Daniel Hiltgen
Daniel Hiltgen
Only use the new cgo tokenizer/detokenizer if we're using the new runners
Support local builds with customized CPU flags for both the CPU runner, and GPU runners. Some users want no vector flags in the GPU runners. Others want ~all the vector...
Complimentary to #6400 for the Go branch
This adds a new entrypoint into the ollama CLI to run the cgo built runner. On Mac arm64, this will have GPU support, but on all other platforms it will...
Advanced users are leveraging innosetup's /DIR switch to target an alternate location, but we get confused by things not existing in the LocalAppData dir. This also hardens the server path...
On windows ensure windows version define is properly set for rocm. Remove duplicate rocm arch flags. Resolve wildcards in the targets so parallel builds don't race. Use readlink to resolve...
Use cosine similarity to make the embeddings tests more robust
We still need a bit more soak time on the new Go server before we're ready to make this switch, so I'm marking this draft for now.
Prepare for removal of the submodule
This resolves a number of problems related to AMD multi-GPU setups on linux. The numeric IDs used by rocm are not the same as the numeric IDs exposed in sysfs...