akshaan
@akshaan
Tempted to yeet $1890 into vast.ai and train a text-to-image diffusion model from scratch https://arxiv.org/abs/2407.15811
2 replies
1 recast
13 reactions
christopher
@christopher
How does Modal compare?
1 reply
0 recast
0 reaction
akshaan
@akshaan
Modal offers much higher level infra. Not sure they support training style workloads, though I haven't tried. Vast just rents out raw GPU boxes in the cloud.
0 reply
0 recast
1 reaction
