From 51cb0a89cfed4212d5d91972ad9fd772af9f17d0 Mon Sep 17 00:00:00 2001 From: Anastasiia Pnevskaia Date: Fri, 20 Dec 2024 17:25:11 +0100 Subject: [PATCH] Fixed error. --- .../speculative_decoding_lm.cpp | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/samples/cpp/speculative_decoding_lm/speculative_decoding_lm.cpp b/samples/cpp/speculative_decoding_lm/speculative_decoding_lm.cpp index ca5a60ec93..aa4ce0eed0 100644 --- a/samples/cpp/speculative_decoding_lm/speculative_decoding_lm.cpp +++ b/samples/cpp/speculative_decoding_lm/speculative_decoding_lm.cpp @@ -6,9 +6,9 @@ #include "openvino/genai/llm_pipeline.hpp" int main(int argc, char* argv[]) try { - if (4 != argc) { - throw std::runtime_error(std::string{"Usage: "} + argv[0] + " ''"); - } + // if (4 != argc) { + // throw std::runtime_error(std::string{"Usage: "} + argv[0] + " ''"); + // } ov::genai::GenerationConfig config; config.max_new_tokens = 100; @@ -18,9 +18,9 @@ int main(int argc, char* argv[]) try { // add parameter to enable speculative decoding to generate candidates by draft_model while candidate probability is higher than `assistant_confidence_threshold` // config.assistant_confidence_threshold = 0.4; - std::string main_model_path = argv[1]; - std::string draft_model_path = argv[2]; - std::string prompt = argv[3]; + std::string main_model_path = "/home/panas/test_models/spec_dec/Llama-2-7b-chat-hf/"; + std::string draft_model_path = "/home/panas/test_models/spec_dec/tiny-llama-1.1b-chat/"; + std::string prompt = "What is openvino?"; // User can run main and draft model on different devices. // Please, set device for main model in `LLMPipeline` constructor and in in `ov::genai::draft_model` for draft.