Implement a single reverse-denoising step: predict all masked tokens, keep the most confid

Medium

Diffusion & Non-Autoregressive Language Models2 / 4

Chapter 222 / 4

Diffusion & Non-Autoregressive Language Models

Medium

Implement a single reverse-denoising step: predict all masked tokens, keep the most confident, re-mask the rest.

Target interface

reverse_step(model, x_t, mask_id, k)

Complete the starter code in the editor. Any correct approach is accepted.

Hints0 / 2

Reference solution

Your own code stays in the editor.

solution.pyPython

Local draft

import torch
import torch.nn.functional as F

@torch.no_grad()
def reverse_step(model, x_t, mask_id, k):
    raise NotImplementedError

Local draft

⌘/Ctrl↵to submit

AI review

Ready when you are

Submit your solution for a score and feedback on correctness, edge cases, and clarity. Any correct approach is accepted.

Chapter 222 / 4

Diffusion & Non-Autoregressive Language Models

Medium

Implement a single reverse-denoising step: predict all masked tokens, keep the most confident, re-mask the rest.

Target interface

reverse_step(model, x_t, mask_id, k)

Complete the starter code in the editor. Any correct approach is accepted.

Hints0 / 2

Reference solution

Your own code stays in the editor.