Conversation
|
I'll fix CI failures and address comments from users with write access that start with 'Devin'.
|
devin-ai-integration
Bot
force-pushed
the
devin/1790748858-gpt-oss-lm-head
branch
from
September 30, 2026 20:25
e525736 to
31a8853
Compare
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
devin-ai-integration
Bot
force-pushed
the
devin/1790748858-gpt-oss-lm-head
branch
from
September 30, 2026 20:48
31a8853 to
3432108
Compare
kevintli
added this pull request to stack #13
September 30, 2026 22:15
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Adds
lm_headas a target module for thegpt-oss-20b-lora-64kpreset. This makes us consistent with the Tinker SDK defaults and the training recipe in https://github.com/jasper-lu/sec-search-rl.Note: it seems the core issue here is that Miles doesn't support a multi-LoRA setup where different LoRA clients have different settings on which adapters they want to train. This forces us to define presets on the Spindle server side that exactly match what users choose for their training run on the client side (so you can't turn
lm_headon/off on demand for example).Filed a follow-up issue for this here with a proposal from Devin, open to further discussion: #12