3 ms·
> We train our proposed system on 8 NVIDIA V100 GPUs with 32G memory Sounds like openly reproducing this result is within independent researchers’ reach.
by coolspot 4y ago
> We train our proposed system on 8 NVIDIA V100 GPUs with 32G memory
Sounds like openly reproducing this result is within independent researchers’ reach.