vae tiling improvements: encoding support and adaptative overlap #697

Job	Run time
macOS-latest-cmake	1m 36s
windows-latest-cmake (noavx, -DGGML_AVX=OFF -DGGML_AVX2=OFF -DGGML_FMA=OFF -DSD_BUILD_SHARED_LIBS...	2m 17s
ubuntu-latest-cmake	1m 21s
windows-latest-cmake (avx2, -DGGML_AVX2=ON -DSD_BUILD_SHARED_LIBS=ON)	1m 57s
windows-latest-cmake (avx, -DGGML_AVX2=OFF -DSD_BUILD_SHARED_LIBS=ON)	1m 25s
windows-latest-cmake (avx512, -DGGML_AVX512=ON -DSD_BUILD_SHARED_LIBS=ON)	1m 26s
windows-latest-cmake (cuda12, -DSD_CUBLAS=ON -DSD_BUILD_SHARED_LIBS=ON)	25m 53s
windows-latest-cmake (vulkan, -DSD_VULKAN=ON -DSD_BUILD_SHARED_LIBS=ON)	2m 40s
release	0s
	38m 35s

Provide feedback