llama.cpp
ggml-zdnn: fix #15414, activate FP16 and BF16 acceleration and incorrect zTensor free
#15839

Merged

Commits

ggml-zdnn: first rev to fix #15414

taronaeo committed 110 days ago
ggml-zdnn: trying to fix set_tensors without needing additional if guard

taronaeo committed 110 days ago
Revert "ggml-zdnn: trying to fix set_tensors without needing additional if guard"

taronaeo committed 110 days ago
ggml-zdnn: clean up set_tensor

taronaeo committed 110 days ago
ggml-zdnn: remove old code

taronaeo committed 110 days ago
ggml-zdnn: attempt at init bias in init_tensor

taronaeo committed 110 days ago
ggml-zdnn: add assert

taronaeo committed 110 days ago
ggml-zdnn: attempt at fixing sigsegv

taronaeo committed 110 days ago
ggml-zdnn: attempt at fixing double free

taronaeo committed 110 days ago
ggml-zdnn: fix incorrect ztensor free

taronaeo committed 110 days ago
ggml-zdnn: add comments to look back

taronaeo committed 110 days ago
ggml-zdnn: activate fp16 and bf16

taronaeo committed 110 days ago
ggml-zdnn: clean up matmul codepath

taronaeo committed 110 days ago
ggml-zdnn: fix compiler warnings

taronaeo committed 109 days ago
ggml-zdnn: fix more compiler warnings

taronaeo committed 109 days ago
ggml-zdnn: rm origtensor from .get_tensor

taronaeo committed 109 days ago
ggml-cpu: clean up s390x simd

taronaeo committed 109 days ago
Revert "ggml-cpu: clean up s390x simd"

taronaeo committed 109 days ago
docs: update ops and build for zDNN

taronaeo committed 109 days ago
ggml-zdnn: rm ggml_backend_zdnn_init from header files

taronaeo committed 106 days ago