feat: add openclaw-training-free — zero-GPU experience loop for LLMs#6
Open
zcyynl wants to merge 1 commit into
Open
feat: add openclaw-training-free — zero-GPU experience loop for LLMs#6zcyynl wants to merge 1 commit into
zcyynl wants to merge 1 commit into
Conversation
openclaw-training-free improves LLM responses over time without any model parameter updates, forming a self-improving experience loop: intercept → extract hint → store → inject → repeat Every user follow-up is an implicit training signal. A Judge LLM makes it explicit; the experience store makes it persistent. Unlike openclaw-rl and openclaw-opd which require a self-hosted model with GPU, this module works as a drop-in proxy for any OpenAI-compatible endpoint (cloud APIs included) — zero GPU, zero training cost. Cost comparison: - openclaw-rl (GRPO): ~$10,000/run — parameter updates + GPU - openclaw-opd (OPD): ~$18/run — parameter updates - openclaw-training-free: ~$0 — no parameter updates Especially suited for open-ended dialogue where reward functions are hard to define and RL training is unstable.
zcyynl
force-pushed
the
feat/openclaw-training-free
branch
from
March 5, 2026 12:55
cc3f76e to
5d42c00
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
openclaw-training-free improves LLM responses over time without any model parameter updates, forming a self-improving experience loop:
intercept → extract hint → store → inject → repeat
Every user follow-up is an implicit training signal. A Judge LLM makes it explicit; the experience store makes it persistent.
Components: