Thomas Dybdahl Ahle
Thomas Dybdahl Ahle
I'm running the `jit.py` file using `torch==2.0.1` and cuda `11.7`. It seems that tensor.type has been deprecated, so I'm getting a lot of warnings like this: ``` /lltm_cuda.cpp:28:40: warning: ‘at::DeprecatedTypeProperties&...
Hopefully the new LM backend will allow us to make better use of models that are trained for "Chat". Below is a good example of how even good models like...
I want to reduce Hadamard Transform every column in my matrix. Is there currently a way to do this with FFHT without using a python outer loop?
I installed FFHT with Python 3.9 and got this error when trying to run `example.py`: $ py example.py Traceback (most recent call last): File "/Users/ahle/Dropbox (Meta)/kode/ts/FFHT/example.py", line 14, in ffht.fht(a)...
Maybe the "1 neighbor for endpoints, 2 for midpoints" rule is too harsh. Janko.at has an interesting version of Numberlink called Arukone^3, which provides the rule "A single line must...
This follows https://github.com/HazyResearch/structured-nets/issues/7 in which we discussed updating the code to use PyTorch's complex tensors. The pull request also extracts all tests into a `tests/` directory, so it's easy to...
It appears the images in the Readme, as well as the poster/slide links, are broken.
I would love to see the slides and poster. Any way to get the links back up?
It seems there's an easy way to compute `G∩H = G.subgroup_search(H.contains)` on two PermutationGroups. Would it make sense to implement `__and__` with this function. It should satisfy `(G & H).elements...
## Description of proposed feature It would be useful if using Transform on VDicts matched items according to their keys. Basically this: ```python def transform_vdicts(scene, vdict1, vdict2, run_time=1): k1 =...