Formatting Multiple Datasets in the --train_file/--valid_file Argument: Syntax and Best Practices

[Byebyjack]

Hello !
I have datasets from three different temperatures (e.g., data100K_train.xyz, data500K_train.xyz, data1000K_train.xyz). How should I format them in the --train_file/--valid_file argument?

Using --train_file="data100K_train.xyz, data500K_train.xyz, data1000K_train.xyz" doesn't seem to work.

[gabor1]

Concatenate them into a single file

[ilyes319]

If you download the latest main you can use the following format: https://mace-docs.readthedocs.io/en/latest/guide/heterogeneous_data.html

[Byebyjack]

Thank you for your responses. The issue was resolved when I concatenated them into a single file. However, an error occurred when I tried the method described in https://mace-docs.readthedocs.io/en/latest/guide/heterogeneous_data.html.

INFO: MACE version: 0.3.10
INFO: CUDA version: 11.8, CUDA device: 0
INFO: ===========LOADING INPUT DATA===========
INFO: Using heads: ['DFT']
INFO: ============= Processing head DFT ===========
Traceback (most recent call last):
File "/mace/cli/run_train.py", line 858, in
main()
File "/mace/cli/run_train.py", line 66, in main
run(args)
File "/mace/cli/run_train.py", line 222, in run
if check_path_ase_read(head_config.train_file):
File "/miniconda3/envs/mace/lib/python3.10/site-packages/mace/tools/scripts_utils.py", line 777, in check_path_ase_read
filepath = Path(filename)
File "/miniconda3/envs/mace/lib/python3.10/pathlib.py", line 960, in new
self = cls._from_parts(args)
File "/miniconda3/envs/mace/lib/python3.10/pathlib.py", line 594, in _from_parts
drv, root, parts = self._parse_args(args)
File "/miniconda3/envs/mace/lib/python3.10/pathlib.py", line 578, in _parse_args
a = os.fspath(a)
TypeError: expected str, bytes or os.PathLike object, not list

[ilyes319]

can you try on the latest develop branch?

[Byebyjack]

After trying the develop branch, the problem was solved. Thank you !