Fix formatting and workflow

Author: vshampor

.github/workflows/llama_cpp_plugin_build_and_test.yml

@@ -29,7 +29,7 @@ jobs:
         with:
           submodules: recursive
           repository: vshampor/openvino
-          branch: llama_cpp_mod
+          ref: llama_cpp_mod
           path: openvino
 
       - name: CMake - configure

modules/llama_cpp_plugin/.clang-format

@@ -0,0 +1,28 @@
+BasedOnStyle: Google
+IndentWidth: 4
+UseTab: Never
+ColumnLimit: 120
+
+Language: Cpp
+Standard: Cpp11
+
+AccessModifierOffset: -4
+AlignConsecutiveMacros: true
+AllowAllArgumentsOnNextLine: false
+AllowAllConstructorInitializersOnNextLine: false
+AllowAllParametersOfDeclarationOnNextLine: false
+AllowShortFunctionsOnASingleLine: Empty
+AllowShortIfStatementsOnASingleLine: Never
+AllowShortLambdasOnASingleLine: Empty
+AllowShortLoopsOnASingleLine: false
+AlwaysBreakBeforeMultilineStrings: false
+BinPackArguments: false
+BinPackParameters: false
+CommentPragmas: '^#'
+DerivePointerAlignment: false
+FixNamespaceComments: true
+IndentCaseLabels: false
+IndentPPDirectives: AfterHash
+ForEachMacros:
+  - foreach
+  - FOREACH_CHILD

modules/llama_cpp_plugin/src/compiled_model.cpp

@@ -18,14 +18,16 @@ LlamaCppModel::LlamaCppModel(const std::shared_ptr<ov::Model>& model,
                              const ov::SoPtr<ov::IRemoteContext>& context,
                              const std::shared_ptr<ov::threading::ITaskExecutor>& task_executor)
     : ICompiledModel(model, plugin, context, task_executor) {
-    OPENVINO_THROW_NOT_IMPLEMENTED("Currently only direct GGUF file loading is supported for the LLAMA_CPP* plugins");
+    OPENVINO_THROW_NOT_IMPLEMENTED("Currently only direct GGUF file loading is "
+                                   "supported for the LLAMA_CPP* plugins");
 }
 
 LlamaCppModel::LlamaCppModel(const std::shared_ptr<ov::Model>& ov_model,
                              std::istream& input_stream,
                              const std::shared_ptr<const IPlugin>& plugin)
     : ICompiledModel(ov_model, plugin) {
-    OPENVINO_THROW_NOT_IMPLEMENTED("Currently only direct GGUF file loading is supported for the LLAMA_CPP* plugins");
+    OPENVINO_THROW_NOT_IMPLEMENTED("Currently only direct GGUF file loading is "
+                                   "supported for the LLAMA_CPP* plugins");
 }
 
 LlamaCppModel::~LlamaCppModel() {

modules/llama_cpp_plugin/src/infer_request.cpp

@@ -60,11 +60,11 @@ void llama_batch_add_reimpl(struct llama_batch& batch,
 }
 
 void LlamaCppSyncInferRequest::infer() {
-    auto input_ids_tensor_ptr = get_tensor(get_inputs()[0]);  // TODO (vshampor) correctly identify input_ids among
-                                                              // all inputs without hardcode
-                                                              //
+    auto input_ids_tensor_ptr = get_tensor(get_inputs()[0]);     // TODO (vshampor) correctly identify input_ids among
+                                                                 // all inputs without hardcode
+                                                                 //
     auto position_ids_tensor_ptr = get_tensor(get_inputs()[2]);  // TODO (vshampor) correctly identify input_ids among
-                                                              // all inputs without hardcode
+                                                                 // all inputs without hardcode
     OPENVINO_ASSERT(input_ids_tensor_ptr->get_element_type() == ov::element::Type_t::i64);
     OPENVINO_ASSERT(input_ids_tensor_ptr->get_shape().size() == 2);
     size_t sequence_length = input_ids_tensor_ptr->get_shape()[1];

modules/llama_cpp_plugin/tests/e2e/prompt_response.cpp

@@ -7,7 +7,8 @@ const std::string TEST_FILES_DIR = "test_data";
 
 // "Why is the Sun yellow?"
 const std::vector<int64_t> GPT2_PROMPT_TOKEN_IDS = {5195, 318, 262, 3825, 7872, 30};
-// "The Sun is a bright red, which means it is a bright red. The Sun is a bright red because it is a bright red."
+// "The Sun is a bright red, which means it is a bright red. The Sun is a bright
+// red because it is a bright red."
 const std::vector<int64_t> GPT2_REFERENCE_RESPONSE_TOKEN_IDS = {
     198, 464,  3825, 318, 257,  6016, 2266, 11,  543, 1724, 340,  318,  257, 6016, 2266, 13,
     383, 3825, 318,  257, 6016, 2266, 780,  340, 318, 257,  6016, 2266, 13,  198,  198,  464};