feat: Add s_off as a parameter in the args struct

This may not be necessary, but it more closely mirrors the CUDA kernel

Branch: GraniteFourPerf

Signed-off-by: Gabe Goodhart <[email protected]>

Author: gabe-l-hart

ggml/src/ggml-metal/ggml-metal-impl.h

@@ -519,6 +519,7 @@ typedef struct {
     int64_t  n_group;
     int64_t  n_seq_tokens;
     int64_t  n_seqs;
+    int64_t  s_off;
     uint64_t nb01;
     uint64_t nb02;
     uint64_t nb03;