# MIT License

# Copyright (c) Microsoft

# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:

# The above copyright notice and this permission notice shall be included in all
# copies or substantial portions of the Software.

# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.

# Copyright (c) [2025] [Microsoft]
# SPDX-License-Identifier: MIT
import torch
import torch.nn as nn


class LayerNorm32(nn.LayerNorm):
    def forward(self, x: torch.Tensor) -> torch.Tensor:
        return super().forward(x.float()).type(x.dtype)
    

class GroupNorm32(nn.GroupNorm):
    """
    A GroupNorm layer that converts to float32 before the forward pass.
    """
    def forward(self, x: torch.Tensor) -> torch.Tensor:
        return super().forward(x.float()).type(x.dtype)
    
    
class ChannelLayerNorm32(LayerNorm32):
    def forward(self, x: torch.Tensor) -> torch.Tensor:
        DIM = x.dim()
        x = x.permute(0, *range(2, DIM), 1).contiguous()
        x = super().forward(x)
        x = x.permute(0, DIM-1, *range(1, DIM-1)).contiguous()
        return x