Add tokenizer test + revert to C++11 (#355)

* Add test-tokenizer-0 to do a few tokenizations - feel free to expand
* Added option to convert-pth-to-ggml.py script to dump just the vocabulary
* Added ./models/ggml-vocab.bin containing just LLaMA vocab data (used for tests)
* Added utility to load vocabulary file from previous point (temporary implementation)
* Avoid using std::string_view and drop back to C++11 (hope I didn't break something)
* Rename gpt_vocab -> llama_vocab
* All CMake binaries go into ./bin/ now

Author: ggerganov

.github/workflows/build.yml

@@ -54,6 +54,7 @@ jobs:
           cd build
           cmake ..
           cmake --build . --config Release
+          ctest --output-on-failure
 
   macOS-latest-make:
     runs-on: macos-latest
@@ -90,6 +91,7 @@ jobs:
           cd build
           cmake ..
           cmake --build . --config Release
+          ctest --output-on-failure
 
   windows-latest-cmake:
     runs-on: windows-latest
@@ -106,6 +108,7 @@ jobs:
           cd build
           cmake ..
           cmake --build . --config Release
+          ctest --output-on-failure
 
       - name: Get commit hash
         id: commit

CMakeLists.txt

@@ -1,11 +1,37 @@
-cmake_minimum_required(VERSION 3.12)
+cmake_minimum_required(VERSION 3.12) # Don't bump this version for no reason
 project("llama.cpp" C CXX)
 
+set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
+
 if (NOT XCODE AND NOT MSVC AND NOT CMAKE_BUILD_TYPE)
     set(CMAKE_BUILD_TYPE Release CACHE STRING "Build type" FORCE)
     set_property(CACHE CMAKE_BUILD_TYPE PROPERTY STRINGS "Debug" "Release" "MinSizeRel" "RelWithDebInfo")
 endif()
 
+set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
+
+if(CMAKE_SOURCE_DIR STREQUAL CMAKE_CURRENT_SOURCE_DIR)
+    set(LLAMA_STANDALONE ON)
+
+    # configure project version
+    # TODO
+else()
+    set(LLAMA_STANDALONE OFF)
+endif()
+
+if (EMSCRIPTEN)
+    set(BUILD_SHARED_LIBS_DEFAULT OFF)
+
+    option(LLAMA_WASM_SINGLE_FILE "llama: embed WASM inside the generated llama.js" ON)
+else()
+    if (MINGW)
+        set(BUILD_SHARED_LIBS_DEFAULT OFF)
+    else()
+        set(BUILD_SHARED_LIBS_DEFAULT ON)
+    endif()
+endif()
+
+
 #
 # Option list
 #
@@ -34,6 +60,9 @@ option(LLAMA_FMA                    "llama: enable FMA"
 option(LLAMA_ACCELERATE             "llama: enable Accelerate framework"                    ON)
 option(LLAMA_OPENBLAS               "llama: use OpenBLAS"                                   OFF)
 
+option(LLAMA_BUILD_TESTS            "llama: build tests"    ${LLAMA_STANDALONE})
+option(LLAMA_BUILD_EXAMPLES         "llama: build examples" ${LLAMA_STANDALONE})
+
 #
 # Compile flags
 #
@@ -187,17 +216,19 @@ add_executable(llama main.cpp)
 
 add_executable(quantize quantize.cpp)
 
-add_library(ggml OBJECT
-            ggml.c
-            ggml.h)
-
 add_library(utils OBJECT
             utils.cpp
             utils.h)
 
+target_include_directories(utils PUBLIC .)
+target_compile_features(utils PUBLIC cxx_std_11) # don't bump
+
+add_library(ggml OBJECT
+            ggml.c
+            ggml.h)
+
 target_include_directories(ggml PUBLIC .)
-target_compile_features(ggml PUBLIC c_std_11)
-target_compile_features(utils PUBLIC cxx_std_17)
+target_compile_features(ggml PUBLIC c_std_11) # don't bump
 
 #
 # Linking
@@ -206,3 +237,16 @@ target_compile_features(utils PUBLIC cxx_std_17)
 target_link_libraries(ggml PRIVATE Threads::Threads ${LLAMA_EXTRA_LIBS})
 target_link_libraries(llama PRIVATE ggml utils)
 target_link_libraries(quantize PRIVATE ggml utils)
+
+#
+# programs, examples and tests
+#
+
+if (LLAMA_BUILD_TESTS AND NOT CMAKE_JS_VERSION)
+    enable_testing()
+    add_subdirectory(tests)
+endif ()
+
+#if (LLAMA_BUILD_EXAMPLES)
+#    add_subdirectory(examples)
+#endif()

Makefile

@@ -30,8 +30,9 @@ endif
 # Compile flags
 #
 
+# keep standard at C11 and C++11
 CFLAGS   = -I.              -O3 -DNDEBUG -std=c11   -fPIC
-CXXFLAGS = -I. -I./examples -O3 -DNDEBUG -std=c++17 -fPIC
+CXXFLAGS = -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC
 LDFLAGS  =
 
 # OS specific

convert-pth-to-ggml.py

@@ -10,26 +10,26 @@
 #   - Name (char[name_length])
 #   - Data (float[n_dims])
 #
-# By default, the bigger matrices are converted to 16-bit floats.
-# This can be disabled by adding the "use-f32" CLI argument.
-#
 # At the start of the ggml file we write the model parameters
 # and vocabulary.
 #
+
 import argparse
 import os
 import sys
 import json
 import struct
 import numpy as np
 import torch
+
 from sentencepiece import SentencePieceProcessor
 
 def parse_args():
 
     parser = argparse.ArgumentParser(description='Convert a LLaMA model checkpoint to a ggml compatible file')
     parser.add_argument('dir_model', help='directory containing the model checkpoint')
     parser.add_argument('ftype', type=int, choices=[0, 1], default=1, help='file type (0: float32, 1: float16)')
+    parser.add_argument('vocab_only', type=bool, default=False, help='only write vocab to file')
     return parser.parse_args()
 
 def get_n_parts(dim):
@@ -134,6 +134,27 @@ def main():
     ftype_str = ["f32", "f16"]
 
     hparams, tokenizer = load_hparams_and_tokenizer(dir_model)
+
+    # if only writing vocab to file
+    if args.vocab_only:
+
+        fname_model = f"{dir_model}/consolidated.00.pth"
+        fname_out = f"{dir_model}/ggml-vocab.bin"
+
+        print(f"Extracting only the vocab from '{fname_model}'\n")
+
+        model = torch.load(fname_model, map_location="cpu")
+
+        with open(fname_out, "wb") as fout:
+            fout.write(struct.pack("i", hparams["vocab_size"]))
+            write_tokens(fout, tokenizer)
+
+        del model
+
+        print(f"Done. Output file: {fname_out}\n")
+
+        return
+
     n_parts = get_n_parts(hparams["dim"])
 
     for p in range(n_parts):
@@ -151,6 +172,7 @@ def main():
             process_and_write_variables(fout, model, ftype)
 
         del model
+
         print(f"Done. Output file: {fname_out}, (part {p})\n")
 
 if __name__ == "__main__":

main.cpp

@@ -90,7 +90,7 @@ struct llama_model {
 };
 
 // load the model's weights from a file
-bool llama_model_load(const std::string & fname, llama_model & model, gpt_vocab & vocab, int n_ctx, ggml_type memory_type = GGML_TYPE_F32) {
+bool llama_model_load(const std::string & fname, llama_model & model, llama_vocab & vocab, int n_ctx, ggml_type memory_type = GGML_TYPE_F32) {
     fprintf(stderr, "%s: loading model from '%s' - please wait ...\n", __func__, fname.c_str());
 
     std::vector<char> f_buf(1024*1024);
@@ -544,9 +544,9 @@ bool llama_eval(
         const llama_model & model,
         const int n_threads,
         const int n_past,
-        const std::vector<gpt_vocab::id> & embd_inp,
-              std::vector<float>         & embd_w,
-              size_t                     & mem_per_token) {
+        const std::vector<llama_vocab::id> & embd_inp,
+              std::vector<float>           & embd_w,
+              size_t                       & mem_per_token) {
     const int N = embd_inp.size();
 
     const auto & hparams = model.hparams;
@@ -832,7 +832,7 @@ int main(int argc, char ** argv) {
 
     int64_t t_load_us = 0;
 
-    gpt_vocab vocab;
+    llama_vocab vocab;
     llama_model model;
 
     // load the model
@@ -864,13 +864,13 @@ int main(int argc, char ** argv) {
     // Add a space in front of the first character to match OG llama tokenizer behavior
     params.prompt.insert(0, 1, ' ');
     // tokenize the prompt
-    std::vector<gpt_vocab::id> embd_inp = ::llama_tokenize(vocab, params.prompt, true);
+    std::vector<llama_vocab::id> embd_inp = ::llama_tokenize(vocab, params.prompt, true);
 
     params.n_predict = std::min(params.n_predict, model.hparams.n_ctx - (int) embd_inp.size());
 
     // prefix & suffix for instruct mode
-    const std::vector<gpt_vocab::id> inp_pfx = ::llama_tokenize(vocab, "\n\n### Instruction:\n\n", true);
-    const std::vector<gpt_vocab::id> inp_sfx = ::llama_tokenize(vocab, "\n\n### Response:\n\n", false);
+    const std::vector<llama_vocab::id> inp_pfx = ::llama_tokenize(vocab, "\n\n### Instruction:\n\n", true);
+    const std::vector<llama_vocab::id> inp_sfx = ::llama_tokenize(vocab, "\n\n### Response:\n\n", false);
 
     // in instruct mode, we inject a prefix and a suffix to each input by the user
     if (params.instruct) {
@@ -879,8 +879,8 @@ int main(int argc, char ** argv) {
     }
 
     // tokenize the reverse prompt
-    std::vector<std::vector<gpt_vocab::id>> antipromptv_inp;
-    
+    std::vector<std::vector<llama_vocab::id>> antipromptv_inp;
+
     for (auto antiprompt : params.antiprompt) {
         antipromptv_inp.push_back(::llama_tokenize(vocab, antiprompt, false));
     }
@@ -925,14 +925,14 @@ int main(int argc, char ** argv) {
     fprintf(stderr, "sampling parameters: temp = %f, top_k = %d, top_p = %f, repeat_last_n = %i, repeat_penalty = %f\n", params.temp, params.top_k, params.top_p, params.repeat_last_n, params.repeat_penalty);
     fprintf(stderr, "\n\n");
 
-    std::vector<gpt_vocab::id> embd;
+    std::vector<llama_vocab::id> embd;
 
     // determine the required inference memory per token:
     size_t mem_per_token = 0;
     llama_eval(model, params.n_threads, 0, { 0, 1, 2, 3 }, logits, mem_per_token);
 
     int last_n_size = params.repeat_last_n;
-    std::vector<gpt_vocab::id> last_n_tokens(last_n_size);
+    std::vector<llama_vocab::id> last_n_tokens(last_n_size);
     std::fill(last_n_tokens.begin(), last_n_tokens.end(), 0);
 
     if (params.interactive) {
@@ -980,7 +980,7 @@ int main(int argc, char ** argv) {
 
             const int n_vocab = model.hparams.n_vocab;
 
-            gpt_vocab::id id = 0;
+            llama_vocab::id id = 0;
 
             {
                 const int64_t t_start_sample_us = ggml_time_us();
@@ -1066,7 +1066,7 @@ int main(int argc, char ** argv) {
                 } while (another_line);
                 if (params.use_color) printf(ANSI_COLOR_RESET);
 
-                std::vector<gpt_vocab::id> line_inp = ::llama_tokenize(vocab, buffer, false);
+                std::vector<llama_vocab::id> line_inp = ::llama_tokenize(vocab, buffer, false);
                 embd_inp.insert(embd_inp.end(), line_inp.begin(), line_inp.end());
 
                 if (params.instruct) {

models/ggml-vocab.bin

@@ -44,7 +44,7 @@ bool llama_model_quantize(const std::string & fname_inp, const std::string & fna
         return false;
     }
 
-    gpt_vocab vocab;
+    llama_vocab vocab;
 
     printf("%s: loading model from '%s'\n", __func__, fname_inp.c_str());
 

quantize.cpp

@@ -0,0 +1,4 @@
+set(TEST_TARGET test-tokenizer-0)
+add_executable(${TEST_TARGET} ${TEST_TARGET}.cpp)
+target_link_libraries(${TEST_TARGET} PRIVATE utils)
+add_test(NAME ${TEST_TARGET} COMMAND $<TARGET_FILE:${TEST_TARGET}> ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab.bin)

tests/CMakeLists.txt

@@ -0,0 +1,69 @@
+#include "utils.h"
+
+#include <cstdio>
+#include <string>
+#include <map>
+
+static const std::map<std::string, std::vector<llama_vocab::id>> k_tests = {
+    { "Hello World",        { 1,  10994,   2787, }, },
+    { " Hello World",       { 1,  15043,   2787, }, },
+    { " Hello World!",      { 1,  15043,   2787,  29991, }, },
+    { " this is 🦙.cpp",    { 1,    445,    338,  29871,    243,    162,    169,    156,  29889,   8223, }, },
+    { "w048 7tuijk dsdfhu", { 1,  29893,  29900,  29946,  29947,  29871,  29955,   9161,  13535,  18031,   2176,   6905, }, },
+    { "нещо на Български",  { 1,    821,   4851,    665,   1386,  29713,   1305, }, },
+};
+
+int main(int argc, char **argv) {
+    if (argc < 2) {
+        fprintf(stderr, "Usage: %s <vocab-file>\n", argv[0]);
+        return 1;
+    }
+
+    const std::string fname = argv[1];
+
+    fprintf(stderr, "%s : reading vocab from: '%s'\n", __func__, fname.c_str());
+
+    llama_vocab vocab;
+
+    if (!llama_vocab_load(fname, vocab)) {
+        fprintf(stderr, "%s : failed to load vocab from: '%s'\n", __func__, fname.c_str());
+        return 1;
+    }
+
+    const int n_vocab = vocab.id_to_token.size();
+
+    if (n_vocab != 32000) {
+        fprintf(stderr, "%s : expected 32000 tokens, got %d\n", __func__, n_vocab);
+        return 2;
+    }
+
+    for (const auto & test_kv : k_tests) {
+        const auto res = llama_tokenize(vocab, test_kv.first, true);
+
+        bool correct = res.size() == test_kv.second.size();
+
+        for (int i = 0; i < (int) res.size() && correct; ++i) {
+            if (res[i] != test_kv.second[i]) {
+                correct = false;
+            }
+        }
+
+        if (!correct) {
+            fprintf(stderr, "%s : failed test: '%s'\n", __func__, test_kv.first.c_str());
+            fprintf(stderr, "%s : expected tokens: ", __func__);
+            for (const auto & t : test_kv.second) {
+                fprintf(stderr, "%6d, ", t);
+            }
+            fprintf(stderr, "\n");
+            fprintf(stderr, "%s : got tokens:      ", __func__);
+            for (const auto & t : res) {
+                fprintf(stderr, "%6d, ", t);
+            }
+            fprintf(stderr, "\n");
+
+            return 3;
+        }
+    }
+
+    return 0;
+}