
Signed-off-by: Eric Curtin <ecurtin@redhat.com>
The migration is run on each command to import all models
from the old store to the new one. It also removes the old
directories and creating the old structure is prevented.
Signed-off-by: Michael Engel <mengel@redhat.com>
Users should be able to list URLs and pull them to the host to
be processed by doc2rag command.
Also should force building of AI Data images to --network=none.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Summary by Sourcery
Add support for pulling models using http, https, and file URLs, enabling direct
execution of models from web or local sources. Update documentation and add tes
ts to cover the new functionality.
New Features:
Add support for pulling models using http, https, and file URLs, allowing mo
dels to be run directly from web sources or local files.
Enhancements:
Refactor the symlink creation process to use os.symlink directly instead of
run_cmd.
Documentation:
Update documentation to include new URL syntax support for http, https, and
file protocols, explaining how models can be pulled from these sources.
Tests:
Add system tests to verify the functionality of pulling models using file an
d https URLs, ensuring they can be listed and removed correctly.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
This is a redesign of run and serve to not run RamaLama within
the container but only the AI Model Runtime llama.cpp or vllm.
This should simplify the operation, although potentially causes
us risk with versions of Python on the host.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
When installing via pipx, requirement executables of RamaLama are
stored in the venvs/ramalama/bin directory. Have to modify the
system PATH so ramalama can find them.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Want to try and install via pipx and wrote a script to remove old
install. Also converted python install script back to shell, the
python experience in a new mac is just not nice in comparison to
bash.
Signed-off-by: Eric Curtin <ecurtin@redhat.com>
Currently PyPi and make install is creating a distinfo file which is
needed in order to retrieve the version of ramalama in use
inside of the container.
Also handle shortnames.conf file created via PyPi which is installed in
$HOME/.local/share/ramalama/shortnames.conf
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Add pyproject.toml, setup.cfg
Add new target make pypi.
fix up requirements.txt
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Update
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
