Issues · deepmodeling/deepmd-kit · GitHub

List of critical bugs giving incorrect results without error messages
#2866 · njzjz opened on Sep 26, 2023
lmp in 2.1.3 make unrecognized pair error
#1809 · pee8379 opened on Jul 2, 2022
13

Labels Milestones

[Feature Request] Improve pip installation path to handle CUDA Toolkit discovery for PyTorch/Inductor workflows

#5731

· qchempku2017 opened

on Jul 4, 2026

[Feature Request] DPA-4 / PyTorch backend import path requires `e3nn`, but `e3nn` is only declared as an optional `dpa-adapt` dependency

#5730

· qchempku2017 opened

on Jul 4, 2026

[BUG] Source installation with PyTorch + CUDA backend fails due to undocumented build-system dependencies and incomplete CUDA path detection

#5729

· qchempku2017 opened

on Jul 4, 2026

for the total across systems we'd need to sum all matches;

#5704

· github-actions opened

on Jul 1, 2026

replace with dedicated DescriptorExtractor class after refactor.

#5705

· github-actions opened

on Jul 1, 2026

[Code scan] Symmetrize ASE calculator stress after reshaping virial

#5689

· njzjz opened

on Jun 30, 2026

[Code scan] Preserve dtype in Paddle RepFlow dynamic aggregation

#5688

· njzjz opened

on Jun 30, 2026

[Code scan] Preserve Paddle fparam/aparam inputs when freezing models

#5687

· njzjz opened

on Jun 30, 2026

[Code scan] Freeze Paddle DPA3 parameters with stop_gradient

#5686

· njzjz opened

on Jun 30, 2026

[Code scan] Close TensorFlow EwaldRecp sessions

#5685

· njzjz opened

on Jun 30, 2026

[Code scan] Map TensorFlow spin virtual types by use_spin

#5680

· njzjz opened

on Jun 30, 2026

[Code scan] Inspect TensorFlow --init-model checkpoints before graph build

#5679

· njzjz opened

on Jun 30, 2026