Repository navigation
Conversation
Signed-off-by: 0z5a <dezhen.lu@student.uni-tuebingen.de>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A prefix graph could be accepted after restoring different weights, or paired with a different actor/reference readout, because parameter versions were first captured when replay opened. Bind the producing actor, readout and parameter versions when the program is created and enforce that binding through backward.
Incremental draft on #25. Applies to Ouro/Nanbeige/Huginn prefix components; the learner schedule remains Ouro-only. No objective, sampling or publication changes.
Validation: 12 regressions fail on the parent because invalid bindings are accepted. The final CPU run has 141 passes and 12 existing strict BF16 numerical expected failures, with unchanged numerical gates. Raw before/after logs, XML and source hashes are committed; actual CUDA/official-model qualification, speedup and convergence remain unmeasured. See docs/prefix_actor_binding.md for the result table and scope.