-
буду хранить тут дамп статей про трансформеры, которые читаю, либо которые хочу прочитать
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale - статья где предложили ViT, иде…
-
In the tests we only check that the fourier transform is unitary,, not that the adjoint is correct. In fact, it is of by rather large factors in many cases. See e.g. this example:
```python
>>> od…
-
## GPU 优势
GPU达到CPU最高帧率时的功率消耗只有CPU的一半。这段话来自[An Independent Evaluation of
Implementing Computer Vision Functions with OpenCL on the Qualcomm Adreno 420 | Berkeley Design Technology, Inc. July 2015](…
-
I don't see a reason not to implement an FFFT that works for any number of modes (not just powers of 2). Has anyone considered this? I think the T-count would also be much better than using Givens rot…
-
Thanks for this great crate!
It would be great to have a faster way for doing a real to complex fft (such as from time to frequency domain), which is a very common use case.
It would approximately h…
-
Hi, very excited to see the work.
I wonder if there are any results about the disentanglement results?
I tried and revised the repo [https://github.com/johndpope/MegaPortrait-hack](url)
but …
-
Subscribe to this issue and stay notified about new [daily trending repos in Jupyter Notebook](https://github.com/trending/jupyter-notebook?since=daily).
-
-
Anyone has interest to utilize the sparsity to accelerate DNNs?
I am working on the fork https://github.com/wenwei202/caffe/tree/scnn and currently, on average, achieve ~5x CPU and ~3x GPU layer-wi…
-
Context: https://twitter.com/rickbrewPDN/status/1562499054952534018
The kernel parameters listed at https://github.com/mikepound/convolve/blob/master/complex_kernels.py are a set of 6 hardcoded 4-v…