[posttrain] Add Nemotron agentic SFT datasets#6270
Open
taivu1998 wants to merge 2 commits into
Open
Conversation
Register the SWE, Cascade, OpenCode, and CUDA Nemotron SFT datasets through the instruction dataset path. Add a rich multi-turn adapter so tool calls, reasoning content, and tool context survive transformation for agentic SFT rows.
Fix the all-files trailing whitespace failure reported by the PR lint job so the branch can pass repository-wide pre-commit checks.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Add a rich multi-turn conversation adapter for agentic SFT rows so tool calls, tool observations, reasoning content, and top-level tools are preserved instead of being flattened to role/content only.
Register nvidia/Nemotron-SFT-SWE-v3, nvidia/Nemotron-Cascade-SFT-SWE, nvidia/Nemotron-SFT-CUDA-v1, and split-specific nvidia/Nemotron-SFT-OpenCode-v1 views with pinned revisions. CUDA, SWE-v3, and OpenCode use the rich adapter; Cascade remains on the plain multi-turn path because its rows are ordinary messages plus task metadata.