Add LIBERO-10 action-policy finetune cookbook#242
Merged
lfengad merged 9 commits intoJul 2, 2026
Conversation
Mirror the DROID action-policy cookbook for LIBERO-10: launch_sft_action_policy_libero.sh + action_policy_libero_repro.toml + finetune README section + action README link. Stages nvidia/LIBERO_LeRobot_v3 libero_10, no keep-ranges filter, lr 5e-5/wu500/cyc16k. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…er results); recipe HSDP 2x8
…only Two presets to match the Cosmos3 LIBERO-10 result: (A) libero_10-only (launch_sft_action_policy_libero.sh) and (B) libero-all equal 4-suite mix (launch_sft_action_policy_libero_all.sh + action_policy_libero_all_repro.toml). README documents both. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
mli0603
approved these changes
Jul 1, 2026
pengcuo
approved these changes
Jul 2, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Adds a LIBERO action-policy SFT cookbook, mirroring the existing DROID action-policy finetune cookbook.
To match the LIBERO-10 results reported in Cosmos3, it ships two presets (both lr 5e-5, warmup 500, cycle 16000, global batch 2048 / HSDP 2x8):
launch_sft_action_policy_libero.sh+toml/sft_config/action_policy_libero_repro.toml(max_iter 2000); stagesnvidia/LIBERO_LeRobot_v3libero_10.launch_sft_action_policy_libero_all.sh+toml/sft_config/action_policy_libero_all_repro.toml(max_iter 5000; needs longer to converge); stages all 4 suites.Each launcher is a local wrapper: stages the dataset, downloads the Wan VAE, converts the Cosmos3-Nano base to DCP, and launches training. Extends
finetune/README.md(recipe table + LIBERO quick start for both presets) and links it from the actionREADME.md.Delegates the canonical training implementation to cosmos-framework (companion PR #61 adds the
action_policy_libero_nano/action_policy_libero_all_nanoexperiments + eval harness).Mirrors the DROID cookbook structure. Based on
main.🤖 Generated with Claude Code