helion.language.cumprod

helion.language.cumprod(input_tensor, dim, reverse=False)

Compute the cumulative product along a specified dimension.

Equivalent to hl.associative_scan(torch.mul, input_tensor, dim, reverse).

Parameters:

• input_tensor (Tensor) – Input tensor to compute cumulative product

• dim (int) – The dimension along which to compute cumulative product

• reverse (bool) – If True, performs the cumprod in reverse order

Returns:

Tensor with same shape as input containing cumulative product

Return type:

torch.Tensor

See also

• associative_scan(): For custom scan operations

• cumsum(): For cumulative sum

• reduce(): For dimension-reducing operations

Note

• Output has same shape as input

• Reverse=True computes cumprod from right to left

• Equivalent to torch.cumprod