metal : use FA-vec kernel up to batch size 20

ggml-ci

ggml/src/ggml-metal/ggml-metal.m

@@ -4358,7 +4358,7 @@ static bool ggml_metal_encode_node(
                 // TODO: add vec kernels for (ne00%64 == 0) and maybe also for (ne00%32 == 0)
                 //       for now avoiding mainly to keep the number of templates/kernels a bit lower
                 //       these are now trivial to add after: https://github.com/ggml-org/llama.cpp/pull/12612
-                if (ne01 >= 4 || (ne00%128 != 0 && ne00 != 96 && ne00 != 192 && ne00 != 576)) {
+                if (ne01 >= 20 || (ne00%128 != 0 && ne00 != 96 && ne00 != 192 && ne00 != 576)) {
                     switch (src1->type) {
                         case GGML_TYPE_F16:
                             {

ggml/src/ggml-metal/ggml-metal.metal

@@ -3887,6 +3887,11 @@ kernel void kernel_flash_attn_ext_vec(
                 sm[tiisg] = pm[ic + tiisg];
             }
 
+            // skip -INF blocks
+            if (simd_max(sm[tiisg]) == -INFINITY) {
+                continue;
+            }
+
             // Q*K^T
             {
                 // each simdgroup processes 1 query and NE (NW/NL) head elements

tools/batched-bench/batched-bench.cpp

@@ -123,8 +123,8 @@ int main(int argc, char ** argv) {
 
                 common_batch_clear(batch);
 
-                for (int i = 0; i < pp; ++i) {
-                    for (int j = 0; j < (is_pp_shared ? 1 : pl); ++j) {
+                for (int j = 0; j < (is_pp_shared ? 1 : pl); ++j) {
+                    for (int i = 0; i < pp; ++i) {
                         common_batch_add(batch, 0, i, { j }, false);
                     }
                 }