Skip to content

Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

Commits on main

additional review comments
Jesse Gross committed 1 year ago
98272fb
ml/backend/ggml: use backend buffer type
Michael Yang committed 1 year ago
b27e8f3
comments
Michael Yang committed 1 year ago
45df786
ml/backend/ggml: clean up
Michael Yang committed 1 year ago
daaf42e
ml/backend/ggml: offload vision to cpu
Michael Yang committed 1 year ago
2dc60d4
ml/backend/ggml: handle tensor split
Michael Yang committed 1 year ago
b5312f3
26c2e0b
ml/backend/ggml: set cpu n_threads
Michael Yang committed 1 year ago
bf92088
kvcache: update tests
Michael Yang committed 1 year ago
58b9ec1
ml/backend/ggml: create tensor on specific backend
Michael Yang committed 1 year ago
7bae7fa
kvcache: create cache ctx per layer
Michael Yang committed 1 year ago
764e199
bfce55d
0682dae
1f6986e
4289c74
Better WantedBy declaration
‮rekcäH nitraM‮ committed 1 year ago
25248f4
ollamarunner: Improve multimodal input handling
Jesse Gross committed 1 year ago
a7e63b8
model: Don't unconditionally add special tokens
Jesse Gross committed 1 year ago
b70fc4d
Win: doc new rocm zip file (#9367)
Daniel Hiltgen committed 1 year ago
cae5d4d
ml/backend/ggml: consolidate system info logging
Michael Yang committed 1 year ago
05a01fd
docs: add granite-3.2 to the readme
aritra saha committed 1 year ago
8fe6f69
New engine: vision models and auto-fallback (#9113)
Daniel Hiltgen committed 1 year ago
1fdb351
fefbf8f
docker: use go version from go.mod
Michael Yang committed 1 year ago
b428ddd
fix: own lib/ollama directory
Michael Yang committed 1 year ago
ba7d312