llama-cpp-python 0.2.35__tar.gz → 0.2.36__tar.gz

llama_cpp_python-0.2.36/.git/FETCH_HEAD

@@ -0,0 +1 @@
+464af5b39fea3cf1ba16e755a9df85f09bbb25ac		'464af5b39fea3cf1ba16e755a9df85f09bbb25ac' of https://github.com/abetlen/llama-cpp-python

llama_cpp_python-0.2.36/.git/HEAD

@@ -0,0 +1 @@
+464af5b39fea3cf1ba16e755a9df85f09bbb25ac

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/.git/config

@@ -9,7 +9,7 @@
 [gc]
 	auto = 0
 [http "https://github.com/"]
-	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX0QwUkJscUZWb0tUcW80YmZabVZwSDJXcUtGVUw5RzFXZkZLaw==
+	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX1Y2cUFITVRmZ0hoM1VTanhaR2VKRTRhUGxXUHJjRzQwRWhTSg==
 [submodule "vendor/llama.cpp"]
 	active = true
 	url = https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.2.36/.git/logs/HEAD

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 464af5b39fea3cf1ba16e755a9df85f09bbb25ac runner <runner@fv-az695-903.gmayhch2cacunn5b1cp1zlasod.dx.internal.cloudapp.net> 1706543240 +0000	checkout: moving from master to refs/tags/v0.2.36

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/HEAD

@@ -0,0 +1 @@
+2aed77eb06a329f0d82bb1c467f4244904d4073f

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/.git/modules/vendor/llama.cpp/config

@@ -13,7 +13,7 @@
 [gc]
 	auto = 0
 [http "https://github.com/"]
-	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX0QwUkJscUZWb0tUcW80YmZabVZwSDJXcUtGVUw5RzFXZkZLaw==
+	extraheader = AUTHORIZATION: basic eC1hY2Nlc3MtdG9rZW46Z2hzX1Y2cUFITVRmZ0hoM1VTanhaR2VKRTRhUGxXUHJjRzQwRWhTSg==
 [url "https://github.com/"]
 	insteadOf = git@github.com:
 	insteadOf = org-6826477@github.com:

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/logs/HEAD

@@ -0,0 +1,2 @@
+0000000000000000000000000000000000000000 2aed77eb06a329f0d82bb1c467f4244904d4073f runner <runner@fv-az695-903.gmayhch2cacunn5b1cp1zlasod.dx.internal.cloudapp.net> 1706543241 +0000	clone: from https://github.com/ggerganov/llama.cpp.git
+2aed77eb06a329f0d82bb1c467f4244904d4073f 2aed77eb06a329f0d82bb1c467f4244904d4073f runner <runner@fv-az695-903.gmayhch2cacunn5b1cp1zlasod.dx.internal.cloudapp.net> 1706543241 +0000	checkout: moving from master to 2aed77eb06a329f0d82bb1c467f4244904d4073f

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/logs/refs/heads/master

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 2aed77eb06a329f0d82bb1c467f4244904d4073f runner <runner@fv-az695-903.gmayhch2cacunn5b1cp1zlasod.dx.internal.cloudapp.net> 1706543241 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/logs/refs/remotes/origin/HEAD

@@ -0,0 +1 @@
+0000000000000000000000000000000000000000 2aed77eb06a329f0d82bb1c467f4244904d4073f runner <runner@fv-az695-903.gmayhch2cacunn5b1cp1zlasod.dx.internal.cloudapp.net> 1706543241 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/packed-refs

@@ -0,0 +1,2 @@
+# pack-refs with: peeled fully-peeled sorted 
+2aed77eb06a329f0d82bb1c467f4244904d4073f refs/remotes/origin/master

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/refs/heads/master

@@ -0,0 +1 @@
+2aed77eb06a329f0d82bb1c467f4244904d4073f

llama_cpp_python-0.2.36/.git/modules/vendor/llama.cpp/shallow

@@ -0,0 +1 @@
+2aed77eb06a329f0d82bb1c467f4244904d4073f

llama_cpp_python-0.2.36/.git/objects/46/4af5b39fea3cf1ba16e755a9df85f09bbb25ac

@@ -0,0 +1,3 @@
+x
��]
+�0�}�)�J�6M@D{�M��B���z~�G�m���i-e��p��s�$�F"g��AYJ�hU�|�5��QPӔ)3{b)�)Ԟ�b<,���Z<jj<��}�
+W��p��˴��A:�F9g�R��;�5�^6�p{�k_c�H+

llama_cpp_python-0.2.36/.git/objects/80/6b120c5e62a7be5fecc631e341d1856dac79a5

@@ -0,0 +1,2 @@
+x
�TMo1����r��E����6�F��HU[!�ڀ�wmٻT(��ػҲ������<ϛT�>|���]e8+E�t��݃,\ɔ"��,�Ui�y��4�T�SjEVY'w��z+ډ�k�B�z�S��,ʌ��+�z�R
+��kY@�\),!\�ߗ]|�����T��e\x��%����u	I+���w¦�	���egx�fߍ����|Х�$������k���Bk��$�E��D/�x�j��/��dߪ�*٦9�M���(l�8����?~a}r�����h'�gW�-+Z4��xr���ҕ�y���Ű7�Vha������W�����Ҍ���]��

llama_cpp_python-0.2.36/.git/refs/tags/v0.2.36

@@ -0,0 +1 @@
+464af5b39fea3cf1ba16e755a9df85f09bbb25ac

llama_cpp_python-0.2.36/.git/shallow

@@ -0,0 +1 @@
+464af5b39fea3cf1ba16e755a9df85f09bbb25ac

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/CHANGELOG.md

@@ -7,6 +7,11 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.2.36]
+
+- feat: Update llama.cpp to ggerganov/llama.cpp@2aed77eb06a329f0d82bb1c467f4244904d4073f
+- feat: Add mistral instruct chat format as "mistral-instruct" by @Rafaelblsilva in #799
+
 ## [0.2.35]
 
 - feat: Update llama.cpp to ggerganov/llama.cpp@d2f650cb5b04ee2726663e79b47da5efe196ce00

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/Makefile

@@ -27,6 +27,9 @@ build.blis:
 build.metal:
 	CMAKE_ARGS="-DLLAMA_METAL=on" python3 -m pip install --verbose -e .
 
+build.vulkan:
+	CMAKE_ARGS="-DLLAMA_VULKAN=on" python3 -m pip install --verbose -e .
+
 build.sdist:
 	python3 -m build --sdist
 

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: llama_cpp_python
-Version: 0.2.35
+Version: 0.2.36
 Summary: Python bindings for the llama.cpp library
 Author-Email: Andrei Betlen <abetlen@gmail.com>
 License: MIT

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/llama_cpp/__init__.py

@@ -1,4 +1,4 @@
 from .llama_cpp import *
 from .llama import *
 
-__version__ = "0.2.35"
+__version__ = "0.2.36"

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/llama_cpp/llama_chat_format.py

@@ -877,6 +877,24 @@ def format_chatml(
     return ChatFormatterResponse(prompt=_prompt, stop=_sep)
 
 
+@register_chat_format("mistral-instruct")
+def format_mistral_instruct(
+    messages: List[llama_types.ChatCompletionRequestMessage],
+    **kwargs: Any,
+) -> ChatFormatterResponse:
+    bos = "<s>"
+    eos = "</s>"
+    stop = eos
+    prompt = bos
+    for message in messages:
+        if message["role"] == "user" and message["content"] is not None and isinstance(message["content"], str):
+            prompt += "[INST] " + message["content"]
+        elif message["role"] == "assistant" and message["content"] is not None and isinstance(message["content"], str):
+            prompt += " [/INST]" + message["content"] + eos
+    prompt += " [/INST]"
+    return ChatFormatterResponse(prompt=prompt, stop=stop)
+
+
 @register_chat_format("chatglm3")
 def format_chatglm3(
     messages: List[llama_types.ChatCompletionRequestMessage],

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/tests/test_llama_chat_format.py

@@ -1,10 +1,33 @@
 import json
 
+import jinja2
+
 from llama_cpp import (
     ChatCompletionRequestUserMessage,
 )
+import llama_cpp.llama_types as llama_types
+import llama_cpp.llama_chat_format as llama_chat_format
+
 from llama_cpp.llama_chat_format import hf_tokenizer_config_to_chat_formatter
 
+def test_mistral_instruct():
+    chat_template = "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token}}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}"
+    chat_formatter = jinja2.Template(chat_template)
+    messages = [
+        llama_types.ChatCompletionRequestUserMessage(role="user", content="Instruction"),
+        llama_types.ChatCompletionRequestAssistantMessage(role="assistant", content="Model answer"),
+        llama_types.ChatCompletionRequestUserMessage(role="user", content="Follow-up instruction"),
+    ]
+    response = llama_chat_format.format_mistral_instruct(
+        messages=messages,
+    )
+    reference = chat_formatter.render(
+        messages=messages,
+        bos_token="<s>",
+        eos_token="</s>",
+    )
+    assert response.prompt == reference
+
 
 mistral_7b_tokenizer_config = """{
   "add_bos_token": true,

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/CMakeLists.txt

@@ -422,7 +422,13 @@ if (LLAMA_VULKAN)
     if (Vulkan_FOUND)
         message(STATUS "Vulkan found")
 
+        set(GGML_HEADERS_VULKAN ggml-vulkan.h)
+        set(GGML_SOURCES_VULKAN ggml-vulkan.cpp)
+
         add_library(ggml-vulkan STATIC ggml-vulkan.cpp ggml-vulkan.h)
+        if (BUILD_SHARED_LIBS)
+            set_target_properties(ggml-vulkan PROPERTIES POSITION_INDEPENDENT_CODE ON)
+        endif()
         target_link_libraries(ggml-vulkan PRIVATE Vulkan::Vulkan)
 
         add_compile_definitions(GGML_USE_VULKAN)
@@ -848,6 +854,7 @@ add_library(ggml OBJECT
             ggml-quants.h
             ${GGML_SOURCES_CUDA}   ${GGML_HEADERS_CUDA}
             ${GGML_SOURCES_OPENCL} ${GGML_HEADERS_OPENCL}
+            ${GGML_SOURCES_VULKAN} ${GGML_HEADERS_VULKAN}
             ${GGML_SOURCES_METAL}  ${GGML_HEADERS_METAL}
             ${GGML_SOURCES_MPI}    ${GGML_HEADERS_MPI}
             ${GGML_SOURCES_EXTRA}  ${GGML_HEADERS_EXTRA}
@@ -928,7 +935,7 @@ install(FILES ${CMAKE_CURRENT_BINARY_DIR}/LlamaConfig.cmake
         DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/Llama)
 
 set(GGML_PUBLIC_HEADERS "ggml.h" "ggml-alloc.h" "ggml-backend.h"
-        "${GGML_HEADERS_CUDA}" "${GGML_HEADERS_OPENCL}"
+        "${GGML_HEADERS_CUDA}" "${GGML_HEADERS_OPENCL}" "${GGML_HEADERS_VULKAN}"
         "${GGML_HEADERS_METAL}" "${GGML_HEADERS_MPI}" "${GGML_HEADERS_EXTRA}")
 
 set_target_properties(ggml PROPERTIES PUBLIC_HEADER "${GGML_PUBLIC_HEADERS}")

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/common/build-info.cpp

@@ -1,4 +1,4 @@
 int LLAMA_BUILD_NUMBER = 1;
-char const *LLAMA_COMMIT = "d2f650c";
+char const *LLAMA_COMMIT = "2aed77e";
 char const *LLAMA_COMPILER = "cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0";
 char const *LLAMA_BUILD_TARGET = "x86_64-linux-gnu";

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/convert.py

@@ -334,7 +334,10 @@ class Params:
 class BpeVocab:
     def __init__(self, fname_tokenizer: Path, fname_added_tokens: Path | None) -> None:
         self.bpe_tokenizer = json.loads(open(str(fname_tokenizer), encoding="utf-8").read())
-        self.vocab = self.bpe_tokenizer["model"]["vocab"]
+        try:
+            self.vocab = self.bpe_tokenizer["model"]["vocab"]
+        except KeyError:
+            self.vocab = self.bpe_tokenizer
         added_tokens: dict[str, int]
         if fname_added_tokens is not None:
             # FIXME: Verify that added tokens here _cannot_ overlap with the main vocab.

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/examples/server/oai.hpp

@@ -206,3 +206,18 @@ inline static std::vector<json> format_partial_response_oaicompat(const task_res
 
     return std::vector<json>({ret});
 }
+
+inline static json format_embeddings_response_oaicompat(const json &request, const json &embeddings)
+{
+    json res =
+        json{
+            {"model", json_value(request, "model", std::string(DEFAULT_OAICOMPAT_MODEL))},
+            {"object", "list"},
+            {"usage",
+                json{{"prompt_tokens", 0},
+                     {"total_tokens", 0}}},
+            {"data", embeddings}
+        };
+    return res;
+}
+

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/examples/server/server.cpp

@@ -2929,6 +2929,66 @@ int main(int argc, char **argv)
                 return res.set_content(result.result_json.dump(), "application/json; charset=utf-8");
             });
 
+    svr.Post("/v1/embeddings", [&llama](const httplib::Request &req, httplib::Response &res)
+            {
+                res.set_header("Access-Control-Allow-Origin", req.get_header_value("Origin"));
+                const json body = json::parse(req.body);
+
+                json prompt;
+                if (body.count("input") != 0)
+                {
+                    prompt = body["input"];
+                    // batch
+                    if(prompt.is_array()) {
+                        json data = json::array();
+                        int i = 0;
+                        for (const json &elem : prompt) {
+                            const int task_id = llama.queue_tasks.get_new_id();
+                            llama.queue_results.add_waiting_task_id(task_id);
+                            llama.request_completion(task_id, { {"prompt", elem}, { "n_predict", 0} }, false, true, -1);
+
+                            // get the result
+                            task_result result = llama.queue_results.recv(task_id);
+                            llama.queue_results.remove_waiting_task_id(task_id);
+
+                            json embedding = json{
+                                {"embedding", json_value(result.result_json, "embedding", json::array())},
+                                {"index", i++},
+                                {"object", "embedding"}
+                            };
+                            data.push_back(embedding);
+                        }
+                        json result = format_embeddings_response_oaicompat(body, data);
+                        return res.set_content(result.dump(), "application/json; charset=utf-8");
+                    }
+                }
+                else
+                {
+                    prompt = "";
+                }
+
+                // create and queue the task
+                const int task_id = llama.queue_tasks.get_new_id();
+                llama.queue_results.add_waiting_task_id(task_id);
+                llama.request_completion(task_id, { {"prompt", prompt}, { "n_predict", 0}}, false, true, -1);
+
+                // get the result
+                task_result result = llama.queue_results.recv(task_id);
+                llama.queue_results.remove_waiting_task_id(task_id);
+
+                json data = json::array({json{
+                        {"embedding", json_value(result.result_json, "embedding", json::array())},
+                        {"index", 0},
+                        {"object", "embedding"}
+                    }}
+                );
+
+                json root = format_embeddings_response_oaicompat(body, data);
+
+                // send the result
+                return res.set_content(root.dump(), "application/json; charset=utf-8");
+            });
+
     // GG: if I put the main loop inside a thread, it crashes on the first request when build in Debug!?
     //     "Bus error: 10" - this is on macOS, it does not crash on Linux
     //std::thread t2([&]()

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/ggml-metal.m

@@ -2375,6 +2375,16 @@ GGML_CALL static size_t ggml_backend_metal_buffer_type_get_alignment(ggml_backen
     UNUSED(buft);
 }
 
+GGML_CALL static size_t ggml_backend_metal_buffer_type_get_max_size(ggml_backend_buffer_type_t buft) {
+    id<MTLDevice> device = ggml_backend_metal_get_device();
+    size_t max_size = device.maxBufferLength;
+    ggml_backend_metal_free_device();
+
+    return max_size;
+
+    UNUSED(buft);
+}
+
 GGML_CALL static bool ggml_backend_metal_buffer_type_supports_backend(ggml_backend_buffer_type_t buft, ggml_backend_t backend) {
     return ggml_backend_is_metal(backend) || ggml_backend_is_cpu(backend);
 
@@ -2393,7 +2403,7 @@ GGML_CALL ggml_backend_buffer_type_t ggml_backend_metal_buffer_type(void) {
             /* .get_name         = */ ggml_backend_metal_buffer_type_get_name,
             /* .alloc_buffer     = */ ggml_backend_metal_buffer_type_alloc_buffer,
             /* .get_alignment    = */ ggml_backend_metal_buffer_type_get_alignment,
-            /* .get_max_size     = */ NULL, // TODO: return device.maxBufferLength
+            /* .get_max_size     = */ ggml_backend_metal_buffer_type_get_max_size,
             /* .get_alloc_size   = */ NULL, // defaults to ggml_nbytes
             /* .supports_backend = */ ggml_backend_metal_buffer_type_supports_backend,
             /* .is_host          = */ ggml_backend_metal_buffer_type_is_host,

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/ggml-opencl.cpp

@@ -2125,6 +2125,15 @@ static size_t ggml_backend_opencl_buffer_type_get_alignment(ggml_backend_buffer_
     GGML_UNUSED(buffer_type);
 }
 
+static size_t ggml_backend_opencl_buffer_type_get_max_size(ggml_backend_buffer_type_t buffer_type) {
+    static size_t max_size = -1;
+    if (max_size == (size_t)-1) {
+        ggml_cl_init();
+        clGetDeviceInfo(device, CL_DEVICE_MAX_MEM_ALLOC_SIZE, sizeof(size_t), &max_size, NULL);
+    }
+    return max_size;
+}
+
 static bool ggml_backend_opencl_buffer_type_supports_backend(ggml_backend_buffer_type_t buffer_type, ggml_backend_t backend) {
     //return ggml_backend_is_opencl(backend); // opencl must be used through the cpu backend
     return ggml_backend_is_cpu(backend);
@@ -2136,7 +2145,7 @@ static ggml_backend_buffer_type_i ggml_backend_opencl_buffer_type_interface = {
     /* .get_name         = */ ggml_backend_opencl_buffer_type_name,
     /* .alloc_buffer     = */ ggml_backend_opencl_buffer_type_alloc_buffer,
     /* .get_alignment    = */ ggml_backend_opencl_buffer_type_get_alignment,
-    /* .get_max_size     = */ NULL, // TODO: return from device info
+    /* .get_max_size     = */ ggml_backend_opencl_buffer_type_get_max_size,
     /* .get_alloc_size   = */ NULL,
     /* .supports_backend = */ ggml_backend_opencl_buffer_type_supports_backend,
     /* .is_host          = */ NULL,

{llama_cpp_python-0.2.35 → llama_cpp_python-0.2.36}/vendor/llama.cpp/llama.cpp

@@ -1158,10 +1158,10 @@ struct llama_mlock {
     #ifdef __APPLE__
         #define MLOCK_SUGGESTION \
             "Try increasing the sysctl values 'vm.user_wire_limit' and 'vm.global_user_wire_limit' and/or " \
-            "decreasing 'vm.global_no_user_wire_amount'.  Also try increasing RLIMIT_MLOCK (ulimit -l).\n"
+            "decreasing 'vm.global_no_user_wire_amount'.  Also try increasing RLIMIT_MEMLOCK (ulimit -l).\n"
     #else
         #define MLOCK_SUGGESTION \
-            "Try increasing RLIMIT_MLOCK ('ulimit -l' as root).\n"
+            "Try increasing RLIMIT_MEMLOCK ('ulimit -l' as root).\n"
     #endif
 
     bool raw_lock(const void * addr, size_t size) const {

llama_cpp_python-0.2.35/.git/FETCH_HEAD

@@ -1 +0,0 @@
-52c4a84faf07d92565e0c4f77ba884a48fc24b52		'52c4a84faf07d92565e0c4f77ba884a48fc24b52' of https://github.com/abetlen/llama-cpp-python

llama_cpp_python-0.2.35/.git/HEAD

@@ -1 +0,0 @@
-52c4a84faf07d92565e0c4f77ba884a48fc24b52

llama_cpp_python-0.2.35/.git/logs/HEAD

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 52c4a84faf07d92565e0c4f77ba884a48fc24b52 runner <runner@fv-az1498-961.rffsckg10wduxcxr53tve02m4e.cx.internal.cloudapp.net> 1706488582 +0000	checkout: moving from master to refs/tags/v0.2.35

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/HEAD

@@ -1 +0,0 @@
-d2f650cb5b04ee2726663e79b47da5efe196ce00

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/logs/HEAD

@@ -1,2 +0,0 @@
-0000000000000000000000000000000000000000 d2f650cb5b04ee2726663e79b47da5efe196ce00 runner <runner@fv-az1498-961.rffsckg10wduxcxr53tve02m4e.cx.internal.cloudapp.net> 1706488583 +0000	clone: from https://github.com/ggerganov/llama.cpp.git
-d2f650cb5b04ee2726663e79b47da5efe196ce00 d2f650cb5b04ee2726663e79b47da5efe196ce00 runner <runner@fv-az1498-961.rffsckg10wduxcxr53tve02m4e.cx.internal.cloudapp.net> 1706488583 +0000	checkout: moving from master to d2f650cb5b04ee2726663e79b47da5efe196ce00

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/logs/refs/heads/master

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 d2f650cb5b04ee2726663e79b47da5efe196ce00 runner <runner@fv-az1498-961.rffsckg10wduxcxr53tve02m4e.cx.internal.cloudapp.net> 1706488583 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/logs/refs/remotes/origin/HEAD

@@ -1 +0,0 @@
-0000000000000000000000000000000000000000 d2f650cb5b04ee2726663e79b47da5efe196ce00 runner <runner@fv-az1498-961.rffsckg10wduxcxr53tve02m4e.cx.internal.cloudapp.net> 1706488583 +0000	clone: from https://github.com/ggerganov/llama.cpp.git

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/packed-refs

@@ -1,2 +0,0 @@
-# pack-refs with: peeled fully-peeled sorted 
-d2f650cb5b04ee2726663e79b47da5efe196ce00 refs/remotes/origin/master

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/refs/heads/master

@@ -1 +0,0 @@
-d2f650cb5b04ee2726663e79b47da5efe196ce00

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/refs/tags/b1999

@@ -1 +0,0 @@
-d2f650cb5b04ee2726663e79b47da5efe196ce00

llama_cpp_python-0.2.35/.git/modules/vendor/llama.cpp/shallow

@@ -1 +0,0 @@
-d2f650cb5b04ee2726663e79b47da5efe196ce00

llama_cpp_python-0.2.35/.git/refs/tags/v0.2.35

@@ -1 +0,0 @@
-52c4a84faf07d92565e0c4f77ba884a48fc24b52

llama_cpp_python-0.2.35/.git/shallow

@@ -1 +0,0 @@
-52c4a84faf07d92565e0c4f77ba884a48fc24b52