Devel update #22
Merged
Devel update #22
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Merge recent updates on devel to master.
* bump the testing Python version to 3.10 Test whether deepmd-kit is compatible with the latest Python. TF 2.8 was just released with Python 3.10 support. * add cp310 to wheel building * the version is not a float...
* do some small optimization to ops 1. avoid concat or add in loops. Instead, append tensors to a list, and concat or accumulate_n after loops 2. remove a duplicated reshape * revert unnecessary changes * revert wfc.py as it has been decrepated
* fix bug of mixed precision training * Update test_mixed_prec_training.py * fix UT error * fix UT error * fix UT error * address comment
* assign energy shift stats if atomic energies are assigned Atomic energies stats may be incorrect. Here, we use assigned atomic energies instead. I am training a model against multiple systems, where some atoms should not contribute energies. Before this commit, the loss of my model could not decrease. After this commit, it decreased quickly. * fix tests * add ut * keep and test the sum energy * fix typo
* calculate neighbor statistics from CLI In many cases, we need to confirm neighbor statistics before training. This command provides such CLI option. * add UT and docs * fix typo * correct name of the class * remove data_requirement
Before, only NN parameters were recovered.
* test: move loading graphs to setUpClass Loading graphs is expensive. * small fix * fix a typo * fix typo
* add an interface to eval descriptors Fix #1393. * move eval descriptor out of eval * bugfix and remove unnecessary changes * fix test * merge duplicated codes * bugfix * imap * Update deep_pot.py * do not return tuple * fix typo * make the code more readable
* run test_python on pre-built container * pin mistune to <2 See miyakogi/m2r#66 * I don't understand why we need m2r... * install m2r2 instead * try to encode with utf-8 instead.. * add back pip cache * Update test_python.yml * bump importlib_metadata version * Revert "bump importlib_metadata version" This reverts commit b8b1727. * Revert "Update test_python.yml" This reverts commit d9a906e. * set PYTHONPATH * Revert "set PYTHONPATH" This reverts commit d35bcf0.
* typo: 'should be in consistent' -> 'should be consistent' * typo: 'notice' -> 'important' Co-authored-by: Han Wang <amcadmus@gmail.com>
Co-authored-by: Han Wang <amcadmus@gmail.com>
`memset`, `cudaMemset`, and `hipMemset` only accept integer as value. It's meaningless to pass a float, whether the compiler optimizes or not. References: https://www.cplusplus.com/reference/cstring/memset/ https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__MEMORY.html#group__CUDART__MEMORY_1gf7338650f7683c51ee26aadc6973c63a https://rocmdocs.amd.com/en/latest/ROCm_API_References/HIP_API/Memory-Management.html#hipmemset
Co-authored-by: Han Wang <amcadmus@gmail.com>
Merge master into devel
Merge devel into master
The library type was changed from SHARED to MODULE in #1384. Fixes errors in conda-forge/deepmd-kit-feedstock#31
* support `type_one_side` with `exclude_types` * add support for se_r * fix compress issues * rewrite comments * fix compatibility with Python 3.6 * add unittests * rename model name to prevent conflict Co-authored-by: Jinzhe Zeng <jz748@syrah.rutgers.edu>
* save checkpoints after training if it's not saved * fix a typo in #1482 * support recovering sel<original_sel * supports increasing sel * add unittests * adjust variable name * add doc
#1573 has an issue: when the frozen graph is recovered again, `original_sel` is not the initial one (but the new one). In this commit, `original_sel` will be saved into the graph.
replace `stable_29Sep2021_update2` with `stable_29Sep2021_update3`
* add atom energy bias to type embedding energy Fix #684, where systems have different `atom_numb`. After this fix, RMSE should be quickly decreased in the very beginning. * looks like tf.repeat is unavaiable in old TF... * add statement `self.bias_atom_e is not None`
A tensor can be reused in a graph but a numpy array can't. This avoids saving duplicated constants in the case #1423.
A graph has a size limitation of 2 GB, so if we do not clean the graph, the limitation will be indeed 1GB...
* fix rcut in hybrid model compression
The current way `self.graph.get_operation_by_name('ProdEnvMatA')` to get `rcut` is incorrect for hybrid models. There may be several ProdEnvMatA ops in a graph.
(cherry picked from commit 34a4e9b)
* fix se_r attr
* rewrite data doc Current data doc was wrttien in the inital version. It's out-of-date. * provide the name of raw file * add raw file to input properties and labeled properties
jsdelivr, which is the default mathjax path in the sphinx, has been blocked in China. See jsdelivr/jsdelivr#18392. Of course, we hope our rendered equations can be shown in China.
* Update training.md * Fix typo * Update overall.md * Update overall.md * Update train-energy.md * Update train-fitting-tensor.md * Update train-fitting-tensor.md * Update train-fitting-tensor.md * Update README.md * Update README.md
…1686) * fix bug of model compression training with se_e2_r type descriptor * Update source/op/tabulate_multi_device.cc Co-authored-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu> Co-authored-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>
* Change Typo "short" --> "shot" * Update index.md
* set default fparam and aparam stat and recover from graph fix #1692 * check avg and std separately
1. Compare square value instead of square root value. Square root is expensive. 2. Compare float instead of double, considering rcut is provided as float. `ProdEnvMatA` (CPU) is observed to improve over 5%.
* implement parallelism for neighbor stat * bugfix * fix typo in docstring
* render equations in markdown files * enable dollarmath in the sphinx * fix Angstrom sybol; it's not supported by mathjax * fix equations of the learning rate * fix equations * fix equations * fix equations * fix equations * Update lammps-command.md * use \cdots for ... * add the detailed keys
Windows platform has different slashes..
* in model_devi, assumes nopbc if box is set to None * fix UT * Update deepmd/infer/model_devi.py Co-authored-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu> * fix make_model_devi Co-authored-by: Han Wang <wang_han@iapcm.ac.cn> Co-authored-by: Han Wang <amcadmus@gmail.com> Co-authored-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>
Fix #1724. (cherry picked from commit 4e41520c0a0351a44875ae7b78958734b0c0d4d9) Signed-off-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.