Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3444 commits
Select commit Hold shift + click to select a range
abf9575
fix: unify compute_output_stats across dp, pt, and pd backends (#5267)
wanghan-iapcm Feb 26, 2026
6b351f4
docs: add mathematical formulas to descriptor classes (#5255)
njzjz-bot Feb 26, 2026
f959a53
build(deps-dev): update scikit-build-core requirement from !=0.6.0,<0…
dependabot[bot] Feb 26, 2026
21d2b1a
fix(CI): pin protobuf<7.34 (#5272)
njzjz Feb 27, 2026
622179e
feat(pt): support spin virial (Rebased and updated version of #4545 b…
OutisLi Feb 27, 2026
f0a966b
feat(pt): log pre-clip gradient total_norm and per-param norm to tens…
OutisLi Feb 27, 2026
314d03a
feat(pt_expt): add dipole, polar, dos, property and dp-zbl models wit…
wanghan-iapcm Mar 1, 2026
a3db25a
CI(deps): bump torch (cpu) to 2.10.0 (#5273)
njzjz Mar 1, 2026
41757f2
refactor(training): remove unused learning_rate_dict multitask handli…
OutisLi Mar 3, 2026
b30bfcb
feat(skill): add skill for debug gradient flow in the pt expt backend…
wanghan-iapcm Mar 3, 2026
2deb70c
[pre-commit.ci] pre-commit autoupdate (#5282)
pre-commit-ci[bot] Mar 3, 2026
3b24cf6
refactor(optimizer): finalize optimizer schema and backend handling (…
OutisLi Mar 3, 2026
51fbd76
feat(pt, dpmodel): use data stat for observed type (#5269)
iProzd Mar 4, 2026
7e648f0
fix(c++): fix NULL type in custom op (#4889)
iProzd Mar 4, 2026
fdeff2b
feat(pt_expt): add training infrastructure (#5270)
wanghan-iapcm Mar 5, 2026
4a29836
feat(pt_expt): implement .pte inference pipeline with dynamic shapes …
wanghan-iapcm Mar 5, 2026
d331c9e
build(deps): bump docker/login-action from 3 to 4 (#5289)
dependabot[bot] Mar 5, 2026
c28419c
build(deps): bump actions/download-artifact from 7 to 8 (#5288)
dependabot[bot] Mar 5, 2026
8c5cb86
build(deps): bump actions/upload-artifact from 6 to 7 (#5291)
dependabot[bot] Mar 5, 2026
73bb1b7
feat(pt_expt): implementing energy hessian model (#5287)
wanghan-iapcm Mar 6, 2026
8f2b3c9
fix(pt, dpmodel): spin model finetune (#5281)
iProzd Mar 6, 2026
084edd6
ci: align package_c TensorFlow image and wheel to 2.20 (#5297)
njzjz-bot Mar 7, 2026
24e54bf
build(deps): bump pypa/cibuildwheel from 3.3 to 3.4 (#5290)
dependabot[bot] Mar 8, 2026
9322168
[pre-commit.ci] pre-commit autoupdate (#5300)
pre-commit-ci[bot] Mar 10, 2026
dabb0ca
feat(pretrained): add built-in pretrained downloader and alias backen…
njzjz-bot Mar 10, 2026
f22456f
ci(workflows): trust paddlepaddle host to bypass TLS outage (#5305)
njzjz-bot Mar 11, 2026
53cd6d3
build(deps-dev): bump tensorflow-cpu from 2.20.0 to 2.21.0 (#5304)
dependabot[bot] Mar 11, 2026
a8d2a8d
doc(pd): update PaddlePaddle version to 3.3.0 and 3.4.0(develop) (#5306)
HydrogenSulfate Mar 11, 2026
3e83c04
build(deps): bump docker/build-push-action from 6 to 7 (#5309)
dependabot[bot] Mar 12, 2026
387ee25
build(deps): bump docker/metadata-action from 5 to 6 (#5310)
dependabot[bot] Mar 12, 2026
c1a50ed
feat(dp, pt): add charge spin embedding (#5295)
iProzd Mar 12, 2026
5f25774
feat(pretrained): add DPA-2.4-7M to registry (#5307)
njzjz-bot Mar 13, 2026
ba7e289
fix(pd): fix adamw for paddle backend (#5308)
HydrogenSulfate Mar 13, 2026
09345bf
fix(lmp): default fparam support (#5311)
iProzd Mar 15, 2026
4ce5c73
fix: statisitcs -> statistics (#5315)
kmu Mar 16, 2026
87b1fff
chore(pre-commit): remove flake8 hook (#5324)
njzjz-bot Mar 17, 2026
b2805fb
[pre-commit.ci] pre-commit autoupdate (#5320)
pre-commit-ci[bot] Mar 17, 2026
3ab3779
feat(pt_expt): add dp freeze support and dp test tests for .pte model…
wanghan-iapcm Mar 18, 2026
a9bbf74
fix: `nan` error when all `fparam`/`aparam` have equal values (#5321)
Chengqian-Zhang Mar 18, 2026
08394be
fix(dpmodel): fix natoms[0] bug, einsum, and return type in EnergyLos…
wanghan-iapcm Mar 18, 2026
9ed2d48
feat(pt_expt): add frozen model (#5318)
wanghan-iapcm Mar 18, 2026
1ce26cf
refactor(pt): fully refactor of HybridMuon optimizer (#5275)
OutisLi Mar 18, 2026
b2c8511
feat(pretrained): add DPA3-Omol-Large (#5327)
njzjz-bot Mar 18, 2026
3f52fa9
feat(dp, pt): add force l2 norm loss & mae loss (#5294)
iProzd Mar 19, 2026
91e3d62
feat(pt_expt): add linear energy model (#5317)
wanghan-iapcm Mar 20, 2026
7f01a21
feat(pt_expt): implement DeepSpin model in pt_expt backend (#5293)
wanghan-iapcm Mar 21, 2026
b2923e8
docs: restructure and clarify learning-rate.md documentation (#5276)
OutisLi Mar 21, 2026
b97ad98
feat(train): show finish time in ETA logs (#5328)
OutisLi Mar 21, 2026
6122d97
feat(pt_expt): add dp compress support for pt_expt backend (#5323)
wanghan-iapcm Mar 23, 2026
034e613
feat(pt_expt): add dp finetune support (#5331)
wanghan-iapcm Mar 23, 2026
97e9c9e
feat(c++,pt-expt): add .pt2 (AOTInductor) C/C++ inference with DPA1/D…
wanghan-iapcm Mar 24, 2026
e97967b
feat(pt_expt): add dp change-bias support (#5330)
wanghan-iapcm Mar 24, 2026
08ceb95
[pre-commit.ci] pre-commit autoupdate (#5335)
pre-commit-ci[bot] Mar 27, 2026
acd4278
build(deps-dev): bump torch from 2.10.0 to 2.11.0 (#5347)
dependabot[bot] Mar 27, 2026
8fea5ab
build(deps): bump actions/deploy-pages from 4 to 5 (#5346)
dependabot[bot] Mar 27, 2026
a7e9fed
fix(tests): move shared test mixins to tests.common to fix DeviceCont…
wanghan-iapcm Mar 27, 2026
206ace0
fix(c++,pt-expt): substitute default_fparam in .pt2 compute (#5343)
wanghan-iapcm Mar 28, 2026
2a82988
feat(c++,pt-expt): add DeepPotModelDevi C++ tests for .pt2 backend (#…
wanghan-iapcm Mar 28, 2026
2a8fc21
refactor: extract shared tabulate math, remove pt_expt→pt dependency …
wanghan-iapcm Mar 29, 2026
cdd9b1a
feat(pt_expt): add missing losses (spin, DOS, tensor, property) (#5345)
wanghan-iapcm Mar 29, 2026
8060865
fix(pt): use strict=False when loading .pt checkpoints for inference …
wanghan-iapcm Mar 29, 2026
0828604
feat: add WSD LR Scheduler (#5326)
OutisLi Mar 29, 2026
7cd0d19
feat(property): Support element type-wise bias in property fitting (#…
Chengqian-Zhang Mar 30, 2026
14c349b
feat(loss): support three-value huber delta (#5329)
OutisLi Mar 30, 2026
f7124db
[pre-commit.ci] pre-commit autoupdate (#5357)
pre-commit-ci[bot] Mar 31, 2026
afb97f6
fix(pt): fix NaN Hessian in DPA2 and DPA3 (#5351)
njzjz Mar 31, 2026
bf3e483
Fix: dptest label mismatch (#5082)
anyangml Apr 1, 2026
dd8e492
refactor(pt_expt): use model API for inference, consistent file namin…
wanghan-iapcm Apr 2, 2026
6984e4c
feat(pd): add add_chg_spin_ebd parameter to DescrptDPA3 (#5333)
HydrogenSulfate Apr 2, 2026
d0cf8df
build(deps): bump codecov/codecov-action from 5 to 6 (#5363)
dependabot[bot] Apr 2, 2026
b02fd91
test(pt_expt): add .pt2 (AOTInductor) unit tests and bug fixes (#5334)
wanghan-iapcm Apr 3, 2026
efc27cf
refactor: use Array API-compatible tabulate math (#5366)
njzjz-bot Apr 5, 2026
adf5b83
perf(test): reduce redundant .pt2 compilations in GPU CI (#5379)
wanghan-iapcm Apr 7, 2026
52792f0
[pre-commit.ci] pre-commit autoupdate (#5381)
pre-commit-ci[bot] Apr 7, 2026
565f4be
feat(pt/dpmodel): add lmdb dataloader (#5283)
iProzd Apr 7, 2026
229cd57
docs(argcheck): fix typos and wording in doc strings (#5369)
njzjz-bot Apr 7, 2026
4880f21
feat(pt_expt): enhance per-module export tests + fix nframes symbol c…
wanghan-iapcm Apr 8, 2026
9d0275f
breaking(gmx): remove in-tree GROMACS patch integration (#5377)
njzjz-bot Apr 8, 2026
a9bc4a2
fix: missing log in multitask training (#5382)
iProzd Apr 8, 2026
e5b0f21
test(pt-expt): add .pte and .pt2 tests for dp convert-backend (#5384)
wanghan-iapcm Apr 8, 2026
59ee054
docs: fix spelling typos in documentation (#5387)
njzjz-bot Apr 9, 2026
d7ee913
build(pyproject): adopt PEP 639 license metadata (#5380)
njzjz-bot Apr 10, 2026
80fc539
feat(pt_expt): add DeepSpin support for pt_expt backend (#5370)
wanghan-iapcm Apr 10, 2026
67a30f2
docs(discussions): add lightweight q-and-a discussion template (#5385)
njzjz-bot Apr 10, 2026
baab3e8
test(consistent): replace Cartesian product with curated matrices for…
njzjz-bot Apr 10, 2026
345d162
feat(pt_expt): add eval_typeebd, eval_descriptor, eval_fitting_last_l…
wanghan-iapcm Apr 12, 2026
11156ad
CI: Update OS version in Read the Docs configuration (#5392)
njzjz Apr 13, 2026
0a9b4b6
test(consistent): curate DPA3 descriptor matrix (#5390)
njzjz-bot Apr 13, 2026
c005479
[pre-commit.ci] pre-commit autoupdate (#5395)
pre-commit-ci[bot] Apr 14, 2026
6e18c8f
build(deps): bump softprops/action-gh-release from 2 to 3 (#5399)
dependabot[bot] Apr 16, 2026
d42732e
build(deps): bump actions/upload-pages-artifact from 4 to 5 (#5398)
dependabot[bot] Apr 16, 2026
3f91293
feat(pt_expt): multi-task training support (#5397)
wanghan-iapcm Apr 20, 2026
1a1dc59
[pre-commit.ci] pre-commit autoupdate (#5409)
pre-commit-ci[bot] Apr 21, 2026
5d9cbdf
fix(pt): remove meaningless error raising (#5411)
OutisLi Apr 22, 2026
e98e1c3
fix(pt): typo in epoch training (#5410)
OutisLi Apr 22, 2026
54f42d9
feat(pt_expt): add LMDB dataset support (#5408)
wanghan-iapcm Apr 23, 2026
aa44999
fix(pt_expt): respect model default_fparam in data requirement (#5417)
wanghan-iapcm Apr 24, 2026
dadb45c
test(lmp): avoid duplicate pb conversions in tests (#5400)
njzjz-bot Apr 24, 2026
5c22e17
feat(pt): add full validation (#5336)
OutisLi Apr 24, 2026
c454e49
fix(pt): fsdp unavailable in older version of pytorch (≦2.5) (#5415)
OutisLi Apr 25, 2026
8419749
Fix: Square atom_norm in non-Huber energy and virial loss calculation…
anyangml Apr 25, 2026
6ec852d
ci: make Codecov upload non-blocking (#5421)
njzjz-bot Apr 25, 2026
d14233e
perf(pt2): optimize .pt2 C++ inference path (#5407)
wanghan-iapcm Apr 25, 2026
c6ca671
fix(pt): recognize AOTInductor-wrapped CUDA OOM in AutoBatchSize (#5418)
OutisLi Apr 26, 2026
1f4db6f
perf(pt): optimize HybridMuon optimizer (#5412)
OutisLi Apr 26, 2026
a3f548d
feat(pt): full validation support lmdb format (#5419)
OutisLi Apr 26, 2026
9d63816
feat(pt/dpmodel): add max and filter mode for lmdb (#5413)
OutisLi Apr 26, 2026
8841b5e
[pre-commit.ci] pre-commit autoupdate (#5426)
pre-commit-ci[bot] Apr 28, 2026
c38dc00
feat(pt_expt): support .pt training checkpoints in DeepEval (#5423)
wanghan-iapcm Apr 29, 2026
36d4137
feat(pt_expt): make model.json optional in .pt2/.pte loading (#5416)
OutisLi Apr 30, 2026
a094624
feat(pt): optimze HybridMuon by borrowing some ideas from deepseek v4…
OutisLi May 1, 2026
0a481de
fix(dpmodel): align NativeLayer bias/idt init with pt MLPLayer (#5428)
wanghan-iapcm May 3, 2026
5bd0889
fix(pt): replace in-place nlist masking to enable CUDA graph capture …
wanghan-iapcm May 7, 2026
be9d87b
sync(skills): feat: add LAMMPS DeePMD skill (#13)
njzjz-bot Mar 10, 2026
c3b2ff9
sync(skills): style: add mdformat hook (#26)
njzjz Mar 14, 2026
373ce6d
sync(skills): fix(lammps-deepmd): pin lammps version with `lmp` extra…
njzjz Mar 15, 2026
8946483
sync(skills): imp(description): update all SKILL.md files (#37)
light-cyan Mar 21, 2026
9f0efaf
sync(skills): Add skills for deepmd training/finetune/inference (#35)
iProzd Mar 21, 2026
2091918
sync(skills): chore(skills): unify licenses to LGPL-3.0-or-later (#52)
njzjz-bot Mar 23, 2026
b30b50d
sync(skills): improve deepmd skills (#56)
iProzd Mar 26, 2026
5156c40
sync(skills): Add a description about `mass` (#73)
light-cyan May 7, 2026
8e4ea33
chore(skills): import DeePMD-kit agent skills (#5437)
njzjz May 8, 2026
78bffde
fix(pt): base LambdaLR on configured start_lr (#5434)
OutisLi May 8, 2026
eab3419
build(jax): split pinned cpu and gpu groups (#5436)
njzjz-bot May 9, 2026
57f870f
feat(pt): add ema shadow model (#5420)
OutisLi May 10, 2026
7ca7baf
[pre-commit.ci] pre-commit autoupdate (#5441)
pre-commit-ci[bot] May 11, 2026
d67e75b
feat(pt_expt): multi-rank LAMMPS support for GNN models (DPA3 / DPA2 …
wanghan-iapcm May 13, 2026
f5c6394
fix(pt): Treat cuBLAS allocation failures as PyTorch OOM during auto …
OutisLi May 13, 2026
9847afb
CI(deps): Update jax (cpu) version to 0.10.0 in pyproject.toml (#5442)
njzjz May 14, 2026
5215c53
test(infer): swap se_e2_a fixture to deeppot_sea (type_one_side=True)…
wanghan-iapcm May 17, 2026
36ba4ff
feat(pt/dpmodel): add sequential_update for dpa3 (#5355)
iProzd May 17, 2026
01bcf47
feat(pt/dpmodel): add use_default_pf (#5356)
iProzd May 18, 2026
d3f08f3
[pre-commit.ci] pre-commit autoupdate (#5447)
pre-commit-ci[bot] May 19, 2026
4604131
fix(pt_expt): fail-fast on .pt2 GNN inference without LAMMPS atom-map…
wanghan-iapcm May 22, 2026
e7108b6
build(deps): bump @tootallnate/once from 2.0.0 to 2.0.1 in /source/no…
dependabot[bot] May 22, 2026
3384efb
test(infer,lmp): drop stale DPA2 .pth fixtures, repoint tests to deep…
wanghan-iapcm May 23, 2026
9245a7b
fix(test): tear down LAMMPS before MPI.Finalize() in mpirun test runn…
wanghan-iapcm May 23, 2026
f39a081
feat(dpa3): decouple charge_spin from fparam (#5431)
iProzd May 24, 2026
4e64f8b
[pre-commit.ci] pre-commit autoupdate (#5456)
pre-commit-ci[bot] May 26, 2026
a7c597d
test(pt_expt): shrink change-bias water dataset to 5 frames (#5467)
wanghan-iapcm May 27, 2026
2087416
feat(jax): add training (#5460)
njzjz-bot May 28, 2026
93f5580
test(universal): replace Cartesian product with curated matrices for …
iProzd May 28, 2026
e679b8d
fix(pt_expt): let deepmd.pt import errors propagate in comm op check …
wanghan-iapcm May 29, 2026
0ce2351
set default timeout for pytest (#5265)
link89 May 29, 2026
ccaa551
feat(pt): Add DPA4/SeZM descriptor & model 🎉🎉🎉 (#5448)
OutisLi May 30, 2026
016141f
fix(cmake): install DeePMD config for C library builds (#5475)
njzjz-bot May 30, 2026
967e525
fix(pt_expt): Fix compiled ckpt load key mismatch (#5468)
anyangml Jun 1, 2026
c6f2479
fix: fake border op import (#5451)
anyangml Jun 1, 2026
c67b7f7
fix(pt/pd): fix incompatibility between AutoBatchSize and eval hooks …
njzjz Jun 1, 2026
1de6de9
docs: Update JAX backend documentation on training support (#5482)
njzjz Jun 2, 2026
27a18b6
perf(pt_expt): share compiled forward_lower across multi-task shared-…
anyangml Jun 2, 2026
65c0420
chore: add DPA-3.3-1M (#5489)
anyangml Jun 3, 2026
fb6ff93
fix(cc): handle nloc==0 in DeepSpinPTExpt with phantom-atom padding (…
wanghan-iapcm Jun 4, 2026
1b54dc5
[pre-commit.ci] pre-commit autoupdate (#5484)
pre-commit-ci[bot] Jun 4, 2026
dd428db
Chore: update DPA-2.4-7M ckpt (#5496)
anyangml Jun 5, 2026
87d6e4e
fix: dpa4 compile aliasing collision (#5483)
anyangml Jun 5, 2026
b77223e
fix(pt): clone inference coords before enabling grad (#5476)
njzjz-bot Jun 6, 2026
2362236
feat(pt_expt): pluggable NeighborList strategy + O(N) vesin neighbor …
wanghan-iapcm Jun 6, 2026
2e3117e
fix: miscalculation of num_steps when using num_epoch and lmdb (#5488)
OutisLi Jun 6, 2026
99c1ece
build: move vesin[torch] to the torch extra (fix conda-forge) (#5501)
wanghan-iapcm Jun 7, 2026
e05af21
fix(pt_expt): report average step time in the normal training log (#5…
wanghan-iapcm Jun 8, 2026
5f848ad
test(pt_expt): tolerance for test_default_fallback (CUDA scatter nond…
wanghan-iapcm Jun 9, 2026
c76c14e
docs(agent-skills): add DeePMD-kit skill guide (#5497)
njzjz-bot Jun 9, 2026
3c44661
[pre-commit.ci] pre-commit autoupdate (#5506)
pre-commit-ci[bot] Jun 9, 2026
890e38a
feat(dpa4): multiple updates for DPA4/SeZM (#5503)
OutisLi Jun 10, 2026
0de53e9
feat(dpmodel): backend-agnostic DPA4 descriptor and fitting (core pat…
wanghan-iapcm Jun 12, 2026
5d94bd6
build(deps): bump codecov/codecov-action from 6 to 7 (#5513)
dependabot[bot] Jun 11, 2026
be6af67
fix(test): unpack charge_spin in test_pt2_dual_artifact_for_gnn (#5523)
wanghan-iapcm Jun 13, 2026
c0b0319
refactor(jax): auto-convert dpmodel modules (#5527)
njzjz Jun 14, 2026
87d8557
refactor(tests): auto-convert array-api-strict modules (#5528)
njzjz Jun 14, 2026
16df629
feat(dpa4): use edge force and atomic virial (#5518)
OutisLi Jun 14, 2026
c89199d
feat(pt/ptexpt): freeze parameters when infer only (#5512)
OutisLi Jun 14, 2026
84f8541
feat(pt): avoid grad clip sync (#5519)
OutisLi Jun 14, 2026
beb50da
fix(pt): sort nlist for compressed se_e2_a in forward_lower (#5524)
wanghan-iapcm Jun 15, 2026
2cbef36
feat(pt_expt): DPA4 descriptor and fitting (training + export) (#5522)
wanghan-iapcm Jun 14, 2026
d3834e2
CI: update `test_cuda` `runs-on` to `gpu` (#5541)
njzjz Jun 16, 2026
0dbb2a7
fix(pt): HybridMuon ZeRO checkpoint slowdown (#5525)
OutisLi Jun 16, 2026
69409c5
fix(pt,dpmodel): drop out-of-rcut neighbors in _format_nlist pad bran…
wanghan-iapcm Jun 16, 2026
3ca4e49
docs(skills): consolidate training skill disclosure (#5530)
njzjz-bot Jun 16, 2026
370935e
[pre-commit.ci] pre-commit autoupdate (#5539)
pre-commit-ci[bot] Jun 16, 2026
3bb17d5
Chore: refactor pt-expt compile (#5504)
anyangml Jun 16, 2026
ca0ce63
fix(jax): restore batched neighbor statistics (#5545)
njzjz Jun 17, 2026
b57aef7
feat(pt_expt): DPA4 inference (.pt2 freeze, DeepEval, C++, LAMMPS) (#…
wanghan-iapcm Jun 17, 2026
f56ec2e
fix(api): trim PyTorch type map separators (#5544)
njzjz-bot Jun 17, 2026
69b56f8
test(pt): document se_e3_tebd gr output (#5536)
njzjz-bot Jun 17, 2026
9aaa9e7
CI: drop container in `test-cuda` (#5543)
njzjz Jun 17, 2026
4a552e3
fix(dpa4): use triton_op to enable triton when freezing (#5521)
OutisLi Jun 17, 2026
a4c5592
fix(pt): stop plain pt dp test from eager-loading pt_expt custom-op f…
wanghan-iapcm Jun 17, 2026
53c5968
fix(pt/pt-expt): preserve shared tensors in change-bias (#5534)
njzjz Jun 17, 2026
e3fda45
feat: add charge spin to cpp runtime (#5509)
anyangml Jun 17, 2026
d0a1959
perf(dpa4): opt so3grid (with pt_expt GridProduct wrapping fix) (#5552)
wanghan-iapcm Jun 18, 2026
1050e46
build(deps): bump pypa/cibuildwheel from 3.4 to 4.1 (#5554)
dependabot[bot] Jun 18, 2026
873e29a
feat(dpmodel): complete DPA4/SeZM SO3 grid projection (mirror current…
wanghan-iapcm Jun 19, 2026
69aa3e1
feat(lmp): provide dE/dfparam to LAMMPS via finite differences (#5498)
Lily200202 Jun 19, 2026
5a0d505
fix(jax): stabilize repflow dynamic selection export (#5533)
njzjz Jun 19, 2026
b1cd6dc
fix(pt): enable second-order autograd for tabulate descriptors (#5537)
njzjz Jun 20, 2026
4b6506d
feat(dpa4): so3_readout across pt + dpmodel/pt_expt backends (#5561)
wanghan-iapcm Jun 20, 2026
f0cf2b9
fix(train): allow zero-step training with bias adjustment (#5477)
njzjz-bot Jun 20, 2026
c5c57d6
fix(lmp): use forward comm for PPPM DPLR AD (#5567)
Yi-FanLi Jun 21, 2026
aae3485
feat(pt_expt): load plugin entry points (#5559)
njzjz-bot Jun 21, 2026
2651de9
fix(pt/train): keep epoch-derived num_steps consistent across ranks (…
OutisLi Jun 21, 2026
03682bf
refactor(pt/dpa4): use edge schema for dpa4 and ignore sel (#5562)
OutisLi Jun 22, 2026
ef45bdf
fix(pt-expt): detach tensors before numpy serialization (#5566)
njzjz Jun 22, 2026
f9554f1
build: bypass Torch CUDA discovery without nvcc (#5558)
njzjz-bot Jun 23, 2026
95fa690
fix(gpu): avoid neighbor-list prefix scan deadlock (#5575)
njzjz Jun 23, 2026
7fed814
feat(pt): add embedding sub cli for pt models (#5571)
OutisLi Jun 23, 2026
dae86a9
docs: add DPA3 cyclohexane distillation tutorial (#5564)
njzjz-bot Jun 23, 2026
b52c359
test(api): add universal DeepPot C/C++ API tests (#5538)
njzjz-bot Jun 23, 2026
99523dc
feat(dpmodel): NeighborGraph foundation — contract, segment toolkit, …
wanghan-iapcm Jun 25, 2026
5733301
feat(pt/dpa4): support multi-rank lammps inference (#5580)
OutisLi Jun 25, 2026
3ff55bf
build(deps): bump actions/cache from 5 to 6 (#5586)
dependabot[bot] Jun 25, 2026
01c75f4
build(deps): bump actions/checkout from 6 to 7 (#5585)
dependabot[bot] Jun 25, 2026
c272661
fix(finetune): warn on descriptor config mismatches (#5549)
njzjz-bot Jun 25, 2026
330fa75
feat: dp test stress (#5588)
OutisLi Jun 26, 2026
adfa278
[pre-commit.ci] pre-commit autoupdate (#5576)
pre-commit-ci[bot] Jun 26, 2026
2b3bb05
feat(tf): support training stat_file (#5551)
njzjz-bot Jun 27, 2026
2d4c615
feat: enable PyTorch custom ops by default (#5578)
njzjz-bot Jun 27, 2026
dbdc9a3
feat(pt): add atomic charge population fitting and model (#5494)
chrisahart Jun 27, 2026
1f0ca6c
feat(pt/dpa4): Support property fitting in DPA4/SeZM models (#5587)
OutisLi Jun 27, 2026
a9bcbc5
feat(pt): add custom save behaviors (#5589)
OutisLi Jun 28, 2026
ac8e430
feat(jax): add JAX-MD interface (#5590)
njzjz Jun 28, 2026
5082854
feat(pt_expt): dpa1(attn_layer=0) graph-native NeighborGraph forward …
wanghan-iapcm Jun 29, 2026
58bef11
fix(tf): add InvalidArgument compatibility wrapper (#5600)
njzjz Jun 29, 2026
1550599
fix(tabulate): add c1 extrapolation outside tables (#5601)
njzjz Jun 29, 2026
f143171
feat(dpmodel): add descriptor compression (#5592)
njzjz Jun 29, 2026
73de44b
feat(tf2): add eager TensorFlow array backend (#5598)
njzjz Jun 29, 2026
63845a5
fix(jax): reuse atomic forward outputs in force path (#5605)
njzjz Jun 30, 2026
92c6a9a
fix(pt): fail early when compress lacks custom ops (#5610)
njzjz-bot Jun 30, 2026
ee3bda4
ci: limit Python PR test matrix (#5611)
njzjz-bot Jun 30, 2026
5d4ad01
test(consistent): use curated cases throughout (#5606)
njzjz Jun 30, 2026
54c64ad
feat(pt): support shared_dict in linear energy model (#5548)
njzjz-bot Jun 30, 2026
0e5c170
fix(jax): skip padding for non-XLA SavedModels (#5602)
njzjz Jun 30, 2026
b22a3c3
refactor(api_cc): load backends as plugins (#5531)
njzjz Jun 30, 2026
e582360
fix(jax): restore jax2tf savedmodel export (#5613)
njzjz Jul 1, 2026
8f18143
refactor(dpmodel/dpa4) (#5599)
OutisLi Jul 1, 2026
44e5007
fix(data): preserve global dipole/polarizability in raw_to_set.sh (#5…
wanghan-iapcm Jul 1, 2026
366e014
fix(dev): fix devcontainer LAMMPS pytest wrapper paths (#5697)
wanghan-iapcm Jul 1, 2026
ee30956
fix(examples): use runnable local data paths in zinc_protein example …
wanghan-iapcm Jul 1, 2026
e19bd18
feat(show): add serialization-tree via DeepEval.serialize (#5316)
njzjz-bot Jul 1, 2026
5e80487
feat(core): batch prod env mat over frames (#5582)
njzjz Jul 1, 2026
a49b6f7
add DPA-ADAPT toolkit for downstream property adaptation (#5572)
zhaiwenxi Jul 1, 2026
c93c84d
docs: fix stale documentation links and placeholders (#5700)
wanghan-iapcm Jul 1, 2026
d0e531d
fix(tf): keep linear-model loss configs after frozen components (#5701)
wanghan-iapcm Jul 1, 2026
42de67e
docs(examples): build and document graph.pb generation in infer_water…
wanghan-iapcm Jul 1, 2026
7d497df
fix(tf): accept standard checkpoint inputs in change-bias (#5702)
wanghan-iapcm Jul 1, 2026
0f19772
fix(tf): honor change-bias --numb-batch (#5703)
wanghan-iapcm Jul 1, 2026
9bd1f16
feat(pt_expt): graph-native dpa1 — .pt2 export + compiled training + …
wanghan-iapcm Jul 1, 2026
4c94171
fix: propagate frame-level data modifier failures (#5709)
wanghan-iapcm Jul 2, 2026
55d7e79
feat(dpmodel): add backend-independent trainer abstraction (#5603)
njzjz Jul 2, 2026
dd38b35
docs: disable Read the Docs PDF build (#5710)
njzjz-bot Jul 2, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
5 changes: 5 additions & 0 deletions .clang-format
Original file line number Diff line number Diff line change
@@ -0,0 +1,5 @@
---
BasedOnStyle: Google
BinPackParameters: false
InsertBraces: true
...
3 changes: 3 additions & 0 deletions .devcontainer/Dockerfile
Original file line number Diff line number Diff line change
@@ -0,0 +1,3 @@
FROM mcr.microsoft.com/devcontainers/cpp:1-ubuntu-24.04

COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
40 changes: 40 additions & 0 deletions .devcontainer/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,40 @@
# DeePMD-kit devcontainer environment

This [devcontainer](https://vscode.js.cn/docs/devcontainers/devcontainer-cli) environment setups Python and C++ environment to develop DeePMD-kit.
One can setup locally or use [GitHub Codespaces](https://docs.github.com/en/codespaces) by clicking the Code button on the DeePMD-kit repository page.
The whole setup process requires about 10 minutes, so one needs to be patient.

## Python environment

The following packages are installed into the Python environment `.venv`:

- DeePMD-kit (in edit mode)
- Backends including TensorFlow, PyTorch, JAX
- LAMMPS
- MPICH
- CMake
- prek (including hooks)
- Test packages including pytest
- Doc packages including sphinx

## C++ interface

The C++ interface with TensorFlow and PyTorch support is installed into `dp` directory.

When calling and debuging LAMMPS with DeePMD-kit, use the following scripts instead of the regular `lmp`:

- `.devcontainer/lmp`
- `.devcontainer/gdb_lmp`

Use the following scripts for `pytest` with LAMMPS:

- `.devcontainer/pytest_lmp`
- `.devcontainer/gdb_pytest_lmp`

## Rebuild

Usually the Python package does not need to reinstall.
But when one wants to recompile the C++ code, the following scripts can be executed.

- `.devcontainer/build_cxx.sh`
- `.devcontainer/build_py.sh`
22 changes: 22 additions & 0 deletions .devcontainer/build_cxx.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,22 @@
#!/bin/bash
set -ev

NPROC=$(nproc --all)
SCRIPT_PATH=$(dirname $(realpath -s $0))

export CMAKE_PREFIX_PATH=${SCRIPT_PATH}/../libtorch
TENSORFLOW_ROOT=$(python -c 'import importlib.util,pathlib;print(pathlib.Path(importlib.util.find_spec("tensorflow").origin).parent)')

mkdir -p ${SCRIPT_PATH}/../buildcxx/
cd ${SCRIPT_PATH}/../buildcxx/
cmake -D ENABLE_TENSORFLOW=ON \
-D ENABLE_PYTORCH=ON \
-D ENABLE_PADDLE=ON \
-D CMAKE_INSTALL_PREFIX=${SCRIPT_PATH}/../dp/ \
-D LAMMPS_VERSION=stable_22Jul2025_update2 \
-D CMAKE_BUILD_TYPE=Debug \
-D BUILD_TESTING:BOOL=TRUE \
-D TENSORFLOW_ROOT=${TENSORFLOW_ROOT} \
${SCRIPT_PATH}/../source
cmake --build . -j${NPROC}
cmake --install .
8 changes: 8 additions & 0 deletions .devcontainer/build_py.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,8 @@
#!/bin/bash
set -ev

SCRIPT_PATH=$(dirname $(realpath -s $0))
cd ${SCRIPT_PATH}/..

uv sync --dev --python 3.12 --extra cpu --extra torch --extra jax --extra lmp --extra test --extra docs
prek install
16 changes: 16 additions & 0 deletions .devcontainer/devcontainer.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
{
"name": "DeePMD-kit",
"build": {
"dockerfile": "Dockerfile"
},
"features": {
"ghcr.io/devcontainers/features/github-cli:1": {}
},
"postCreateCommand": ".devcontainer/build_py.sh && .devcontainer/download_libtorch.sh && .devcontainer/build_cxx.sh && prek install-hooks",
"remoteEnv": {
"PATH": "${containerEnv:PATH}:${containerWorkspaceFolder}/.venv/bin",
"DP_ENABLE_PYTORCH": "1",
"DP_VARIANT": "cpu",
"UV_EXTRA_INDEX_URL": "https://download.pytorch.org/whl/cpu"
}
}
8 changes: 8 additions & 0 deletions .devcontainer/download_libtorch.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,8 @@
#!/bin/bash
set -ev

SCRIPT_PATH=$(dirname $(realpath -s $0))
cd ${SCRIPT_PATH}/..

wget https://download.pytorch.org/libtorch/cpu/libtorch-cxx11-abi-shared-with-deps-2.8.0%2Bcpu.zip -O ~/libtorch.zip
unzip ~/libtorch.zip
9 changes: 9 additions & 0 deletions .devcontainer/gdb_lmp
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
#!/bin/bash
SCRIPT_PATH=$(dirname $(realpath -s $0))

export CMAKE_PREFIX_PATH=${SCRIPT_PATH}/../libtorch
TENSORFLOW_ROOT=$(python -c 'import importlib.util,pathlib;print(pathlib.Path(importlib.util.find_spec("tensorflow").origin).parent)')

env LAMMPS_PLUGIN_PATH=${SCRIPT_PATH}/../dp/lib/deepmd_lmp \
LD_LIBRARY_PATH=${SCRIPT_PATH}/../dp/lib:${CMAKE_PREFIX_PATH}/lib:${TENSORFLOW_ROOT} \
gdb ${SCRIPT_PATH}/../.venv/lib/python3.12/site-packages/lammps/lmp "$@"
9 changes: 9 additions & 0 deletions .devcontainer/gdb_pytest_lmp
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
#!/bin/bash
SCRIPT_PATH=$(dirname $(realpath -s $0))

export CMAKE_PREFIX_PATH=${SCRIPT_PATH}/../libtorch
TENSORFLOW_ROOT=$(python -c 'import importlib.util,pathlib;print(pathlib.Path(importlib.util.find_spec("tensorflow").origin).parent)')

env LAMMPS_PLUGIN_PATH=${SCRIPT_PATH}/../dp/lib/deepmd_lmp \
LD_LIBRARY_PATH=${SCRIPT_PATH}/../dp/lib:${CMAKE_PREFIX_PATH}/lib:${TENSORFLOW_ROOT} \
gdb --args python -m pytest -s "$@"
9 changes: 9 additions & 0 deletions .devcontainer/lmp
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
#!/bin/bash
SCRIPT_PATH=$(dirname $(realpath -s $0))

export CMAKE_PREFIX_PATH=${SCRIPT_PATH}/../libtorch
TENSORFLOW_ROOT=$(python -c 'import importlib.util,pathlib;print(pathlib.Path(importlib.util.find_spec("tensorflow").origin).parent)')

env LAMMPS_PLUGIN_PATH=${SCRIPT_PATH}/../dp/lib/deepmd_lmp \
LD_LIBRARY_PATH=${SCRIPT_PATH}/../dp/lib:${CMAKE_PREFIX_PATH}/lib:${TENSORFLOW_ROOT} \
${SCRIPT_PATH}/../.venv/bin/lmp "$@"
9 changes: 9 additions & 0 deletions .devcontainer/pytest_lmp
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
#!/bin/bash
SCRIPT_PATH=$(dirname $(realpath -s $0))

export CMAKE_PREFIX_PATH=${SCRIPT_PATH}/../libtorch
TENSORFLOW_ROOT=$(python -c 'import importlib.util,pathlib;print(pathlib.Path(importlib.util.find_spec("tensorflow").origin).parent)')

env LAMMPS_PLUGIN_PATH=${SCRIPT_PATH}/../dp/lib/deepmd_lmp \
LD_LIBRARY_PATH=${SCRIPT_PATH}/../dp/lib:${CMAKE_PREFIX_PATH}/lib:${TENSORFLOW_ROOT} \
python -m pytest "$@"
3 changes: 3 additions & 0 deletions .git-blame-ignore-revs
Original file line number Diff line number Diff line change
@@ -0,0 +1,3 @@
# .git-blame-ignore-revs
# pre-commit
97c38bb4235c9a440bd61a2459e330dba78269fa
3 changes: 3 additions & 0 deletions .git_archival.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,3 @@
node: $Format:%H$
node-date: $Format:%cI$
describe-name: $Format:%(describe:tags=true,match=*[0-9]*)$
4 changes: 4 additions & 0 deletions .gitattributes
Original file line number Diff line number Diff line change
@@ -0,0 +1,4 @@
# do not show up detailed difference on GitHub
source/3rdparty/* linguist-generated=true
source/3rdparty/README.md linguist-generated=false
.git_archival.txt export-subst
13 changes: 13 additions & 0 deletions .github/DISCUSSION_TEMPLATE/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
# Discussion Templates

This directory contains GitHub Discussion category forms for DeePMD-kit.

## Active template

- `q-a.yml` → **Q&A**

## Notes

- The filename must match the discussion category slug exactly.
- GitHub Discussion forms only support a limited set of top-level keys, such as `title`, `labels`, and `body`.
- This Q&A template is intentionally lightweight; most fields are optional and should be filled only when relevant.
92 changes: 92 additions & 0 deletions .github/DISCUSSION_TEMPLATE/q-a.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,92 @@
title: "[Q&A] "
labels: ["question"]
body:
- type: markdown
attributes:
value: |
Use this form for Q&A discussions.

This template is intentionally lightweight.
**Only fill the technical fields below if they are relevant to your question.**

If your question is about using DeePMD-kit or diagnosing a problem, version info and a minimal reproducible example can help others answer faster.
- type: textarea
id: question
attributes:
label: Question
description: Please describe your question clearly.
placeholder: What would you like to ask?
value:
validations:
required: true
- type: input
id: version
attributes:
label: DeePMD-kit Version
description: "If applicable: `dp --version`."
placeholder: "e.g., 2.2.7"
validations:
required: false
- type: input
id: backend-version
attributes:
label: Backend and its version
description: "If applicable: the backend and its version will be printed when running DeePMD-kit, e.g. TensorFlow v2.15.0."
placeholder: "e.g., TensorFlow v2.15.0"
validations:
required: false
- type: textarea
id: other-version
attributes:
label: Python Version, CUDA Version, GCC Version, LAMMPS Version, etc
description: "If applicable, specify your platform or environment details."
placeholder:
value:
validations:
required: false
- type: textarea
id: details
attributes:
label: Details
description: "If applicable, add more context such as what you are trying to do, what you have tried, and where you got stuck."
placeholder:
value:
validations:
required: false
- type: textarea
id: reproducible-example
attributes:
label: Reproducible Example, Input Files, and Commands
description: |
If applicable, provide a minimal reproducible example.
For software-usage questions, this is often the fastest way to get a useful answer.
placeholder: |
Commands:
```bash
dp train input.json
```

Input / config:
```json
{
"model": {}
}
```

Expected behavior:
...

Actual behavior / error message:
...
value:
validations:
required: false
- type: textarea
id: further
attributes:
label: Further Information, Files, and Links
description: Put any additional information here, including links to docs, papers, related issues, screenshots, or external resources.
placeholder:
value:
validations:
required: false
70 changes: 70 additions & 0 deletions .github/ISSUE_TEMPLATE/bug-report.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,70 @@
name: Bug report
description: Create a bug report to help us eliminate issues and improve deepmd-kit.
title: "[BUG] _Replace With Suitable Title_"
labels: bug
assignees: []
body:
- type: textarea
id: summary
attributes:
label: Bug summary
description: Please provide a clear and concise description of what the bug is.
placeholder:
value:
validations:
required: true
- type: input
id: version
attributes:
label: DeePMD-kit Version
description: "`dp --version`"
validations:
required: true
- type: input
id: backend-version
attributes:
label: Backend and its version
description: "The backend and its version will be printed when running DeePMD-kit, e.g. TensorFlow v2.15.0."
validations:
required: true
- type: dropdown
id: installation-way
attributes:
label: How did you download the software?
options:
- Offline packages
- conda
- docker
- pip
- Built from source
- dp1s
- Others (write below)
validations:
required: true
- type: textarea
id: log
attributes:
label: Input Files, Running Commands, Error Log, etc.
description: "Please provide necessary information including input file, running commands, error log , etc., AS DETAILED AS POSSIBLE to help locate and reproduce your problem. WARNING: Do not use image to show error log! Paste texts in a code block instead."
placeholder:
value:
validations:
required: true
- type: textarea
id: reproduce
attributes:
label: Steps to Reproduce
description: "Describe the steps required to (quickly) reproduce the issue. You can attach (small) files to the section below or add URLs where to download an archive with all necessary files. Please try to create an input set that is as minimal and small as possible and reproduces the bug as quickly as possible. **NOTE:** the less effort and time it takes to reproduce your reported bug, the more likely it becomes, that somebody will look into it and fix the problem."
placeholder:
value:
validations:
required: true
- type: textarea
id: further
attributes:
label: Further Information, Files, and Links
description: Put any additional information here, attach relevant text or image files and URLs to external sites, e.g. relevant publications
placeholder:
value:
validations:
required: false
5 changes: 5 additions & 0 deletions .github/ISSUE_TEMPLATE/config.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,5 @@
blank_issues_enabled: false
contact_links:
- name: Request for Help
url: https://github.com/deepmodeling/deepmd-kit/discussions/491
about: If you have a usage question
33 changes: 33 additions & 0 deletions .github/ISSUE_TEMPLATE/feature-request.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,33 @@
name: Feature request
description: Suggest an idea for this project.
title: "[Feature Request] _Replace with Title_"
labels: enhancement
assignees: []
body:
- type: textarea
id: summary
attributes:
label: Summary
description: "Please provide a brief and concise description of the suggested feature or change"
placeholder:
value:
validations:
required: true
- type: textarea
id: details
attributes:
label: Detailed Description
description: "Please explain how you would like to see deepmd-kit enhanced, what feature(s) you are looking for, what specific problems this will solve. If possible, provide references to relevant background information like publications or web pages, and whether you are planning to implement the enhancement yourself or would like to participate in the implementation. If applicable add a reference to an existing bug report or issue that this will address."
placeholder:
value:
validations:
required: true
- type: textarea
id: further
attributes:
label: Further Information, Files, and Links
description: Put any additional information here, attach relevant text or image files and URLs to external sites, e.g. relevant publications
placeholder:
value:
validations:
required: false
Loading