PyTorch torch.dsplit() method is **“used to split the input, a tensor with three or more dimensions, into multiple tensors depthwise according to indices_or_sections.”** Each split is a view of input.

**Syntax**

```
torch.dsplit(input, indices_or_sections)
```

**Parameters**

**input (Tensor)**: The input tensor to be split. It should have at least 3 dimensions.

**split_size_or_sections****(int or list)**: Size of each chunk or section to divide. If it’s an integer, the input tensor will be divided into equally sized chunks (if possible). If it’s a list, it specifies the number of chunks of each size.

**Example 1: Splitting a 3D tensor into equally sized chunks**

```
import torch
tensor_1 = torch.rand((2, 2, 6)) # A 3D tensor
chunks_1 = torch.dsplit(tensor_1, 3)
print("Original Tensor Shape:")
print(tensor_1.shape)
print("\nShapes of Chunks:")
for c in chunks_1:
print(c.shape)
```

**Output**

**Original Tensor Shape**:
torch.Size([2, 2, 6])
**Shapes of Chunks**:
torch.Size([2, 2, 2])
torch.Size([2, 2, 2])
torch.Size([2, 2, 2])

**Example 2: Splitting using a list to define chunk sizes**

```
import torch
tensor_2 = torch.rand((2, 2, 8))
chunks_2 = torch.dsplit(tensor_2, [3, 5])
print("Original Tensor Shape:")
print(tensor_2.shape)
print("\nShapes of Chunks:")
for c in chunks_2:
print(c.shape)
```

**Output**

**Original Tensor Shape**:
torch.Size([2, 2, 8])
**Shapes of Chunks**:
torch.Size([2, 2, 3])
torch.Size([2, 2, 2])
torch.Size([2, 2, 3])

**Example 3: Working with a larger 3D tensor**

```
import torch
tensor_3 = torch.rand((4, 4, 10))
chunks_3 = torch.dsplit(tensor_3, 5)
print("Original Tensor Shape:")
print(tensor_3.shape)
print("\nShapes of Chunks:")
for c in chunks_3:
print(c.shape)
```

**Output**

**Original Tensor Shape:**
torch.Size([4, 4, 10])
**Shapes of Chunks:**
torch.Size([4, 4, 2])
torch.Size([4, 4, 2])
torch.Size([4, 4, 2])
torch.Size([4, 4, 2])
torch.Size([4, 4, 2])

That’s it!

**Related posts**

Krunal Lathiya is a seasoned Computer Science expert with over eight years in the tech industry. He boasts deep knowledge in Data Science and Machine Learning. Versed in Python, JavaScript, PHP, R, and Golang. Skilled in frameworks like Angular and React and platforms such as Node.js. His expertise spans both front-end and back-end development. His proficiency in the Machine Learning frameworks like PyTorch and Tensorflow is a testament to his versatility and commitment to the craft.