ACEsuit/mace

Issues

• multihead fine-tuning starting from a previously fit model gives a tensor size error (E0 related?)

  #622 opened 3 months ago by bernstei
  16

• MACE with latest version of LAMMPS, giving error

  #587 opened 2 months ago by ShubhangG
  8

• Evaluating multihead model on test set fails unless `head_key` is in the info.

  #680 opened 2 months ago by ThomasWarford
  0

• Release mismatch between PyPi and github

  #676 opened 2 months ago by R1j1t
  1

• Error at end of training

  #672 opened 2 months ago by cecilia-hong
  3

• Stacked at evaluation at the end of training for multi-GPU and patience termination

  #677 opened 2 months ago by dbidoggia
  0

• Issue: NotImplementedError with aten::empty_strided on CPU-Only Machine

  #656 opened 2 months ago by GUANGZChen
  16

• Mat shape missing match for `Multihead` fine-tune

  #615 opened 3 months ago by MengnanCui
  12

• "mace_run_train --help" should list defaults

  #644 opened 2 months ago by bernstei
  0

• Run time Error after training model using tutorial dataset

  #665 opened 2 months ago by spriti523
  10

• multigpu training with torchrun uses only device:0

  #666 opened 2 months ago by alvarovm
  0

• Simulations/Energy minimizations on multiple GPUs with MACE/ASE

  #649 opened 2 months ago by betterwithchem
  4

• Use proper logging instead of print statements

  #650 opened 2 months ago by Andrew-S-Rosen
  0

• Error when specifying more e0s then present in training set

  #605 opened 3 months ago by LarsSchaaf
  1

• Specifying E0s with .json During Multihead Training

  #651 opened 2 months ago by ThomasWarford
  0

• A error on installing MACE in LAMMPS

  #660 opened 2 months ago by WeJear
  3

• Cannot pass model paths to mace_mp

  #653 opened 2 months ago by ElliottKasoar
  0

• How to apply it to my DFT dataset produced by VASP?

  #659 opened 2 months ago by zhoufy20
  0

• mace_run_train loss output needs more digits

  #646 opened 2 months ago by bernstei
  3

• Is it possible to develop ReaxFF force fields in LAMMPS-MACE?

  #641 opened 2 months ago by Matt-A-Thompson
  7

• NaCl cube example missing in drive (example-xyz-movie/aqueous interface/NaCl_cube)

  #640 opened 2 months ago by dgvalcarce
  1

• Model architecture with --model="ScaleShiftMACE"

  #633 opened 3 months ago by wcwitt
  1

• Performance and scalability benchmarks for MACE

  #645 opened 2 months ago by hatemhelal
  1

• Unit of get_potential_energy()

  #630 opened 2 months ago by and-tos
  2

• MACE compile mode test failures with InternalTorchDynamoError and Unsupported exceptions in main

  #639 opened 2 months ago by yixian929
  1

• better default for device used to _save_ model in run_train

  #635 opened 3 months ago by bernstei
  2

• ASE Calculator example fails when using compile_mode

  #631 opened 3 months ago by hatemhelal
  1

• Reproducing Readme

  #629 opened 3 months ago by and-tos
  1

• load_from_xyz() got an unexpected keyword argument 'head_key'

  #627 opened 3 months ago by cecilia-hong
  5

• An error occurs when running tutorial: T01-MACE-Practice-I

  #600 opened 3 months ago by mast1986
  12

• MD22 MD17 train and test data split

  #623 opened 3 months ago by jhmlam
  2

• Problem when finalizing the training of a pure water model

  #581 opened 3 months ago by nihil39
  23

• preprocess_data.py error: 'Namespace' object has no attribute 'work_dir'

  #614 opened 3 months ago by axgmz
  5

• Issue creating LAMMPS model from EnergyDipoleMACE model

  #582 opened 3 months ago by lemessuriernb
  3

• Loosen the dependency to avoid conflict with other package

  #588 opened 3 months ago by chiang-yuan
  8

• How to transfer the mace pre-trained model from dtype_float64 to dtype_float32 while running lammps

  #603 opened 3 months ago by junqili
  4

• python 3.9 support foiled by trivial syntax

  #617 opened 3 months ago by bernstei
  2

• ASE dipole calculation

  #610 opened 3 months ago by J-Fabila
  1

• Errors when using mace in combination with lammps.

  #611 opened 3 months ago by fhfeng21
  2

• Is Structural Optimization Possible with MACE? Looking for Sample Code

  #608 opened 3 months ago by bishopfunc
  3

• Segmentation fault

  #592 opened 3 months ago by alfonso2166
  4

• Fine-tuning raises `RuntimeError: both inputs should have same dtype`

  #607 opened 3 months ago by MoseyQAQ
  2

• Is there a tag I can use to save most recent or last epoch as my checkpoint?

  #606 opened 3 months ago by ankitmathanker
  1

• MD Simulation with MACE Results in NaN Temperature and Unchanging Energy

  #594 opened 3 months ago by bishopfunc
  7

• aarch64 support for MACE?

  #598 opened 3 months ago by ShubhangG
  1

• mace_create_lammps_model fails to load mace_agnesi_small.model

  #596 opened 3 months ago by stenczelt
  0

• Fine-tuning of foundation model failure on main branch

  #590 opened 3 months ago by stenczelt
  2

• Example of using pretrained foundation models in LAMMPS

  #593 opened 3 months ago by ChiahsinChu
  2

• Having trouble reproducing Large Model Training

  #591 opened 3 months ago by daniel-sintef
  0

• PytorchStreamReader failed locating file constants.pkl

  #585 opened 4 months ago by Olivier7017
  1