happyz
/
gemma.cpp
mirror of https://github.com/google/gemma.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge 1dfd43a520 into 1d18c5a129

This commit is contained in:
Charles Chan 2024-04-30 12:11:04 +08:00 committed by GitHub
parent 1d18c5a129 1dfd43a520
commit 94ba6ca957
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 343 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
CMakeLists.txt
View File

@ -88,6 +88,15 @@ target_compile_options(libgemma PRIVATE $<$<PLATFORM_ID:Windows>:-Wno-deprecated
add_executable(gemma gemma/run.cc)
target_link_libraries(gemma libgemma hwy hwy_contrib)
option (WITH_WEBSOCKET "Run with websocket" OFF)
IF(WITH_WEBSOCKET)
FetchContent_Declare(IXWebSocket GIT_REPOSITORY https://github.com/machinezone/IXWebSocket GIT_TAG dc8807ec9d25f602d94f023e40a896de820d2bee EXCLUDE_FROM_ALL)
FetchContent_MakeAvailable(IXWebSocket)
add_definitions(-DWITH_WEBSOCKET)
add_executable(gemma.ws gemma/run-ws.cc)
target_link_libraries(gemma.ws libgemma hwy hwy_contrib ixwebsocket)
ENDIF(WITH_WEBSOCKET)
add_executable(benchmark gemma/benchmark.cc)
target_link_libraries(benchmark libgemma hwy hwy_contrib nlohmann_json::nlohmann_json)

328
gemma/run-ws.cc Normal file
View File

@ -0,0 +1,328 @@
// Copyright 2024 Google LLC
// SPDX-License-Identifier: Apache-2.0
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// https://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
// Command line text interface to gemma.
#include <ctime>
#include <iostream>
#include <random>
#include <sstream>
#include <string>
#include <thread> // NOLINT
#include <vector>
// Placeholder for internal header, do not modify.
#include <ixwebsocket/IXSocket.h>
#include <ixwebsocket/IXSocketFactory.h>
#include <ixwebsocket/IXWebSocket.h>
#include <ixwebsocket/IXWebSocketServer.h>
#include "compression/compress.h"
#include "gemma/gemma.h" // Gemma
#include "hwy/base.h"
#include "hwy/contrib/thread_pool/thread_pool.h"
#include "hwy/highway.h"
#include "hwy/per_target.h"
#include "hwy/profiler.h"
#include "hwy/timer.h"
#include "util/app.h"
#include "util/args.h" // HasHelp
/** test client by python
#!/usr/bin/python
#!pip install websocket-client
import websocket
ws = websocket.WebSocket()
ws.connect("ws://127.0.0.1:9999")
print(ws.recv())
ws.send("Hello, Gemma!")
while True:
token = ws.recv()
if token == "%EOS%":
break
print(token, end='')
ws.close()
*/
static constexpr bool kVerboseLogTokens = false;
namespace gcpp {
static constexpr std::string_view kAsciiArtBanner = R""(
__ _ ___ _ __ ___ _ __ ___ __ _ ___ _ __ _ __
/ _` |/ _ \ '_ ` _ \| '_ ` _ \ / _` | / __| '_ \| '_ \
| (_| | __/ | | | | | | | | | | (_| || (__| |_) | |_) |
\__, |\___|_| |_| |_|_| |_| |_|\__,_(_)___| .__/| .__/
__/ | | | | |
|___/ |_| |_|
)"";
void ShowHelp(gcpp::LoaderArgs& loader, gcpp::InferenceArgs& inference,
gcpp::AppArgs& app) {
std::cerr
<< kAsciiArtBanner
<< "\n\ngemma.cpp : a lightweight, standalone C++ inference engine\n"
"==========================================================\n\n"
"To run gemma.cpp, you need to "
"specify 3 required model loading arguments:\n --tokenizer\n "
"--compressed_weights\n"
" --model.\n";
std::cerr << "\n*Example Usage*\n\n./gemma --tokenizer tokenizer.spm "
"--compressed_weights 2b-it-sfp.sbs --model 2b-it\n";
std::cerr << "\n*Model Loading Arguments*\n\n";
loader.Help();
std::cerr << "\n*Inference Arguments*\n\n";
inference.Help();
std::cerr << "\n*Application Arguments*\n\n";
app.Help();
std::cerr << "\n";
}
void ReplGemma(gcpp::Gemma& model, ModelTraining training,
gcpp::KVCache& kv_cache, hwy::ThreadPool& pool,
hwy::ThreadPool& inner_pool, const InferenceArgs& args,
int verbosity, const gcpp::AcceptFunc& accept_token,
std::string& eot_line, int port) {
PROFILER_ZONE("Gen.misc");
size_t abs_pos = 0; // absolute token index over all turns
std::mt19937 gen;
if (args.deterministic) {
gen.seed(42);
} else {
std::random_device rd;
gen.seed(rd());
}
ix::WebSocketServer ws(port);
ws.setOnClientMessageCallback(
[&](std::shared_ptr<ix::ConnectionState> connectionState,
ix::WebSocket& webSocket, const ix::WebSocketMessagePtr& msg) {
auto toClient = [&](std::string token_text) {
for (auto&& client : ws.getClients()) {
if (client.get() == &webSocket) {
client->send(token_text, false);
}
}
};
if (msg->type == ix::WebSocketMessageType::Open) {
if (verbosity >= 2) {
std::cout << "New connection" << std::endl;
}
std::stringstream ntf;
time_t now = time(nullptr);
char* dt = ctime(&now); // NOLINT
ntf << "*Usage*\n"
" - Enter an instruction and press enter (%C resets "
"conversation, "
"%Q quits).\n\n"
"*Examples*\n"
" - Write an email to grandma thanking her for the cookies.\n"
" - What are some historical attractions to visit around "
"Massachusetts?\n"
" - Compute the nth fibonacci number in javascript.\n"
" - Write a standup comedy bit about GPU programming.\n"
"\n";
ntf << "*Config*\n"
<< "- Date & Time : " << dt
<< "- Prefill Token Batch Size : " << gcpp::kPrefillBatchSize
<< "\n"
<< "- Hardware concurrency : "
<< std::thread::hardware_concurrency() << "\n"
<< "- Instruction set : "
<< hwy::TargetName(hwy::DispatchedTarget()) << " ("
<< hwy::VectorBytes() * 8 << " bits)"
<< "\n"
<< "- Compiled config : " << CompiledConfig()
<< "\n"
<< "- Weight Type : "
<< gcpp::TypeName(gcpp::GemmaWeightT()) << "\n"
<< "- EmbedderInput Type : "
<< gcpp::TypeName(gcpp::EmbedderInputT()) << "\n";
toClient(ntf.str());
} else if (msg->type == ix::WebSocketMessageType::Close) {
if (verbosity >= 2) {
std::cout << "Closed connection" << std::endl;
}
} else if (msg->type == ix::WebSocketMessageType::Message) {
int current_pos = 0; // token index within the current turn
int prompt_size{};
// callback function invoked for each generated token.
auto stream_token = [&abs_pos, &current_pos, &args, &gen,
&prompt_size, tokenizer = model.Tokenizer(),
verbosity, &ws, &webSocket,
&toClient](int token, float) {
++abs_pos;
++current_pos;
// <= since position is incremented before
if (current_pos <= prompt_size) {
std::cerr << "." << std::flush;
toClient(".");
} else if (token == gcpp::EOS_ID) {
if (!args.multiturn) {
abs_pos = 0;
if (args.deterministic) {
gen.seed(42);
}
}
if (verbosity >= 2) {
std::cout << "\n[ End ]\n";
}
toClient("%EOS%");
} else {
std::string token_text;
HWY_ASSERT(
tokenizer->Decode(std::vector<int>{token}, &token_text));
// +1 since position is incremented above
if (current_pos == prompt_size + 1) {
// first token of response
token_text.erase(0, token_text.find_first_not_of(" \t\n"));
if (verbosity >= 1) {
std::cout << "\n\n";
}
}
// std::cout << token_text << std::flush;
toClient(token_text);
}
return true;
};
std::string prompt_string = msg->str;
if (abs_pos >= args.max_tokens) {
ws.stop();
} else if (prompt_string == "%q" || prompt_string == "%Q") {
ws.stop();
} else if (prompt_string == "%c" || prompt_string == "%C") {
abs_pos = 0;
} else {
std::vector<int> prompt;
current_pos = 0;
if (training == ModelTraining::GEMMA_IT) {
// For instruction-tuned models: add control tokens.
prompt_string = "<start_of_turn>user\n" + prompt_string +
"<end_of_turn>\n<start_of_turn>model\n";
if (abs_pos != 0) {
// Prepend "<end_of_turn>" token if this is a multi-turn
// dialogue continuation.
prompt_string = "<end_of_turn>\n" + prompt_string;
}
}
HWY_ASSERT(model.Tokenizer()->Encode(prompt_string, &prompt));
// For both pre-trained and instruction-tuned models: prepend
// "<bos>" token if needed.
if (abs_pos == 0) {
prompt.insert(prompt.begin(), 2);
}
prompt_size = prompt.size();
std::cerr << "\n"
<< "[ Reading prompt ] " << std::flush;
if constexpr (kVerboseLogTokens) {
for (int i = 0; i < static_cast<int>(prompt.size()); ++i) {
fprintf(stderr, "DDD TOKEN %3d: %6d\n", i, prompt[i]);
}
}
const double time_start = hwy::platform::Now();
GenerateGemma(model, args.max_tokens, args.max_generated_tokens,
args.temperature, prompt, abs_pos, kv_cache, pool,
inner_pool, stream_token, accept_token, gen,
verbosity);
const double time_end = hwy::platform::Now();
const double tok_sec = current_pos / (time_end - time_start);
if (verbosity >= 2) {
std::cout << current_pos << " tokens (" << abs_pos
<< " total tokens)"
<< "\n"
<< tok_sec << " tokens / sec"
<< "\n";
}
std::cout << "\n\n";
}
}
});
ws.listenAndStart();
std::cout << "Listening on " << port << std::endl;
ws.wait();
std::cout
<< "max_tokens (" << args.max_tokens
<< ") exceeded. Use a larger value if desired using the --max_tokens "
<< "command line flag.\n";
}
void Run(LoaderArgs& loader, InferenceArgs& inference, AppArgs& app) {
PROFILER_ZONE("Run.misc");
hwy::ThreadPool inner_pool(0);
hwy::ThreadPool pool(app.num_threads);
// For many-core, pinning threads to cores helps.
if (app.num_threads > 10) {
PinThreadToCore(app.num_threads - 1); // Main thread
pool.Run(0, pool.NumThreads(),
[](uint64_t /*task*/, size_t thread) { PinThreadToCore(thread); });
}
gcpp::Gemma model(loader.tokenizer, loader.weights, loader.ModelType(), pool);
auto kv_cache = CreateKVCache(loader.ModelType());
if (const char* error = inference.Validate()) {
ShowHelp(loader, inference, app);
HWY_ABORT("\nInvalid args: %s", error);
}
ReplGemma(
model, loader.ModelTraining(), kv_cache, pool, inner_pool, inference,
app.verbosity,
/*accept_token=*/[](int) { return true; }, app.eot_line, app.port);
}
} // namespace gcpp
int main(int argc, char** argv) {
{
PROFILER_ZONE("Startup.misc");
// Placeholder for internal init, do not modify.
gcpp::LoaderArgs loader(argc, argv);
gcpp::InferenceArgs inference(argc, argv);
gcpp::AppArgs app(argc, argv);
if (gcpp::HasHelp(argc, argv)) {
ShowHelp(loader, inference, app);
return 0;
}
if (const char* error = loader.Validate()) {
ShowHelp(loader, inference, app);
HWY_ABORT("\nInvalid args: %s", error);
}
gcpp::Run(loader, inference, app);
}
PROFILER_PRINT_RESULTS(); // Must call outside the zone above.
return 0;
}

6
util/app.h
View File

@ -104,6 +104,9 @@ class AppArgs : public ArgsBase<AppArgs> {
int verbosity;
size_t num_threads;
std::string eot_line;
#ifdef WITH_WEBSOCKET
int port;
#endif
template <class Visitor>
void ForEach(const Visitor& visitor) {
@ -124,6 +127,9 @@ class AppArgs : public ArgsBase<AppArgs> {
"before the line where only the given string appears.\n Default = "
"When a newline is encountered, that signals the end of the turn.",
2);
#ifdef WITH_WEBSOCKET
visitor(port, "port", 9999, "Port for WebSocket.\n Default = 9999.", 2);
#endif
}
};