Description
Please make sure that this is a feature request. As per our GitHub Policy, we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template
System information
- TensorFlow.js version (you are using): 4.22.0
- Are you willing to contribute it (Yes/No): Maybe if I get bored
Describe the feature and the current behavior/state.
The cumprod function was added in #6211, with the reviewer @pyu10055 saying
Thank you for contributing the cumprod op, the gradient function can come later, for reference the similar grad def for cumsum is in src/gradients/Cumsum_grad.ts
and the author @iislucas writing
I'll do gradients in a followup.
but the gradient was never implemented. This is a feature request for that gradient implementation.
Will this change the current api? How?
No.
Who will benefit with this feature?
Anyone using cumprod who wants gradients will benefit, but more specifically, I was hoping to use the cumprod function to use TensorFlow.js to implement a benchmark in gradbench/gradbench#561. But I don't mind implementing it in a worse way if TensorFlow.js doesn't support the better way, since the goal of this benchmarking project is merely to assess different autodiff tools as they actually exist.
Any Other info.