PDEBench icon indicating copy to clipboard operation
PDEBench copied to clipboard

PDEBench: An Extensive Benchmark for Scientific Machine Learning

Results 30 PDEBench issues
Sort by recently updated
recently updated
newest added
trafficstars

Bumps [pypa/gh-action-pypi-publish](https://github.com/pypa/gh-action-pypi-publish) from 1.12.3 to 1.12.4. Release notes Sourced from pypa/gh-action-pypi-publish's releases. v1.12.4 ✨ What's Changed The main theme of this patch release that the support for uploading PEP 639...

dependencies

When I ran the command `python gen_ns_incomp.py`, I encountered the below error: File "/home/liyezhang/hzh/PDEBench/pdebench/data_gen/src/data_io.py", line 68, in to_centre_grid if isinstance(field, CenteredGrid): TypeError: isinstance() arg 2 must be a type or...

Thank you for the great work! 👍 I installed the environment following the instruction, and am attempting to generate 2D compressible NS data. With `numbers=1` in `2D_multi_Rand.yaml`, I run `sh...

The solutions in provided reaction--diffusion datasets appear to converge to approximately $0.5$ uniformly in space, rather than to $1$ as expected for the Fisher--KPP equation $u_t = \nu u_{xx} +...

Network problems often cause errors and interruptions in downloads. Each time you download again, you will recheck the existing files. Repeated re-downloads will cause repeated checks, which is extremely time-consuming....

In the UNet training script, when `training_type="single"`, I think the model might only be training on one batch element per epoch due to an indentation error. Specifically [this](https://github.com/pdebench/PDEBench/blob/9754b4c0ee2bcb967233a89d1c8a6a58bf7f7bc9/pdebench/models/unet/train.py#L315) `if`-clause should...

cf #70. This package is incompatible with most currently supported versions of it's pinned dependencies python (`3.11+`), torch/torchvision (`2.x.x`), numpy (`2.x.x`). Would it be possible to support more modern versions...

When I look at this site: https://github.com/pdebench/PDEBench And I click on the release link which goes here: https://github.com/pdebench/PDEBench/releases/tag/v0.1.0 It gives this for the version v0.1.0 But when I look in...

Actually, it's quite important to know whether the datas are DNS or LES. I didn't see any detaile description in the paper and the main page.