2 ms·
Hmm yeah I figure it’s better to just rent a monster machine in the cloud for that. But I haven’t dipped into that much yet
by coffeebeqn 4y ago
Hmm yeah I figure it’s better to just rent a monster machine in the cloud for that. But I haven’t dipped into that much yet
- Our_Benefactors 4y agoIt’s not because then you’re trading graphical fidelity for input lag.
- Nullabillity 4y agoInput lag doesn't seem relevant for playing around with AI models for 5 minutes until you get bored.
- coffeebeqn 4y agoYeah just the “compilation” (sorry not a ML engineer term) would need meaningful resources. You can run a python repl and editor on a raspberry pi