Fix too stringent check on CUDA "fast copy" (can_be_transposed) condition and extend with one more case#17759

bssrdf wants to merge 3 commits intoggml-org:masterggml-org/llama.cpp:masterfrom

bssrdf:cuda-transposed-copy-fix-and-extensionbssrdf/llama.cpp:cuda-transposed-copy-fix-and-extensionCopy head branch name to clipboard