Merge remote-tracking branch 'origin/master'

* origin/master:
  bench.py : add different large models (ggerganov#1655)
  wchess : update README.md
  release : v1.5.2
  wchess : update readme
  wchess : whisper assisted chess (ggerganov#1595)
  sync : ggml (Metal fixes, new ops, tests) (ggerganov#1633)
  cmake : target windows 8 or above for prefetchVirtualMemory in llama-talk (ggerganov#1617)
  cmake : Fix bug in httplib.h for mingw (ggerganov#1615)
  metal : fix `ggml_metal_log` vargs (ggerganov#1606)
  whisper.objc : disable timestamps for real-time transcription
  whisper : more debug messages + fix fallback logic
  metal : fix soft_max kernel src1 argument (ggerganov#1602)
  sync : ggml (new ops, new backend, etc) (ggerganov#1602)
  server : pass max-len argument to the server (ggerganov#1574)
  ios : Remove `#if arch(arm)` check for using Metal (ggerganov#1561)
  ggml : Fix 32-bit compiler warning (ggerganov#1575)
  ggml : re-enable blas for src0 != F32 (ggerganov#1583)

CMakeLists.txt

@@ -1,6 +1,6 @@
 cmake_minimum_required (VERSION 3.5)
 
-project(whisper.cpp VERSION 1.5.1)
+project(whisper.cpp VERSION 1.5.2)
 
 # Add path to modules
 list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake/")

Package.swift

@@ -2,41 +2,22 @@
 
 import PackageDescription
 
-#if arch(arm) || arch(arm64)
-let platforms: [SupportedPlatform]? = [
-    .macOS(.v12),
-    .iOS(.v14),
-    .watchOS(.v4),
-    .tvOS(.v14)
-]
-let exclude: [String] = []
-let resources: [Resource] = [
-    .process("ggml-metal.metal")
-]
-let additionalSources: [String] = ["ggml-metal.m"]
-let additionalSettings: [CSetting] = [
-    .unsafeFlags(["-fno-objc-arc"]),
-    .define("GGML_USE_METAL")
-]
-#else
-let platforms: [SupportedPlatform]? = nil
-let exclude: [String] = ["ggml-metal.metal"]
-let resources: [Resource] = []
-let additionalSources: [String] = []
-let additionalSettings: [CSetting] = []
-#endif
-
 let package = Package(
     name: "whisper",
-    platforms: platforms,
+    platforms: [
+        .macOS(.v12),
+        .iOS(.v14),
+        .watchOS(.v4),
+        .tvOS(.v14)
+    ],
     products: [
         .library(name: "whisper", targets: ["whisper"]),
     ],
     targets: [
         .target(
             name: "whisper",
             path: ".",
-            exclude: exclude + [
+            exclude: [
                "bindings",
                "cmake",
                "coreml",
@@ -55,19 +36,22 @@ let package = Package(
                 "whisper.cpp",
                 "ggml-alloc.c",
                 "ggml-backend.c",
-                "ggml-quants.c"
-            ] + additionalSources,
-            resources: resources,
+                "ggml-quants.c",
+                "ggml-metal.m"
+            ],
+            resources: [.process("ggml-metal.metal")],
             publicHeadersPath: "spm-headers",
             cSettings: [
                 .unsafeFlags(["-Wno-shorten-64-to-32", "-O3", "-DNDEBUG"]),
-                .define("GGML_USE_ACCELERATE")
+                .define("GGML_USE_ACCELERATE"),
+                .unsafeFlags(["-fno-objc-arc"]),
+                .define("GGML_USE_METAL")
                 // NOTE: NEW_LAPACK will required iOS version 16.4+
                 // We should consider add this in the future when we drop support for iOS 14
                 // (ref: ref: https://developer.apple.com/documentation/accelerate/1513264-cblas_sgemm?language=objc)
                 // .define("ACCELERATE_NEW_LAPACK"),
                 // .define("ACCELERATE_LAPACK_ILP64")
-            ] + additionalSettings,
+            ],
             linkerSettings: [
                 .linkedFramework("Accelerate")
             ]

README.md

@@ -6,7 +6,7 @@
 [![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](https://opensource.org/licenses/MIT)
 [![npm](https://img.shields.io/npm/v/whisper.cpp.svg)](https://www.npmjs.com/package/whisper.cpp/)
 
-Stable: [v1.5.1](https://github.com/ggerganov/whisper.cpp/releases/tag/v1.5.1) / [Roadmap | F.A.Q.](https://github.com/ggerganov/whisper.cpp/discussions/126)
+Stable: [v1.5.2](https://github.com/ggerganov/whisper.cpp/releases/tag/v1.5.2) / [Roadmap | F.A.Q.](https://github.com/ggerganov/whisper.cpp/discussions/126)
 
 High-performance inference of [OpenAI's Whisper](https://github.com/openai/whisper) automatic speech recognition (ASR) model:
 
@@ -770,6 +770,7 @@ Some of the examples are even ported to run in the browser using WebAssembly. Ch
 | [bench](examples/bench) | [bench.wasm](examples/bench.wasm) | Benchmark the performance of Whisper on your machine |
 | [stream](examples/stream) | [stream.wasm](examples/stream.wasm) | Real-time transcription of raw microphone capture |
 | [command](examples/command) | [command.wasm](examples/command.wasm) | Basic voice assistant example for receiving voice commands from the mic |
+| [wchess](examples/wchess) | [wchess.wasm](examples/wchess) | Voice-controlled chess |
 | [talk](examples/talk) | [talk.wasm](examples/talk.wasm) | Talk with a GPT-2 bot |
 | [talk-llama](examples/talk-llama) | | Talk with a LLaMA bot |
 | [whisper.objc](examples/whisper.objc) | | iOS mobile application using whisper.cpp |

bindings/javascript/package.json

@@ -1,6 +1,6 @@
 {
   "name": "whisper.cpp",
-  "version": "1.5.1",
+  "version": "1.5.2",
   "description": "Whisper speech recognition",
   "main": "whisper.js",
   "scripts": {

examples/CMakeLists.txt

@@ -187,3 +187,5 @@ else()
     #add_subdirectory(talk-llama)
     #add_subdirectory(lsp)
 endif()
+
+add_subdirectory(wchess)

examples/helpers.js

@@ -22,6 +22,7 @@ var printTextarea = (function() {
 async function clearCache() {
     if (confirm('Are you sure you want to clear the cache?\nAll the models will be downloaded again.')) {
         indexedDB.deleteDatabase(dbName);
+        location.reload();
     }
 }
 

examples/server/CMakeLists.txt

@@ -4,3 +4,9 @@ add_executable(${TARGET} server.cpp httplib.h json.hpp)
 include(DefaultTargetOptions)
 
 target_link_libraries(${TARGET} PRIVATE common whisper ${CMAKE_THREAD_LIBS_INIT})
+
+# Check if the compiler is MinGW
+if(MINGW)
+    # Link the necessary libraries for SSL and Winsock
+    target_link_libraries(${TARGET} PRIVATE -lcrypt32 -lssl -lcrypto -lws2_32)
+endif()

examples/server/server.cpp

@@ -589,6 +589,7 @@ int main(int argc, char ** argv) {
             wparams.duration_ms      = params.duration_ms;
 
             wparams.thold_pt         = params.word_thold;
+            wparams.max_len          = params.max_len == 0 ? 60 : params.max_len;
             wparams.split_on_word    = params.split_on_word;
 
             wparams.speed_up         = params.speed_up;

examples/talk-llama/CMakeLists.txt

@@ -18,6 +18,11 @@ if (WHISPER_SDL2)
         ../../ggml-quants.c
         ../../whisper.cpp)
 
+    if(WIN32)
+    # It requires Windows 8.1 or later for PrefetchVirtualMemory
+    target_compile_definitions(${TARGET} PRIVATE -D_WIN32_WINNT=0x0602)
+    endif()
+
     target_include_directories(${TARGET} PRIVATE ${SDL2_INCLUDE_DIRS} ../../)
     target_link_libraries(${TARGET} PRIVATE ${SDL2_LIBRARIES} ${CMAKE_THREAD_LIBS_INIT})
 

examples/wchess/CMakeLists.txt

@@ -0,0 +1,9 @@
+set(CMAKE_CXX_STANDARD 11)
+
+add_subdirectory(libwchess)
+
+if (EMSCRIPTEN)
+    add_subdirectory(wchess.wasm)
+else()
+    add_subdirectory(wchess.cmd)
+endif()

examples/wchess/README.md

@@ -0,0 +1,45 @@
+# wchess
+
+Voice-controlled chess using Whisper
+
+Online demo: https://whisper.ggerganov.com/wchess/
+
+https://github.com/ggerganov/whisper.cpp/assets/1991296/c2b2f03c-9684-49f3-8106-357d2d4e67fa
+
+## Command-line tool
+
+```bash
+mkdir build && cd build
+cmake -DWHISPER_SDL2=1 ..
+make -j
+
+./bin/wchess -m ../models/ggml-base.en.bin
+
+Move: start
+
+a b c d e f g h
+r n b q k b n r 8
+p p p p p p p p 7
+. * . * . * . * 6
+* . * . * . * . 5
+. * . * . * . * 4
+* . * . * . * . 3
+P P P P P P P P 2
+R N B Q K B N R 1
+
+White's turn
+[(l)isten/(p)ause/(q)uit]: 
+```
+
+## TODO
+
+- Fix bugs in the chess moves logic
+- Improve web-browser audio capture - sometimes it does not record the voice properly
+- Add support for more languages by making the generated grammar string multilingual
+- Explore ways to improve the dynamic grammar to be narrower
+
+PRs welcome!
+
+## Thanks
+
+- [chessboardjs](https://chessboardjs.com) for the neat chessboard JS library used in this demo

examples/wchess/libwchess/CMakeLists.txt

@@ -0,0 +1,19 @@
+add_library(wchess-core STATIC
+    WChess.cpp
+    WChess.h
+    Chessboard.cpp
+    Chessboard.h
+)
+
+target_link_libraries(wchess-core
+    PUBLIC
+    whisper
+    common
+)
+
+target_include_directories(wchess-core
+    PUBLIC
+    "$<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}>"
+)
+
+# add_executable(test-chessboard test-chessboard.cpp Chessboard.cpp)