prior-fitting/decoders.py

import torch
from torch import nn
import random


class ScaledDecoder(nn.Module):
    def __init__(self, ninp, nhid, nout):
        super().__init__()
        self.linear = nn.Linear(ninp, nhid)
        self.linear1 = nn.Linear(nhid, nout)
        self.linear2 = nn.Linear(nhid, 10)

    def forward(self, x):
        #return torch.cat([self.linear1(x), self.linear2(x)], -1)
        x = self.linear(x)
        x = nn.GELU()(x)
        temps = self.linear2(x).softmax(-1) @ torch.tensor([1.,1.4,1.7,2.,5.,10.,20.,40.,80.,160.], device=x.device)
        if random.random() > .99:
            print(temps.shape,temps[:,:2])
        return self.linear1(x) / temps.unsqueeze(-1)

class FixedScaledDecoder(nn.Module):
    def __init__(self, ninp, nhid, nout):
        super().__init__()
        self.mapper = nn.Sequential(nn.Linear(ninp, nhid), nn.GELU(), nn.Linear(nhid, nout))
        self.T = nn.Parameter(torch.ones(10000)/10000)

    def forward(self, x):
        return self.mapper(x)/self.T.sum()