FAQs and Common Errors/Warnings

FAQs

  • Q: How does logging work? How do I use Tensorboard or Weights and Biases?

    A: Logging is configured under the trainer section of the config file by specifying the logger argument of the lightning.Trainer (see API). Compatible loggers are found here. Read the Config docs for a more complete description.

  • Q: What units do nequip framework models use?

    A: nequip has no prefered system of units and uses the units of the data provided. Model inputs, outputs, error metrics, and all other quantities follow the units of the dataset. Users must use consistent input and output units. For example, if the length unit is Å and the energy labels are in eV, the force predictions from the model will be in eV/Å. The provided force labels should hence also be in eV/Å.

    Warning

    nequip cannot and does not check the consistency of units in inputs you provide, and it is your responsibility to ensure consistent treatment of input and output units

  • Q: What floating point precision (torch.dtype) is used in the nequip framework?

    A: float64 precision is used for data (inputs to model and reference labels). Either float32 or float64 precision can be used as the model_dtype (which is a mandatory hyperparameter of models in the nequip framework). If float32 precision is used for model_dtype, the model will cast down from the float64 inputs (e.g. positions) and cast up the outputs (e.g. energy) to float64. A major change in the post-revamp nequip framework is that NequIP or Allegro models keep the initial embeddings in float64 before casting down if model_dtype=float32 for better numerics.

Commons Errors

  • Problem: Trying to run nequip-train as follows fails.

    nequip-train config.yaml

    Solution: Read the workflow docs and follow hydra’s command line options, e.g.

    nequip-train -cn config.yaml

Common Warnings

Other

  • Setting global configuration settings

    Warnings of the form

    Setting the GLOBAL value for ...

    nequip manages a number of global configuration settings of PyTorch and e3nn and correctly restores those values when a deployed model is loaded. These settings, however, must be set at a global level and thus may affect a host application; for this reason nequip by default will warn whenever overriding global configuration options. If you need to, these warnings can be silenced with set_global_options=True. (Setting set_global_options=False is strongly discouraged and might lead to strange issues or incorrect numerical results.)