about summary refs log tree commit diff
path: root/pkgs/tools/misc/ollama
AgeCommit message (Collapse)AuthorFilesLines
2024-04-01ollama: 0.1.29 -> 0.1.30abysssol5-74/+57
2024-03-24ollama: fix a symlink and use a smaller rocm pathabysssol1-1/+5
2024-03-22ollama: Fix rocm runtime issuesUlrik Strid1-1/+2
2024-03-21ollama: Add needed rocm libraries to rocmPathUlrik Strid1-0/+5
2024-03-15ollama: add new patches and update patch hashesabysssol1-2/+5
2024-03-15ollama: 0.1.28 -> 0.1.29Pol Dellaiera1-6/+14
diff: https://github.com/ollama/ollama/compare/v0.1.28..v0.1.29 changelog: https://github.com/ollama/ollama/releases/tag/v0.1.29
2024-03-11ollama: add nixos test to `passthru.tests`abysssol1-2/+3
2024-03-06ollama: 0.1.27 -> 0.1.28R. Ryantm1-3/+3
2024-02-27ollama: 0.1.26 -> 0.1.27abysssol1-9/+8
2024-02-25ollama: test that rocm and cuda build successfullyabysssol1-0/+7
2024-02-25ollama: refactorabysssol1-125/+117
2024-02-25ollama: 0.1.24 -> 0.1.26abysssol1-5/+5
2024-02-16ollama: add support for gpu accelerationabysssol1-2/+106
2024-02-16ollama: 0.1.17 -> 0.1.24abysssol6-74/+103
Remove outdated patches and substitution of llama.cpp server. Patch source build script to remove git. Ollama's `go generate` script assumes the source is an initialized git repository, which it is not in nix. So, remove usage of `git` and replace it with `patch`. Also replace call to `g++` with `$CXX`. Run `go generate` to build dependencies using ollama's custom build script.
2023-12-19ollama: 0.1.11 -> 0.1.17happysalada1-4/+4
2023-11-22ollama: 0.1.7 -> 0.1.11Rafael Fernández López1-3/+3
2023-11-02ollama: 0.0.17 -> 0.1.7, use llama-cppEnno Richter3-11/+64
2023-09-01ollama: 0.0.15 -> 0.0.17R. Ryantm1-2/+2
2023-08-21ollama: 0.0.14 -> 0.0.15R. Ryantm1-3/+3
2023-08-16ollama: 0.0.13 -> 0.0.14R. Ryantm1-3/+3
2023-08-06ollama: 0.0.12 -> 0.0.13R. Ryantm1-3/+3
2023-08-01ollama: init at 0.0.12Mostly Void1-0/+35