FunFact
FunFact copied to clipboard
yhtang
Tensor decomposition with arbitrary expressions: inner, outer, elementwise operators; nonlinear transformations; and more.
In the readme, in the ``ff.factorize(tsrex, T, max_steps=1000, nvec=8, penalty_weight=10)`` the ``n_vec`` should be ``vec_size``
As indicated by @fabian-sp in #241, following the installation instructions on the GitHub README page and landing page on readthedocs only installs the NumPy backend which does not enable autograd...
This is related to the review of FunFact for JOSS (see https://github.com/openjournals/joss-reviews/issues/4502) I have a few questions regarding your optimization procedures. - If I understand correctly, i could use any...
This is related to the review of `FunFact` for JOSS (see https://github.com/openjournals/joss-reviews/issues/4502) I understand that your package can do (or approximate) arbitrary tensor decompositions with optimization. As the underlying properties...
This is related to the review of `FunFact` for JOSS (see https://github.com/openjournals/joss-reviews/issues/4502) I ran the unit tests (on the branch `tags/1.0` which I hope is the correct one) and this...
For a `historic_convergence` plugin, we would want to save the `loss`, `penalty` and combination of both for *every* vector instance separately. This is currently not possible externally as the `GradientDescentState`...
**Is your feature request related to a problem? Please describe.** Add a method that calls `factorize` internally, but saves the boilerplate code for converting between circuits and tensor expressions. **Describe...
**Is your feature request related to a problem? Please describe.** The FunFact tensor expression syntax lacks a way to express convolutions and tensor reshaping in a convenient manner. **Describe the...
Showcase using FF to implement compressed layers in deep neural networks. Outline: - [ ] Introduction on NN compression/tensorization. - [ ] Example problem description (MNIST) - [ ] NN...