Commit Graph

  • 20d5f3ca34 Fix Windows force deserialization by writing ofstream in binary mode (#166) master v1.5.1 Stefan Doerr 2025-02-20 18:01:17 +01:00
  • d1861af768 Fix windows build (#164) Stefan Doerr 2025-02-19 20:52:09 +01:00
  • d8632dbce1 Fix CI failures (#165) Peter Eastman 2025-02-19 09:09:06 -08:00
  • 9ef733c071 Changes for PyPI wheels (#161) v1.5 Peter Eastman 2024-11-11 15:41:08 -08:00
  • 53e8786311 Don't assume SSE is present (#160) Peter Eastman 2024-11-11 14:26:25 -08:00
  • 5510c23811 Add pytorch lib dir to rpath (#158) Peter Eastman 2024-11-05 15:37:17 -08:00
  • b659e3638a Created HIP platform (#157) Peter Eastman 2024-11-05 11:39:24 -08:00
  • 681adc2329 Update Mac build to newer versions (#153) Peter Eastman 2024-08-23 14:20:06 -07:00
  • 9ceb9a0a8c Serialize property values (#152) Peter Eastman 2024-08-23 12:26:08 -07:00
  • 8893c0f2e2 Add energy derivatives to README (#145) Raul 2024-08-16 18:00:01 +02:00
  • 3aa2054f42 Merge pull request #148 from peastman/freeze Raul 2024-07-09 08:18:50 +02:00
  • 8d6cc53f3b Freeze pytorch model Peter Eastman 2024-06-28 12:51:28 -07:00
  • d4476435f5 Attempts at fixing CI (#144) Peter Eastman 2024-05-27 09:26:16 -07:00
  • e17b5d0060 Merge pull request #143 from peastman/derivatives Raul 2024-05-22 08:59:03 +02:00
  • 1d83b867a1 Handle CUDA graphs correctly Peter Eastman 2024-05-16 13:18:43 -07:00
  • fde51ffaad Improvements to tests peastman 2024-05-15 10:54:59 -07:00
  • e801c10e24 Serialization of parameter derivatives Peter Eastman 2024-05-15 10:46:09 -07:00
  • 0fc454b37f Python interface for parameter derivatives Peter Eastman 2024-05-15 10:35:37 -07:00
  • f751c65402 Fixes to CUDA implemenetation Peter Eastman 2024-05-15 10:31:12 -07:00
  • 151c9e2516 Can compute parameter derivatives peastman 2024-05-15 09:17:58 -07:00
  • 4fe23f6869 Use setuptools instead of distutils (#123) v1.4 Peter Eastman 2023-10-03 09:21:43 -07:00
  • a8f467b8d6 Use the same stream as the OpenMM context when replaying a CUDA graph. (#122) Raul 2023-09-27 10:41:57 +02:00
  • 2270256fed Fix automatic CUDA graphing not working when requiring backwards (#120) v1.3 Raul 2023-09-20 15:46:14 +02:00
  • a0b51a9435 Add AWS GPU Runner (#107) Mike Henry 2023-09-06 10:13:11 -07:00
  • 0e733e3e5b only run on master branch feat/add_aws_gpu_runner Mike Henry 2023-09-06 09:57:52 -07:00
  • ff2bf34fbb see if xdist speeds things up Mike Henry 2023-09-06 09:41:54 -07:00
  • 77df4f8756 run on push to master + weekly Mike Henry 2023-09-05 13:16:36 -07:00
  • 74798bc214 Clone the module on TorchForceImpl initialization to avoid several (#116) v1.2 Raul 2023-09-05 14:37:00 +02:00
  • e9f2ae24f0 Document the constructor that takes a scripted module (#104) Raul 2023-07-27 16:33:41 +01:00
  • e196c8d527 Torch2 compatibility (#106) v1.1 Raul 2023-07-24 10:11:26 +01:00
  • 49ee1fc139 now testing torch2 test_torch2_on_gpu_v2 Mike Henry 2023-05-22 16:01:55 -07:00
  • 42e3cc0466 first lets test torrch 1.12 Mike Henry 2023-05-22 15:31:42 -07:00
  • 84662ad7cc see if we get segfaults using pytorch 1.12 test_torch2_on_gpu Mike Henry 2023-05-19 09:55:02 -07:00
  • 2147e11780 hardcode sitepackages path for now Mike Henry 2023-05-18 17:00:00 -07:00
  • 8ef848afbd hardcode sitepackages path for now Mike Henry 2023-05-18 16:39:58 -07:00
  • c18997b796 c++ tests pass but python tests are segfaulting, trying newer version of python 3.10 Mike Henry 2023-05-18 16:25:01 -07:00
  • 40cee46241 only have cuda 11.7 on aws runner Mike Henry 2023-05-18 16:10:40 -07:00
  • 2332f4195a actually test on GPU Mike Henry 2023-05-18 16:03:14 -07:00
  • 49fe27c0d7 Actually test pytorch2 Mike Henry 2023-05-18 15:39:14 -07:00
  • 4c2025d995 use micromamba to test torch2 Mike Henry 2023-05-18 15:37:27 -07:00
  • cd89ea4233 Revert "Merge remote-tracking branch 'origin/feat/add_aws_gpu_runner' into torch2" Mike Henry 2023-05-18 14:08:53 -07:00
  • 5dc26f66ae Revert "actually test with pytorch2" Mike Henry 2023-05-18 13:40:59 -07:00
  • 4e7a4fbc1e actually test with pytorch2 Mike Henry 2023-05-18 13:40:21 -07:00
  • 5e01c8ad82 run gpu tests on this branch Mike Henry 2023-05-18 13:25:21 -07:00
  • 98e42ef3a1 Merge remote-tracking branch 'origin/feat/add_aws_gpu_runner' into torch2 Mike Henry 2023-05-18 13:24:04 -07:00
  • f51492dc95 Revert "Merge remote-tracking branch 'origin/feat/add_aws_gpu_runner' into torch2" RaulPPealez 2023-05-02 08:49:35 +02:00
  • 0a58de459b Merge remote-tracking branch 'origin/feat/add_aws_gpu_runner' into torch2 RaulPPealez 2023-05-02 08:48:12 +02:00
  • a1c16c6f56 Update swig typemaps to not use a temp file RaulPPealez 2023-04-27 13:05:24 +02:00
  • 4159b77f46 Make sure temporary files are deleted even if there was an error RaulPPealez 2023-04-26 09:08:51 +02:00
  • 552c4f0fce Switch to using micromamba Mike Henry 2023-04-25 15:45:19 -07:00
  • 9a2cbf42c8 Fix formatting RaulPPealez 2023-04-25 20:28:23 +02:00
  • 8b33e6d247 Update TorchForce swig input typemap to avoid as_module RaulPPealez 2023-04-25 20:02:00 +02:00
  • e488b36430 run CUDA tests on GPU runner Mike Henry 2023-04-25 10:00:06 -07:00
  • 784b770a6f Update swig typecheck to use py::isinstance instead of as_module RaulPPealez 2023-04-25 18:12:26 +02:00
  • 8778ae1b33 Merge remote-tracking branch 'origin/master' into torch2 RaulPPealez 2023-04-25 13:19:50 +02:00
  • b76deb4df8 Making TorchForce CUDA-graph aware (#103) Raul 2023-04-21 13:52:39 +01:00
  • 2c4dde68ae see if cuda 11.7 works Mike Henry 2023-04-20 03:13:04 -07:00
  • 712af05932 forgot to remove bit from version I copy Mike Henry 2023-04-20 03:00:51 -07:00
  • 6d4e00e3ca port the latest CI versions to the self-hosted runner Mike Henry 2023-04-20 02:49:01 -07:00
  • ebcfc35839 Update ci RaulPPealez 2023-04-20 10:02:05 +02:00
  • f8055c3dad Use Ubuntu-22 instead of deprecated Ubuntu 18 in CI (#105) Raul 2023-04-14 11:26:38 +02:00
  • 769302afd3 Add a constructor to TorchForce that takes a torch::jit::Module (#97) Raul 2023-02-10 23:58:03 +01:00
  • c76684f737 Attempt at fixing SWIG issues (#96) Peter Eastman 2023-01-31 11:04:03 -08:00
  • 994f92fde4 Pin Swig to <4.1 (#86) v1.0 Raimondas Galvelis 2022-10-25 18:53:28 +02:00
  • 1410331145 Update CI for PyTorch 1.12 (#83) Raimondas Galvelis 2022-09-29 18:29:17 +02:00
  • 0a60e830cd Add a comment pt_112 Raimondas Galvelis 2022-09-27 14:02:53 +02:00
  • c09b171d81 Properly set CUDA version Raimondas Galvelis 2022-09-27 13:03:57 +02:00
  • 6282e85547 Force CUDA version Raimondas Galvelis 2022-09-27 12:25:14 +02:00
  • 1bd2af052c Update CI for PyTorch 1.12 Raimondas Galvelis 2022-09-27 11:36:42 +02:00
  • 5b917a3d1a Use Mamba for CI (#81) Raimondas Galvelis 2022-08-01 17:16:11 +02:00
  • 5dc727951b Fix interoperability with CustomCVForce (#80) v0.8 Raimondas Galvelis 2022-07-08 17:02:19 +02:00
  • 661b004eb1 Revert "Push primary context before invoking PyTorch (#78)" (#79) Raimondas Galvelis 2022-06-15 15:54:56 +02:00
  • b5d758913b Revert "Push primary context before invoking PyTorch (#78)" revert-78-primary Raimondas Galvelis 2022-06-15 15:52:41 +02:00
  • a74c8883c2 Push primary context before invoking PyTorch (#78) Peter Eastman 2022-06-14 17:12:59 -07:00
  • 25fad7f36e Fix the tutorial (#77) Raimondas Galvelis 2022-06-10 12:15:15 +02:00
  • fa102e69f4 Update the test models and CI dependencies (#73) v0.7 Raimondas Galvelis 2022-05-26 20:47:40 +02:00
  • 84f7d884ec Add PyTorch lib directory to library path (#72) Peter Eastman 2022-03-07 13:18:57 -08:00
  • e6a3179054 The first tutorial of OpenMM-Torch with NNPOps (#62) v0.6 Raimondas Galvelis 2022-03-01 14:45:16 +01:00
  • db7c009e47 Move the PyTorch module to a correct device (#70) Raimondas Galvelis 2022-03-01 01:08:55 +01:00
  • 13e87155b5 Fix the Python wrapper of "setOutputsForces" and "getOutputsForces" (#60) Raimondas Galvelis 2022-01-21 23:41:19 +01:00
  • e2a9302fab Fix the synchronisation between the CUDA contexts (#49) v0.5 Raimondas Galvelis 2022-01-12 12:03:43 +01:00
  • 1316c05b87 Models can directly output forces (#52) Peter Eastman 2022-01-10 08:42:06 -08:00
  • 4614813608 Test if a PyTorch module receives corrects arguments (#50) Raimondas Galvelis 2022-01-08 01:10:10 +01:00
  • 77d6ac32a2 Update CI (#53) Raimondas Galvelis 2022-01-07 21:02:35 +01:00
  • 02f99351d9 Improved coordination of CUDA contexts with PyTorch (#47) Peter Eastman 2021-11-09 09:44:43 -08:00
  • bca1b1d48b Fix the force accumulation in the reference platform (#46) v0.4 Raimondas Galvelis 2021-09-27 20:59:10 +02:00
  • 76f55e7f43 Added missing information to serialization proxy (#38) v0.3 Peter Eastman 2021-05-28 15:57:22 -07:00
  • 6472507abd Added global parameters to Python wrapper (#29) v0.2 Peter Eastman 2021-04-11 13:24:44 -07:00
  • 51615ebb87 add CI (#25) Jaime Rodríguez-Guerra 2021-02-02 18:17:31 +01:00
  • 3ef5aa2b16 Merge pull request #23 from openmm/update-readme John Chodera 2021-01-22 08:33:09 -08:00
  • da95ece81b Fix docstrings John Chodera 2021-01-22 08:32:09 -08:00
  • 07313e73f6 Fix precision in README John Chodera 2021-01-21 17:41:56 -08:00
  • 9ccde5ebc9 Update README to be more comprehensible. John Chodera 2021-01-19 11:24:07 -08:00
  • 792ff7c0ad Merge pull request #21 from jaimergp/patch-1 John Chodera 2021-01-19 09:28:44 -08:00
  • 9771b5467f Add instructions for conda installation Jaime Rodríguez-Guerra 2021-01-18 18:14:40 +01:00
  • 00f45698da Merge pull request #18 from peastman/global Peter Eastman 2021-01-15 15:08:17 -08:00
  • a3dca1dba2 Merge pull request #19 from peastman/path v0.1 Peter Eastman 2021-01-14 16:41:19 -08:00
  • d79dc420cb README describes how to identify OpenMM directory peastman 2021-01-14 16:39:08 -08:00
  • e0d3b6ecb5 TorchForce supports global parameters Peter Eastman 2021-01-05 16:14:59 -08:00
  • d00a0fe2a9 Merge pull request #17 from yaoyic/upstream-master peastman 2020-10-06 17:20:07 -07:00