tasks prim_bwd # task (prim_fwd, prim_bwd, prim_ibp, ode1, ode2) Python main.py -export_data true # main parameters This can easily be done by setting -export_data true: However, the generation process can take a while, so we recommend to first generate data, and export it into a dataset that can be used for training. If you want to use your own dataset / generator, it is possible to train a model by generating data on the fly. Note that these datasets and models slightly differ from the ones used in the paper. We also provide models trained on the above datasets, for integration: Model training dataĪnd for differential equations: Model training dataĪll accuracies above are given using a beam search of size 10. We provide datasets for each task considered in the paper: Dataset An ipython notebook with an interactive demo of the model on function integration.Models trained with different configurations of training data.Train / Valid / Test sets for all tasks considered in the paper.Ordinary differential equations with their solutions.PyTorch original implementation of Deep Learning for Symbolic Mathematics (ICLR 2020).
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |