Get up and running with Llama 2, Mistral, Gemma, and other large language models.
https://ollama.com
Get up and running with Llama 2, Mistral, Gemma, and other large language models.
You can find a list of models available for use at https://ollama.com/library.
- Links to science:machinelearning / ollama
- Has a link diff
- Download package
-
Checkout Package
osc -A https://api.opensuse.org checkout home:birdwatcher:machinelearning/ollama && cd $_ - Create Badge
Refresh
Source Files (show merged sources derived from linked package)
| Filename | Size | Changed |
|---|---|---|
| _link | 0000000131 131 Bytes | |
| _service | 0000000226 226 Bytes | |
| ollama-0.9.5.tar.gz | 0008396484 8.01 MB | |
| ollama-avoid-recomputing-special-vocabulary.patch | 0000012001 11.7 KB | |
| ollama-lib64-runner-path.patch | 0000002494 2.44 KB | |
| ollama-no-hip-redist.patch | 0000001349 1.32 KB | |
| ollama-user.conf | 0000000159 159 Bytes | |
| ollama.changes | 0000056191 54.9 KB | |
| ollama.service | 0000000232 232 Bytes | |
| ollama.spec | 0000006016 5.88 KB | |
| sysconfig.ollama | 0000001410 1.38 KB | |
| vendor.tar.zstd | 0005615103 5.35 MB |
Comments 1
AMD users, install this one. This is the one you want with proper ROCm support. Thanks birdwatcher for taking the time to properly make ollama and ROCm modules fully available on Tumbleweed.
For Radeon 780M, there's no need to modify anything to get it running. However, due to some limitations imposed by ROCm and perhaps Ollama as well, you might be limited to 4096 MiB of VRAM. My GTT says I have 7000+ MiB of memory. However, ROCm only detects 4096 and will crash on most 7B models even though I have set UMA in BIOS to 16G.
As a workaround, you need to set a custom GTT size as well as TTM pool and page pool sizes to use your whole available VRAM. Instructions here: https://www.reddit.com/r/ROCm/comments/1g3lnuj/rocm_apu_680m_and_gtt_memory_on_arch/
There's an open PR in Ollama's repository as well: https://github.com/ollama/ollama/pull/6282