Skip to content

[CI] assert train_rollout_logprob_abs_diff <= 0.1 (regression guard) #94

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
CalvinXKY merged 2 commits into from
Jun 2, 2026
Merged

[CI] assert train_rollout_logprob_abs_diff <= 0.1 (regression guard) #94

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
765122d
[CI] assert train_rollout_logprob_abs_diff <= 0.1 (regression guard)
aoshen02 May 31, 2026
96e0e22
Merge remote-tracking branch 'origin/main' into sync-slime-logprob-ab…
aoshen02 Jun 2, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions vime/backends/megatron_utils/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -769,6 +769,9 @@ def train(
log_dict["train/step"] = accumulated_step_id
logging_utils.log(args, log_dict, step_key="train/step")

if args.ci_test and "train/train_rollout_logprob_abs_diff" in log_dict:
assert log_dict["train/train_rollout_logprob_abs_diff"] <= 0.1, f"{log_dict=}"

if args.ci_test and not args.ci_disable_kl_checker:
if step_id == 0 and "train/ppo_kl" in log_dict and "train/pg_clipfrac" in log_dict:
# TODO: figure out why KL is not exactly zero when using PPO loss with KL clipping, and whether this is expected behavior or a bug.
Expand Down
Loading