Jaivardhan Kapoor
Jaivardhan Kapoor
I might be a little late, but the simplest way seems to use the dot notation here to pass a flattened dict from wandb and then unflatten it in the...
I tested the compiled binary and I could not see any oiutouts from the transformer, even though the GPU showed that the file was loaded. I ran ```nvcc run.cu -o...
Following up on this question -- I don't understand the adding of noise either.
Hey @pcicales, could you please share any insights you got from your experiment? I've been thinking about this too, and would really appreciate any pointers!