What's changed in this repo:
Upgrade readme.md with more relevant info about GPU timeout and shared memory assigment.
What's changed from the main Ollama and its contributors
Relevant TLDR:
- Gemma 4 model various fixes and improvments.
- More claw things.
- Some relevant ROCm parity with CUDA
More about the release on https://github.com/ollama/ollama/releases/tag/v0.20.6
Full Changelog: v0.20.0...v0.20.6