From 5f81588780339fd58c172591aa8273a198a20bca Mon Sep 17 00:00:00 2001
From: Zack Li <zhiyuanli0718@gmail.com>
Date: Tue, 10 Sep 2024 20:50:54 +0000
Subject: [PATCH] support ggml

---
 CMakeLists.txt                                | 12 ++++----
 Makefile                                      | 14 +++++-----
 examples/gguf-hash/CMakeLists.txt             |  2 +-
 examples/gguf/CMakeLists.txt                  |  2 +-
 examples/llava/CMakeLists.txt                 |  4 +--
 examples/rpc/CMakeLists.txt                   |  2 +-
 ggml/src/kompute                              |  1 -
 {ggml => ggml_llama}/.gitignore               |  0
 {ggml => ggml_llama}/CMakeLists.txt           | 10 +++----
 {ggml => ggml_llama}/cmake/FindSIMD.cmake     |  0
 {ggml => ggml_llama}/include/ggml-alloc.h     |  0
 {ggml => ggml_llama}/include/ggml-backend.h   |  0
 {ggml => ggml_llama}/include/ggml-blas.h      |  0
 {ggml => ggml_llama}/include/ggml-cann.h      |  0
 {ggml => ggml_llama}/include/ggml-cuda.h      |  0
 {ggml => ggml_llama}/include/ggml-kompute.h   |  0
 {ggml => ggml_llama}/include/ggml-metal.h     |  0
 {ggml => ggml_llama}/include/ggml-rpc.h       |  0
 {ggml => ggml_llama}/include/ggml-sycl.h      |  0
 {ggml => ggml_llama}/include/ggml-vulkan.h    |  0
 {ggml => ggml_llama}/include/ggml.h           |  0
 {ggml => ggml_llama}/src/CMakeLists.txt       | 28 +++++++++----------
 {ggml => ggml_llama}/src/ggml-aarch64.c       |  0
 {ggml => ggml_llama}/src/ggml-aarch64.h       |  0
 {ggml => ggml_llama}/src/ggml-alloc.c         |  0
 {ggml => ggml_llama}/src/ggml-backend-impl.h  |  0
 {ggml => ggml_llama}/src/ggml-backend.c       |  0
 {ggml => ggml_llama}/src/ggml-blas.cpp        |  0
 {ggml => ggml_llama}/src/ggml-cann.cpp        |  0
 .../src/ggml-cann/.clang-format               |  0
 {ggml => ggml_llama}/src/ggml-cann/Doxyfile   |  0
 .../src/ggml-cann/acl_tensor.cpp              |  0
 .../src/ggml-cann/acl_tensor.h                |  0
 .../src/ggml-cann/aclnn_ops.cpp               |  0
 .../src/ggml-cann/aclnn_ops.h                 |  0
 {ggml => ggml_llama}/src/ggml-cann/common.h   |  0
 .../src/ggml-cann/kernels/CMakeLists.txt      |  0
 .../src/ggml-cann/kernels/ascendc_kernels.h   |  0
 .../src/ggml-cann/kernels/dup.cpp             |  0
 .../src/ggml-cann/kernels/get_row_f16.cpp     |  0
 .../src/ggml-cann/kernels/get_row_f32.cpp     |  0
 .../src/ggml-cann/kernels/get_row_q4_0.cpp    |  0
 .../src/ggml-cann/kernels/get_row_q8_0.cpp    |  0
 .../ggml-cann/kernels/quantize_f16_q8_0.cpp   |  0
 .../ggml-cann/kernels/quantize_f32_q8_0.cpp   |  0
 .../kernels/quantize_float_to_q4_0.cpp        |  0
 {ggml => ggml_llama}/src/ggml-common.h        |  0
 {ggml => ggml_llama}/src/ggml-cuda.cu         |  0
 {ggml => ggml_llama}/src/ggml-cuda/acc.cu     |  0
 {ggml => ggml_llama}/src/ggml-cuda/acc.cuh    |  0
 {ggml => ggml_llama}/src/ggml-cuda/arange.cu  |  0
 {ggml => ggml_llama}/src/ggml-cuda/arange.cuh |  0
 {ggml => ggml_llama}/src/ggml-cuda/argsort.cu |  0
 .../src/ggml-cuda/argsort.cuh                 |  0
 .../src/ggml-cuda/binbcast.cu                 |  0
 .../src/ggml-cuda/binbcast.cuh                |  0
 {ggml => ggml_llama}/src/ggml-cuda/clamp.cu   |  0
 {ggml => ggml_llama}/src/ggml-cuda/clamp.cuh  |  0
 {ggml => ggml_llama}/src/ggml-cuda/common.cuh |  0
 {ggml => ggml_llama}/src/ggml-cuda/concat.cu  |  0
 {ggml => ggml_llama}/src/ggml-cuda/concat.cuh |  0
 .../src/ggml-cuda/conv-transpose-1d.cu        |  0
 .../src/ggml-cuda/conv-transpose-1d.cuh       |  0
 {ggml => ggml_llama}/src/ggml-cuda/convert.cu |  0
 .../src/ggml-cuda/convert.cuh                 |  0
 {ggml => ggml_llama}/src/ggml-cuda/cpy.cu     |  0
 {ggml => ggml_llama}/src/ggml-cuda/cpy.cuh    |  0
 .../src/ggml-cuda/cross-entropy-loss.cu       |  0
 .../src/ggml-cuda/cross-entropy-loss.cuh      |  0
 .../src/ggml-cuda/dequantize.cuh              |  0
 .../src/ggml-cuda/diagmask.cu                 |  0
 .../src/ggml-cuda/diagmask.cuh                |  0
 {ggml => ggml_llama}/src/ggml-cuda/dmmv.cu    |  0
 {ggml => ggml_llama}/src/ggml-cuda/dmmv.cuh   |  0
 .../src/ggml-cuda/fattn-common.cuh            |  0
 .../src/ggml-cuda/fattn-tile-f16.cu           |  0
 .../src/ggml-cuda/fattn-tile-f16.cuh          |  0
 .../src/ggml-cuda/fattn-tile-f32.cu           |  0
 .../src/ggml-cuda/fattn-tile-f32.cuh          |  0
 .../src/ggml-cuda/fattn-vec-f16.cuh           |  0
 .../src/ggml-cuda/fattn-vec-f32.cuh           |  0
 .../src/ggml-cuda/fattn-wmma-f16.cuh          |  0
 {ggml => ggml_llama}/src/ggml-cuda/fattn.cu   |  0
 {ggml => ggml_llama}/src/ggml-cuda/fattn.cuh  |  0
 {ggml => ggml_llama}/src/ggml-cuda/getrows.cu |  0
 .../src/ggml-cuda/getrows.cuh                 |  0
 {ggml => ggml_llama}/src/ggml-cuda/im2col.cu  |  0
 {ggml => ggml_llama}/src/ggml-cuda/im2col.cuh |  0
 {ggml => ggml_llama}/src/ggml-cuda/mma.cuh    |  0
 {ggml => ggml_llama}/src/ggml-cuda/mmq.cu     |  0
 {ggml => ggml_llama}/src/ggml-cuda/mmq.cuh    |  0
 {ggml => ggml_llama}/src/ggml-cuda/mmvq.cu    |  0
 {ggml => ggml_llama}/src/ggml-cuda/mmvq.cuh   |  0
 {ggml => ggml_llama}/src/ggml-cuda/norm.cu    |  0
 {ggml => ggml_llama}/src/ggml-cuda/norm.cuh   |  0
 {ggml => ggml_llama}/src/ggml-cuda/pad.cu     |  0
 {ggml => ggml_llama}/src/ggml-cuda/pad.cuh    |  0
 {ggml => ggml_llama}/src/ggml-cuda/pool2d.cu  |  0
 {ggml => ggml_llama}/src/ggml-cuda/pool2d.cuh |  0
 .../src/ggml-cuda/quantize.cu                 |  0
 .../src/ggml-cuda/quantize.cuh                |  0
 {ggml => ggml_llama}/src/ggml-cuda/rope.cu    |  0
 {ggml => ggml_llama}/src/ggml-cuda/rope.cuh   |  0
 {ggml => ggml_llama}/src/ggml-cuda/scale.cu   |  0
 {ggml => ggml_llama}/src/ggml-cuda/scale.cuh  |  0
 {ggml => ggml_llama}/src/ggml-cuda/softmax.cu |  0
 .../src/ggml-cuda/softmax.cuh                 |  0
 {ggml => ggml_llama}/src/ggml-cuda/sumrows.cu |  0
 .../src/ggml-cuda/sumrows.cuh                 |  0
 .../fattn-vec-f16-instance-hs128-f16-f16.cu   |  0
 .../fattn-vec-f16-instance-hs128-f16-q4_0.cu  |  0
 .../fattn-vec-f16-instance-hs128-f16-q4_1.cu  |  0
 .../fattn-vec-f16-instance-hs128-f16-q5_0.cu  |  0
 .../fattn-vec-f16-instance-hs128-f16-q5_1.cu  |  0
 .../fattn-vec-f16-instance-hs128-f16-q8_0.cu  |  0
 .../fattn-vec-f16-instance-hs128-q4_0-f16.cu  |  0
 .../fattn-vec-f16-instance-hs128-q4_0-q4_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_0-q4_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_0-q5_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_0-q5_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_0-q8_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_1-f16.cu  |  0
 .../fattn-vec-f16-instance-hs128-q4_1-q4_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_1-q4_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_1-q5_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_1-q5_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q4_1-q8_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_0-f16.cu  |  0
 .../fattn-vec-f16-instance-hs128-q5_0-q4_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_0-q4_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_0-q5_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_0-q5_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_0-q8_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_1-f16.cu  |  0
 .../fattn-vec-f16-instance-hs128-q5_1-q4_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_1-q4_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_1-q5_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_1-q5_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q5_1-q8_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q8_0-f16.cu  |  0
 .../fattn-vec-f16-instance-hs128-q8_0-q4_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q8_0-q4_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q8_0-q5_0.cu |  0
 .../fattn-vec-f16-instance-hs128-q8_0-q5_1.cu |  0
 .../fattn-vec-f16-instance-hs128-q8_0-q8_0.cu |  0
 .../fattn-vec-f16-instance-hs256-f16-f16.cu   |  0
 .../fattn-vec-f16-instance-hs64-f16-f16.cu    |  0
 .../fattn-vec-f16-instance-hs64-f16-q4_0.cu   |  0
 .../fattn-vec-f16-instance-hs64-f16-q4_1.cu   |  0
 .../fattn-vec-f16-instance-hs64-f16-q5_0.cu   |  0
 .../fattn-vec-f16-instance-hs64-f16-q5_1.cu   |  0
 .../fattn-vec-f16-instance-hs64-f16-q8_0.cu   |  0
 .../fattn-vec-f32-instance-hs128-f16-f16.cu   |  0
 .../fattn-vec-f32-instance-hs128-f16-q4_0.cu  |  0
 .../fattn-vec-f32-instance-hs128-f16-q4_1.cu  |  0
 .../fattn-vec-f32-instance-hs128-f16-q5_0.cu  |  0
 .../fattn-vec-f32-instance-hs128-f16-q5_1.cu  |  0
 .../fattn-vec-f32-instance-hs128-f16-q8_0.cu  |  0
 .../fattn-vec-f32-instance-hs128-q4_0-f16.cu  |  0
 .../fattn-vec-f32-instance-hs128-q4_0-q4_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_0-q4_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_0-q5_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_0-q5_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_0-q8_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_1-f16.cu  |  0
 .../fattn-vec-f32-instance-hs128-q4_1-q4_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_1-q4_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_1-q5_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_1-q5_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q4_1-q8_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_0-f16.cu  |  0
 .../fattn-vec-f32-instance-hs128-q5_0-q4_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_0-q4_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_0-q5_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_0-q5_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_0-q8_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_1-f16.cu  |  0
 .../fattn-vec-f32-instance-hs128-q5_1-q4_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_1-q4_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_1-q5_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_1-q5_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q5_1-q8_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q8_0-f16.cu  |  0
 .../fattn-vec-f32-instance-hs128-q8_0-q4_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q8_0-q4_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q8_0-q5_0.cu |  0
 .../fattn-vec-f32-instance-hs128-q8_0-q5_1.cu |  0
 .../fattn-vec-f32-instance-hs128-q8_0-q8_0.cu |  0
 .../fattn-vec-f32-instance-hs256-f16-f16.cu   |  0
 .../fattn-vec-f32-instance-hs64-f16-f16.cu    |  0
 .../fattn-vec-f32-instance-hs64-f16-q4_0.cu   |  0
 .../fattn-vec-f32-instance-hs64-f16-q4_1.cu   |  0
 .../fattn-vec-f32-instance-hs64-f16-q5_0.cu   |  0
 .../fattn-vec-f32-instance-hs64-f16-q5_1.cu   |  0
 .../fattn-vec-f32-instance-hs64-f16-q8_0.cu   |  0
 .../fattn-wmma-f16-instance-kqfloat-cpb16.cu  |  0
 .../fattn-wmma-f16-instance-kqfloat-cpb32.cu  |  0
 .../fattn-wmma-f16-instance-kqhalf-cpb16.cu   |  0
 .../fattn-wmma-f16-instance-kqhalf-cpb32.cu   |  0
 .../fattn-wmma-f16-instance-kqhalf-cpb8.cu    |  0
 .../template-instances/generate_cu_files.py   |  0
 .../template-instances/mmq-instance-iq1_s.cu  |  0
 .../template-instances/mmq-instance-iq2_s.cu  |  0
 .../template-instances/mmq-instance-iq2_xs.cu |  0
 .../mmq-instance-iq2_xxs.cu                   |  0
 .../template-instances/mmq-instance-iq3_s.cu  |  0
 .../mmq-instance-iq3_xxs.cu                   |  0
 .../template-instances/mmq-instance-iq4_nl.cu |  0
 .../template-instances/mmq-instance-iq4_xs.cu |  0
 .../template-instances/mmq-instance-q2_k.cu   |  0
 .../template-instances/mmq-instance-q3_k.cu   |  0
 .../template-instances/mmq-instance-q4_0.cu   |  0
 .../template-instances/mmq-instance-q4_1.cu   |  0
 .../template-instances/mmq-instance-q4_k.cu   |  0
 .../template-instances/mmq-instance-q5_0.cu   |  0
 .../template-instances/mmq-instance-q5_1.cu   |  0
 .../template-instances/mmq-instance-q5_k.cu   |  0
 .../template-instances/mmq-instance-q6_k.cu   |  0
 .../template-instances/mmq-instance-q8_0.cu   |  0
 {ggml => ggml_llama}/src/ggml-cuda/tsembd.cu  |  0
 {ggml => ggml_llama}/src/ggml-cuda/tsembd.cuh |  0
 {ggml => ggml_llama}/src/ggml-cuda/unary.cu   |  0
 {ggml => ggml_llama}/src/ggml-cuda/unary.cuh  |  0
 {ggml => ggml_llama}/src/ggml-cuda/upscale.cu |  0
 .../src/ggml-cuda/upscale.cuh                 |  0
 .../src/ggml-cuda/vecdotq.cuh                 |  0
 .../src/ggml-cuda/vendors/cuda.h              |  0
 .../src/ggml-cuda/vendors/hip.h               |  0
 .../src/ggml-cuda/vendors/musa.h              |  0
 {ggml => ggml_llama}/src/ggml-impl.h          |  0
 {ggml => ggml_llama}/src/ggml-kompute.cpp     |  0
 {ggml => ggml_llama}/src/ggml-metal.m         |  0
 {ggml => ggml_llama}/src/ggml-metal.metal     |  0
 {ggml => ggml_llama}/src/ggml-quants.c        |  0
 {ggml => ggml_llama}/src/ggml-quants.h        |  0
 {ggml => ggml_llama}/src/ggml-rpc.cpp         |  0
 {ggml => ggml_llama}/src/ggml-sycl.cpp        |  0
 .../src/ggml-sycl/backend.hpp                 |  0
 {ggml => ggml_llama}/src/ggml-sycl/common.cpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/common.hpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/concat.cpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/concat.hpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/conv.cpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/conv.hpp   |  0
 .../src/ggml-sycl/convert.cpp                 |  0
 .../src/ggml-sycl/convert.hpp                 |  0
 .../src/ggml-sycl/dequantize.hpp              |  0
 {ggml => ggml_llama}/src/ggml-sycl/dmmv.cpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/dmmv.hpp   |  0
 .../src/ggml-sycl/dpct/helper.hpp             |  0
 {ggml => ggml_llama}/src/ggml-sycl/gemm.hpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/im2col.cpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/im2col.hpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/mmq.cpp    |  0
 {ggml => ggml_llama}/src/ggml-sycl/mmq.hpp    |  0
 {ggml => ggml_llama}/src/ggml-sycl/mmvq.cpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/mmvq.hpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/norm.cpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/norm.hpp   |  0
 .../src/ggml-sycl/presets.hpp                 |  0
 {ggml => ggml_llama}/src/ggml-sycl/rope.cpp   |  0
 {ggml => ggml_llama}/src/ggml-sycl/rope.hpp   |  0
 .../src/ggml-sycl/softmax.cpp                 |  0
 .../src/ggml-sycl/softmax.hpp                 |  0
 {ggml => ggml_llama}/src/ggml-sycl/tsembd.cpp |  0
 {ggml => ggml_llama}/src/ggml-sycl/tsembd.hpp |  0
 .../src/ggml-sycl/vecdotq.hpp                 |  0
 {ggml => ggml_llama}/src/ggml-vulkan.cpp      |  0
 {ggml => ggml_llama}/src/ggml.c               |  0
 .../src/kompute-shaders/common.comp           |  0
 .../src/kompute-shaders/op_add.comp           |  0
 .../src/kompute-shaders/op_addrow.comp        |  0
 .../src/kompute-shaders/op_cpy_f16_f16.comp   |  0
 .../src/kompute-shaders/op_cpy_f16_f32.comp   |  0
 .../src/kompute-shaders/op_cpy_f32_f16.comp   |  0
 .../src/kompute-shaders/op_cpy_f32_f32.comp   |  0
 .../src/kompute-shaders/op_diagmask.comp      |  0
 .../src/kompute-shaders/op_gelu.comp          |  0
 .../src/kompute-shaders/op_getrows.comp       |  0
 .../src/kompute-shaders/op_getrows_f16.comp   |  0
 .../src/kompute-shaders/op_getrows_f32.comp   |  0
 .../src/kompute-shaders/op_getrows_q4_0.comp  |  0
 .../src/kompute-shaders/op_getrows_q4_1.comp  |  0
 .../src/kompute-shaders/op_getrows_q6_k.comp  |  0
 .../src/kompute-shaders/op_mul.comp           |  0
 .../src/kompute-shaders/op_mul_mat_f16.comp   |  0
 .../kompute-shaders/op_mul_mat_mat_f32.comp   |  0
 .../src/kompute-shaders/op_mul_mat_q4_0.comp  |  0
 .../src/kompute-shaders/op_mul_mat_q4_1.comp  |  0
 .../src/kompute-shaders/op_mul_mat_q6_k.comp  |  0
 .../src/kompute-shaders/op_mul_mat_q8_0.comp  |  0
 .../src/kompute-shaders/op_mul_mv_q_n.comp    |  0
 .../kompute-shaders/op_mul_mv_q_n_pre.comp    |  0
 .../src/kompute-shaders/op_norm.comp          |  0
 .../src/kompute-shaders/op_relu.comp          |  0
 .../src/kompute-shaders/op_rmsnorm.comp       |  0
 .../src/kompute-shaders/op_rope_f16.comp      |  0
 .../src/kompute-shaders/op_rope_f32.comp      |  0
 .../src/kompute-shaders/op_scale.comp         |  0
 .../src/kompute-shaders/op_scale_8.comp       |  0
 .../src/kompute-shaders/op_silu.comp          |  0
 .../src/kompute-shaders/op_softmax.comp       |  0
 .../src/kompute-shaders/rope_common.comp      |  0
 {ggml => ggml_llama}/src/llamafile/sgemm.cpp  |  0
 {ggml => ggml_llama}/src/llamafile/sgemm.h    |  0
 .../src/vulkan-shaders/CMakeLists.txt         |  0
 .../src/vulkan-shaders/acc.comp               |  0
 .../src/vulkan-shaders/add.comp               |  0
 .../src/vulkan-shaders/argsort.comp           |  0
 .../src/vulkan-shaders/clamp.comp             |  0
 .../src/vulkan-shaders/concat.comp            |  0
 .../src/vulkan-shaders/copy.comp              |  0
 .../src/vulkan-shaders/cos.comp               |  0
 .../src/vulkan-shaders/dequant_f32.comp       |  0
 .../src/vulkan-shaders/dequant_funcs.comp     |  0
 .../src/vulkan-shaders/dequant_head.comp      |  0
 .../src/vulkan-shaders/dequant_iq4_nl.comp    |  0
 .../src/vulkan-shaders/dequant_q2_k.comp      |  0
 .../src/vulkan-shaders/dequant_q3_k.comp      |  0
 .../src/vulkan-shaders/dequant_q4_0.comp      |  0
 .../src/vulkan-shaders/dequant_q4_1.comp      |  0
 .../src/vulkan-shaders/dequant_q4_k.comp      |  0
 .../src/vulkan-shaders/dequant_q5_0.comp      |  0
 .../src/vulkan-shaders/dequant_q5_1.comp      |  0
 .../src/vulkan-shaders/dequant_q5_k.comp      |  0
 .../src/vulkan-shaders/dequant_q6_k.comp      |  0
 .../src/vulkan-shaders/dequant_q8_0.comp      |  0
 .../src/vulkan-shaders/diag_mask_inf.comp     |  0
 .../src/vulkan-shaders/div.comp               |  0
 .../src/vulkan-shaders/gelu.comp              |  0
 .../src/vulkan-shaders/gelu_quick.comp        |  0
 .../vulkan-shaders/generic_binary_head.comp   |  0
 .../src/vulkan-shaders/generic_head.comp      |  0
 .../vulkan-shaders/generic_unary_head.comp    |  0
 .../src/vulkan-shaders/get_rows.comp          |  0
 .../src/vulkan-shaders/get_rows_quant.comp    |  0
 .../src/vulkan-shaders/group_norm.comp        |  0
 .../src/vulkan-shaders/im2col.comp            |  0
 .../src/vulkan-shaders/leaky_relu.comp        |  0
 .../src/vulkan-shaders/mul.comp               |  0
 .../mul_mat_split_k_reduce.comp               |  0
 .../src/vulkan-shaders/mul_mat_vec.comp       |  0
 .../src/vulkan-shaders/mul_mat_vec_base.comp  |  0
 .../src/vulkan-shaders/mul_mat_vec_nc.comp    |  0
 .../src/vulkan-shaders/mul_mat_vec_p021.comp  |  0
 .../src/vulkan-shaders/mul_mat_vec_q2_k.comp  |  0
 .../src/vulkan-shaders/mul_mat_vec_q3_k.comp  |  0
 .../src/vulkan-shaders/mul_mat_vec_q4_k.comp  |  0
 .../src/vulkan-shaders/mul_mat_vec_q5_k.comp  |  0
 .../src/vulkan-shaders/mul_mat_vec_q6_k.comp  |  0
 .../src/vulkan-shaders/mul_mm.comp            |  0
 .../src/vulkan-shaders/norm.comp              |  0
 .../src/vulkan-shaders/pad.comp               |  0
 .../src/vulkan-shaders/relu.comp              |  0
 .../src/vulkan-shaders/repeat.comp            |  0
 .../src/vulkan-shaders/rms_norm.comp          |  0
 .../src/vulkan-shaders/rope_head.comp         |  0
 .../src/vulkan-shaders/rope_neox.comp         |  0
 .../src/vulkan-shaders/rope_norm.comp         |  0
 .../src/vulkan-shaders/scale.comp             |  0
 .../src/vulkan-shaders/silu.comp              |  0
 .../src/vulkan-shaders/sin.comp               |  0
 .../src/vulkan-shaders/soft_max.comp          |  0
 .../src/vulkan-shaders/square.comp            |  0
 .../src/vulkan-shaders/sum_rows.comp          |  0
 .../src/vulkan-shaders/tanh.comp              |  0
 .../vulkan-shaders/timestep_embedding.comp    |  0
 .../src/vulkan-shaders/types.comp             |  0
 .../src/vulkan-shaders/upscale.comp           |  0
 .../src/vulkan-shaders/vulkan-shaders-gen.cpp |  0
 src/CMakeLists.txt                            |  2 +-
 371 files changed, 38 insertions(+), 39 deletions(-)
 delete mode 160000 ggml/src/kompute
 rename {ggml => ggml_llama}/.gitignore (100%)
 rename {ggml => ggml_llama}/CMakeLists.txt (96%)
 rename {ggml => ggml_llama}/cmake/FindSIMD.cmake (100%)
 rename {ggml => ggml_llama}/include/ggml-alloc.h (100%)
 rename {ggml => ggml_llama}/include/ggml-backend.h (100%)
 rename {ggml => ggml_llama}/include/ggml-blas.h (100%)
 rename {ggml => ggml_llama}/include/ggml-cann.h (100%)
 rename {ggml => ggml_llama}/include/ggml-cuda.h (100%)
 rename {ggml => ggml_llama}/include/ggml-kompute.h (100%)
 rename {ggml => ggml_llama}/include/ggml-metal.h (100%)
 rename {ggml => ggml_llama}/include/ggml-rpc.h (100%)
 rename {ggml => ggml_llama}/include/ggml-sycl.h (100%)
 rename {ggml => ggml_llama}/include/ggml-vulkan.h (100%)
 rename {ggml => ggml_llama}/include/ggml.h (100%)
 rename {ggml => ggml_llama}/src/CMakeLists.txt (98%)
 rename {ggml => ggml_llama}/src/ggml-aarch64.c (100%)
 rename {ggml => ggml_llama}/src/ggml-aarch64.h (100%)
 rename {ggml => ggml_llama}/src/ggml-alloc.c (100%)
 rename {ggml => ggml_llama}/src/ggml-backend-impl.h (100%)
 rename {ggml => ggml_llama}/src/ggml-backend.c (100%)
 rename {ggml => ggml_llama}/src/ggml-blas.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/.clang-format (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/Doxyfile (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/acl_tensor.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/acl_tensor.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/aclnn_ops.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/aclnn_ops.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/common.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/CMakeLists.txt (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/ascendc_kernels.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/dup.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/get_row_f16.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/get_row_f32.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/get_row_q4_0.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/get_row_q8_0.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/quantize_f16_q8_0.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/quantize_f32_q8_0.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-common.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/acc.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/acc.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/arange.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/arange.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/argsort.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/argsort.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/binbcast.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/binbcast.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/clamp.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/clamp.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/common.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/concat.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/concat.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/conv-transpose-1d.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/conv-transpose-1d.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/convert.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/convert.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/cpy.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/cpy.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/cross-entropy-loss.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/cross-entropy-loss.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/dequantize.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/diagmask.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/diagmask.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/dmmv.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/dmmv.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-common.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-tile-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-tile-f16.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-tile-f32.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-tile-f32.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-vec-f16.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-vec-f32.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn-wmma-f16.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/fattn.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/getrows.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/getrows.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/im2col.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/im2col.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/mma.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/mmq.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/mmq.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/mmvq.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/mmvq.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/norm.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/norm.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/pad.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/pad.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/pool2d.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/pool2d.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/quantize.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/quantize.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/rope.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/rope.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/scale.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/scale.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/softmax.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/softmax.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/sumrows.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/sumrows.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb32.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb16.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb32.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb8.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/generate_cu_files.py (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq2_xxs.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq3_s.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q4_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q4_k.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q5_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q5_1.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q6_k.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/template-instances/mmq-instance-q8_0.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/tsembd.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/tsembd.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/unary.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/unary.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/upscale.cu (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/upscale.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/vecdotq.cuh (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/vendors/cuda.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/vendors/hip.h (100%)
 rename {ggml => ggml_llama}/src/ggml-cuda/vendors/musa.h (100%)
 rename {ggml => ggml_llama}/src/ggml-impl.h (100%)
 rename {ggml => ggml_llama}/src/ggml-kompute.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-metal.m (100%)
 rename {ggml => ggml_llama}/src/ggml-metal.metal (100%)
 rename {ggml => ggml_llama}/src/ggml-quants.c (100%)
 rename {ggml => ggml_llama}/src/ggml-quants.h (100%)
 rename {ggml => ggml_llama}/src/ggml-rpc.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/backend.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/common.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/common.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/concat.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/concat.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/conv.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/conv.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/convert.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/convert.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/dequantize.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/dmmv.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/dmmv.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/dpct/helper.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/gemm.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/im2col.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/im2col.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/mmq.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/mmq.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/mmvq.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/mmvq.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/norm.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/norm.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/presets.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/rope.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/rope.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/softmax.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/softmax.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/tsembd.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/tsembd.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-sycl/vecdotq.hpp (100%)
 rename {ggml => ggml_llama}/src/ggml-vulkan.cpp (100%)
 rename {ggml => ggml_llama}/src/ggml.c (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/common.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_add.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_addrow.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_cpy_f16_f16.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_cpy_f16_f32.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_cpy_f32_f16.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_cpy_f32_f32.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_diagmask.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_gelu.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_getrows.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_getrows_f16.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_getrows_f32.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_getrows_q4_0.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_getrows_q4_1.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_getrows_q6_k.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mat_f16.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mat_mat_f32.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mat_q4_0.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mat_q4_1.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mat_q6_k.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mat_q8_0.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mv_q_n.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_mul_mv_q_n_pre.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_norm.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_relu.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_rmsnorm.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_rope_f16.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_rope_f32.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_scale.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_scale_8.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_silu.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/op_softmax.comp (100%)
 rename {ggml => ggml_llama}/src/kompute-shaders/rope_common.comp (100%)
 rename {ggml => ggml_llama}/src/llamafile/sgemm.cpp (100%)
 rename {ggml => ggml_llama}/src/llamafile/sgemm.h (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/CMakeLists.txt (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/acc.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/add.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/argsort.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/clamp.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/concat.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/copy.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/cos.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_f32.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_funcs.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_head.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_iq4_nl.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q2_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q3_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q4_0.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q4_1.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q4_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q5_0.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q5_1.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q5_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q6_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/dequant_q8_0.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/diag_mask_inf.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/div.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/gelu.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/gelu_quick.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/generic_binary_head.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/generic_head.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/generic_unary_head.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/get_rows.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/get_rows_quant.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/group_norm.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/im2col.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/leaky_relu.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_split_k_reduce.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_base.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_nc.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_p021.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_q2_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_q3_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_q4_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_q5_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mat_vec_q6_k.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/mul_mm.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/norm.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/pad.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/relu.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/repeat.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/rms_norm.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/rope_head.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/rope_neox.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/rope_norm.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/scale.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/silu.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/sin.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/soft_max.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/square.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/sum_rows.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/tanh.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/timestep_embedding.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/types.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/upscale.comp (100%)
 rename {ggml => ggml_llama}/src/vulkan-shaders/vulkan-shaders-gen.cpp (100%)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index a31320635..cf4a48f43 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -112,9 +112,9 @@ llama_option_depr(WARNING     LLAMA_CANN                GGML_CANN)
 # build the library
 #
 
-if (NOT TARGET ggml)
-    add_subdirectory(ggml)
-    # ... otherwise assume ggml is added by a parent CMakeLists.txt
+if (NOT TARGET ggml_llama)
+    add_subdirectory(ggml_llama)
+    # ... otherwise assume ggml_llama is added by a parent CMakeLists.txt
 endif()
 add_subdirectory(src)
 
@@ -139,11 +139,11 @@ set(LLAMA_BIN_INSTALL_DIR     ${CMAKE_INSTALL_BINDIR}     CACHE PATH "Location o
 # determining _precisely_ which defines are necessary for the llama-config
 # package.
 #
-get_target_property(GGML_DIRECTORY ggml SOURCE_DIR)
+get_target_property(GGML_DIRECTORY ggml_llama SOURCE_DIR)
 get_directory_property(GGML_DIR_DEFINES DIRECTORY ${GGML_DIRECTORY} COMPILE_DEFINITIONS)
-get_target_property(GGML_TARGET_DEFINES ggml COMPILE_DEFINITIONS)
+get_target_property(GGML_TARGET_DEFINES ggml_llama COMPILE_DEFINITIONS)
 set(GGML_TRANSIENT_DEFINES ${GGML_TARGET_DEFINES} ${GGML_DIR_DEFINES})
-get_target_property(GGML_LINK_LIBRARIES ggml LINK_LIBRARIES)
+get_target_property(GGML_LINK_LIBRARIES ggml_llama LINK_LIBRARIES)
 
 set_target_properties(llama PROPERTIES PUBLIC_HEADER ${CMAKE_CURRENT_SOURCE_DIR}/include/llama.h)
 install(TARGETS llama LIBRARY PUBLIC_HEADER)
diff --git a/Makefile b/Makefile
index 332496cfc..90364d0b9 100644
--- a/Makefile
+++ b/Makefile
@@ -876,7 +876,7 @@ ifdef GGML_METAL_NDEBUG
 endif
 ifdef GGML_METAL_EMBED_LIBRARY
 	MK_CPPFLAGS += -DGGML_METAL_EMBED_LIBRARY
-	OBJ_GGML   += ggml/src/ggml-metal-embed.o
+	OBJ_GGML   += ggml/src/ggml-metal-embed-llama.o
 endif
 endif # GGML_METAL
 
@@ -888,20 +888,20 @@ ggml/src/ggml-metal.o: \
 	$(CC) $(CFLAGS) -c $< -o $@
 
 ifdef GGML_METAL_EMBED_LIBRARY
-ggml/src/ggml-metal-embed.o: \
+ggml/src/ggml-metal-embed-llama.o: \
 	ggml/src/ggml-metal.metal \
 	ggml/src/ggml-common.h
 	@echo "Embedding Metal library"
-	@sed -e '/#include "ggml-common.h"/r ggml/src/ggml-common.h' -e '/#include "ggml-common.h"/d' < ggml/src/ggml-metal.metal > ggml/src/ggml-metal-embed.metal
+	@sed -e '/#include "ggml-common.h"/r ggml/src/ggml-common.h' -e '/#include "ggml-common.h"/d' < ggml/src/ggml-metal.metal > ggml/src/ggml-metal-embed-llama.metal
 	$(eval TEMP_ASSEMBLY=$(shell mktemp -d))
 	@echo ".section __DATA, __ggml_metallib"            >  $(TEMP_ASSEMBLY)/ggml-metal-embed.s
 	@echo ".globl _ggml_metallib_start"                 >> $(TEMP_ASSEMBLY)/ggml-metal-embed.s
 	@echo "_ggml_metallib_start:"                       >> $(TEMP_ASSEMBLY)/ggml-metal-embed.s
-	@echo ".incbin \"ggml/src/ggml-metal-embed.metal\"" >> $(TEMP_ASSEMBLY)/ggml-metal-embed.s
+	@echo ".incbin \"ggml/src/ggml-metal-embed-llama.metal\"" >> $(TEMP_ASSEMBLY)/ggml-metal-embed.s
 	@echo ".globl _ggml_metallib_end"                   >> $(TEMP_ASSEMBLY)/ggml-metal-embed.s
 	@echo "_ggml_metallib_end:"                         >> $(TEMP_ASSEMBLY)/ggml-metal-embed.s
-	$(CC) $(CFLAGS) -c $(TEMP_ASSEMBLY)/ggml-metal-embed.s -o $@
-	@rm -f ${TEMP_ASSEMBLY}/ggml-metal-embed.s
+	$(CC) $(CFLAGS) -c $(TEMP_ASSEMBLY)/ggml-metal-embed-llama.s -o $@
+	@rm -f ${TEMP_ASSEMBLY}/ggml-metal-embed-llama.s
 	@rmdir ${TEMP_ASSEMBLY}
 endif
 endif # GGML_METAL
@@ -1213,7 +1213,7 @@ clean:
 	rm -vrf ggml/src/*.o
 	rm -rvf ggml/src/llamafile/*.o
 	rm -rvf common/build-info.cpp
-	rm -vrf ggml/src/ggml-metal-embed.metal
+	rm -vrf ggml/src/ggml-metal-embed-llama.metal
 	rm -vrf ggml/src/ggml-cuda/*.o
 	rm -vrf ggml/src/ggml-cuda/template-instances/*.o
 	rm -rvf $(BUILD_TARGETS)
diff --git a/examples/gguf-hash/CMakeLists.txt b/examples/gguf-hash/CMakeLists.txt
index 633f45535..6ef93bcfb 100644
--- a/examples/gguf-hash/CMakeLists.txt
+++ b/examples/gguf-hash/CMakeLists.txt
@@ -11,5 +11,5 @@ target_link_libraries(${TARGET} PRIVATE sha1)
 add_library(sha256 OBJECT deps/sha256/sha256.c deps/sha256/sha256.h)
 target_link_libraries(${TARGET} PRIVATE sha256)
 
-target_link_libraries(${TARGET} PRIVATE ggml ${CMAKE_THREAD_LIBS_INIT})
+target_link_libraries(${TARGET} PRIVATE ggml_llama ${CMAKE_THREAD_LIBS_INIT})
 target_compile_features(${TARGET} PRIVATE cxx_std_11)
diff --git a/examples/gguf/CMakeLists.txt b/examples/gguf/CMakeLists.txt
index a9569b411..dcdbb0f14 100644
--- a/examples/gguf/CMakeLists.txt
+++ b/examples/gguf/CMakeLists.txt
@@ -1,5 +1,5 @@
 set(TARGET llama-gguf)
 add_executable(${TARGET} gguf.cpp)
 install(TARGETS ${TARGET} RUNTIME)
-target_link_libraries(${TARGET} PRIVATE ggml ${CMAKE_THREAD_LIBS_INIT})
+target_link_libraries(${TARGET} PRIVATE ggml_llama ${CMAKE_THREAD_LIBS_INIT})
 target_compile_features(${TARGET} PRIVATE cxx_std_11)
diff --git a/examples/llava/CMakeLists.txt b/examples/llava/CMakeLists.txt
index bbf5fec58..ab992e047 100644
--- a/examples/llava/CMakeLists.txt
+++ b/examples/llava/CMakeLists.txt
@@ -5,7 +5,7 @@ add_library(llava OBJECT
             clip.h
             )
 
-target_link_libraries(llava PRIVATE ggml llama ${CMAKE_THREAD_LIBS_INIT})
+target_link_libraries(llava PRIVATE ggml_llama llama ${CMAKE_THREAD_LIBS_INIT})
 
 target_include_directories(llava PUBLIC .)
 target_include_directories(llava PUBLIC ../..)
@@ -18,7 +18,7 @@ if (BUILD_SHARED_LIBS)
     set_target_properties(llava PROPERTIES POSITION_INDEPENDENT_CODE ON)
     target_compile_definitions(llava PRIVATE LLAMA_SHARED LLAMA_BUILD)
     add_library(llava_shared SHARED $<TARGET_OBJECTS:llava>)
-    target_link_libraries(llava_shared PRIVATE ggml llama ${CMAKE_THREAD_LIBS_INIT})
+    target_link_libraries(llava_shared PRIVATE ggml_llama llama ${CMAKE_THREAD_LIBS_INIT})
     install(TARGETS llava_shared LIBRARY)
 endif()
 
diff --git a/examples/rpc/CMakeLists.txt b/examples/rpc/CMakeLists.txt
index ae48fb98d..aa1f0175b 100644
--- a/examples/rpc/CMakeLists.txt
+++ b/examples/rpc/CMakeLists.txt
@@ -1,2 +1,2 @@
 add_executable(rpc-server rpc-server.cpp)
-target_link_libraries(rpc-server PRIVATE ggml llama)
+target_link_libraries(rpc-server PRIVATE ggml_llama llama)
diff --git a/ggml/src/kompute b/ggml/src/kompute
deleted file mode 160000
index 4565194ed..000000000
--- a/ggml/src/kompute
+++ /dev/null
@@ -1 +0,0 @@
-Subproject commit 4565194ed7c32d1d2efa32ceab4d3c6cae006306
diff --git a/ggml/.gitignore b/ggml_llama/.gitignore
similarity index 100%
rename from ggml/.gitignore
rename to ggml_llama/.gitignore
diff --git a/ggml/CMakeLists.txt b/ggml_llama/CMakeLists.txt
similarity index 96%
rename from ggml/CMakeLists.txt
rename to ggml_llama/CMakeLists.txt
index cc1685884..df37a54cc 100644
--- a/ggml/CMakeLists.txt
+++ b/ggml_llama/CMakeLists.txt
@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.14) # for add_link_options and implicit target directories.
-project("ggml" C CXX)
+project("ggml_llama" C CXX)
 include(CheckIncludeFileCXX)
 
 set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
@@ -216,14 +216,14 @@ set(GGML_PUBLIC_HEADERS
     include/ggml-sycl.h
     include/ggml-vulkan.h)
 
-set_target_properties(ggml PROPERTIES PUBLIC_HEADER "${GGML_PUBLIC_HEADERS}")
+set_target_properties(ggml_llama PROPERTIES PUBLIC_HEADER "${GGML_PUBLIC_HEADERS}")
 #if (GGML_METAL)
-#    set_target_properties(ggml PROPERTIES RESOURCE "${CMAKE_CURRENT_SOURCE_DIR}/src/ggml-metal.metal")
+#    set_target_properties(ggml_llama PROPERTIES RESOURCE "${CMAKE_CURRENT_SOURCE_DIR}/src/ggml-metal.metal")
 #endif()
-install(TARGETS ggml PUBLIC_HEADER)
+install(TARGETS ggml_llama PUBLIC_HEADER)
 
 if (BUILD_SHARED_LIBS)
-    install(TARGETS ggml LIBRARY)
+    install(TARGETS ggml_llama LIBRARY)
 endif()
 
 if (GGML_METAL)
diff --git a/ggml/cmake/FindSIMD.cmake b/ggml_llama/cmake/FindSIMD.cmake
similarity index 100%
rename from ggml/cmake/FindSIMD.cmake
rename to ggml_llama/cmake/FindSIMD.cmake
diff --git a/ggml/include/ggml-alloc.h b/ggml_llama/include/ggml-alloc.h
similarity index 100%
rename from ggml/include/ggml-alloc.h
rename to ggml_llama/include/ggml-alloc.h
diff --git a/ggml/include/ggml-backend.h b/ggml_llama/include/ggml-backend.h
similarity index 100%
rename from ggml/include/ggml-backend.h
rename to ggml_llama/include/ggml-backend.h
diff --git a/ggml/include/ggml-blas.h b/ggml_llama/include/ggml-blas.h
similarity index 100%
rename from ggml/include/ggml-blas.h
rename to ggml_llama/include/ggml-blas.h
diff --git a/ggml/include/ggml-cann.h b/ggml_llama/include/ggml-cann.h
similarity index 100%
rename from ggml/include/ggml-cann.h
rename to ggml_llama/include/ggml-cann.h
diff --git a/ggml/include/ggml-cuda.h b/ggml_llama/include/ggml-cuda.h
similarity index 100%
rename from ggml/include/ggml-cuda.h
rename to ggml_llama/include/ggml-cuda.h
diff --git a/ggml/include/ggml-kompute.h b/ggml_llama/include/ggml-kompute.h
similarity index 100%
rename from ggml/include/ggml-kompute.h
rename to ggml_llama/include/ggml-kompute.h
diff --git a/ggml/include/ggml-metal.h b/ggml_llama/include/ggml-metal.h
similarity index 100%
rename from ggml/include/ggml-metal.h
rename to ggml_llama/include/ggml-metal.h
diff --git a/ggml/include/ggml-rpc.h b/ggml_llama/include/ggml-rpc.h
similarity index 100%
rename from ggml/include/ggml-rpc.h
rename to ggml_llama/include/ggml-rpc.h
diff --git a/ggml/include/ggml-sycl.h b/ggml_llama/include/ggml-sycl.h
similarity index 100%
rename from ggml/include/ggml-sycl.h
rename to ggml_llama/include/ggml-sycl.h
diff --git a/ggml/include/ggml-vulkan.h b/ggml_llama/include/ggml-vulkan.h
similarity index 100%
rename from ggml/include/ggml-vulkan.h
rename to ggml_llama/include/ggml-vulkan.h
diff --git a/ggml/include/ggml.h b/ggml_llama/include/ggml.h
similarity index 100%
rename from ggml/include/ggml.h
rename to ggml_llama/include/ggml.h
diff --git a/ggml/src/CMakeLists.txt b/ggml_llama/src/CMakeLists.txt
similarity index 98%
rename from ggml/src/CMakeLists.txt
rename to ggml_llama/src/CMakeLists.txt
index ff84b9bb5..e867c7d74 100644
--- a/ggml/src/CMakeLists.txt
+++ b/ggml_llama/src/CMakeLists.txt
@@ -70,8 +70,8 @@ if (GGML_METAL)
         file(MAKE_DIRECTORY "${CMAKE_BINARY_DIR}/autogenerated")
 
         # merge ggml-common.h and ggml-metal.metal into a single file
-        set(METALLIB_EMBED_ASM    "${CMAKE_BINARY_DIR}/autogenerated/ggml-metal-embed.s")
-        set(METALLIB_SOURCE_EMBED "${CMAKE_BINARY_DIR}/autogenerated/ggml-metal-embed.metal")
+        set(METALLIB_EMBED_ASM    "${CMAKE_BINARY_DIR}/autogenerated/ggml-metal-embed-llama.s")
+        set(METALLIB_SOURCE_EMBED "${CMAKE_BINARY_DIR}/autogenerated/ggml-metal-embed-llama.metal")
 
         add_custom_command(
             OUTPUT ${METALLIB_EMBED_ASM}
@@ -827,7 +827,7 @@ if (GGML_CPU_HBM)
 
     add_compile_definitions(GGML_USE_CPU_HBM)
 
-    target_link_libraries(ggml PUBLIC memkind)
+    target_link_libraries(ggml_llama PUBLIC memkind)
 endif()
 
 if (GGML_CANN)
@@ -1291,7 +1291,7 @@ endif()
 
 # ggml
 
-add_library(ggml
+add_library(ggml_llama
             ../include/ggml.h
             ../include/ggml-alloc.h
             ../include/ggml-backend.h
@@ -1315,25 +1315,25 @@ add_library(ggml
             )
 
 if (EMSCRIPTEN)
-    set_target_properties(ggml PROPERTIES COMPILE_FLAGS "-msimd128")
+    set_target_properties(ggml_llama PROPERTIES COMPILE_FLAGS "-msimd128")
 endif()
 
-target_compile_definitions(ggml PUBLIC  ${GGML_CDEF_PUBLIC})
-target_include_directories(ggml PUBLIC ../include)
-target_include_directories(ggml PRIVATE . ${GGML_EXTRA_INCLUDES})
-target_link_directories(ggml PRIVATE ${GGML_EXTRA_LIBDIRS})
-target_compile_features   (ggml PRIVATE c_std_11) # don't bump
+target_compile_definitions(ggml_llama PUBLIC  ${GGML_CDEF_PUBLIC})
+target_include_directories(ggml_llama PUBLIC ../include)
+target_include_directories(ggml_llama PRIVATE . ${GGML_EXTRA_INCLUDES})
+target_link_directories(ggml_llama PRIVATE ${GGML_EXTRA_LIBDIRS})
+target_compile_features   (ggml_llama PRIVATE c_std_11) # don't bump
 
-target_link_libraries(ggml PRIVATE Threads::Threads ${GGML_EXTRA_LIBS})
+target_link_libraries(ggml_llama PRIVATE Threads::Threads ${GGML_EXTRA_LIBS})
 
 find_library(MATH_LIBRARY m)
 if (MATH_LIBRARY)
     if (NOT WIN32 OR NOT GGML_SYCL)
-        target_link_libraries(ggml PRIVATE ${MATH_LIBRARY})
+        target_link_libraries(ggml_llama PRIVATE ${MATH_LIBRARY})
     endif()
 endif()
 
 if (BUILD_SHARED_LIBS)
-    set_target_properties(ggml PROPERTIES POSITION_INDEPENDENT_CODE ON)
-    target_compile_definitions(ggml PRIVATE GGML_SHARED GGML_BUILD)
+    set_target_properties(ggml_llama PROPERTIES POSITION_INDEPENDENT_CODE ON)
+    target_compile_definitions(ggml_llama PRIVATE GGML_SHARED GGML_BUILD)
 endif()
diff --git a/ggml/src/ggml-aarch64.c b/ggml_llama/src/ggml-aarch64.c
similarity index 100%
rename from ggml/src/ggml-aarch64.c
rename to ggml_llama/src/ggml-aarch64.c
diff --git a/ggml/src/ggml-aarch64.h b/ggml_llama/src/ggml-aarch64.h
similarity index 100%
rename from ggml/src/ggml-aarch64.h
rename to ggml_llama/src/ggml-aarch64.h
diff --git a/ggml/src/ggml-alloc.c b/ggml_llama/src/ggml-alloc.c
similarity index 100%
rename from ggml/src/ggml-alloc.c
rename to ggml_llama/src/ggml-alloc.c
diff --git a/ggml/src/ggml-backend-impl.h b/ggml_llama/src/ggml-backend-impl.h
similarity index 100%
rename from ggml/src/ggml-backend-impl.h
rename to ggml_llama/src/ggml-backend-impl.h
diff --git a/ggml/src/ggml-backend.c b/ggml_llama/src/ggml-backend.c
similarity index 100%
rename from ggml/src/ggml-backend.c
rename to ggml_llama/src/ggml-backend.c
diff --git a/ggml/src/ggml-blas.cpp b/ggml_llama/src/ggml-blas.cpp
similarity index 100%
rename from ggml/src/ggml-blas.cpp
rename to ggml_llama/src/ggml-blas.cpp
diff --git a/ggml/src/ggml-cann.cpp b/ggml_llama/src/ggml-cann.cpp
similarity index 100%
rename from ggml/src/ggml-cann.cpp
rename to ggml_llama/src/ggml-cann.cpp
diff --git a/ggml/src/ggml-cann/.clang-format b/ggml_llama/src/ggml-cann/.clang-format
similarity index 100%
rename from ggml/src/ggml-cann/.clang-format
rename to ggml_llama/src/ggml-cann/.clang-format
diff --git a/ggml/src/ggml-cann/Doxyfile b/ggml_llama/src/ggml-cann/Doxyfile
similarity index 100%
rename from ggml/src/ggml-cann/Doxyfile
rename to ggml_llama/src/ggml-cann/Doxyfile
diff --git a/ggml/src/ggml-cann/acl_tensor.cpp b/ggml_llama/src/ggml-cann/acl_tensor.cpp
similarity index 100%
rename from ggml/src/ggml-cann/acl_tensor.cpp
rename to ggml_llama/src/ggml-cann/acl_tensor.cpp
diff --git a/ggml/src/ggml-cann/acl_tensor.h b/ggml_llama/src/ggml-cann/acl_tensor.h
similarity index 100%
rename from ggml/src/ggml-cann/acl_tensor.h
rename to ggml_llama/src/ggml-cann/acl_tensor.h
diff --git a/ggml/src/ggml-cann/aclnn_ops.cpp b/ggml_llama/src/ggml-cann/aclnn_ops.cpp
similarity index 100%
rename from ggml/src/ggml-cann/aclnn_ops.cpp
rename to ggml_llama/src/ggml-cann/aclnn_ops.cpp
diff --git a/ggml/src/ggml-cann/aclnn_ops.h b/ggml_llama/src/ggml-cann/aclnn_ops.h
similarity index 100%
rename from ggml/src/ggml-cann/aclnn_ops.h
rename to ggml_llama/src/ggml-cann/aclnn_ops.h
diff --git a/ggml/src/ggml-cann/common.h b/ggml_llama/src/ggml-cann/common.h
similarity index 100%
rename from ggml/src/ggml-cann/common.h
rename to ggml_llama/src/ggml-cann/common.h
diff --git a/ggml/src/ggml-cann/kernels/CMakeLists.txt b/ggml_llama/src/ggml-cann/kernels/CMakeLists.txt
similarity index 100%
rename from ggml/src/ggml-cann/kernels/CMakeLists.txt
rename to ggml_llama/src/ggml-cann/kernels/CMakeLists.txt
diff --git a/ggml/src/ggml-cann/kernels/ascendc_kernels.h b/ggml_llama/src/ggml-cann/kernels/ascendc_kernels.h
similarity index 100%
rename from ggml/src/ggml-cann/kernels/ascendc_kernels.h
rename to ggml_llama/src/ggml-cann/kernels/ascendc_kernels.h
diff --git a/ggml/src/ggml-cann/kernels/dup.cpp b/ggml_llama/src/ggml-cann/kernels/dup.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/dup.cpp
rename to ggml_llama/src/ggml-cann/kernels/dup.cpp
diff --git a/ggml/src/ggml-cann/kernels/get_row_f16.cpp b/ggml_llama/src/ggml-cann/kernels/get_row_f16.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/get_row_f16.cpp
rename to ggml_llama/src/ggml-cann/kernels/get_row_f16.cpp
diff --git a/ggml/src/ggml-cann/kernels/get_row_f32.cpp b/ggml_llama/src/ggml-cann/kernels/get_row_f32.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/get_row_f32.cpp
rename to ggml_llama/src/ggml-cann/kernels/get_row_f32.cpp
diff --git a/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp b/ggml_llama/src/ggml-cann/kernels/get_row_q4_0.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/get_row_q4_0.cpp
rename to ggml_llama/src/ggml-cann/kernels/get_row_q4_0.cpp
diff --git a/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp b/ggml_llama/src/ggml-cann/kernels/get_row_q8_0.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/get_row_q8_0.cpp
rename to ggml_llama/src/ggml-cann/kernels/get_row_q8_0.cpp
diff --git a/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp b/ggml_llama/src/ggml-cann/kernels/quantize_f16_q8_0.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp
rename to ggml_llama/src/ggml-cann/kernels/quantize_f16_q8_0.cpp
diff --git a/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp b/ggml_llama/src/ggml-cann/kernels/quantize_f32_q8_0.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp
rename to ggml_llama/src/ggml-cann/kernels/quantize_f32_q8_0.cpp
diff --git a/ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp b/ggml_llama/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp
similarity index 100%
rename from ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp
rename to ggml_llama/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp
diff --git a/ggml/src/ggml-common.h b/ggml_llama/src/ggml-common.h
similarity index 100%
rename from ggml/src/ggml-common.h
rename to ggml_llama/src/ggml-common.h
diff --git a/ggml/src/ggml-cuda.cu b/ggml_llama/src/ggml-cuda.cu
similarity index 100%
rename from ggml/src/ggml-cuda.cu
rename to ggml_llama/src/ggml-cuda.cu
diff --git a/ggml/src/ggml-cuda/acc.cu b/ggml_llama/src/ggml-cuda/acc.cu
similarity index 100%
rename from ggml/src/ggml-cuda/acc.cu
rename to ggml_llama/src/ggml-cuda/acc.cu
diff --git a/ggml/src/ggml-cuda/acc.cuh b/ggml_llama/src/ggml-cuda/acc.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/acc.cuh
rename to ggml_llama/src/ggml-cuda/acc.cuh
diff --git a/ggml/src/ggml-cuda/arange.cu b/ggml_llama/src/ggml-cuda/arange.cu
similarity index 100%
rename from ggml/src/ggml-cuda/arange.cu
rename to ggml_llama/src/ggml-cuda/arange.cu
diff --git a/ggml/src/ggml-cuda/arange.cuh b/ggml_llama/src/ggml-cuda/arange.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/arange.cuh
rename to ggml_llama/src/ggml-cuda/arange.cuh
diff --git a/ggml/src/ggml-cuda/argsort.cu b/ggml_llama/src/ggml-cuda/argsort.cu
similarity index 100%
rename from ggml/src/ggml-cuda/argsort.cu
rename to ggml_llama/src/ggml-cuda/argsort.cu
diff --git a/ggml/src/ggml-cuda/argsort.cuh b/ggml_llama/src/ggml-cuda/argsort.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/argsort.cuh
rename to ggml_llama/src/ggml-cuda/argsort.cuh
diff --git a/ggml/src/ggml-cuda/binbcast.cu b/ggml_llama/src/ggml-cuda/binbcast.cu
similarity index 100%
rename from ggml/src/ggml-cuda/binbcast.cu
rename to ggml_llama/src/ggml-cuda/binbcast.cu
diff --git a/ggml/src/ggml-cuda/binbcast.cuh b/ggml_llama/src/ggml-cuda/binbcast.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/binbcast.cuh
rename to ggml_llama/src/ggml-cuda/binbcast.cuh
diff --git a/ggml/src/ggml-cuda/clamp.cu b/ggml_llama/src/ggml-cuda/clamp.cu
similarity index 100%
rename from ggml/src/ggml-cuda/clamp.cu
rename to ggml_llama/src/ggml-cuda/clamp.cu
diff --git a/ggml/src/ggml-cuda/clamp.cuh b/ggml_llama/src/ggml-cuda/clamp.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/clamp.cuh
rename to ggml_llama/src/ggml-cuda/clamp.cuh
diff --git a/ggml/src/ggml-cuda/common.cuh b/ggml_llama/src/ggml-cuda/common.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/common.cuh
rename to ggml_llama/src/ggml-cuda/common.cuh
diff --git a/ggml/src/ggml-cuda/concat.cu b/ggml_llama/src/ggml-cuda/concat.cu
similarity index 100%
rename from ggml/src/ggml-cuda/concat.cu
rename to ggml_llama/src/ggml-cuda/concat.cu
diff --git a/ggml/src/ggml-cuda/concat.cuh b/ggml_llama/src/ggml-cuda/concat.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/concat.cuh
rename to ggml_llama/src/ggml-cuda/concat.cuh
diff --git a/ggml/src/ggml-cuda/conv-transpose-1d.cu b/ggml_llama/src/ggml-cuda/conv-transpose-1d.cu
similarity index 100%
rename from ggml/src/ggml-cuda/conv-transpose-1d.cu
rename to ggml_llama/src/ggml-cuda/conv-transpose-1d.cu
diff --git a/ggml/src/ggml-cuda/conv-transpose-1d.cuh b/ggml_llama/src/ggml-cuda/conv-transpose-1d.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/conv-transpose-1d.cuh
rename to ggml_llama/src/ggml-cuda/conv-transpose-1d.cuh
diff --git a/ggml/src/ggml-cuda/convert.cu b/ggml_llama/src/ggml-cuda/convert.cu
similarity index 100%
rename from ggml/src/ggml-cuda/convert.cu
rename to ggml_llama/src/ggml-cuda/convert.cu
diff --git a/ggml/src/ggml-cuda/convert.cuh b/ggml_llama/src/ggml-cuda/convert.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/convert.cuh
rename to ggml_llama/src/ggml-cuda/convert.cuh
diff --git a/ggml/src/ggml-cuda/cpy.cu b/ggml_llama/src/ggml-cuda/cpy.cu
similarity index 100%
rename from ggml/src/ggml-cuda/cpy.cu
rename to ggml_llama/src/ggml-cuda/cpy.cu
diff --git a/ggml/src/ggml-cuda/cpy.cuh b/ggml_llama/src/ggml-cuda/cpy.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/cpy.cuh
rename to ggml_llama/src/ggml-cuda/cpy.cuh
diff --git a/ggml/src/ggml-cuda/cross-entropy-loss.cu b/ggml_llama/src/ggml-cuda/cross-entropy-loss.cu
similarity index 100%
rename from ggml/src/ggml-cuda/cross-entropy-loss.cu
rename to ggml_llama/src/ggml-cuda/cross-entropy-loss.cu
diff --git a/ggml/src/ggml-cuda/cross-entropy-loss.cuh b/ggml_llama/src/ggml-cuda/cross-entropy-loss.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/cross-entropy-loss.cuh
rename to ggml_llama/src/ggml-cuda/cross-entropy-loss.cuh
diff --git a/ggml/src/ggml-cuda/dequantize.cuh b/ggml_llama/src/ggml-cuda/dequantize.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/dequantize.cuh
rename to ggml_llama/src/ggml-cuda/dequantize.cuh
diff --git a/ggml/src/ggml-cuda/diagmask.cu b/ggml_llama/src/ggml-cuda/diagmask.cu
similarity index 100%
rename from ggml/src/ggml-cuda/diagmask.cu
rename to ggml_llama/src/ggml-cuda/diagmask.cu
diff --git a/ggml/src/ggml-cuda/diagmask.cuh b/ggml_llama/src/ggml-cuda/diagmask.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/diagmask.cuh
rename to ggml_llama/src/ggml-cuda/diagmask.cuh
diff --git a/ggml/src/ggml-cuda/dmmv.cu b/ggml_llama/src/ggml-cuda/dmmv.cu
similarity index 100%
rename from ggml/src/ggml-cuda/dmmv.cu
rename to ggml_llama/src/ggml-cuda/dmmv.cu
diff --git a/ggml/src/ggml-cuda/dmmv.cuh b/ggml_llama/src/ggml-cuda/dmmv.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/dmmv.cuh
rename to ggml_llama/src/ggml-cuda/dmmv.cuh
diff --git a/ggml/src/ggml-cuda/fattn-common.cuh b/ggml_llama/src/ggml-cuda/fattn-common.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-common.cuh
rename to ggml_llama/src/ggml-cuda/fattn-common.cuh
diff --git a/ggml/src/ggml-cuda/fattn-tile-f16.cu b/ggml_llama/src/ggml-cuda/fattn-tile-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-tile-f16.cu
rename to ggml_llama/src/ggml-cuda/fattn-tile-f16.cu
diff --git a/ggml/src/ggml-cuda/fattn-tile-f16.cuh b/ggml_llama/src/ggml-cuda/fattn-tile-f16.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-tile-f16.cuh
rename to ggml_llama/src/ggml-cuda/fattn-tile-f16.cuh
diff --git a/ggml/src/ggml-cuda/fattn-tile-f32.cu b/ggml_llama/src/ggml-cuda/fattn-tile-f32.cu
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-tile-f32.cu
rename to ggml_llama/src/ggml-cuda/fattn-tile-f32.cu
diff --git a/ggml/src/ggml-cuda/fattn-tile-f32.cuh b/ggml_llama/src/ggml-cuda/fattn-tile-f32.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-tile-f32.cuh
rename to ggml_llama/src/ggml-cuda/fattn-tile-f32.cuh
diff --git a/ggml/src/ggml-cuda/fattn-vec-f16.cuh b/ggml_llama/src/ggml-cuda/fattn-vec-f16.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-vec-f16.cuh
rename to ggml_llama/src/ggml-cuda/fattn-vec-f16.cuh
diff --git a/ggml/src/ggml-cuda/fattn-vec-f32.cuh b/ggml_llama/src/ggml-cuda/fattn-vec-f32.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-vec-f32.cuh
rename to ggml_llama/src/ggml-cuda/fattn-vec-f32.cuh
diff --git a/ggml/src/ggml-cuda/fattn-wmma-f16.cuh b/ggml_llama/src/ggml-cuda/fattn-wmma-f16.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn-wmma-f16.cuh
rename to ggml_llama/src/ggml-cuda/fattn-wmma-f16.cuh
diff --git a/ggml/src/ggml-cuda/fattn.cu b/ggml_llama/src/ggml-cuda/fattn.cu
similarity index 100%
rename from ggml/src/ggml-cuda/fattn.cu
rename to ggml_llama/src/ggml-cuda/fattn.cu
diff --git a/ggml/src/ggml-cuda/fattn.cuh b/ggml_llama/src/ggml-cuda/fattn.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/fattn.cuh
rename to ggml_llama/src/ggml-cuda/fattn.cuh
diff --git a/ggml/src/ggml-cuda/getrows.cu b/ggml_llama/src/ggml-cuda/getrows.cu
similarity index 100%
rename from ggml/src/ggml-cuda/getrows.cu
rename to ggml_llama/src/ggml-cuda/getrows.cu
diff --git a/ggml/src/ggml-cuda/getrows.cuh b/ggml_llama/src/ggml-cuda/getrows.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/getrows.cuh
rename to ggml_llama/src/ggml-cuda/getrows.cuh
diff --git a/ggml/src/ggml-cuda/im2col.cu b/ggml_llama/src/ggml-cuda/im2col.cu
similarity index 100%
rename from ggml/src/ggml-cuda/im2col.cu
rename to ggml_llama/src/ggml-cuda/im2col.cu
diff --git a/ggml/src/ggml-cuda/im2col.cuh b/ggml_llama/src/ggml-cuda/im2col.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/im2col.cuh
rename to ggml_llama/src/ggml-cuda/im2col.cuh
diff --git a/ggml/src/ggml-cuda/mma.cuh b/ggml_llama/src/ggml-cuda/mma.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/mma.cuh
rename to ggml_llama/src/ggml-cuda/mma.cuh
diff --git a/ggml/src/ggml-cuda/mmq.cu b/ggml_llama/src/ggml-cuda/mmq.cu
similarity index 100%
rename from ggml/src/ggml-cuda/mmq.cu
rename to ggml_llama/src/ggml-cuda/mmq.cu
diff --git a/ggml/src/ggml-cuda/mmq.cuh b/ggml_llama/src/ggml-cuda/mmq.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/mmq.cuh
rename to ggml_llama/src/ggml-cuda/mmq.cuh
diff --git a/ggml/src/ggml-cuda/mmvq.cu b/ggml_llama/src/ggml-cuda/mmvq.cu
similarity index 100%
rename from ggml/src/ggml-cuda/mmvq.cu
rename to ggml_llama/src/ggml-cuda/mmvq.cu
diff --git a/ggml/src/ggml-cuda/mmvq.cuh b/ggml_llama/src/ggml-cuda/mmvq.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/mmvq.cuh
rename to ggml_llama/src/ggml-cuda/mmvq.cuh
diff --git a/ggml/src/ggml-cuda/norm.cu b/ggml_llama/src/ggml-cuda/norm.cu
similarity index 100%
rename from ggml/src/ggml-cuda/norm.cu
rename to ggml_llama/src/ggml-cuda/norm.cu
diff --git a/ggml/src/ggml-cuda/norm.cuh b/ggml_llama/src/ggml-cuda/norm.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/norm.cuh
rename to ggml_llama/src/ggml-cuda/norm.cuh
diff --git a/ggml/src/ggml-cuda/pad.cu b/ggml_llama/src/ggml-cuda/pad.cu
similarity index 100%
rename from ggml/src/ggml-cuda/pad.cu
rename to ggml_llama/src/ggml-cuda/pad.cu
diff --git a/ggml/src/ggml-cuda/pad.cuh b/ggml_llama/src/ggml-cuda/pad.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/pad.cuh
rename to ggml_llama/src/ggml-cuda/pad.cuh
diff --git a/ggml/src/ggml-cuda/pool2d.cu b/ggml_llama/src/ggml-cuda/pool2d.cu
similarity index 100%
rename from ggml/src/ggml-cuda/pool2d.cu
rename to ggml_llama/src/ggml-cuda/pool2d.cu
diff --git a/ggml/src/ggml-cuda/pool2d.cuh b/ggml_llama/src/ggml-cuda/pool2d.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/pool2d.cuh
rename to ggml_llama/src/ggml-cuda/pool2d.cuh
diff --git a/ggml/src/ggml-cuda/quantize.cu b/ggml_llama/src/ggml-cuda/quantize.cu
similarity index 100%
rename from ggml/src/ggml-cuda/quantize.cu
rename to ggml_llama/src/ggml-cuda/quantize.cu
diff --git a/ggml/src/ggml-cuda/quantize.cuh b/ggml_llama/src/ggml-cuda/quantize.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/quantize.cuh
rename to ggml_llama/src/ggml-cuda/quantize.cuh
diff --git a/ggml/src/ggml-cuda/rope.cu b/ggml_llama/src/ggml-cuda/rope.cu
similarity index 100%
rename from ggml/src/ggml-cuda/rope.cu
rename to ggml_llama/src/ggml-cuda/rope.cu
diff --git a/ggml/src/ggml-cuda/rope.cuh b/ggml_llama/src/ggml-cuda/rope.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/rope.cuh
rename to ggml_llama/src/ggml-cuda/rope.cuh
diff --git a/ggml/src/ggml-cuda/scale.cu b/ggml_llama/src/ggml-cuda/scale.cu
similarity index 100%
rename from ggml/src/ggml-cuda/scale.cu
rename to ggml_llama/src/ggml-cuda/scale.cu
diff --git a/ggml/src/ggml-cuda/scale.cuh b/ggml_llama/src/ggml-cuda/scale.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/scale.cuh
rename to ggml_llama/src/ggml-cuda/scale.cuh
diff --git a/ggml/src/ggml-cuda/softmax.cu b/ggml_llama/src/ggml-cuda/softmax.cu
similarity index 100%
rename from ggml/src/ggml-cuda/softmax.cu
rename to ggml_llama/src/ggml-cuda/softmax.cu
diff --git a/ggml/src/ggml-cuda/softmax.cuh b/ggml_llama/src/ggml-cuda/softmax.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/softmax.cuh
rename to ggml_llama/src/ggml-cuda/softmax.cuh
diff --git a/ggml/src/ggml-cuda/sumrows.cu b/ggml_llama/src/ggml-cuda/sumrows.cu
similarity index 100%
rename from ggml/src/ggml-cuda/sumrows.cu
rename to ggml_llama/src/ggml-cuda/sumrows.cu
diff --git a/ggml/src/ggml-cuda/sumrows.cuh b/ggml_llama/src/ggml-cuda/sumrows.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/sumrows.cuh
rename to ggml_llama/src/ggml-cuda/sumrows.cuh
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-f16-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_0-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs256-f16-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f16-instance-hs64-f16-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-f16-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_0-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q4_1-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_0-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q5_1-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs128-q8_0-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs256-f16-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-f16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-vec-f32-instance-hs64-f16-q8_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb32.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb32.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb32.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqfloat-cpb32.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb16.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb16.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb16.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb16.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb32.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb32.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb32.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb32.cu
diff --git a/ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb8.cu b/ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb8.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb8.cu
rename to ggml_llama/src/ggml-cuda/template-instances/fattn-wmma-f16-instance-kqhalf-cpb8.cu
diff --git a/ggml/src/ggml-cuda/template-instances/generate_cu_files.py b/ggml_llama/src/ggml-cuda/template-instances/generate_cu_files.py
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/generate_cu_files.py
rename to ggml_llama/src/ggml-cuda/template-instances/generate_cu_files.py
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xxs.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq2_xxs.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xxs.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq2_xxs.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_s.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq3_s.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_s.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq3_s.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq4_nl.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-iq4_xs.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q2_k.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q3_k.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q4_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q4_1.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q4_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q4_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q4_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q4_k.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q4_k.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q4_k.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q4_k.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q5_0.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q5_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q5_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q5_0.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q5_1.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q5_1.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q5_1.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q5_1.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q5_k.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q6_k.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q6_k.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q6_k.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q6_k.cu
diff --git a/ggml/src/ggml-cuda/template-instances/mmq-instance-q8_0.cu b/ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q8_0.cu
similarity index 100%
rename from ggml/src/ggml-cuda/template-instances/mmq-instance-q8_0.cu
rename to ggml_llama/src/ggml-cuda/template-instances/mmq-instance-q8_0.cu
diff --git a/ggml/src/ggml-cuda/tsembd.cu b/ggml_llama/src/ggml-cuda/tsembd.cu
similarity index 100%
rename from ggml/src/ggml-cuda/tsembd.cu
rename to ggml_llama/src/ggml-cuda/tsembd.cu
diff --git a/ggml/src/ggml-cuda/tsembd.cuh b/ggml_llama/src/ggml-cuda/tsembd.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/tsembd.cuh
rename to ggml_llama/src/ggml-cuda/tsembd.cuh
diff --git a/ggml/src/ggml-cuda/unary.cu b/ggml_llama/src/ggml-cuda/unary.cu
similarity index 100%
rename from ggml/src/ggml-cuda/unary.cu
rename to ggml_llama/src/ggml-cuda/unary.cu
diff --git a/ggml/src/ggml-cuda/unary.cuh b/ggml_llama/src/ggml-cuda/unary.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/unary.cuh
rename to ggml_llama/src/ggml-cuda/unary.cuh
diff --git a/ggml/src/ggml-cuda/upscale.cu b/ggml_llama/src/ggml-cuda/upscale.cu
similarity index 100%
rename from ggml/src/ggml-cuda/upscale.cu
rename to ggml_llama/src/ggml-cuda/upscale.cu
diff --git a/ggml/src/ggml-cuda/upscale.cuh b/ggml_llama/src/ggml-cuda/upscale.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/upscale.cuh
rename to ggml_llama/src/ggml-cuda/upscale.cuh
diff --git a/ggml/src/ggml-cuda/vecdotq.cuh b/ggml_llama/src/ggml-cuda/vecdotq.cuh
similarity index 100%
rename from ggml/src/ggml-cuda/vecdotq.cuh
rename to ggml_llama/src/ggml-cuda/vecdotq.cuh
diff --git a/ggml/src/ggml-cuda/vendors/cuda.h b/ggml_llama/src/ggml-cuda/vendors/cuda.h
similarity index 100%
rename from ggml/src/ggml-cuda/vendors/cuda.h
rename to ggml_llama/src/ggml-cuda/vendors/cuda.h
diff --git a/ggml/src/ggml-cuda/vendors/hip.h b/ggml_llama/src/ggml-cuda/vendors/hip.h
similarity index 100%
rename from ggml/src/ggml-cuda/vendors/hip.h
rename to ggml_llama/src/ggml-cuda/vendors/hip.h
diff --git a/ggml/src/ggml-cuda/vendors/musa.h b/ggml_llama/src/ggml-cuda/vendors/musa.h
similarity index 100%
rename from ggml/src/ggml-cuda/vendors/musa.h
rename to ggml_llama/src/ggml-cuda/vendors/musa.h
diff --git a/ggml/src/ggml-impl.h b/ggml_llama/src/ggml-impl.h
similarity index 100%
rename from ggml/src/ggml-impl.h
rename to ggml_llama/src/ggml-impl.h
diff --git a/ggml/src/ggml-kompute.cpp b/ggml_llama/src/ggml-kompute.cpp
similarity index 100%
rename from ggml/src/ggml-kompute.cpp
rename to ggml_llama/src/ggml-kompute.cpp
diff --git a/ggml/src/ggml-metal.m b/ggml_llama/src/ggml-metal.m
similarity index 100%
rename from ggml/src/ggml-metal.m
rename to ggml_llama/src/ggml-metal.m
diff --git a/ggml/src/ggml-metal.metal b/ggml_llama/src/ggml-metal.metal
similarity index 100%
rename from ggml/src/ggml-metal.metal
rename to ggml_llama/src/ggml-metal.metal
diff --git a/ggml/src/ggml-quants.c b/ggml_llama/src/ggml-quants.c
similarity index 100%
rename from ggml/src/ggml-quants.c
rename to ggml_llama/src/ggml-quants.c
diff --git a/ggml/src/ggml-quants.h b/ggml_llama/src/ggml-quants.h
similarity index 100%
rename from ggml/src/ggml-quants.h
rename to ggml_llama/src/ggml-quants.h
diff --git a/ggml/src/ggml-rpc.cpp b/ggml_llama/src/ggml-rpc.cpp
similarity index 100%
rename from ggml/src/ggml-rpc.cpp
rename to ggml_llama/src/ggml-rpc.cpp
diff --git a/ggml/src/ggml-sycl.cpp b/ggml_llama/src/ggml-sycl.cpp
similarity index 100%
rename from ggml/src/ggml-sycl.cpp
rename to ggml_llama/src/ggml-sycl.cpp
diff --git a/ggml/src/ggml-sycl/backend.hpp b/ggml_llama/src/ggml-sycl/backend.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/backend.hpp
rename to ggml_llama/src/ggml-sycl/backend.hpp
diff --git a/ggml/src/ggml-sycl/common.cpp b/ggml_llama/src/ggml-sycl/common.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/common.cpp
rename to ggml_llama/src/ggml-sycl/common.cpp
diff --git a/ggml/src/ggml-sycl/common.hpp b/ggml_llama/src/ggml-sycl/common.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/common.hpp
rename to ggml_llama/src/ggml-sycl/common.hpp
diff --git a/ggml/src/ggml-sycl/concat.cpp b/ggml_llama/src/ggml-sycl/concat.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/concat.cpp
rename to ggml_llama/src/ggml-sycl/concat.cpp
diff --git a/ggml/src/ggml-sycl/concat.hpp b/ggml_llama/src/ggml-sycl/concat.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/concat.hpp
rename to ggml_llama/src/ggml-sycl/concat.hpp
diff --git a/ggml/src/ggml-sycl/conv.cpp b/ggml_llama/src/ggml-sycl/conv.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/conv.cpp
rename to ggml_llama/src/ggml-sycl/conv.cpp
diff --git a/ggml/src/ggml-sycl/conv.hpp b/ggml_llama/src/ggml-sycl/conv.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/conv.hpp
rename to ggml_llama/src/ggml-sycl/conv.hpp
diff --git a/ggml/src/ggml-sycl/convert.cpp b/ggml_llama/src/ggml-sycl/convert.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/convert.cpp
rename to ggml_llama/src/ggml-sycl/convert.cpp
diff --git a/ggml/src/ggml-sycl/convert.hpp b/ggml_llama/src/ggml-sycl/convert.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/convert.hpp
rename to ggml_llama/src/ggml-sycl/convert.hpp
diff --git a/ggml/src/ggml-sycl/dequantize.hpp b/ggml_llama/src/ggml-sycl/dequantize.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/dequantize.hpp
rename to ggml_llama/src/ggml-sycl/dequantize.hpp
diff --git a/ggml/src/ggml-sycl/dmmv.cpp b/ggml_llama/src/ggml-sycl/dmmv.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/dmmv.cpp
rename to ggml_llama/src/ggml-sycl/dmmv.cpp
diff --git a/ggml/src/ggml-sycl/dmmv.hpp b/ggml_llama/src/ggml-sycl/dmmv.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/dmmv.hpp
rename to ggml_llama/src/ggml-sycl/dmmv.hpp
diff --git a/ggml/src/ggml-sycl/dpct/helper.hpp b/ggml_llama/src/ggml-sycl/dpct/helper.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/dpct/helper.hpp
rename to ggml_llama/src/ggml-sycl/dpct/helper.hpp
diff --git a/ggml/src/ggml-sycl/gemm.hpp b/ggml_llama/src/ggml-sycl/gemm.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/gemm.hpp
rename to ggml_llama/src/ggml-sycl/gemm.hpp
diff --git a/ggml/src/ggml-sycl/im2col.cpp b/ggml_llama/src/ggml-sycl/im2col.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/im2col.cpp
rename to ggml_llama/src/ggml-sycl/im2col.cpp
diff --git a/ggml/src/ggml-sycl/im2col.hpp b/ggml_llama/src/ggml-sycl/im2col.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/im2col.hpp
rename to ggml_llama/src/ggml-sycl/im2col.hpp
diff --git a/ggml/src/ggml-sycl/mmq.cpp b/ggml_llama/src/ggml-sycl/mmq.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/mmq.cpp
rename to ggml_llama/src/ggml-sycl/mmq.cpp
diff --git a/ggml/src/ggml-sycl/mmq.hpp b/ggml_llama/src/ggml-sycl/mmq.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/mmq.hpp
rename to ggml_llama/src/ggml-sycl/mmq.hpp
diff --git a/ggml/src/ggml-sycl/mmvq.cpp b/ggml_llama/src/ggml-sycl/mmvq.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/mmvq.cpp
rename to ggml_llama/src/ggml-sycl/mmvq.cpp
diff --git a/ggml/src/ggml-sycl/mmvq.hpp b/ggml_llama/src/ggml-sycl/mmvq.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/mmvq.hpp
rename to ggml_llama/src/ggml-sycl/mmvq.hpp
diff --git a/ggml/src/ggml-sycl/norm.cpp b/ggml_llama/src/ggml-sycl/norm.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/norm.cpp
rename to ggml_llama/src/ggml-sycl/norm.cpp
diff --git a/ggml/src/ggml-sycl/norm.hpp b/ggml_llama/src/ggml-sycl/norm.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/norm.hpp
rename to ggml_llama/src/ggml-sycl/norm.hpp
diff --git a/ggml/src/ggml-sycl/presets.hpp b/ggml_llama/src/ggml-sycl/presets.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/presets.hpp
rename to ggml_llama/src/ggml-sycl/presets.hpp
diff --git a/ggml/src/ggml-sycl/rope.cpp b/ggml_llama/src/ggml-sycl/rope.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/rope.cpp
rename to ggml_llama/src/ggml-sycl/rope.cpp
diff --git a/ggml/src/ggml-sycl/rope.hpp b/ggml_llama/src/ggml-sycl/rope.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/rope.hpp
rename to ggml_llama/src/ggml-sycl/rope.hpp
diff --git a/ggml/src/ggml-sycl/softmax.cpp b/ggml_llama/src/ggml-sycl/softmax.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/softmax.cpp
rename to ggml_llama/src/ggml-sycl/softmax.cpp
diff --git a/ggml/src/ggml-sycl/softmax.hpp b/ggml_llama/src/ggml-sycl/softmax.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/softmax.hpp
rename to ggml_llama/src/ggml-sycl/softmax.hpp
diff --git a/ggml/src/ggml-sycl/tsembd.cpp b/ggml_llama/src/ggml-sycl/tsembd.cpp
similarity index 100%
rename from ggml/src/ggml-sycl/tsembd.cpp
rename to ggml_llama/src/ggml-sycl/tsembd.cpp
diff --git a/ggml/src/ggml-sycl/tsembd.hpp b/ggml_llama/src/ggml-sycl/tsembd.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/tsembd.hpp
rename to ggml_llama/src/ggml-sycl/tsembd.hpp
diff --git a/ggml/src/ggml-sycl/vecdotq.hpp b/ggml_llama/src/ggml-sycl/vecdotq.hpp
similarity index 100%
rename from ggml/src/ggml-sycl/vecdotq.hpp
rename to ggml_llama/src/ggml-sycl/vecdotq.hpp
diff --git a/ggml/src/ggml-vulkan.cpp b/ggml_llama/src/ggml-vulkan.cpp
similarity index 100%
rename from ggml/src/ggml-vulkan.cpp
rename to ggml_llama/src/ggml-vulkan.cpp
diff --git a/ggml/src/ggml.c b/ggml_llama/src/ggml.c
similarity index 100%
rename from ggml/src/ggml.c
rename to ggml_llama/src/ggml.c
diff --git a/ggml/src/kompute-shaders/common.comp b/ggml_llama/src/kompute-shaders/common.comp
similarity index 100%
rename from ggml/src/kompute-shaders/common.comp
rename to ggml_llama/src/kompute-shaders/common.comp
diff --git a/ggml/src/kompute-shaders/op_add.comp b/ggml_llama/src/kompute-shaders/op_add.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_add.comp
rename to ggml_llama/src/kompute-shaders/op_add.comp
diff --git a/ggml/src/kompute-shaders/op_addrow.comp b/ggml_llama/src/kompute-shaders/op_addrow.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_addrow.comp
rename to ggml_llama/src/kompute-shaders/op_addrow.comp
diff --git a/ggml/src/kompute-shaders/op_cpy_f16_f16.comp b/ggml_llama/src/kompute-shaders/op_cpy_f16_f16.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_cpy_f16_f16.comp
rename to ggml_llama/src/kompute-shaders/op_cpy_f16_f16.comp
diff --git a/ggml/src/kompute-shaders/op_cpy_f16_f32.comp b/ggml_llama/src/kompute-shaders/op_cpy_f16_f32.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_cpy_f16_f32.comp
rename to ggml_llama/src/kompute-shaders/op_cpy_f16_f32.comp
diff --git a/ggml/src/kompute-shaders/op_cpy_f32_f16.comp b/ggml_llama/src/kompute-shaders/op_cpy_f32_f16.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_cpy_f32_f16.comp
rename to ggml_llama/src/kompute-shaders/op_cpy_f32_f16.comp
diff --git a/ggml/src/kompute-shaders/op_cpy_f32_f32.comp b/ggml_llama/src/kompute-shaders/op_cpy_f32_f32.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_cpy_f32_f32.comp
rename to ggml_llama/src/kompute-shaders/op_cpy_f32_f32.comp
diff --git a/ggml/src/kompute-shaders/op_diagmask.comp b/ggml_llama/src/kompute-shaders/op_diagmask.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_diagmask.comp
rename to ggml_llama/src/kompute-shaders/op_diagmask.comp
diff --git a/ggml/src/kompute-shaders/op_gelu.comp b/ggml_llama/src/kompute-shaders/op_gelu.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_gelu.comp
rename to ggml_llama/src/kompute-shaders/op_gelu.comp
diff --git a/ggml/src/kompute-shaders/op_getrows.comp b/ggml_llama/src/kompute-shaders/op_getrows.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_getrows.comp
rename to ggml_llama/src/kompute-shaders/op_getrows.comp
diff --git a/ggml/src/kompute-shaders/op_getrows_f16.comp b/ggml_llama/src/kompute-shaders/op_getrows_f16.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_getrows_f16.comp
rename to ggml_llama/src/kompute-shaders/op_getrows_f16.comp
diff --git a/ggml/src/kompute-shaders/op_getrows_f32.comp b/ggml_llama/src/kompute-shaders/op_getrows_f32.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_getrows_f32.comp
rename to ggml_llama/src/kompute-shaders/op_getrows_f32.comp
diff --git a/ggml/src/kompute-shaders/op_getrows_q4_0.comp b/ggml_llama/src/kompute-shaders/op_getrows_q4_0.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_getrows_q4_0.comp
rename to ggml_llama/src/kompute-shaders/op_getrows_q4_0.comp
diff --git a/ggml/src/kompute-shaders/op_getrows_q4_1.comp b/ggml_llama/src/kompute-shaders/op_getrows_q4_1.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_getrows_q4_1.comp
rename to ggml_llama/src/kompute-shaders/op_getrows_q4_1.comp
diff --git a/ggml/src/kompute-shaders/op_getrows_q6_k.comp b/ggml_llama/src/kompute-shaders/op_getrows_q6_k.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_getrows_q6_k.comp
rename to ggml_llama/src/kompute-shaders/op_getrows_q6_k.comp
diff --git a/ggml/src/kompute-shaders/op_mul.comp b/ggml_llama/src/kompute-shaders/op_mul.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul.comp
rename to ggml_llama/src/kompute-shaders/op_mul.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mat_f16.comp b/ggml_llama/src/kompute-shaders/op_mul_mat_f16.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mat_f16.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mat_f16.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mat_mat_f32.comp b/ggml_llama/src/kompute-shaders/op_mul_mat_mat_f32.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mat_mat_f32.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mat_mat_f32.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mat_q4_0.comp b/ggml_llama/src/kompute-shaders/op_mul_mat_q4_0.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mat_q4_0.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mat_q4_0.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mat_q4_1.comp b/ggml_llama/src/kompute-shaders/op_mul_mat_q4_1.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mat_q4_1.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mat_q4_1.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mat_q6_k.comp b/ggml_llama/src/kompute-shaders/op_mul_mat_q6_k.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mat_q6_k.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mat_q6_k.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mat_q8_0.comp b/ggml_llama/src/kompute-shaders/op_mul_mat_q8_0.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mat_q8_0.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mat_q8_0.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mv_q_n.comp b/ggml_llama/src/kompute-shaders/op_mul_mv_q_n.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mv_q_n.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mv_q_n.comp
diff --git a/ggml/src/kompute-shaders/op_mul_mv_q_n_pre.comp b/ggml_llama/src/kompute-shaders/op_mul_mv_q_n_pre.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_mul_mv_q_n_pre.comp
rename to ggml_llama/src/kompute-shaders/op_mul_mv_q_n_pre.comp
diff --git a/ggml/src/kompute-shaders/op_norm.comp b/ggml_llama/src/kompute-shaders/op_norm.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_norm.comp
rename to ggml_llama/src/kompute-shaders/op_norm.comp
diff --git a/ggml/src/kompute-shaders/op_relu.comp b/ggml_llama/src/kompute-shaders/op_relu.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_relu.comp
rename to ggml_llama/src/kompute-shaders/op_relu.comp
diff --git a/ggml/src/kompute-shaders/op_rmsnorm.comp b/ggml_llama/src/kompute-shaders/op_rmsnorm.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_rmsnorm.comp
rename to ggml_llama/src/kompute-shaders/op_rmsnorm.comp
diff --git a/ggml/src/kompute-shaders/op_rope_f16.comp b/ggml_llama/src/kompute-shaders/op_rope_f16.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_rope_f16.comp
rename to ggml_llama/src/kompute-shaders/op_rope_f16.comp
diff --git a/ggml/src/kompute-shaders/op_rope_f32.comp b/ggml_llama/src/kompute-shaders/op_rope_f32.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_rope_f32.comp
rename to ggml_llama/src/kompute-shaders/op_rope_f32.comp
diff --git a/ggml/src/kompute-shaders/op_scale.comp b/ggml_llama/src/kompute-shaders/op_scale.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_scale.comp
rename to ggml_llama/src/kompute-shaders/op_scale.comp
diff --git a/ggml/src/kompute-shaders/op_scale_8.comp b/ggml_llama/src/kompute-shaders/op_scale_8.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_scale_8.comp
rename to ggml_llama/src/kompute-shaders/op_scale_8.comp
diff --git a/ggml/src/kompute-shaders/op_silu.comp b/ggml_llama/src/kompute-shaders/op_silu.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_silu.comp
rename to ggml_llama/src/kompute-shaders/op_silu.comp
diff --git a/ggml/src/kompute-shaders/op_softmax.comp b/ggml_llama/src/kompute-shaders/op_softmax.comp
similarity index 100%
rename from ggml/src/kompute-shaders/op_softmax.comp
rename to ggml_llama/src/kompute-shaders/op_softmax.comp
diff --git a/ggml/src/kompute-shaders/rope_common.comp b/ggml_llama/src/kompute-shaders/rope_common.comp
similarity index 100%
rename from ggml/src/kompute-shaders/rope_common.comp
rename to ggml_llama/src/kompute-shaders/rope_common.comp
diff --git a/ggml/src/llamafile/sgemm.cpp b/ggml_llama/src/llamafile/sgemm.cpp
similarity index 100%
rename from ggml/src/llamafile/sgemm.cpp
rename to ggml_llama/src/llamafile/sgemm.cpp
diff --git a/ggml/src/llamafile/sgemm.h b/ggml_llama/src/llamafile/sgemm.h
similarity index 100%
rename from ggml/src/llamafile/sgemm.h
rename to ggml_llama/src/llamafile/sgemm.h
diff --git a/ggml/src/vulkan-shaders/CMakeLists.txt b/ggml_llama/src/vulkan-shaders/CMakeLists.txt
similarity index 100%
rename from ggml/src/vulkan-shaders/CMakeLists.txt
rename to ggml_llama/src/vulkan-shaders/CMakeLists.txt
diff --git a/ggml/src/vulkan-shaders/acc.comp b/ggml_llama/src/vulkan-shaders/acc.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/acc.comp
rename to ggml_llama/src/vulkan-shaders/acc.comp
diff --git a/ggml/src/vulkan-shaders/add.comp b/ggml_llama/src/vulkan-shaders/add.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/add.comp
rename to ggml_llama/src/vulkan-shaders/add.comp
diff --git a/ggml/src/vulkan-shaders/argsort.comp b/ggml_llama/src/vulkan-shaders/argsort.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/argsort.comp
rename to ggml_llama/src/vulkan-shaders/argsort.comp
diff --git a/ggml/src/vulkan-shaders/clamp.comp b/ggml_llama/src/vulkan-shaders/clamp.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/clamp.comp
rename to ggml_llama/src/vulkan-shaders/clamp.comp
diff --git a/ggml/src/vulkan-shaders/concat.comp b/ggml_llama/src/vulkan-shaders/concat.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/concat.comp
rename to ggml_llama/src/vulkan-shaders/concat.comp
diff --git a/ggml/src/vulkan-shaders/copy.comp b/ggml_llama/src/vulkan-shaders/copy.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/copy.comp
rename to ggml_llama/src/vulkan-shaders/copy.comp
diff --git a/ggml/src/vulkan-shaders/cos.comp b/ggml_llama/src/vulkan-shaders/cos.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/cos.comp
rename to ggml_llama/src/vulkan-shaders/cos.comp
diff --git a/ggml/src/vulkan-shaders/dequant_f32.comp b/ggml_llama/src/vulkan-shaders/dequant_f32.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_f32.comp
rename to ggml_llama/src/vulkan-shaders/dequant_f32.comp
diff --git a/ggml/src/vulkan-shaders/dequant_funcs.comp b/ggml_llama/src/vulkan-shaders/dequant_funcs.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_funcs.comp
rename to ggml_llama/src/vulkan-shaders/dequant_funcs.comp
diff --git a/ggml/src/vulkan-shaders/dequant_head.comp b/ggml_llama/src/vulkan-shaders/dequant_head.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_head.comp
rename to ggml_llama/src/vulkan-shaders/dequant_head.comp
diff --git a/ggml/src/vulkan-shaders/dequant_iq4_nl.comp b/ggml_llama/src/vulkan-shaders/dequant_iq4_nl.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_iq4_nl.comp
rename to ggml_llama/src/vulkan-shaders/dequant_iq4_nl.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q2_k.comp b/ggml_llama/src/vulkan-shaders/dequant_q2_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q2_k.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q2_k.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q3_k.comp b/ggml_llama/src/vulkan-shaders/dequant_q3_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q3_k.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q3_k.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q4_0.comp b/ggml_llama/src/vulkan-shaders/dequant_q4_0.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q4_0.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q4_0.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q4_1.comp b/ggml_llama/src/vulkan-shaders/dequant_q4_1.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q4_1.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q4_1.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q4_k.comp b/ggml_llama/src/vulkan-shaders/dequant_q4_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q4_k.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q4_k.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q5_0.comp b/ggml_llama/src/vulkan-shaders/dequant_q5_0.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q5_0.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q5_0.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q5_1.comp b/ggml_llama/src/vulkan-shaders/dequant_q5_1.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q5_1.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q5_1.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q5_k.comp b/ggml_llama/src/vulkan-shaders/dequant_q5_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q5_k.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q5_k.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q6_k.comp b/ggml_llama/src/vulkan-shaders/dequant_q6_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q6_k.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q6_k.comp
diff --git a/ggml/src/vulkan-shaders/dequant_q8_0.comp b/ggml_llama/src/vulkan-shaders/dequant_q8_0.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/dequant_q8_0.comp
rename to ggml_llama/src/vulkan-shaders/dequant_q8_0.comp
diff --git a/ggml/src/vulkan-shaders/diag_mask_inf.comp b/ggml_llama/src/vulkan-shaders/diag_mask_inf.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/diag_mask_inf.comp
rename to ggml_llama/src/vulkan-shaders/diag_mask_inf.comp
diff --git a/ggml/src/vulkan-shaders/div.comp b/ggml_llama/src/vulkan-shaders/div.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/div.comp
rename to ggml_llama/src/vulkan-shaders/div.comp
diff --git a/ggml/src/vulkan-shaders/gelu.comp b/ggml_llama/src/vulkan-shaders/gelu.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/gelu.comp
rename to ggml_llama/src/vulkan-shaders/gelu.comp
diff --git a/ggml/src/vulkan-shaders/gelu_quick.comp b/ggml_llama/src/vulkan-shaders/gelu_quick.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/gelu_quick.comp
rename to ggml_llama/src/vulkan-shaders/gelu_quick.comp
diff --git a/ggml/src/vulkan-shaders/generic_binary_head.comp b/ggml_llama/src/vulkan-shaders/generic_binary_head.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/generic_binary_head.comp
rename to ggml_llama/src/vulkan-shaders/generic_binary_head.comp
diff --git a/ggml/src/vulkan-shaders/generic_head.comp b/ggml_llama/src/vulkan-shaders/generic_head.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/generic_head.comp
rename to ggml_llama/src/vulkan-shaders/generic_head.comp
diff --git a/ggml/src/vulkan-shaders/generic_unary_head.comp b/ggml_llama/src/vulkan-shaders/generic_unary_head.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/generic_unary_head.comp
rename to ggml_llama/src/vulkan-shaders/generic_unary_head.comp
diff --git a/ggml/src/vulkan-shaders/get_rows.comp b/ggml_llama/src/vulkan-shaders/get_rows.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/get_rows.comp
rename to ggml_llama/src/vulkan-shaders/get_rows.comp
diff --git a/ggml/src/vulkan-shaders/get_rows_quant.comp b/ggml_llama/src/vulkan-shaders/get_rows_quant.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/get_rows_quant.comp
rename to ggml_llama/src/vulkan-shaders/get_rows_quant.comp
diff --git a/ggml/src/vulkan-shaders/group_norm.comp b/ggml_llama/src/vulkan-shaders/group_norm.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/group_norm.comp
rename to ggml_llama/src/vulkan-shaders/group_norm.comp
diff --git a/ggml/src/vulkan-shaders/im2col.comp b/ggml_llama/src/vulkan-shaders/im2col.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/im2col.comp
rename to ggml_llama/src/vulkan-shaders/im2col.comp
diff --git a/ggml/src/vulkan-shaders/leaky_relu.comp b/ggml_llama/src/vulkan-shaders/leaky_relu.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/leaky_relu.comp
rename to ggml_llama/src/vulkan-shaders/leaky_relu.comp
diff --git a/ggml/src/vulkan-shaders/mul.comp b/ggml_llama/src/vulkan-shaders/mul.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul.comp
rename to ggml_llama/src/vulkan-shaders/mul.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_split_k_reduce.comp b/ggml_llama/src/vulkan-shaders/mul_mat_split_k_reduce.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_split_k_reduce.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_split_k_reduce.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_base.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_base.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_base.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_base.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_nc.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_nc.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_nc.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_nc.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_p021.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_p021.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_p021.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_p021.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_q2_k.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_q2_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_q2_k.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_q2_k.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_q3_k.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_q3_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_q3_k.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_q3_k.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_q4_k.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_q4_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_q4_k.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_q4_k.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_q5_k.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_q5_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_q5_k.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_q5_k.comp
diff --git a/ggml/src/vulkan-shaders/mul_mat_vec_q6_k.comp b/ggml_llama/src/vulkan-shaders/mul_mat_vec_q6_k.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mat_vec_q6_k.comp
rename to ggml_llama/src/vulkan-shaders/mul_mat_vec_q6_k.comp
diff --git a/ggml/src/vulkan-shaders/mul_mm.comp b/ggml_llama/src/vulkan-shaders/mul_mm.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/mul_mm.comp
rename to ggml_llama/src/vulkan-shaders/mul_mm.comp
diff --git a/ggml/src/vulkan-shaders/norm.comp b/ggml_llama/src/vulkan-shaders/norm.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/norm.comp
rename to ggml_llama/src/vulkan-shaders/norm.comp
diff --git a/ggml/src/vulkan-shaders/pad.comp b/ggml_llama/src/vulkan-shaders/pad.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/pad.comp
rename to ggml_llama/src/vulkan-shaders/pad.comp
diff --git a/ggml/src/vulkan-shaders/relu.comp b/ggml_llama/src/vulkan-shaders/relu.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/relu.comp
rename to ggml_llama/src/vulkan-shaders/relu.comp
diff --git a/ggml/src/vulkan-shaders/repeat.comp b/ggml_llama/src/vulkan-shaders/repeat.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/repeat.comp
rename to ggml_llama/src/vulkan-shaders/repeat.comp
diff --git a/ggml/src/vulkan-shaders/rms_norm.comp b/ggml_llama/src/vulkan-shaders/rms_norm.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/rms_norm.comp
rename to ggml_llama/src/vulkan-shaders/rms_norm.comp
diff --git a/ggml/src/vulkan-shaders/rope_head.comp b/ggml_llama/src/vulkan-shaders/rope_head.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/rope_head.comp
rename to ggml_llama/src/vulkan-shaders/rope_head.comp
diff --git a/ggml/src/vulkan-shaders/rope_neox.comp b/ggml_llama/src/vulkan-shaders/rope_neox.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/rope_neox.comp
rename to ggml_llama/src/vulkan-shaders/rope_neox.comp
diff --git a/ggml/src/vulkan-shaders/rope_norm.comp b/ggml_llama/src/vulkan-shaders/rope_norm.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/rope_norm.comp
rename to ggml_llama/src/vulkan-shaders/rope_norm.comp
diff --git a/ggml/src/vulkan-shaders/scale.comp b/ggml_llama/src/vulkan-shaders/scale.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/scale.comp
rename to ggml_llama/src/vulkan-shaders/scale.comp
diff --git a/ggml/src/vulkan-shaders/silu.comp b/ggml_llama/src/vulkan-shaders/silu.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/silu.comp
rename to ggml_llama/src/vulkan-shaders/silu.comp
diff --git a/ggml/src/vulkan-shaders/sin.comp b/ggml_llama/src/vulkan-shaders/sin.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/sin.comp
rename to ggml_llama/src/vulkan-shaders/sin.comp
diff --git a/ggml/src/vulkan-shaders/soft_max.comp b/ggml_llama/src/vulkan-shaders/soft_max.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/soft_max.comp
rename to ggml_llama/src/vulkan-shaders/soft_max.comp
diff --git a/ggml/src/vulkan-shaders/square.comp b/ggml_llama/src/vulkan-shaders/square.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/square.comp
rename to ggml_llama/src/vulkan-shaders/square.comp
diff --git a/ggml/src/vulkan-shaders/sum_rows.comp b/ggml_llama/src/vulkan-shaders/sum_rows.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/sum_rows.comp
rename to ggml_llama/src/vulkan-shaders/sum_rows.comp
diff --git a/ggml/src/vulkan-shaders/tanh.comp b/ggml_llama/src/vulkan-shaders/tanh.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/tanh.comp
rename to ggml_llama/src/vulkan-shaders/tanh.comp
diff --git a/ggml/src/vulkan-shaders/timestep_embedding.comp b/ggml_llama/src/vulkan-shaders/timestep_embedding.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/timestep_embedding.comp
rename to ggml_llama/src/vulkan-shaders/timestep_embedding.comp
diff --git a/ggml/src/vulkan-shaders/types.comp b/ggml_llama/src/vulkan-shaders/types.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/types.comp
rename to ggml_llama/src/vulkan-shaders/types.comp
diff --git a/ggml/src/vulkan-shaders/upscale.comp b/ggml_llama/src/vulkan-shaders/upscale.comp
similarity index 100%
rename from ggml/src/vulkan-shaders/upscale.comp
rename to ggml_llama/src/vulkan-shaders/upscale.comp
diff --git a/ggml/src/vulkan-shaders/vulkan-shaders-gen.cpp b/ggml_llama/src/vulkan-shaders/vulkan-shaders-gen.cpp
similarity index 100%
rename from ggml/src/vulkan-shaders/vulkan-shaders-gen.cpp
rename to ggml_llama/src/vulkan-shaders/vulkan-shaders-gen.cpp
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index 46a6ad562..caf0a334b 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -25,7 +25,7 @@ add_library(llama
 target_include_directories(llama PUBLIC . ../include)
 target_compile_features   (llama PUBLIC cxx_std_11) # don't bump
 
-target_link_libraries(llama PUBLIC ggml)
+target_link_libraries(llama PUBLIC ggml_llama)
 
 if (BUILD_SHARED_LIBS)
     set_target_properties(llama PROPERTIES POSITION_INDEPENDENT_CODE ON)