Comment Comment by htrp how do you justify the compute investment for something like nemotron ? especially if all the labs are willing to pay for those same GPU clusters for inference or training runs? parent htrp 6 months ago
