ttnn.leaky_relu

ttnn.leaky_relu(input_tensor: ttnn.Tensor, negative_slope: float, *, memory_config: ttnn.MemoryConfig = None, output_tensor: ttnn.Tensor = None, sub_core_grids: ttnn.CoreRangeSet = None) → ttnn.Tensor

Applies leaky_relu to input_tensor element-wise with negative_slope.

Passes non-negative inputs through unchanged and scales negative inputs by negative_slope.

\[\mathrm{output\_tensor}_i = \max(0, \mathrm{input\_tensor}_i) + \verb|negative_slope| \cdot \min(0, \mathrm{input\_tensor}_i)\]

Parameters:

input_tensor (ttnn.Tensor) – the input tensor.
negative_slope (float) – The slope parameter for the Leaky ReLU function.

Keyword Arguments:

memory_config (ttnn.MemoryConfig, optional) – Memory configuration for the operation. Defaults to None.
output_tensor (ttnn.Tensor, optional) – preallocated output tensor. Defaults to None.
sub_core_grids (ttnn.CoreRangeSet, optional) – sub core grids for the operation. Defaults to None.

Returns:

ttnn.Tensor – the output tensor.

Note

Supported dtypes and layouts:

Dtypes	Layouts
FLOAT32, BFLOAT16, BFLOAT8_B, UINT32, UINT16, UINT8	TILE, ROW_MAJOR

Example

# Create a tensor with specific values
tensor = ttnn.from_torch(
    torch.tensor([[1, 2], [3, 4]], dtype=torch.bfloat16),
    dtype=ttnn.bfloat16,
    layout=ttnn.TILE_LAYOUT,
    device=device,
)
negative_slope = 3

# Apply Leaky ReLU activation function
output = ttnn.leaky_relu(tensor, negative_slope)
logger.info(f"Leaky ReLU: {output}")