5 commits

Author	SHA1	Message	Date
Ettore Di Giacinto	2e2a0dffbc	fix(diffusers-hipblas): pin to rocm6.1 ... As per https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/3rd-party/pytorch-install.html Signed-off-by: Ettore Di Giacinto <mudler@users.noreply.github.com>	2024-08-09 22:36:10 +02:00
Ettore Di Giacinto	a507c13f8e	fix(diffusers): do not specify --pre as with pip ... drop --pre as it is not supported by `uv` Signed-off-by: Ettore Di Giacinto <mudler@users.noreply.github.com>	2024-08-08 22:21:05 +02:00
Ettore Di Giacinto	8317839ca5	fix(diffusers): use nightly rocm for hipblas builds (#3202) ... Signed-off-by: Ettore Di Giacinto <mudler@users.noreply.github.com>	2024-08-08 17:28:07 +02:00
Ettore Di Giacinto	61b5602111	fix(python): move accelerate and GPU-specific libs to build-type (#3194) ... Some of the dependencies in `requirements.txt`, even if generic, pulls down the line CUDA libraries. This changes moves mostly all GPU-specific libs to the build-type, and tries a safer approach. In `requirements.txt` now are listed only "first-level" dependencies, for instance, grpc, but libs-dependencies are moved down to the respective build-type `requirements.txt` to avoid any mixin. This should fix #2737 and #1592. Signed-off-by: Ettore Di Giacinto <mudler@localai.io>	2024-08-07 17:02:32 +02:00
cryptk	16433d2e8e	fix: install pytorch from proper index for hipblas builds (#2413) ... Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>	2024-05-26 18:05:52 +00:00