Repository navigation
Pin llama.cpp b8d587e: load Hadamard-folded weights (Ternary Bonsai) - #247
Closed
bong-water-water-bong wants to merge 1 commit into
Closed
bong-water-water-bong wants to merge 1 commit into
bong-water-water-bong wants to merge 1 commit into
Conversation
Fork #34: the loader reads prism.hadamard.* keys and rotates activations before folded weights, so Ternary-Bonsai-2-27B computes correct output on HRX0. Registry regenerated for the new pin. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Docs7 for 1bit-monster/engine
Commit |
Collaborator
Author
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Moves
third_party/llama.cppfrom acf9c74 to b8d587e (fork #34):prism.hadamard.*keys and rotates activations before folded weights, with the inverse ontoken_embd.src/llama-hadamard.{h,cpp}, which carries the MIT notice of the PrismML code it follows; upstream files get hook lines only.Registry regenerated for the new pin (counts unchanged).
Checked on strixhalo: b8d587e builds with GGML_HRX (amdclang, Ninja). The HRX0 perplexity check (ZAYA1-8B and Qwen3-0.6B, new pin vs acf9c74) is queued behind the shared GPU lock; results will be posted here before merge.
🤖 Generated with Claude Code