gnu: Add ctranslate2.

* gnu/packages/machine-learning.scm (ctranslate2): New variable.
* gnu/packages/patches/CTranslate2-local-build.patch: New file.
* gnu/local.mk: Record patch.

Co-authored-by: Nicolas Graves <ngraves@ngraves.fr>
Change-Id: I61f04d92c9f4c7bff41b62867ef43e8cf3ad87ea

gnu/local.mk

@@ -1163,6 +1163,7 @@ dist_patch_DATA =						\
   %D%/packages/patches/crda-optional-gcrypt.patch		\
   %D%/packages/patches/clucene-contribs-lib.patch               \
   %D%/packages/patches/csvkit-set-locale-for-tests.patch			\
+  %D%/packages/patches/ctranslate2-local-build.patch		\
   %D%/packages/patches/cube-nocheck.patch			\
   %D%/packages/patches/cups-minimal-Address-PPD-injection-issues.patch	\
   %D%/packages/patches/curl-CVE-2024-8096.patch			\

gnu/packages/machine-learning.scm

@@ -30,6 +30,7 @@
 ;;; Copyright © 2024 Spencer King <spencer.king@geneoscopy.com>
 ;;; Copyright © 2024, 2025 David Elsing <david.elsing@posteo.net>
 ;;; Copyright © 2024 Andy Tai <atai@atai.org>
+;;; Copyright © 2025 Lapearldot <lapearldot@disroot.org>
 ;;;
 ;;; This file is part of GNU Guix.
 ;;;
@@ -107,6 +108,7 @@
   #:use-module (gnu packages libffi)
   #:use-module (gnu packages linux)
   #:use-module (gnu packages llvm)
+  #:use-module (gnu packages logging)
   #:use-module (gnu packages maths)
   #:use-module (gnu packages mpi)
   #:use-module (gnu packages ninja)
@@ -6291,6 +6293,50 @@ Transformers is backed by the three most popular deep learning libraries —
 Jax, PyTorch and TensorFlow — with a seamless integration between them.")
     (license license:asl2.0)))
 
+(define-public ctranslate2
+  (package
+    (name "ctranslate2")
+    (version "4.6.0")
+    (source
+     (origin
+       (method git-fetch)
+       (uri (git-reference
+              (url "https://github.com/OpenNMT/CTranslate2")
+              (commit (string-append "v" version))))
+       (file-name (git-file-name "CTranslate2" version))
+       (sha256 (base32 "0sngxjq5rmrgjabgxfazdnhfn3wpfi62n226f4k47cx1xjjkllcd"))
+       (patches (search-patches "ctranslate2-local-build.patch"))))
+    (build-system cmake-build-system)
+    (arguments
+     (list
+      #:test-target "ctranslate2_test"
+      ;; XXX: mkl and openblas seem incompatible.
+      #:configure-flags `(list "-DBUILD_TESTS=ON"
+                               "-DWITH_ACCELERATE=OFF"
+                               "-DWITH_RUY=OFF"
+                               "-DWITH_DNNL=ON"
+                               "-DWITH_CUDA=OFF"
+                               "-DWITH_CUDNN=OFF"
+                               "-DWITH_MKL=OFF"
+                               "-DWITH_OPENBLAS=ON")))
+    (native-inputs (list libomp
+                         cxxopts
+                         spdlog
+                         googletest
+                         cpu-features
+                         oneapi-dnnl
+                         openblas))
+    (home-page "https://opennmt.net/CTranslate2/")
+    (synopsis "Fast inference engine for Transformer models")
+    (description "CTranslate2 is a C++ and Python library for efficient
+inference with Transformer models.
+
+The project implements a custom runtime that applies many performance
+optimization techniques such as weights quantization, layers fusion, batch
+reordering, etc., to accelerate and reduce the memory usage of Transformer
+models on CPU and GPU.")
+    (license license:expat)))
+
 (define-public python-hmmlearn
   (package
     (name "python-hmmlearn")

gnu/packages/patches/ctranslate2-local-build.patch

@@ -0,0 +1,129 @@
+Patch subprojects to use guix builds instead
+
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 62b99d1..c1ac38a 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -98,7 +98,7 @@ else()
+ endif()
+ 
+ find_package(Threads)
+-add_subdirectory(third_party/spdlog EXCLUDE_FROM_ALL)
++find_package( spdlog)
+ 
+ set(PRIVATE_INCLUDE_DIRECTORIES
+   ${CMAKE_CURRENT_SOURCE_DIR}/src
+@@ -249,9 +249,10 @@ elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "(x86_64)|(amd64)|(AMD64)")
+   set(BUILD_SHARED_LIBS_SAVED "${BUILD_SHARED_LIBS}")
+   set(BUILD_SHARED_LIBS OFF)
+   set(BUILD_TESTING OFF)
+-  add_subdirectory(third_party/cpu_features EXCLUDE_FROM_ALL)
++  find_package( CpuFeatures)
+   set(BUILD_SHARED_LIBS "${BUILD_SHARED_LIBS_SAVED}")
+-  list(APPEND LIBRARIES cpu_features)
++  find_library( CPU_FEATURES_LIBRARIES cpu_features)
++  list(APPEND LIBRARIES ${CPU_FEATURES_LIBRARIES})
+ endif()
+ 
+ if(ENABLE_CPU_DISPATCH)
+@@ -349,8 +350,8 @@ if(WITH_MKL)
+   else()
+     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -m64")
+     set(MKL_LIBRARIES
+-      ${MKL_LIBRARY_DIR}/libmkl_core.a
+-      ${MKL_LIBRARY_DIR}/libmkl_intel_ilp64.a
++      ${MKL_LIBRARY_DIR}/libmkl_core.so
++      ${MKL_LIBRARY_DIR}/libmkl_intel_ilp64.so
+       )
+   endif()
+ 
+@@ -358,19 +359,19 @@ if(WITH_MKL)
+     if(WIN32)
+       list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/mkl_intel_thread.lib)
+     else()
+-      list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/libmkl_intel_thread.a)
++      list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/libmkl_intel_thread.so)
+     endif()
+   elseif(OPENMP_RUNTIME STREQUAL "COMP")
+     if(WIN32)
+       message(FATAL_ERROR "Building with MKL requires Intel OpenMP")
+     else()
+-      list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/libmkl_gnu_thread.a)
++      list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/libmkl_gnu_thread.so)
+     endif()
+   elseif(OPENMP_RUNTIME STREQUAL "NONE")
+     if(WIN32)
+       list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/mkl_sequential.lib)
+     else()
+-      list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/libmkl_sequential.a)
++      list(APPEND MKL_LIBRARIES ${MKL_LIBRARY_DIR}/libmkl_sequential.so)
+     endif()
+   endif()
+   list(APPEND PRIVATE_INCLUDE_DIRECTORIES ${MKL_INCLUDE_DIR})
+@@ -439,6 +440,7 @@ if (WITH_RUY)
+   add_definitions(-DCT2_WITH_RUY)
+   set(CMAKE_POSITION_INDEPENDENT_CODE ON)
+   set(CPUINFO_LIBRARY_TYPE static CACHE STRING "cpuinfo library type")
++  # if package found and appended to libraries, issue with linking
+   add_subdirectory(third_party/ruy EXCLUDE_FROM_ALL)
+   unset(CMAKE_POSITION_INDEPENDENT_CODE)
+   list(APPEND LIBRARIES ruy)
+diff --git a/cli/CMakeLists.txt b/cli/CMakeLists.txt
+index 3311ad3..7d90693 100644
+--- a/cli/CMakeLists.txt
++++ b/cli/CMakeLists.txt
+@@ -1,14 +1,12 @@
+-if (NOT IS_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}/../third_party/cxxopts/include")
+-  message(FATAL_ERROR "The client dependency repository (cxxopts) is missing! "
+-    "You probably didn't clone the project with --recursive. You can include it "
+-    "by calling \"git submodule update --init --recursive\"")
+-endif()
+-
+ add_executable(translator
+   translator.cc
+   )
++find_path(
++  CXXOPTS_ROOT
++  include/cxxopts.hpp
++)
+ target_include_directories(translator
+-  PRIVATE ${CMAKE_CURRENT_SOURCE_DIR}/../third_party/cxxopts/include
++  PRIVATE ${CXXOPTS_ROOT}/include
+   )
+ target_link_libraries(translator
+   PRIVATE ${PROJECT_NAME}
+diff --git a/src/cpu/cpu_info.cc b/src/cpu/cpu_info.cc
+index 9030ac7..6949644 100644
+--- a/src/cpu/cpu_info.cc
++++ b/src/cpu/cpu_info.cc
+@@ -4,7 +4,7 @@
+ 
+ #include <cstring>
+ 
+-#include <cpuinfo_x86.h>
++#include <cpu_features/cpuinfo_x86.h>
+ 
+ namespace ctranslate2 {
+   namespace cpu {
+diff --git a/tests/CMakeLists.txt b/tests/CMakeLists.txt
+index 283c49d..9ca6cc8 100644
+--- a/tests/CMakeLists.txt
++++ b/tests/CMakeLists.txt
+@@ -2,7 +2,7 @@ enable_testing()
+ 
+ option(BUILD_GMOCK "" OFF)
+ option(INSTALL_GTEST "" OFF)
+-add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/../third_party/googletest ${CMAKE_CURRENT_BINARY_DIR}/googletest)
++find_package( GTest)
+ 
+ add_executable(ctranslate2_test
+   batching_test.cc
+@@ -20,7 +20,7 @@ target_include_directories(ctranslate2_test PRIVATE
+   )
+ target_link_libraries(ctranslate2_test
+   ${PROJECT_NAME}
+-  gtest_main
++  gtest
+   )
+ 
+ add_executable(benchmark_ops