3bc7103d2e
This makes the weight buft detection in src/llama.cpp simpler. * convert : transpose Mamba-2 A, D and reshape SSM_NORM This breaks existing conversions of Mamba-2 models to avoid some reshapes. Not sure if it's a good idea, but it makes the graph slightly cleaner. * llama : more appropriate SSM_SCAN and SSM_CONV buft support checks |
||
|---|---|---|
| .. | ||
| ggml-alloc.h | ||
| ggml-amx.h | ||
| ggml-backend.h | ||
| ggml-blas.h | ||
| ggml-cann.h | ||
| ggml-cuda.h | ||
| ggml-kompute.h | ||
| ggml-metal.h | ||
| ggml-rpc.h | ||
| ggml-sycl.h | ||
| ggml-vulkan.h | ||
| ggml.h | ||