Andres Nowak
Andres Nowak
### Bug description When running a function as an alias and inside that function we modify a struct that modifies a dynamic vector it owns. And then we try to...
### Bug description When using the load store tensor it see we don't get the correct values. ### Steps to reproduce ```py from tensor import Tensor, TensorShape from random import...
### Bug description This problem doesn't happen if you just create a simple compile time vector and then try to access it's values it happens in a more complex example...
This branch is the implementation of an idea that @mikowals wanted to implement, changing the vectorize function for tile in the Batch_matmul function. Based on benchmarking with lamatune, changing vectorize...
I implemented an idea for matmul tiling, but i didn't see any performance improvement i get the same results as the original matmul_parallel implementation, so maybe it seems i don't...
There seems to be a memory leak in the program, at the end of the test_infermo, the program uses sometimes 6gb of memory.
This is a pr that is making the upsample ops native instead of using the index op and the operation for the mode. - [x] nearest - [x] forward -...
## Update to 24.5 **A lot of errors about warning: call argument allows writing a memory location previously writable through another aliased argument**