diff --git a/ggml/src/ggml-cuda/fattn-new-mma.cu b/ggml/src/ggml-cuda/fattn-new-mma.cu
index 21253e0b..72a5f2d6 100644
--- a/ggml/src/ggml-cuda/fattn-new-mma.cu
+++ b/ggml/src/ggml-cuda/fattn-new-mma.cu
@@ -1999,9 +1999,9 @@ static void ggml_cuda_flash_attn_ext_mma_f16_case(ggml_backend_cuda_context & ct
     const int nstages = cp_async_available(cc) ? c::nstages_target : 0;
 
     constexpr int ncols         = ncols1 * ncols2;
-    constexpr int ntiles        = ncols <= 8 ? 1 : 2; // Number of tiles per warp.
+    constexpr int ntiles        = ncols <= 8 && DKQ < 576 ? 1 : 2; // Number of tiles per warp.
     constexpr int cols_per_warp = ntiles * tile_B::I;
-    constexpr int nwarps_max_x  = ncols / cols_per_warp;
+    constexpr int nwarps_max_x  = (ncols + cols_per_warp - 1) / cols_per_warp;
     constexpr int nwarps_max_y  = c::nbatch_fa / tile_A::I;
     constexpr int nwarps        = nwarps_max_x*nwarps_max_y <= c::nwarps_max ? nwarps_max_x*nwarps_max_y : c::nwarps_max;
 
@@ -2063,6 +2063,10 @@ template <int DKQ, int DV, int ncols2>
 static void ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
     const ggml_tensor * Q = dst->src[0];
 
+    if constexpr (DKQ == 576 && ncols2 <= 4) {
+        ggml_cuda_flash_attn_ext_mma_f16_case<DKQ, DV, 4, ncols2>(ctx, dst);
+    } else {
+
     if constexpr (ncols2 <= 8) {
         if (Q->ne[1] <= 8/ncols2) {
             ggml_cuda_flash_attn_ext_mma_f16_case<DKQ, DV, 8/ncols2, ncols2>(ctx, dst);
@@ -2081,6 +2085,7 @@ static void ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1(ggml_backend_cuda_con
     }
 
     ggml_cuda_flash_attn_ext_mma_f16_case<DKQ, DV, 64/ncols2, ncols2>(ctx, dst);
+    }
 }
 
 template <int DKQ, int DV>
@@ -2156,8 +2161,15 @@ void ggml_cuda_flash_attn_ext_mma_new(ggml_backend_cuda_context & ctx, ggml_tens
         return;
     }
     GGML_ASSERT(Q->ne[0] == 576 && K->ne[0] == 576 && V->ne[0] == 512);
-    GGML_ASSERT(gqa_ratio % 16 == 0);
-    ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1<576, 512, 16>(ctx, dst);
+    if (gqa_ratio % 16 == 0) {
+        ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1<576, 512, 16>(ctx, dst);
+    } else if (gqa_ratio % 4 == 0) {
+        ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1<576, 512, 4>(ctx, dst);
+    } else {
+        GGML_ABORT("Unsupported GQA 576 x 512 case");
+    }
+    //GGML_ASSERT(gqa_ratio % 16 == 0);
+    //ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1<576, 512, 16>(ctx, dst);
 
     //switch (Q->ne[0]) {
     //    case 64:
diff --git a/ggml/src/ggml-cuda/fattn.cu b/ggml/src/ggml-cuda/fattn.cu
index ab3353db..ae5a3507 100644
--- a/ggml/src/ggml-cuda/fattn.cu
+++ b/ggml/src/ggml-cuda/fattn.cu
@@ -178,6 +178,10 @@ bool ggml_cuda_fattn_is_supported(ggml_backend_cuda_context & ctx, const ggml_te
     }
 
     if (new_mma_available(cc) && (Q->ne[0] == 576 || (K->ne[0] == 192 && V->ne[0] == 128 && mma_better_than_turing(cc)))) {
+        if (Q->ne[0] == 576) {
+            int gqa_ratio = Q->ne[2]/K->ne[2];
+            return (gqa_ratio % 4) == 0;
+        }
         return true;
     }
 
diff --git a/src/llama-hparams.cpp b/src/llama-hparams.cpp
index 4cef9236..79e0a1d0 100644
--- a/src/llama-hparams.cpp
+++ b/src/llama-hparams.cpp
@@ -750,7 +750,7 @@ void llm_load_hparams(
             {
                 if (hparams.n_head_kv() == 1) {
                     int n_nead_kv = hparams.n_gqa();
-                    if (n_nead_kv%16 != 0 || hparams.n_embd_head_k != 576 || hparams.n_embd_head_v != 512 ||
+                    if (n_nead_kv%4 != 0 || hparams.n_embd_head_k != 576 || hparams.n_embd_head_v != 512 ||
                         hparams.n_rot != 64) {
                         printf("==========================================================================\n");
                         printf("Detected incompatible DeepSeek model without a known way to fixc it.\n");
@@ -788,6 +788,7 @@ void llm_load_hparams(
 
                 switch (hparams.n_layer) {
                     case 27: model.type = e_model::MODEL_16B; break;
+                    case 47: model.type = e_model::MODEL_30B_A3B; break; // GLM-4.7-Flash
                     case 60: model.type = e_model::MODEL_236B; break;
                     case 61: model.type = e_model::MODEL_671B; break;
                     default: model.type = e_model::MODEL_UNKNOWN;