Skip to content

geo3k VLM multi-turn megatron update #1378

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

Jump to bottom
Merged
zhaochenyang20 merged 4 commits into from
Jan 11, 2026
Merged

geo3k VLM multi-turn megatron update #1378

zhaochenyang20 merged 4 commits into from
Jan 11, 2026
+10 −8

Conversation

@gxlvera
Copy link
Contributor

@gxlvera gxlvera commented Jan 11, 2026

In this PR is related to PR #1141 , three changes are made

  1. update the experiment result using megatron as backend in README.md
  2. change obs_log_probs to be 0.0 instead of '-inf' to prevent NaN value for logging
  3. change dummy_messages to a static variable

zhaochenyang20
zhaochenyang20 approved these changes Jan 11, 2026
View reviewed changes
Copy link
Collaborator

@zhaochenyang20 zhaochenyang20 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

NB

@zhaochenyang20 zhaochenyang20 merged commit f68dc4a into THUDM:main Jan 11, 2026
@zhaochenyang20
Copy link
Collaborator

zhaochenyang20 commented Jan 11, 2026

/gemini review

Beichen-Ma pushed a commit to Beichen-Ma/slime that referenced this pull request Jan 12, 2026
@gxlvera @Beichen-Ma
geo3k VLM multi-turn megatron update (THUDM#1378)
d83a0dc
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@zhaochenyang20 zhaochenyang20 zhaochenyang20 approved these changes

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@gxlvera @zhaochenyang20