3 ms·
How big are the matrixes in some modern training pipelines? We always talk of absurdly large parameter spaces.
by taeric 3y ago
How big are the matrixes in some modern training pipelines? We always talk of absurdly large parameter spaces.
- ogogmad 3y agoSGD keeps the matrices small, I think.