RLHF & Preference Optimization5 / 7
RLHF & Preference Optimization
Implement Dr. GRPO (GRPO with the length and std-normalization terms removed) and show, on a toy batch, that it changes the relative weighting vs GRPO. [DeepSeek]
grpo_adv(rewards)Implement the function/class skeleton in the editor. Any correct approach is accepted.
import torch
def grpo_adv(rewards):
raise NotImplementedError
def dr_grpo_adv(rewards):
raise NotImplementedError
def grpo_objective(per_tok_pg, mask):
raise NotImplementedError
def dr_grpo_objective(per_tok_pg, mask):
raise NotImplementedError⌘/Ctrl↵to submit
No review yet
Submit your solution for a verdict, a score out of 100 and concrete feedback. Any correct approach passes.
RLHF & Preference Optimization
Implement Dr. GRPO (GRPO with the length and std-normalization terms removed) and show, on a toy batch, that it changes the relative weighting vs GRPO. [DeepSeek]
grpo_adv(rewards)Implement the function/class skeleton in the editor. Any correct approach is accepted.