minihack/agent/polybeast/models/base.py [236:274]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            flags.use_index_select,
        )

        K = flags.embedding_dim  # number of input filters
        F = 3  # filter dimensions
        S = 1  # stride
        P = 1  # padding
        M = 16  # number of intermediate filters
        self.Y = 8  # number of output filters
        L = flags.layers  # number of convnet layers

        in_channels = [K] + [M] * (L - 1)
        out_channels = [M] * (L - 1) + [self.Y]

        def interleave(xs, ys):
            return [val for pair in zip(xs, ys) for val in pair]

        conv_extract = [
            nn.Conv2d(
                in_channels=in_channels[i],
                out_channels=out_channels[i],
                kernel_size=(F, F),
                stride=S,
                padding=P,
            )
            for i in range(L)
        ]

        self.extract_representation = nn.Sequential(
            *interleave(conv_extract, [nn.ELU()] * len(conv_extract))
        )

        if self.crop_model == "transformer":
            self.extract_crop_representation = TransformerEncoder(
                K,
                N=L,
                heads=8,
                height=self.crop_dim,
                width=self.crop_dim,
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



minihack/agent/rllib/models.py [230:268]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            flags.use_index_select,
        )

        K = flags.embedding_dim  # number of input filters
        F = 3  # filter dimensions
        S = 1  # stride
        P = 1  # padding
        M = 16  # number of intermediate filters
        self.Y = 8  # number of output filters
        L = flags.layers  # number of convnet layers

        in_channels = [K] + [M] * (L - 1)
        out_channels = [M] * (L - 1) + [self.Y]

        def interleave(xs, ys):
            return [val for pair in zip(xs, ys) for val in pair]

        conv_extract = [
            nn.Conv2d(
                in_channels=in_channels[i],
                out_channels=out_channels[i],
                kernel_size=(F, F),
                stride=S,
                padding=P,
            )
            for i in range(L)
        ]

        self.extract_representation = nn.Sequential(
            *interleave(conv_extract, [nn.ELU()] * len(conv_extract))
        )

        if self.crop_model == "transformer":
            self.extract_crop_representation = TransformerEncoder(
                K,
                N=L,
                heads=8,
                height=self.crop_dim,
                width=self.crop_dim,
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -