GitLab

Prior refactoring passes accidentally removed the logic to bypass VRAM
checks for CPU loads.  This adds that back, along with test coverage.

This also fixes loaded map access in the unit test to be behind the mutex which was
likely the cause of various flakes in the tests.

* reload model if `num_gpu` changes

* dont reload on -1

* fix tests

Give the go routine a moment to deliver the expired event

This change adds support for multiple concurrent requests, as well as
loading multiple models by spawning multiple runners. The default
settings are currently set at 1 concurrent request per model and only 1
loaded model at a time, but these can be adjusted by setting
OLLAMA_NUM_PARALLEL and OLLAMA_MAX_LOADED_MODELS.

* terminate subprocess if receiving `SIGINT` or `SIGTERM` signals while model is loading

* use `unload` in signal handler

This is a quick fix to help users who are stuck on the "pull" step at
99%.

In the near future we're introducing a new registry client that
should/will hopefully be smarter. In the meantime, this should unblock
the users hitting issue #1736.

This should resolve a number of memory leak and stability defects by allowing
us to isolate llama.cpp in a separate process and shutdown when idle, and
gracefully restart if it has problems.  This also serves as a first step to be
able to run multiple copies to support multiple models concurrently.

count each layer independently when deciding gpu offloading

Co-authored-by: Michael Yang <mxyng@pm.me>

This uplevels the integration tests to run the server which can allow
testing an existing server, or a remote server.

This fixes issues with blob file names that contain ':' characters to be rejected by file systems that do not support them.

---------
Co-authored-by: Chris-AS1 <8493773+Chris-AS1@users.noreply.github.com>

The recent ROCm change partially removed idempotent
payloads, but the ggml-metal.metal file for mac was still
idempotent.  This finishes switching to always extract
the payloads, and now that idempotentcy is gone, the
version directory is no longer useful.

This refines where we extract the LLM libraries to by adding a new
OLLAMA_HOME env var, that defaults to `~/.ollama` The logic was already
idempotenent, so this should speed up startups after the first time a
new release is deployed.  It also cleans up after itself.

We now build only a single ROCm version (latest major) on both windows
and linux.  Given the large size of ROCms tensor files, we split the
dependency out.  It's bundled into the installer on windows, and a
separate download on windows.  The linux install script is now smart and
detects the presence of AMD GPUs and looks to see if rocm v6 is already
present, and if not, then downloads our dependency tar file.

For Linux discovery, we now use sysfs and check each GPU against what
ROCm supports so we can degrade to CPU gracefully instead of having
llama.cpp+rocm assert/crash on us.  For Windows, we now use go's windows
dynamic library loading logic to access the amdhip64.dll APIs to query
the GPU information.

instead of appending image tags, prepend them - this generally produces better results