Build a toy post-training stack with intentional bugs, then write tests that catch every o

Super-hard

Correctness, Debugging, and Performance Experiments7 / 8

Chapter 267 / 8

Correctness, Debugging, and Performance Experiments

Build a toy post-training stack with intentional bugs, then write tests that catch every o

Super-hard

Build a toy post-training stack with intentional bugs, then write tests that catch every one.

Target interface

seq_logprob(model, ids, prompt_len, mask_bug=False)

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 os, torch, torch.nn.functional as F

def seq_logprob(model, ids, prompt_len, mask_bug=False):
    raise NotImplementedError

def dpo_loss(policy, ref, chosen, rejected, prompt_len, beta=0.1):
    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 267 / 8

Correctness, Debugging, and Performance Experiments

Build a toy post-training stack with intentional bugs, then write tests that catch every o

Super-hard

Build a toy post-training stack with intentional bugs, then write tests that catch every one.

Target interface

seq_logprob(model, ids, prompt_len, mask_bug=False)

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

Hints0 / 2

Reference solution

Your own code stays in the editor.