|
|
|
|
|
|
by elliotto
783 days ago
|
|
|
Hi, I work at a startup where we train / fine tune / inference models on a gcp kubernetes cluster on some a100s. There isn't really that much information about how to do this properly because everyone is working it out and it changes month by month. It requires a bunch of DevOps an infrastructure knowledge above and beyond the raw ml knowledge. Your best bet is probably just to tool around and see what you can do. |
|
|