From 0c0b36a33838161b769bd6aa71bbee0419215b32 Mon Sep 17 00:00:00 2001 From: Anastasiya Pronina Date: Tue, 31 Dec 2024 12:37:13 +0000 Subject: [PATCH] Fixed clang-format --- .../intel_npu/src/plugin/npuw/llm_compiled_model.cpp | 8 -------- .../intel_npu/src/plugin/npuw/llm_infer_request.cpp | 9 ++++----- 2 files changed, 4 insertions(+), 13 deletions(-) diff --git a/src/plugins/intel_npu/src/plugin/npuw/llm_compiled_model.cpp b/src/plugins/intel_npu/src/plugin/npuw/llm_compiled_model.cpp index 484b5373d7cb47..496fd8bd8fd8d7 100644 --- a/src/plugins/intel_npu/src/plugin/npuw/llm_compiled_model.cpp +++ b/src/plugins/intel_npu/src/plugin/npuw/llm_compiled_model.cpp @@ -330,14 +330,6 @@ std::optional pop_option(ov::AnyMap& config, const std::string& option_ return std::nullopt; } -template -std::optional get_option(ov::AnyMap& config, const std::string& option_name) { - if (auto it = config.find(option_name); it != config.end()) { - return std::make_optional(it->second.as()); - } - return std::nullopt; -} - ov::AnyMap get_baseline_common_config() { ov::AnyMap config = { {"NPU_COMPILATION_MODE_PARAMS", "compute-layers-with-higher-precision=Sqrt,Power,ReduceMean,Add_RMSNorm"}, diff --git a/src/plugins/intel_npu/src/plugin/npuw/llm_infer_request.cpp b/src/plugins/intel_npu/src/plugin/npuw/llm_infer_request.cpp index dda0c6a2945d74..0bad68a35aa4ba 100644 --- a/src/plugins/intel_npu/src/plugin/npuw/llm_infer_request.cpp +++ b/src/plugins/intel_npu/src/plugin/npuw/llm_infer_request.cpp @@ -148,11 +148,10 @@ void ov::npuw::LLMInferRequest::infer_generate(ov::SoPtr input_ids, ? 3u : kvcache_desc.dim; - auto prefill_out_slice = - make_tensor_slice(prefill_out_tensor, - kv_dim, - kvcache_desc.max_prompt_size - kvcache_desc.num_stored_tokens, - kvcache_desc.max_prompt_size); + auto prefill_out_slice = make_tensor_slice(prefill_out_tensor, + kv_dim, + kvcache_desc.max_prompt_size - kvcache_desc.num_stored_tokens, + kvcache_desc.max_prompt_size); auto kvcache_in_slice = make_tensor_slice(kvcache_in_tensor, kv_dim, 0u, kvcache_desc.num_stored_tokens);