q6k, q4_1 mat*mat

11 months ago · 2490977f89
parent a35f1ab784
commit 2490977f89
2 changed files with 5 additions and 1 deletions
--- a/gpt4all-backend/llama.cpp-mainline
+++ b/gpt4all-backend/llama.cpp-mainline
@ -1 +1 @@
-Subproject commit 3742085b0429cbe0ede49bcb9f891e4a5e25a724
+Subproject commit 500689ad356a81a471a7fb68cc70f7aee5a5f56e
--- a/gpt4all-backend/llama.cpp.cmake
+++ b/gpt4all-backend/llama.cpp.cmake
@ -242,6 +242,8 @@ if (LLAMA_KOMPUTE)
          kompute/op_mul_mat_mat_f16.comp
          kompute/op_mul_mat_mat_q8_0.comp
          kompute/op_mul_mat_mat_q4_0.comp
+          kompute/op_mul_mat_mat_q4_1.comp
+          kompute/op_mul_mat_mat_q6_k.comp
          kompute/op_mul_mat_f16.comp
          kompute/op_mul_mat_q8_0.comp
          kompute/op_mul_mat_q4_0.comp
@ -276,6 +278,8 @@ if (LLAMA_KOMPUTE)
          shaderop_mul_mat_mat_f16.h
          shaderop_mul_mat_mat_q8_0.h
          shaderop_mul_mat_mat_q4_0.h
+          shaderop_mul_mat_mat_q4_1.h
+          shaderop_mul_mat_mat_q6_k.h
          shaderop_mul_mat_f16.h
          shaderop_mul_mat_q8_0.h
          shaderop_mul_mat_q4_0.h