1) change c++14 request from global to grpc-server only

2) change proto to llama/v1 dir according to lint suggestion

CMakeLists.txt

@@ -116,7 +116,11 @@ endif()
 # Compile flags
 #
 
-set(CMAKE_CXX_STANDARD 14)
+if(LLAMA_BUILD_GRPC_SERVER)
+    set(CMAKE_CXX_STANDARD 11)
+else()
+    set(CMAKE_CXX_STANDARD 14)
+endif()
 set(CMAKE_CXX_STANDARD_REQUIRED true)
 set(CMAKE_C_STANDARD 11)
 set(CMAKE_C_STANDARD_REQUIRED true)

examples/grpc-server/CMakeLists.txt

@@ -15,7 +15,7 @@ message(STATUS "Using protobuf ${Protobuf_VERSION} ${Protobuf_INCLUDE_DIRS} ${CM
 
 
 # Proto file
-get_filename_component(hw_proto "./message.proto" ABSOLUTE)
+get_filename_component(hw_proto "./llama/v1/message.proto" ABSOLUTE)
 get_filename_component(hw_proto_path "${hw_proto}" PATH)
 
 # Generated sources

examples/grpc-server/README.md

@@ -0,0 +1,14 @@
+# llama grpc server
+
+service as a grpc server to completion and embedding (when `--embedding` argument is given) based on examples/server.
+
+## running service
+
+run grpc-server command using argument like main program of llama.cpp with the following change:
+
+* add `--host` argument to set the listening host
+* add `--port` argument to set the listening port
+
+### behaving differences with examples/server
+
+* grpc-server will always break when <eos> is the predicted token.

examples/grpc-server/grpc-server.cpp

@@ -41,9 +41,10 @@ using grpc::ServerContext;
 using grpc::ServerUnaryReactor;
 using grpc::ServerWriteReactor;
 using grpc::Status;
-using llama::Job;
-using llama::LlamaGoService;
-using llama::Output;
+using llama::v1::Request;
+using llama::v1::LlamaService;
+using llama::v1::EmbedResponse;
+using llama::v1::CompletionResponse;
 
 struct server_params
 {
@@ -497,13 +498,13 @@ private:
 };
 
 // Logic and data behind the server's behavior.
-class LlamaServiceImpl final : public LlamaGoService::CallbackService
+class LlamaServiceImpl final : public LlamaService::CallbackService
 {
 
-  class Reactor : public grpc::ServerWriteReactor<Output>
+  class Reactor : public grpc::ServerWriteReactor<CompletionResponse>
   {
   public:
-    Reactor(CallbackServerContext *ctx, LlamaServerContext *llama, const Job *request)
+    Reactor(CallbackServerContext *ctx, LlamaServerContext *llama, const Request *request)
         : ctx_(ctx), request_(request), llama_(llama)
     {
       if (llama->loadPrompt(request->prompt()))
@@ -534,22 +535,22 @@ class LlamaServiceImpl final : public LlamaGoService::CallbackService
   private:
     CallbackServerContext *const ctx_;
     LlamaServerContext *llama_;
-    const Job *const request_;
+    const Request *const request_;
     int n_remain{0};
     std::mutex finish_mu_;
     bool finished_{false};
-    Output *response;
+    CompletionResponse *response;
 
     void NextWrite()
     {
-      response = new Output();
+      response = new CompletionResponse();
       // loop inference until finish completion
       if (llama_->has_next_token)
       {
         std::lock_guard<std::mutex> l(finish_mu_);
         auto result = llama_->doCompletion();
         fprintf(stderr, "%s", result.c_str());
-        response->set_status(llama::Status::RUNNING);
+        response->set_status(llama::v1::Status::RUNNING);
         response->set_output(result);
         StartWrite(response);
       }
@@ -560,7 +561,7 @@ class LlamaServiceImpl final : public LlamaGoService::CallbackService
               l(finish_mu_);
           if (!finished_)
           {
-            response->set_status(llama::Status::FINISHED);
+            response->set_status(llama::v1::Status::FINISHED);
             StartWriteLast(response, grpc::WriteOptions());
           }
         }
@@ -585,8 +586,8 @@ public:
   {
   }
 
-  ServerWriteReactor<Output> *Answer(
-      CallbackServerContext *context, const Job *request)
+  ServerWriteReactor<CompletionResponse> *Complete(
+      CallbackServerContext *context, const Request *request)
   {
     fprintf(stderr, "%s : new answer request: %s\n", __func__, request->prompt().c_str());
     llama->rewind();
@@ -598,7 +599,7 @@ public:
   }
 
   ServerUnaryReactor *Embed(
-      CallbackServerContext *context, const Job *request, Output *response)
+      CallbackServerContext *context, const Request *request, EmbedResponse *response)
   {
     fprintf(stderr, "%s : get embed %s\n", __func__, request->prompt().c_str());
     std::vector<float> embeded = llama->embedding(request->prompt());

examples/grpc-server/llama/v1/message.proto

@@ -0,0 +1,33 @@
+syntax = "proto3";
+
+package llama.v1;
+
+option go_package = "./pkg/grpc";
+
+service LlamaService {
+    rpc Complete(Request) returns (stream CompletionResponse){}
+    rpc Embed(Request) returns (EmbedResponse){}
+}
+
+message Request {
+	string id = 1;  
+	string prompt = 2;     
+}
+
+enum Status {
+    PENDING_UNSPECIFIED = 0;
+    RUNNING = 1;
+    FINISHED = 2;
+    FAILED = 3;
+}
+
+message CompletionResponse {
+    string id = 1;
+    Status status = 2;
+    string output = 3;
+}
+
+message EmbedResponse {
+    string id = 1;
+    repeated float embed = 2;
+}

examples/grpc-server/message.proto

@@ -1,29 +0,0 @@
-syntax = "proto3";
-
-package llama;
-
-option go_package = "./pkg/grpc";
-
-service LlamaGoService {
-    rpc Answer(Job) returns (stream Output){}
-    rpc Embed(Job) returns (Output){}
-}
-
-message Job {
-	string id = 1;  
-	string prompt = 2;     
-}
-
-enum Status {
-    PENDING = 0;
-    RUNNING = 1;
-    FINISHED = 2;
-    FAILED = 3;
-}
-
-message Output {
-    string id = 1;
-    Status status = 2;
-    string output = 3;
-    repeated float embed = 4;
-}