Jason Ramapuram
Jason Ramapuram
``` bash (.venv)➜ examples git:(master) ✗ python demo_cifar.py --devtype=gpu Traceback (most recent call last): File "demo_cifar.py", line 90, in main() File "demo_cifar.py", line 60, in main train = cgt.function(inputs=[X, y],...
Pytorch docs don't seem to display latex equations: 
So I tried to get this setup in a venv instead of / because its just cleaner imho. After the dependency install (as well as fixing python3-gi as [such](http://stackoverflow.com/questions/26678457/how-do-i-install-python3-gi-within-virtualenv) )...
There are quite a few problems using `af::array` to handle CPU GPU transitions. A swap to ndarray would be cleaner for host data
Data loading is slow. To resolve this we need a separate thread(s) that prefetch data into a queue which is used then in the actual model to execute (get_batch_*) calls.
When adding layers, check the sizing and throw a clean error message.
Remove the `unwrap()` calls all over the code and change it to a `Result` type
So been trying to use `image` as an alternative to speedup image cropping in python. I tested this against python bindings provided by `PIL-SIMD` and `vips`. `image` + `rayon` provides...
example signature: ``` c af_err af_shuffle_in_place(af_array* out, const int seq_dim); ``` where `const int seq_dim` is the dim onto which shuffling takes place. Furthermore, it would make sense implementing something...
It would be awesome to have arrayfire built as a static library in addition to a shared one. It makes it a lot easier to build in ar files into...