@fugood/llama.node 0.3.16 → 0.4.0

package/src/llama.cpp/tools/rpc/rpc-server.cpp

@@ -0,0 +1,322 @@
+#if defined(_MSC_VER)
+#define _SILENCE_CXX17_CODECVT_HEADER_DEPRECATION_WARNING
+#endif
+
+#include "ggml-rpc.h"
+#ifdef _WIN32
+#  define NOMINMAX
+#  define DIRECTORY_SEPARATOR '\\'
+#  include <locale>
+#  include <windows.h>
+#  include <fcntl.h>
+#  include <io.h>
+#else
+#  define DIRECTORY_SEPARATOR '/'
+#  include <unistd.h>
+#  include <sys/stat.h>
+#endif
+#include <codecvt>
+#include <string>
+#include <stdio.h>
+#include <vector>
+#include <filesystem>
+#include <algorithm>
+#include <thread>
+
+namespace fs = std::filesystem;
+
+// NOTE: this is copied from common.cpp to avoid linking with libcommon
+// returns true if successful, false otherwise
+static bool fs_create_directory_with_parents(const std::string & path) {
+#ifdef _WIN32
+    std::wstring_convert<std::codecvt_utf8<wchar_t>> converter;
+    std::wstring wpath = converter.from_bytes(path);
+
+    // if the path already exists, check whether it's a directory
+    const DWORD attributes = GetFileAttributesW(wpath.c_str());
+    if ((attributes != INVALID_FILE_ATTRIBUTES) && (attributes & FILE_ATTRIBUTE_DIRECTORY)) {
+        return true;
+    }
+
+    size_t pos_slash = 0;
+
+    // process path from front to back, procedurally creating directories
+    while ((pos_slash = path.find('\\', pos_slash)) != std::string::npos) {
+        const std::wstring subpath = wpath.substr(0, pos_slash);
+        const wchar_t * test = subpath.c_str();
+
+        const bool success = CreateDirectoryW(test, NULL);
+        if (!success) {
+            const DWORD error = GetLastError();
+
+            // if the path already exists, ensure that it's a directory
+            if (error == ERROR_ALREADY_EXISTS) {
+                const DWORD attributes = GetFileAttributesW(subpath.c_str());
+                if (attributes == INVALID_FILE_ATTRIBUTES || !(attributes & FILE_ATTRIBUTE_DIRECTORY)) {
+                    return false;
+                }
+            } else {
+                return false;
+            }
+        }
+
+        pos_slash += 1;
+    }
+
+    return true;
+#else
+    // if the path already exists, check whether it's a directory
+    struct stat info;
+    if (stat(path.c_str(), &info) == 0) {
+        return S_ISDIR(info.st_mode);
+    }
+
+    size_t pos_slash = 1; // skip leading slashes for directory creation
+
+    // process path from front to back, procedurally creating directories
+    while ((pos_slash = path.find('/', pos_slash)) != std::string::npos) {
+        const std::string subpath = path.substr(0, pos_slash);
+        struct stat info;
+
+        // if the path already exists, ensure that it's a directory
+        if (stat(subpath.c_str(), &info) == 0) {
+            if (!S_ISDIR(info.st_mode)) {
+                return false;
+            }
+        } else {
+            // create parent directories
+            const int ret = mkdir(subpath.c_str(), 0755);
+            if (ret != 0) {
+                return false;
+            }
+        }
+
+        pos_slash += 1;
+    }
+
+    return true;
+#endif // _WIN32
+}
+
+// NOTE: this is copied from common.cpp to avoid linking with libcommon
+static std::string fs_get_cache_directory() {
+    std::string cache_directory = "";
+    auto ensure_trailing_slash = [](std::string p) {
+        // Make sure to add trailing slash
+        if (p.back() != DIRECTORY_SEPARATOR) {
+            p += DIRECTORY_SEPARATOR;
+        }
+        return p;
+    };
+    if (getenv("LLAMA_CACHE")) {
+        cache_directory = std::getenv("LLAMA_CACHE");
+    } else {
+#if defined(__linux__) || defined(__FreeBSD__) || defined(_AIX)
+        if (std::getenv("XDG_CACHE_HOME")) {
+            cache_directory = std::getenv("XDG_CACHE_HOME");
+        } else {
+            cache_directory = std::getenv("HOME") + std::string("/.cache/");
+        }
+#elif defined(__APPLE__)
+        cache_directory = std::getenv("HOME") + std::string("/Library/Caches/");
+#elif defined(_WIN32)
+        cache_directory = std::getenv("LOCALAPPDATA");
+#else
+#  error Unknown architecture
+#endif
+        cache_directory = ensure_trailing_slash(cache_directory);
+        cache_directory += "llama.cpp";
+    }
+    return ensure_trailing_slash(cache_directory);
+}
+
+struct rpc_server_params {
+    std::string host        = "127.0.0.1";
+    int         port        = 50052;
+    size_t      backend_mem = 0;
+    bool        use_cache   = false;
+    int         n_threads   = std::max(1U, std::thread::hardware_concurrency()/2);
+    std::string device;
+};
+
+static void print_usage(int /*argc*/, char ** argv, rpc_server_params params) {
+    fprintf(stderr, "Usage: %s [options]\n\n", argv[0]);
+    fprintf(stderr, "options:\n");
+    fprintf(stderr, "  -h, --help                show this help message and exit\n");
+    fprintf(stderr, "  -t,      --threads        number of threads for the CPU backend (default: %d)\n", params.n_threads);
+    fprintf(stderr, "  -d DEV,  --device         device to use\n");
+    fprintf(stderr, "  -H HOST, --host HOST      host to bind to (default: %s)\n", params.host.c_str());
+    fprintf(stderr, "  -p PORT, --port PORT      port to bind to (default: %d)\n", params.port);
+    fprintf(stderr, "  -m MEM,  --mem MEM        backend memory size (in MB)\n");
+    fprintf(stderr, "  -c,      --cache          enable local file cache\n");
+    fprintf(stderr, "\n");
+}
+
+static bool rpc_server_params_parse(int argc, char ** argv, rpc_server_params & params) {
+    std::string arg;
+    for (int i = 1; i < argc; i++) {
+        arg = argv[i];
+        if (arg == "-H" || arg == "--host") {
+            if (++i >= argc) {
+                return false;
+            }
+            params.host = argv[i];
+        } else if (arg == "-t" || arg == "--threads") {
+            if (++i >= argc) {
+                return false;
+            }
+            params.n_threads = std::stoi(argv[i]);
+            if (params.n_threads <= 0) {
+                fprintf(stderr, "error: invalid number of threads: %d\n", params.n_threads);
+                return false;
+            }
+        } else if (arg == "-d" || arg == "--device") {
+            if (++i >= argc) {
+                return false;
+            }
+            params.device = argv[i];
+            if (ggml_backend_dev_by_name(params.device.c_str()) == nullptr) {
+                fprintf(stderr, "error: unknown device: %s\n", params.device.c_str());
+                fprintf(stderr, "available devices:\n");
+                for (size_t i = 0; i < ggml_backend_dev_count(); i++) {
+                    auto * dev = ggml_backend_dev_get(i);
+                    size_t free, total;
+                    ggml_backend_dev_memory(dev, &free, &total);
+                    printf("  %s: %s (%zu MiB, %zu MiB free)\n", ggml_backend_dev_name(dev), ggml_backend_dev_description(dev), total / 1024 / 1024, free / 1024 / 1024);
+                }
+                return false;
+            }
+        } else if (arg == "-p" || arg == "--port") {
+            if (++i >= argc) {
+                return false;
+            }
+            params.port = std::stoi(argv[i]);
+            if (params.port <= 0 || params.port > 65535) {
+                return false;
+            }
+        } else if (arg == "-c" || arg == "--cache") {
+            params.use_cache = true;
+        } else if (arg == "-m" || arg == "--mem") {
+            if (++i >= argc) {
+                return false;
+            }
+            params.backend_mem = std::stoul(argv[i]) * 1024 * 1024;
+        } else if (arg == "-h" || arg == "--help") {
+            print_usage(argc, argv, params);
+            exit(0);
+        } else {
+            fprintf(stderr, "error: unknown argument: %s\n", arg.c_str());
+            print_usage(argc, argv, params);
+            exit(0);
+        }
+    }
+    return true;
+}
+
+static ggml_backend_t create_backend(const rpc_server_params & params) {
+    ggml_backend_t backend = nullptr;
+
+    if (!params.device.empty()) {
+        ggml_backend_dev_t dev = ggml_backend_dev_by_name(params.device.c_str());
+        if (dev) {
+            backend = ggml_backend_dev_init(dev, nullptr);
+            if (!backend) {
+                fprintf(stderr, "Failed to create backend for device %s\n", params.device.c_str());
+                return nullptr;
+            }
+        }
+    }
+
+    // try to initialize a GPU backend first
+    if (!backend) {
+        backend = ggml_backend_init_by_type(GGML_BACKEND_DEVICE_TYPE_GPU, nullptr);
+    }
+
+    // if there aren't GPU backends fallback to CPU backend
+    if (!backend) {
+        backend = ggml_backend_init_by_type(GGML_BACKEND_DEVICE_TYPE_CPU, nullptr);
+    }
+
+    if (backend) {
+        fprintf(stderr, "%s: using %s backend\n", __func__, ggml_backend_name(backend));
+
+        // set the number of threads
+        ggml_backend_dev_t dev = ggml_backend_get_device(backend);
+        ggml_backend_reg_t reg = dev ? ggml_backend_dev_backend_reg(dev) : nullptr;
+        if (reg) {
+            auto ggml_backend_set_n_threads_fn = (ggml_backend_set_n_threads_t) ggml_backend_reg_get_proc_address(reg, "ggml_backend_set_n_threads");
+            if (ggml_backend_set_n_threads_fn) {
+                ggml_backend_set_n_threads_fn(backend, params.n_threads);
+            }
+        }
+    }
+
+    return backend;
+}
+
+static void get_backend_memory(ggml_backend_t backend, size_t * free_mem, size_t * total_mem) {
+    ggml_backend_dev_t dev = ggml_backend_get_device(backend);
+    GGML_ASSERT(dev != nullptr);
+    ggml_backend_dev_memory(dev, free_mem, total_mem);
+}
+
+int main(int argc, char * argv[]) {
+    ggml_backend_load_all();
+
+    rpc_server_params params;
+    if (!rpc_server_params_parse(argc, argv, params)) {
+        fprintf(stderr, "Invalid parameters\n");
+        return 1;
+    }
+
+    if (params.host != "127.0.0.1") {
+        fprintf(stderr, "\n");
+        fprintf(stderr, "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n");
+        fprintf(stderr, "WARNING: Host ('%s') is != '127.0.0.1'\n", params.host.c_str());
+        fprintf(stderr, "         Never expose the RPC server to an open network!\n");
+        fprintf(stderr, "         This is an experimental feature and is not secure!\n");
+        fprintf(stderr, "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n");
+        fprintf(stderr, "\n");
+    }
+
+    ggml_backend_t backend = create_backend(params);
+    if (!backend) {
+        fprintf(stderr, "Failed to create backend\n");
+        return 1;
+    }
+    std::string endpoint = params.host + ":" + std::to_string(params.port);
+    size_t free_mem, total_mem;
+    if (params.backend_mem > 0) {
+        free_mem = params.backend_mem;
+        total_mem = params.backend_mem;
+    } else {
+        get_backend_memory(backend, &free_mem, &total_mem);
+    }
+    const char * cache_dir = nullptr;
+    std::string cache_dir_str;
+    if (params.use_cache) {
+        cache_dir_str = fs_get_cache_directory() + "rpc/";
+        if (!fs_create_directory_with_parents(cache_dir_str)) {
+            fprintf(stderr, "Failed to create cache directory: %s\n", cache_dir_str.c_str());
+            return 1;
+        }
+        cache_dir = cache_dir_str.c_str();
+    }
+
+    ggml_backend_reg_t reg = ggml_backend_reg_by_name("RPC");
+    if (!reg) {
+        fprintf(stderr, "Failed to find RPC backend\n");
+        return 1;
+    }
+
+    auto start_server_fn = (decltype(ggml_backend_rpc_start_server)*) ggml_backend_reg_get_proc_address(reg, "ggml_backend_rpc_start_server");
+    if (!start_server_fn) {
+        fprintf(stderr, "Failed to obtain RPC backend start server function\n");
+        return 1;
+    }
+
+    start_server_fn(backend, endpoint.c_str(), cache_dir, free_mem, total_mem);
+
+    ggml_backend_free(backend);
+    return 0;
+}

package/src/llama.cpp/tools/run/CMakeLists.txt

@@ -0,0 +1,16 @@
+set(TARGET llama-run)
+add_executable(${TARGET} run.cpp linenoise.cpp/linenoise.cpp)
+
+# TODO: avoid copying this code block from common/CMakeLists.txt
+set(LLAMA_RUN_EXTRA_LIBS "")
+if (LLAMA_CURL)
+    find_package(CURL REQUIRED)
+    target_compile_definitions(${TARGET} PUBLIC LLAMA_USE_CURL)
+    include_directories(${CURL_INCLUDE_DIRS})
+    find_library(CURL_LIBRARY curl REQUIRED)
+    set(LLAMA_RUN_EXTRA_LIBS ${LLAMA_RUN_EXTRA_LIBS} ${CURL_LIBRARY})
+endif ()
+
+install(TARGETS ${TARGET} RUNTIME)
+target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT} ${LLAMA_RUN_EXTRA_LIBS})
+target_compile_features(${TARGET} PRIVATE cxx_std_17)

package/src/llama.cpp/{examples → tools}/run/run.cpp

@@ -38,24 +38,6 @@
 }
 #endif
 
-GGML_ATTRIBUTE_FORMAT(1, 2)
-static std::string fmt(const char * fmt, ...) {
-    va_list ap;
-    va_list ap2;
-    va_start(ap, fmt);
-    va_copy(ap2, ap);
-    const int size = vsnprintf(NULL, 0, fmt, ap);
-    GGML_ASSERT(size >= 0 && size < INT_MAX); // NOLINT
-    std::string buf;
-    buf.resize(size);
-    const int size2 = vsnprintf(const_cast<char *>(buf.data()), buf.size() + 1, fmt, ap2);
-    GGML_ASSERT(size2 == size);
-    va_end(ap2);
-    va_end(ap);
-
-    return buf;
-}
-
 GGML_ATTRIBUTE_FORMAT(1, 2)
 static int printe(const char * fmt, ...) {
     va_list args;
@@ -285,7 +267,7 @@ class Opt {
             "Commands:\n"
             "  model\n"
             "      Model is a string with an optional prefix of \n"
-            "      huggingface:// (hf://), ollama://, https:// or file://.\n"
+            "      huggingface:// (hf://), modelscope:// (ms://), ollama://, https:// or file://.\n"
             "      If no protocol is specified and a file exists in the specified\n"
             "      path, file:// is assumed, otherwise if a file does not exist in\n"
             "      the specified path, ollama:// is assumed. Models that are being\n"
@@ -300,6 +282,9 @@ class Opt {
             "  llama-run hf://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf\n"
             "  llama-run "
             "huggingface://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf\n"
+            "  llama-run ms://QuantFactory/SmolLM-135M-GGUF/SmolLM-135M.Q2_K.gguf\n"
+            "  llama-run "
+            "modelscope://bartowski/SmolLM-1.7B-Instruct-v0.2-GGUF/SmolLM-1.7B-Instruct-v0.2-IQ3_M.gguf\n"
             "  llama-run https://example.com/some-file1.gguf\n"
             "  llama-run some-file2.gguf\n"
             "  llama-run file://some-file3.gguf\n"
@@ -525,11 +510,11 @@ class HttpClient {
         int secs = static_cast<int>(seconds) % 60;
 
         if (hrs > 0) {
-            return fmt("%dh %02dm %02ds", hrs, mins, secs);
+            return string_format("%dh %02dm %02ds", hrs, mins, secs);
         } else if (mins > 0) {
-            return fmt("%dm %02ds", mins, secs);
+            return string_format("%dm %02ds", mins, secs);
         } else {
-            return fmt("%ds", secs);
+            return string_format("%ds", secs);
         }
     }
 
@@ -544,7 +529,7 @@ class HttpClient {
             }
         }
 
-        return fmt("%.2f %s", dbl_size, suffix[i]);
+        return string_format("%.2f %s", dbl_size, suffix[i]);
     }
 
     static int update_progress(void * ptr, curl_off_t total_to_download, curl_off_t now_downloaded, curl_off_t,
@@ -578,7 +563,9 @@ class HttpClient {
         return (now_downloaded_plus_file_size * 100) / total_to_download;
     }
 
-    static std::string generate_progress_prefix(curl_off_t percentage) { return fmt("%3ld%% |", static_cast<long int>(percentage)); }
+    static std::string generate_progress_prefix(curl_off_t percentage) {
+        return string_format("%3ld%% |", static_cast<long int>(percentage));
+    }
 
     static double calculate_speed(curl_off_t now_downloaded, const std::chrono::steady_clock::time_point & start_time) {
         const auto                          now             = std::chrono::steady_clock::now();
@@ -589,9 +576,9 @@ class HttpClient {
     static std::string generate_progress_suffix(curl_off_t now_downloaded_plus_file_size, curl_off_t total_to_download,
                                                 double speed, double estimated_time) {
         const int width = 10;
-        return fmt("%*s/%*s%*s/s%*s", width, human_readable_size(now_downloaded_plus_file_size).c_str(), width,
-                   human_readable_size(total_to_download).c_str(), width, human_readable_size(speed).c_str(), width,
-                   human_readable_time(estimated_time).c_str());
+        return string_format("%*s/%*s%*s/s%*s", width, human_readable_size(now_downloaded_plus_file_size).c_str(),
+                             width, human_readable_size(total_to_download).c_str(), width,
+                             human_readable_size(speed).c_str(), width, human_readable_time(estimated_time).c_str());
     }
 
     static int calculate_progress_bar_width(const std::string & progress_prefix, const std::string & progress_suffix) {
@@ -705,7 +692,7 @@ class LlamaData {
         return 0;
     }
 
-    int huggingface_dl(std::string & model, const std::string & bn) {
+    int dl_from_endpoint(std::string & model_endpoint, std::string & model, const std::string & bn) {
         // Find the second occurrence of '/' after protocol string
         size_t pos = model.find('/');
         pos        = model.find('/', pos + 1);
@@ -714,7 +701,7 @@ class LlamaData {
         std::string              url;
 
         if (pos == std::string::npos) {
-            auto [model_name, manifest_url] = extract_model_and_tag(model, "https://huggingface.co/v2/");
+            auto [model_name, manifest_url] = extract_model_and_tag(model, model_endpoint + "v2/");
             hfr                             = model_name;
 
             nlohmann::json manifest;
@@ -729,11 +716,21 @@ class LlamaData {
             hff = model.substr(pos + 1);
         }
 
-        url = "https://huggingface.co/" + hfr + "/resolve/main/" + hff;
+        url = model_endpoint + hfr + "/resolve/main/" + hff;
 
         return download(url, bn, true, headers);
     }
 
+    int modelscope_dl(std::string & model, const std::string & bn) {
+        std::string model_endpoint = "https://modelscope.cn/models/";
+        return dl_from_endpoint(model_endpoint, model, bn);
+    }
+
+    int huggingface_dl(std::string & model, const std::string & bn) {
+        std::string model_endpoint = get_model_endpoint();
+        return dl_from_endpoint(model_endpoint, model, bn);
+    }
+
     int ollama_dl(std::string & model, const std::string & bn) {
         const std::vector<std::string> headers = { "Accept: application/vnd.docker.distribution.manifest.v2+json" };
         if (model.find('/') == std::string::npos) {
@@ -851,6 +848,9 @@ class LlamaData {
             rm_until_substring(model_, "hf.co/");
             rm_until_substring(model_, "://");
             ret = huggingface_dl(model_, bn);
+        } else if (string_starts_with(model_, "ms://") || string_starts_with(model_, "modelscope://")) {
+            rm_until_substring(model_, "://");
+            ret = modelscope_dl(model_, bn);
         } else if ((string_starts_with(model_, "https://") || string_starts_with(model_, "http://")) &&
                    !string_starts_with(model_, "https://ollama.com/library/")) {
             ret = download(model_, bn, true);

package/src/llama.cpp/{examples → tools}/server/CMakeLists.txt

@@ -34,8 +34,9 @@ endforeach()
 add_executable(${TARGET} ${TARGET_SRCS})
 install(TARGETS ${TARGET} RUNTIME)
 
+target_include_directories(${TARGET} PRIVATE ../llava)
 target_include_directories(${TARGET} PRIVATE ${CMAKE_SOURCE_DIR})
-target_link_libraries(${TARGET} PRIVATE common ${CMAKE_THREAD_LIBS_INIT})
+target_link_libraries(${TARGET} PRIVATE common mtmd ${CMAKE_THREAD_LIBS_INIT})
 
 if (LLAMA_SERVER_SSL)
     find_package(OpenSSL REQUIRED)