llama-cpp-python 0.3.3__tar.gz → 0.3.5__tar.gz

llama_cpp_python-0.3.5/.git/FETCH_HEAD

@@ -0,0 +1 @@
+803924b02cc8ca7531b70e6497d1ff7b83f75737		'803924b02cc8ca7531b70e6497d1ff7b83f75737' of https://github.com/abetlen/llama-cpp-python

llama_cpp_python-0.3.5/.git/HEAD

@@ -0,0 +1 @@
+803924b02cc8ca7531b70e6497d1ff7b83f75737

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.git/config

@@ -9,7 +9,7 @@
 [gc]
 	auto = 0
 [http "https://github.com/"]
-	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzXzVPWUlOYkl6RzBpbEU5eE9mU2k0RTN5VExjMVRGdzNPRTIxbQ==
+	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX2dVbGl2b0FqTGMyVEF6UXIwQTFwOGJtZE1EY0l3RDNLMktYOA==
 [submodule "vendor/llama.cpp"]
 	active = true
 	url = https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.3.5/.git/logs/HEAD

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 803924b02cc8ca7531b70e6497d1ff7b83f75737 runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789569 +0000	checkout: moving from master to refs/tags/v0.3.5

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/HEAD

@@ -0,0 +1 @@
+26a8406ba9198eb6fdd8329fa717555b4f77f05f

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.git/modules/vendor/llama.cpp/config

@@ -16,7 +16,7 @@
 [gc]
 	auto = 0
 [http "https://github.com/"]
-	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzXzVPWUlOYkl6RzBpbEU5eE9mU2k0RTN5VExjMVRGdzNPRTIxbQ==
+	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX2dVbGl2b0FqTGMyVEF6UXIwQTFwOGJtZE1EY0l3RDNLMktYOA==
 [url "https://github.com/"]
 	insteadOf = git@github.com:
 	insteadOf = org-6826477@github.com:

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/logs/HEAD

@@ -0,0 +1,2 @@
+0000000000000000000000000000000000000000 26a8406ba9198eb6fdd8329fa717555b4f77f05f runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789572 +0000	clone: from https://github.com/ggerganov/llama.cpp.git
+26a8406ba9198eb6fdd8329fa717555b4f77f05f 26a8406ba9198eb6fdd8329fa717555b4f77f05f runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789573 +0000	checkout: moving from master to 26a8406ba9198eb6fdd8329fa717555b4f77f05f

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/logs/refs/heads/master

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 26a8406ba9198eb6fdd8329fa717555b4f77f05f runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789572 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/logs/refs/remotes/origin/HEAD

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 26a8406ba9198eb6fdd8329fa717555b4f77f05f runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789572 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.git/modules/vendor/llama.cpp/modules/kompute/config

@@ -13,7 +13,7 @@
 [gc]
 	auto = 0
 [http "https://github.com/"]
-	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzXzVPWUlOYkl6RzBpbEU5eE9mU2k0RTN5VExjMVRGdzNPRTIxbQ==
+	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX2dVbGl2b0FqTGMyVEF6UXIwQTFwOGJtZE1EY0l3RDNLMktYOA==
 [url "https://github.com/"]
 	insteadOf = git@github.com:
 	insteadOf = org-6826477@github.com:

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/modules/kompute/logs/HEAD

@@ -0,0 +1,2 @@
+0000000000000000000000000000000000000000 7c20efa30bb53d08bf04f84e510275766ebe9923 runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789573 +0000	clone: from https://github.com/nomic-ai/kompute.git
+7c20efa30bb53d08bf04f84e510275766ebe9923 4565194ed7c32d1d2efa32ceab4d3c6cae006306 runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789574 +0000	checkout: moving from master to 4565194ed7c32d1d2efa32ceab4d3c6cae006306

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/modules/kompute/logs/refs/heads/master

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 7c20efa30bb53d08bf04f84e510275766ebe9923 runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789573 +0000	clone: from https://github.com/nomic-ai/kompute.git

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/modules/kompute/logs/refs/remotes/origin/HEAD

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 7c20efa30bb53d08bf04f84e510275766ebe9923 runner <runner@fv-az575-826.txkdwr3xd4ruhezurzmahy12zg.cx.internal.cloudapp.net> 1733789573 +0000	clone: from https://github.com/nomic-ai/kompute.git

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/packed-refs

@@ -0,0 +1,2 @@
+# pack-refs with: peeled fully-peeled sorted 
+26a8406ba9198eb6fdd8329fa717555b4f77f05f refs/remotes/origin/master

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/refs/heads/master

@@ -0,0 +1 @@
+26a8406ba9198eb6fdd8329fa717555b4f77f05f

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/refs/tags/b4295

@@ -0,0 +1 @@
+26a8406ba9198eb6fdd8329fa717555b4f77f05f

llama_cpp_python-0.3.5/.git/modules/vendor/llama.cpp/shallow

@@ -0,0 +1 @@
+26a8406ba9198eb6fdd8329fa717555b4f77f05f

llama_cpp_python-0.3.5/.git/refs/tags/v0.3.5

@@ -0,0 +1 @@
+803924b02cc8ca7531b70e6497d1ff7b83f75737

llama_cpp_python-0.3.5/.git/shallow

@@ -0,0 +1 @@
+803924b02cc8ca7531b70e6497d1ff7b83f75737

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.github/workflows/build-and-release.yaml

@@ -11,7 +11,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        os: [ubuntu-20.04, windows-2019, macos-12]
+        os: [ubuntu-20.04, windows-2019, macos-13]
 
     steps:
       - uses: actions/checkout@v4

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.github/workflows/build-wheels-cuda.yaml

@@ -22,7 +22,7 @@ jobs:
           $matrix = @{
               'os' = @('ubuntu-latest', 'windows-2019')
               'pyver' = @("3.9", "3.10", "3.11", "3.12")
-              'cuda' = @("12.1.1", "12.2.2", "12.3.2", "12.4.1", "12.5.1", "12.6.1")
+              'cuda' = @("12.1.1", "12.2.2", "12.3.2", "12.4.1") #, "12.5.1", "12.6.1")
               'releasetag' = @("basic")
           }
 

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.github/workflows/build-wheels-metal.yaml

@@ -11,7 +11,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        os: [macos-12, macos-13, macos-14]
+        os: [macos-13, macos-14, macos-15]
 
     steps:
       - uses: actions/checkout@v4

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/.github/workflows/generate-index-from-release.yaml

@@ -44,8 +44,8 @@ jobs:
           ./scripts/releases-to-pep-503.sh index/whl/cu122 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu122$'
           ./scripts/releases-to-pep-503.sh index/whl/cu123 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu123$'
           ./scripts/releases-to-pep-503.sh index/whl/cu124 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu124$'
-          ./scripts/releases-to-pep-503.sh index/whl/cu125 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu124$'
-          ./scripts/releases-to-pep-503.sh index/whl/cu126 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu124$'
+          # ./scripts/releases-to-pep-503.sh index/whl/cu125 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu124$'
+          # ./scripts/releases-to-pep-503.sh index/whl/cu126 '^[v]?[0-9]+\.[0-9]+\.[0-9]+-cu124$'
           ./scripts/releases-to-pep-503.sh index/whl/metal '^[v]?[0-9]+\.[0-9]+\.[0-9]+-metal$'
       - name: Upload artifact
         uses: actions/upload-pages-artifact@v3

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/CHANGELOG.md

@@ -7,6 +7,16 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.3.5]
+
+- feat: Update llama.cpp to ggerganov/llama.cpp@26a8406ba9198eb6fdd8329fa717555b4f77f05f
+- fix(ci): Fix release by updating macos runner image to non-deprecated version by @abetlen in afedfc888462f9a6e809dc9455eb3b663764cc3f
+- fix(server): add missing await statements for async exit_stack handling by @gjpower in #1858
+
+## [0.3.4]
+
+- fix(ci): Build wheels for macos 13-15, cuda 12.1-12.4 by @abetlen in ca808028bd16b8327bd84128d48015a4b1304690
+
 ## [0.3.3]
 
 - feat: Update llama.cpp to ggerganov/llama.cpp@ce8784bdb153ff7794dde5a50b0ebfa51baa6171

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: llama_cpp_python
-Version: 0.3.3
+Version: 0.3.5
 Summary: Python bindings for the llama.cpp library
 Author-Email: Andrei Betlen <abetlen@gmail.com>
 License: MIT

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/llama_cpp/__init__.py

@@ -1,4 +1,4 @@
 from .llama_cpp import *
 from .llama import *
 
-__version__ = "0.3.3"
+__version__ = "0.3.5"

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/llama_cpp/server/app.py

@@ -318,7 +318,7 @@ async def create_completion(
             Iterator[llama_cpp.CreateCompletionStreamResponse],
         ] = await run_in_threadpool(llama, **kwargs)
     except Exception as err:
-        exit_stack.close()
+        await exit_stack.aclose()
         raise err
 
     if isinstance(iterator_or_completion, Iterator):
@@ -475,7 +475,7 @@ async def create_chat_completion(
     # is complete.
     # https://github.com/tiangolo/fastapi/issues/11143
     exit_stack = contextlib.AsyncExitStack()
-    llama_proxy = exit_stack.enter_async_context(contextlib.asynccontextmanager(get_llama_proxy)())
+    llama_proxy = await exit_stack.enter_async_context(contextlib.asynccontextmanager(get_llama_proxy)())
     if llama_proxy is None:
         raise HTTPException(
             status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
@@ -513,7 +513,7 @@ async def create_chat_completion(
             llama_cpp.ChatCompletion, Iterator[llama_cpp.ChatCompletionChunk]
         ] = await run_in_threadpool(llama.create_chat_completion, **kwargs)
     except Exception as err:
-        exit_stack.close()
+        await exit_stack.aclose()
         raise err
 
     if isinstance(iterator_or_completion, Iterator):

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/vendor/llama.cpp/CMakeLists.txt

@@ -46,11 +46,9 @@ if (WIN32)
     add_compile_definitions(_CRT_SECURE_NO_WARNINGS)
 endif()
 
-if ("${CMAKE_CXX_COMPILER_ID}" STREQUAL "MSVC")
-    add_compile_options("$<$<COMPILE_LANGUAGE:C>:/source-charset:utf-8>")
-    add_compile_options("$<$<COMPILE_LANGUAGE:CXX>:/source-charset:utf-8>")
-    add_compile_options("$<$<COMPILE_LANGUAGE:C>:/execution-charset:utf-8>")
-    add_compile_options("$<$<COMPILE_LANGUAGE:CXX>:/execution-charset:utf-8>")
+if (MSVC)
+    add_compile_options("$<$<COMPILE_LANGUAGE:C>:/utf-8>")
+    add_compile_options("$<$<COMPILE_LANGUAGE:CXX>:/utf-8>")
 endif()
 
 #

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/vendor/llama.cpp/CMakePresets.json

@@ -31,6 +31,13 @@
     { "name": "sycl_f16", "hidden": true, "cacheVariables": { "GGML_SYCL_F16":    "ON" } },
     { "name": "vulkan",   "hidden": true, "cacheVariables": { "GGML_VULKAN":      "ON" } },
 
+    {
+        "name": "x64-windows-llvm", "hidden": true,
+        "cacheVariables": {
+            "CMAKE_TOOLCHAIN_FILE": "${sourceDir}/cmake/x64-windows-llvm.cmake"
+        }
+    },
+
     {
         "name": "arm64-windows-msvc", "hidden": true,
         "architecture": { "value": "arm64",    "strategy": "external" },
@@ -70,6 +77,11 @@
     { "name": "arm64-windows-msvc-release", "inherits": [ "base", "arm64-windows-msvc",  "reldbg" ] },
     { "name": "arm64-windows-msvc+static-release", "inherits": [ "base", "arm64-windows-msvc",  "reldbg", "static" ] },
 
+    { "name": "x64-windows-llvm-debug", "inherits": [ "base", "x64-windows-llvm", "debug" ] },
+    { "name": "x64-windows-llvm-release", "inherits": [ "base", "x64-windows-llvm", "release" ] },
+    { "name": "x64-windows-llvm-reldbg", "inherits": [ "base", "x64-windows-llvm", "reldbg" ] },
+    { "name": "x64-windows-llvm+static-release", "inherits": [ "base", "x64-windows-llvm", "reldbg", "static" ] },
+
     { "name": "x64-windows-msvc-debug", "inherits": [ "base", "debug" ] },
     { "name": "x64-windows-msvc-release", "inherits": [ "base", "reldbg" ] },
     { "name": "x64-windows-msvc+static-release", "inherits": [ "base", "reldbg", "static" ] },

llama_cpp_python-0.3.5/vendor/llama.cpp/cmake/x64-windows-llvm.cmake

@@ -0,0 +1,11 @@
+set( CMAKE_SYSTEM_NAME Windows )
+set( CMAKE_SYSTEM_PROCESSOR x86_64 )
+
+set( CMAKE_C_COMPILER    clang )
+set( CMAKE_CXX_COMPILER  clang++ )
+
+set( arch_c_flags "-march=native" )
+
+set( CMAKE_C_FLAGS_INIT   "${arch_c_flags}" )
+set( CMAKE_CXX_FLAGS_INIT "${arch_c_flags}" )
+

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/vendor/llama.cpp/common/build-info.cpp

@@ -1,4 +1,4 @@
 int LLAMA_BUILD_NUMBER = 1;
-char const *LLAMA_COMMIT = "ce8784b";
+char const *LLAMA_COMMIT = "26a8406";
 char const *LLAMA_COMPILER = "gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0";
 char const *LLAMA_BUILD_TARGET = "x86_64-linux-gnu";

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/vendor/llama.cpp/docs/build.md

@@ -57,6 +57,13 @@ cmake --build build --config Release
     ```
     Building for arm64 can also be done with the MSVC compiler with the build-arm64-windows-MSVC preset, or the standard CMake build instructions. However, note that the MSVC compiler does not support inline ARM assembly code, used e.g. for the accelerated Q4_0_N_M CPU kernels.
 
+    For building with ninja generator and clang compiler as default:
+      -set path:set LIB=C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\um\x64;C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\MSVC\14.41.34120\lib\x64\uwp;C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\ucrt\x64
+      ```bash
+      cmake --preset x64-windows-llvm-release
+      cmake --build build-x64-windows-llvm-release
+      ```
+
 ## BLAS Build
 
 Building the program with BLAS support may lead to some performance improvements in prompt processing using batch sizes higher than 32 (the default is 512). Using BLAS doesn't affect the generation performance. There are currently several different BLAS implementations available for build and use:

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/vendor/llama.cpp/ggml/src/ggml-cuda/mmv.cu

@@ -57,7 +57,7 @@ static __global__ void mul_mat_vec(
     if (block_size > WARP_SIZE) {
         buf_iw[tid/WARP_SIZE] = sumf;
         __syncthreads();
-        if (tid > WARP_SIZE) {
+        if (tid >= WARP_SIZE) {
             return;
         }
         sumf = buf_iw[tid];

{llama_cpp_python-0.3.3 → llama_cpp_python-0.3.5}/vendor/llama.cpp/ggml/src/ggml-vulkan/ggml-vulkan.cpp

@@ -427,7 +427,7 @@ static_assert(sizeof(vk_op_unary_push_constants) <= 128, "sizeof(vk_op_unary_pus
 // and a shift:
 //
 // n/d = (mulhi(n, mp) + n) >> L;
-void init_fastdiv_values(uint32_t d, uint32_t &mp, uint32_t &L)
+static void init_fastdiv_values(uint32_t d, uint32_t &mp, uint32_t &L)
 {
     // compute L = ceil(log2(d));
     L = 0;
@@ -439,6 +439,7 @@ void init_fastdiv_values(uint32_t d, uint32_t &mp, uint32_t &L)
 }
 
 template <typename T> void init_pushconst_fastdiv(T &p) {
+    GGML_UNUSED(p);
     static_assert(!std::is_const<T>::value, "unexpected type");
 }
 
@@ -3417,7 +3418,7 @@ static uint32_t ggml_vk_guess_split_k(ggml_backend_vk_context * ctx, int m, int
     return split_k;
 }
 
-static vk_pipeline ggml_vk_guess_matmul_pipeline(ggml_backend_vk_context * ctx, vk_matmul_pipeline& mmp, int m, int n, bool aligned, ggml_type type_a) {
+static vk_pipeline ggml_vk_guess_matmul_pipeline(ggml_backend_vk_context * ctx, vk_matmul_pipeline& mmp, int m, int n, bool aligned) {
     VK_LOG_DEBUG("ggml_vk_guess_matmul_pipeline(" << m << ", " << n << ", " << aligned << ")");
 
     if (ctx->device->coopmat2) {
@@ -3439,9 +3440,9 @@ static vk_pipeline ggml_vk_guess_matmul_pipeline(ggml_backend_vk_context * ctx,
     return aligned ? mmp->a_l : mmp->l;
 }
 
-static uint32_t ggml_vk_guess_matmul_pipeline_align(ggml_backend_vk_context * ctx, vk_matmul_pipeline& mmp, int m, int n, ggml_type type_a) {
+static uint32_t ggml_vk_guess_matmul_pipeline_align(ggml_backend_vk_context * ctx, vk_matmul_pipeline& mmp, int m, int n) {
     VK_LOG_DEBUG("ggml_vk_guess_matmul_pipeline_align(" << m << ", " << n << ")");
-    return ggml_vk_guess_matmul_pipeline(ctx, mmp, m, n, true, type_a)->align;
+    return ggml_vk_guess_matmul_pipeline(ctx, mmp, m, n, true)->align;
 }
 
 static void ggml_vk_matmul(
@@ -3571,6 +3572,7 @@ static void ggml_vk_cpy_to_contiguous(ggml_backend_vk_context * ctx, vk_context&
         (uint32_t)tensor->ne[0], (uint32_t)tensor->ne[1], (uint32_t)tensor->ne[2], (uint32_t)tensor->ne[3],                       1                   , (uint32_t)tensor->ne[0]                   , (uint32_t)(tensor->ne[0] * tensor->ne[1]) , (uint32_t)(tensor->ne[0] * tensor->ne[1] * tensor->ne[2]),
         0,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     };
     init_pushconst_fastdiv(pc);
     ggml_vk_sync_buffers(subctx);
@@ -3644,10 +3646,10 @@ static void ggml_vk_mul_mat_q_f16(ggml_backend_vk_context * ctx, vk_context& sub
     const int y_ne = ne11 * ne10;
     const int d_ne = ne11 * ne01;
 
-    const uint32_t kpad = ggml_vk_align_size(ne10, ggml_vk_guess_matmul_pipeline_align(ctx, mmp, ne01, ne11, src0->type));
+    const uint32_t kpad = ggml_vk_align_size(ne10, ggml_vk_guess_matmul_pipeline_align(ctx, mmp, ne01, ne11));
     const bool aligned = ne10 == kpad && ne01 > 8 && ne11 > 8;
 
-    vk_pipeline pipeline = ggml_vk_guess_matmul_pipeline(ctx, mmp, ne01, ne11, aligned, src0->type);
+    vk_pipeline pipeline = ggml_vk_guess_matmul_pipeline(ctx, mmp, ne01, ne11, aligned);
 
     const uint32_t split_k = ggml_vk_guess_split_k(ctx, ne01, ne11, ne10, pipeline);
 
@@ -5351,7 +5353,8 @@ static void ggml_vk_scale(ggml_backend_vk_context * ctx, vk_context& subctx, con
         (uint32_t)src0->ne[0], (uint32_t)src0->ne[1], (uint32_t)src0->ne[2], (uint32_t)src0->ne[3], (uint32_t)src0->nb[0] / src0_type_size, (uint32_t)src0->nb[1] / src0_type_size, (uint32_t)src0->nb[2] / src0_type_size, (uint32_t)src0->nb[3] / src0_type_size,
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
-        op_params[0], 0.0f
+        op_params[0], 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5365,6 +5368,7 @@ static void ggml_vk_sqr(ggml_backend_vk_context * ctx, vk_context& subctx, const
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5378,6 +5382,7 @@ static void ggml_vk_sin(ggml_backend_vk_context * ctx, vk_context& subctx, const
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5391,6 +5396,7 @@ static void ggml_vk_cos(ggml_backend_vk_context * ctx, vk_context& subctx, const
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5405,6 +5411,7 @@ static void ggml_vk_clamp(ggml_backend_vk_context * ctx, vk_context& subctx, con
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
         op_params[0], op_params[1],
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5418,6 +5425,7 @@ static void ggml_vk_pad(ggml_backend_vk_context * ctx, vk_context& subctx, const
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5431,6 +5439,7 @@ static void ggml_vk_repeat(ggml_backend_vk_context * ctx, vk_context& subctx, co
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         0,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 
@@ -5445,6 +5454,7 @@ static void ggml_vk_cpy(ggml_backend_vk_context * ctx, vk_context& subctx, const
         (uint32_t) dst->ne[0], (uint32_t) dst->ne[1], (uint32_t) dst->ne[2], (uint32_t) dst->ne[3], (uint32_t) dst->nb[0] /  dst_type_size, (uint32_t) dst->nb[1] /  dst_type_size, (uint32_t) dst->nb[2] /  dst_type_size, (uint32_t) dst->nb[3] /  dst_type_size,
         d_offset,
         0.0f, 0.0f,
+        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
     }, dryrun);
 }
 

llama_cpp_python-0.3.3/.git/FETCH_HEAD

@@ -1 +0,0 @@
-a9fe0f872527ffc1bf64e125b7996ec4dfa598eb		'a9fe0f872527ffc1bf64e125b7996ec4dfa598eb' of https://github.com/abetlen/llama-cpp-python

llama_cpp_python-0.3.3/.git/HEAD

@@ -1 +0,0 @@
-a9fe0f872527ffc1bf64e125b7996ec4dfa598eb

llama_cpp_python-0.3.3/.git/logs/HEAD

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 a9fe0f872527ffc1bf64e125b7996ec4dfa598eb runner <runner@fv-az711-799.(none)> 1733717754 +0000	checkout: moving from master to refs/tags/v0.3.3

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/HEAD

@@ -1 +0,0 @@
-ce8784bdb153ff7794dde5a50b0ebfa51baa6171

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/logs/HEAD

@@ -1,2 +0,0 @@
-0000000000000000000000000000000000000000 ce8784bdb153ff7794dde5a50b0ebfa51baa6171 runner <runner@fv-az711-799.(none)> 1733717756 +0000	clone: from https://github.com/ggerganov/llama.cpp.git
-ce8784bdb153ff7794dde5a50b0ebfa51baa6171 ce8784bdb153ff7794dde5a50b0ebfa51baa6171 runner <runner@fv-az711-799.(none)> 1733717757 +0000	checkout: moving from master to ce8784bdb153ff7794dde5a50b0ebfa51baa6171

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/logs/refs/heads/master

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 ce8784bdb153ff7794dde5a50b0ebfa51baa6171 runner <runner@fv-az711-799.(none)> 1733717756 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/logs/refs/remotes/origin/HEAD

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 ce8784bdb153ff7794dde5a50b0ebfa51baa6171 runner <runner@fv-az711-799.(none)> 1733717756 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/modules/kompute/logs/HEAD

@@ -1,2 +0,0 @@
-0000000000000000000000000000000000000000 7c20efa30bb53d08bf04f84e510275766ebe9923 runner <runner@fv-az711-799.(none)> 1733717758 +0000	clone: from https://github.com/nomic-ai/kompute.git
-7c20efa30bb53d08bf04f84e510275766ebe9923 4565194ed7c32d1d2efa32ceab4d3c6cae006306 runner <runner@fv-az711-799.(none)> 1733717759 +0000	checkout: moving from master to 4565194ed7c32d1d2efa32ceab4d3c6cae006306

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/modules/kompute/logs/refs/heads/master

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 7c20efa30bb53d08bf04f84e510275766ebe9923 runner <runner@fv-az711-799.(none)> 1733717758 +0000	clone: from https://github.com/nomic-ai/kompute.git

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/modules/kompute/logs/refs/remotes/origin/HEAD

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 7c20efa30bb53d08bf04f84e510275766ebe9923 runner <runner@fv-az711-799.(none)> 1733717758 +0000	clone: from https://github.com/nomic-ai/kompute.git

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/packed-refs

@@ -1,2 +0,0 @@
-# pack-refs with: peeled fully-peeled sorted 
-ce8784bdb153ff7794dde5a50b0ebfa51baa6171 refs/remotes/origin/master

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/refs/heads/master

@@ -1 +0,0 @@
-ce8784bdb153ff7794dde5a50b0ebfa51baa6171

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/refs/tags/b4291

@@ -1 +0,0 @@
-ce8784bdb153ff7794dde5a50b0ebfa51baa6171

llama_cpp_python-0.3.3/.git/modules/vendor/llama.cpp/shallow

@@ -1 +0,0 @@
-ce8784bdb153ff7794dde5a50b0ebfa51baa6171

llama_cpp_python-0.3.3/.git/refs/tags/v0.3.3

@@ -1 +0,0 @@
-a9fe0f872527ffc1bf64e125b7996ec4dfa598eb

llama_cpp_python-0.3.3/.git/shallow

@@ -1 +0,0 @@
-a9fe0f872527ffc1bf64e125b7996ec4dfa598eb