Releases: gpustack/llama-box
Releases · gpustack/llama-box
v0.0.28
chore: bump llama.cpp Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.27
refactor: lock only one slot for embedding at pooling Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.26
fix: cmake Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.25
chore: bump llama.cpp Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.24
chore: bump llama.cpp Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.23
chore: bump llama.cpp Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.22
feat: distinguish embedding-only model Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.21
fix: build musa (#4) * Revert "revert: "feat: support musa"" This reverts commit 775680409df4e9950f4cada1c39bf1d48943824d. * fix: build musa Signed-off-by: Xiaodong Ye <yeahdongcn@gmail.com> * Switch C/CXX to clang/clang++ for building MUSA target Signed-off-by: Xiaodong Ye <yeahdongcn@gmail.com> --------- Signed-off-by: Xiaodong Ye <yeahdongcn@gmail.com>
v0.0.20
fix: embedding Signed-off-by: thxCode <thxcode0824@gmail.com>
v0.0.19
chore: bump llama.cpp Signed-off-by: thxCode <thxcode0824@gmail.com>