FTorch: a library for coupling PyTorch models to Fortran

Version 1.1.0 - Flashy

All Changes: https://github.com/Cambridge-ICCS/FTorch/compare/v1.0.0...v1.1.0 and in CHANGELOG

What's Changed

• Post JOSS publication updates by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/309
• Fortitude v0.7 by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/310
• Add Park and Chung paper and add other details by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/314
• Hook up requires_grad by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/288
• MPI example by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/270
• Temporarily disable Windows CI by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/315
• Reorder examples by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/317
• Rework scalar multiplication and division by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/289
• Fix some broken links in the docs by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/322
• Typo correction of is_training variable description in the ftorch API. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/324
• Versioning and Changelog by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/313
• Add tests for constructing after destructing by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/319
• Mention test changes in changelog by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/326
• Tensor manipulation demo by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/291
• Ignore *.DS_Store and fix pre-commit hook path in the docs by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/330
• Add fypp to requirements by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/331
• Add news section to website by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/333
• Fix static analysis workflow (failing after #331) by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/335
• Implement torch_tensor_backward by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/286
• Spring cleaning by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/339
• Implement zeroing of gradients by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/341
• Implement torch_tensor_zero by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/338
• Use fork of pFUnit with CMake version bumped to resolve failing install in workflows by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/351
• Variable definitions in docstrings for tensor API by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/353
• torch_tensor_from_array interface with default layout by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/348
• Update to use version 4.12 of pFUnit following contributed fix. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/355
• Raise an error if user attempts to assign to tensor with requires_grad=.true. by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/359
• News archive by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/354
• Add error handling by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/347
• Overload the sum and mean over tensors by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/344
• Do not carry over requires_grad with overloaded ops by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/364
• Tidying of wrapped functions in ftorch.F90 by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/363
• Add upcoming events to webpage by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/366
• Adding ability to move torch_tensors between devices and dtypes by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/360
• Add user guide page on offline training by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/358
• Add FTorch presentation in News section by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/372
• requires_grad should be carried through expressions by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/373
• Docs on overloaded assignment operator by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/374
• Convert Reading seminars to presentation by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/376
• Fix bug in implementation of wrapper function torch_from_blob (#365) by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/377
• Add Codespace for running FTorch Online by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/346
• Autograd example updates by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/378
• Update README to add clarity around GPU setup. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/382
• Trivial example text tweaks by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/383
• feat: add HIP support to cmake using CUDA backend by @TomMelt in https://github.com/Cambridge-ICCS/FTorch/pull/385
• Update docs with information about HIP. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/388
• Mention FTorch workshop repo and separate out events by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/380
• Fig bug that made Fortran source missing from the API docs by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/395
• Fix unused-variable warning by @Mikolaj-A-Kowalski in https://github.com/Cambridge-ICCS/FTorch/pull/398
• Remove warning about pre-v1.0 from README by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/399
• Move AOPP tutorial to presentation archive by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/401
• Add slides for the 2025 Oxford AOPP Seminar. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/402
• Update FPTLIB-TODO to bare TODO by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/403
• Make info on citing FTorch more prominent by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/404
• Make FORD fail on warnings by @Mikolaj-A-Kowalski in https://github.com/Cambridge-ICCS/FTorch/pull/405
• Run test suite with ctest by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/357
• Mention graphviz dependency of FORD in the docs by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/408
• Fix cmake bug in building CUDA unit tests by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/411
• Add CI for test suite on GPU runner | Add MULTI_GPU cmake option by @niccolozanotti in https://github.com/Cambridge-ICCS/FTorch/pull/410
• Add reference in README.md for recent JAMES paper on cloud resolving models emulation by @dorchard in https://github.com/Cambridge-ICCS/FTorch/pull/421
• fix a minor bug by @zhenkunl in https://github.com/Cambridge-ICCS/FTorch/pull/422
• Update presentations by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/426
• Update to use inference_mode in preference to no_grad by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/425
• Expose access to tensor stride by @Mikolaj-A-Kowalski in https://github.com/Cambridge-ICCS/FTorch/pull/416
• Prevent fatal error from being thrown during cmake examples by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/431
• find_package(FTorch) should be find_package(FTorch REQUIRED) (fixes #432) by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/434
• Add a note to developer pages to explicitly state use of cmake-lint but not the formatter. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/440
• Use Zizmor action by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/441
• Adopt Green CI good practices by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/442
• Fix website deployment CI job by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/443
• CI for Intel (ifx) and Intel Classic (ifort) each with OpenMPI (fixes #140) by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/438
• link issue in example (fixes #435) by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/437
• Bump CUDA version used in GPU testing CI job by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/445
• Add a CODEOWNERS file for GitHub. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/447
• Support building tests with conda env present by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/450
• added explicit instructions for using PFUNIT with FTorch (fixes #283) by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/452
• Extend test suite timeout and allow running of intel on PRs if required. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/455
• Intel CI now uses Intel oneAPI MPI instead of OpenMPI built with Intel compilers (fixes #140) by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/449
• Support building FTorch as static library by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/448
• fix README example numbers to match folder numbers (#457) by @zka26 in https://github.com/Cambridge-ICCS/FTorch/pull/458
• Pin Python version in test suite. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/461
• Update example 5 to fix mismatching TorchScript and executable names. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/462
• Example 5: fix parallel .mod race by separate Fortran module dirs per target by @zka26 in https://github.com/Cambridge-ICCS/FTorch/pull/460
• corrected RELEASE build type to Debug build type in hpc docs by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/463
• fixed page/hpc.html to render CMAKE_<LANG>FLAGS instead of CMAKE<LANG>_FLAGS by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/465
• pkg-config support by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/464
• Update GitHub username in CODEOWNERS by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/466
• Fix deadlink to gpu user guide by @jfdev001 in https://github.com/Cambridge-ICCS/FTorch/pull/468
• Pedantic Zizmor fixes by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/473
• Documentation clearout and refactor by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/459
• Separate out CI testing workflows by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/475
• docs: generalize device argument docs from CUDA to GPU by @zka26 in https://github.com/Cambridge-ICCS/FTorch/pull/480
• Fix link to dev docs by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/479
• Update autograd worked example summary by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/485
• Add a standalone SimpleNet example build to the CI by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/304
• Bump minimum CMake version to 3.18 by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/491
• Fix link to use cases page in README by @dorchard in https://github.com/Cambridge-ICCS/FTorch/pull/502
• Subroutine for printing model layer weights by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/488
• Exercise typos by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/505
• Mention mailing list in README by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/510
• Add Batching examples and associated documentation by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/500
• Reactivate Windows CI by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/490
• Update Windows installation instructions by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/495
• Update install location of FTorch in GitHub workflow by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/515
• Docs on unit testing by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/489
• Bugfix .to() only works for single tensors, not iterables of tensors in example 5. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/519
• Add macOS testing workflow using gfortran by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/497
• Expunge example numbers from CMake Testing by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/521
• Update installation guidance for MacOS. by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/523
• Create subdirectories to better organise unit testing by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/511
• Better modularisation by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/508
• Use specific modules in unit tests by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/526
• Provide information about tensor finalizer by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/522
• Reduce duplication in unit test names by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/532
• Unit tests for creating and destroying tensors on GPU by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/546
• Use pointers in array tensor constructor functions by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/530
• Apply fypp by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/549
• Recommend defining as much as possible in model by @joewallwork in https://github.com/Cambridge-ICCS/FTorch/pull/535
• Provide a release checklist by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/552
• Make tensor deletion elemental by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/545
• Batching IO unit tests by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/559
• Prepare Release v1.1.0 by @jatkinson1000 in https://github.com/Cambridge-ICCS/FTorch/pull/560

New Contributors

• @danielskatz made their first contribution in https://github.com/Cambridge-ICCS/FTorch/pull/308
• @Mikolaj-A-Kowalski made their first contribution in https://github.com/Cambridge-ICCS/FTorch/pull/398
• @zhenkunl made their first contribution in https://github.com/Cambridge-ICCS/FTorch/pull/422
• @jfdev001 made their first contribution in https://github.com/Cambridge-ICCS/FTorch/pull/431
• @zka26 made their first contribution in https://github.com/Cambridge-ICCS/FTorch/pull/458

Full Changelog: https://github.com/Cambridge-ICCS/FTorch/compare/v1.0.0...v1.1.0