Differences between torch.nn.TransformerDecoderLayer and mindspore.nn.TransformerDecoderLayer

torch.nn.TransformerDecoderLayer

class torch.nn.TransformerDecoderLayer(
    d_model,
    nhead,
    dim_feedforward=2048,
    dropout=0.1,
    activation='relu'
)(tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None)

For more information, see torch.nn.TransformerDecoderLayer.

mindspore.nn.TransformerDecoderLayer

class mindspore.nn.TransformerDecoderLayer(
    d_model,
    nhead,
    dim_feedforward=2048,
    dropout=0.1,
    activation='relu',
    layer_norm_eps=1e-5,
    batch_first=False,
    norm_first=False,
    dtype=mstype.float32
)(tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_padding_mask=None, memory_key_padding_mask=None)

For more information, see mindspore.nn.TransformerDecoderLayer.

Differences

The code implementation and parameter update logic of mindspore.nn.TransformerDecoderLayer optimizer is mostly the same with torch.nn.TransformerDecoderLayer.

Categories	Subcategories	PyTorch	MindSpore	Difference
Parameters	Parameter 1	d_model	d_model	Consistent function
	Parameter 2	nhead	nhead	Consistent function
	Parameter 3	dim_feedforward	dim_feedforward	Consistent function
	Parameter 4	dropout	dropout	Consistent function
	Parameter 5	activation	activation	Consistent function
	Parameter 6		layer_norm_eps	In MindSpore, the value of eps can be set in LayerNorm, PyTorch does not have this function
	Parameter 7		batch_first	In MindSpore, first batch can be set as batch dimension, PyTorch does not have this function
	Parameter 8		norm_first	In MindSpore, LayerNorm can be set in between Multiheadttention Layer and FeedForward Layer or after, PyTorch does not have this function
	Parameter 9		dtype	In MindSpore, dtype can be set in Parameters using ‘dtype’. PyTorch does not have this function.
Input	Input 1	tgt	tgt	Consistent function
	Input 2	memory	memory	Consistent function
	Input 3	tgt_mask	tgt_mask	In MindSpore, dtype can be set as float or bool Tensor; in PyTorch dtype can be set as float, byte or bool Tensor.
	Input 4	memory_mask	memory_mask	In MindSpore, dtype can be set as float or bool Tensor; in PyTorch dtype can be set as float, byte or bool Tensor.
	Input 5	tgt_key_padding_mask	tgt_key_padding_mask	In MindSpore, dtype can be set as float or bool Tensor; in PyTorch dtype can be set as byte or bool Tensor.
	Input 6	memory_key_padding_mask	memory_key_padding_mask	In MindSpore, dtype can be set as float or bool Tensor; in PyTorch dtype can be set as byte or bool Tensor.

Code Example

# PyTorch
import torch
from torch import nn

decoder_layer = nn.TransformerDecoderLayer(d_model=512, nhead=8)
transformer_decoder = nn.TransformerDecoder(decoder_layer, num_layers=6)
memory = torch.rand(10, 32, 512)
tgt = torch.rand(20, 32, 512)
out = transformer_decoder(tgt, memory)
print(out.shape)
#torch.Size([20, 32, 512])

# MindSpore
import mindspore as ms
import numpy as np

decoder_layer = ms.nn.TransformerDecoderLayer(d_model=512, nhead=8)
transformer_decoder = ms.nn.TransformerDecoder(decoder_layer, num_layers=6)
memory = ms.Tensor(np.random.rand(10, 32, 512), ms.float32)
tgt = ms.Tensor(np.random.rand(20, 32, 512), ms.float32)
out = transformer_decoder(tgt, memory)
print(out.shape)
#(20, 32, 512)