Daniel Galvez

icon company @NVIDIA icon location Bay Area, CA icon location I do speech recognition, machine learning, and high performance computing.

Results 42 issues of Daniel Galvez

Update the test checking for cooperative kernels in conditional nodes.

1 comment

Now we conditionally xfail only when a cuda driver version less than 12.6 is installed. CUDA 12.6 fixes this issue. Before it, cooperative kernels could not be used within the...

core
stale
ASR
Run CICD

Copy data from GPU to CPU only once in batched_hyps_to_hypotheses for each tensor of interest.

1 comment

Previously, several small GPU->CPU copies were used, which caused excess latency linearly proportional to the batch size. For small copies, it is much more efficient to do a single memory...

stale
ASR
Run CICD