LAWA-EMA frontier fork (pr198 base, SWA -> LAWA val_bpb=1.1551)#201
Open
machdragon wants to merge 1 commit intoopenai:mainfrom
Open
LAWA-EMA frontier fork (pr198 base, SWA -> LAWA val_bpb=1.1551)#201machdragon wants to merge 1 commit intoopenai:mainfrom
machdragon wants to merge 1 commit intoopenai:mainfrom
Conversation
11-Layer Int6 + LAWA-EMA (decay=0.995) + Overtone Init, based on PR openai#198. Replaces SWA with every-step EMA averaging. Fixes bigram proj zero-init override and sliding window partial-window overlap. 12.7 MB artifact. 8xH100 SXM, 600s, seed=1337, 6715 steps. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
machdragon
force-pushed
the
submission/lawa-frontier-int6-mlp3x
branch
from
3693d95 to
31c16a6
Compare
machdragon
changed the title
machdragon
changed the title
machdragon
changed the title
machdragon
changed the title
machdragon
added a commit
to machdragon/parameter-golf
that referenced
this pull request
Mar 20, 2026
Built on PR openai#201 (LAWA-EMA + Int6 + Overtone + MLP3x, val_bpb=1.1551). Adds four improvements targeting quantization fidelity and eval-time adaptation: - KURE kurtosis regularization + R2 outlier penalty for int6-friendly weights - Tanh weight reparameterization bounding effective weights to [-1,1] - Parallel EMA tracks (0.995/0.999/0.9995) with proxy-eval selection - Causal LoRA TTT (rank 8) ported from PR openai#77 for eval-time adaptation Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
15 tasks
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
SXM, 600s
U-Net skips, WD=0.04)
shadow, every-step update)
quantization)
Single seed run (seed=1337). Additional seed runs pending for statistical validation.
Test plan
train_gpt.pyon 8xH100 with official harnessOur numbers:
┌────────────────────────────┬────────┐
│ Metric │ Value │
├────────────────────────────┼────────┤
│ Pre-quant val_bpb │ 1.1622 │
├────────────────────────────┼────────┤
│ Int6 roundtrip val_bpb │ 1.1779 │
├────────────────────────────┼────────┤
│ Int6 sliding val_bpb (s64) │ 1.1551 │
└────────────────────────────┴────────┘
Note: our 1.1551 is worse than PR #198's 1.1318. The artifact is 3 MB smaller (12.7 vs
15.7 MB), but the BPB regressed.