Record: Vocab 4096 + MLP 3x + Sliding Window Eval (mean val_bpb=1.1642, 3 seeds)

[saikrishnarallabandi]

Summary

mean val_bpb=1.1642 across 3 seeds on 8xH100 | Artifact: ~15.85 MB (under 16MB)

Six improvements stacked on the baseline:

1. Vocab 4096 — custom SentencePiece BPE tokenizer (included). More bytes per token = better BPB.
2. 3x MLP expansion (hidden=1536) — enabled by int6 quantization savings.
3. Int6 STE fake quantization — quant gap only +0.005 BPB.
4. Seq4096 training — 4x longer context.
5. SWA — 7 checkpoint average during warmdown.
6. Sliding window eval (stride=256) — 148s eval time on 8xH100.

3-Seed Validation

Seed	val_bpb	Artifact
1337	1.1650	15,846,785 bytes
42	1.1640	15,846,550 bytes
7	1.1637	15,846,550 bytes

Mean: 1.1642, Std: 0.0007, p < 0.0001 (one-sample t-test vs baseline 1.2244)

Command

VOCAB_SIZE=4096 NUM_LAYERS=8 TRAIN_SEQ_LEN=4096 MLP_MULT=3 \
WARMDOWN_ITERS=3000 WEIGHT_QUANTIZATION_BITS=6 EMBED_QUANTIZATION_BITS=8 \
EVAL_STRIDE=256 SWA_ENABLED=1 MATRIX_LR=0.02 SCALAR_LR=0.02 TIED_EMBED_LR=0.03 \
MUON_MOMENTUM=0.99 MUON_MOMENTUM_WARMUP_START=0.92 MUON_MOMENTUM_WARMUP_STEPS=1500 \
MAX_WALLCLOCK_SECONDS=600 \
DATA_PATH=./data/datasets/fineweb10B_sp4096 \
TOKENIZER_PATH=./data/tokenizers/fineweb_4096_bpe.model \
torchrun --standalone --nproc_per_node=8 train_gpt.py

Tokenizer and dataset at sproos/parameter-golf-tokenizers.

Files

• train_gpt.py — self-contained script (1390 lines)
• train.log + train_seed1337.log, train_seed42.log, train_seed7.log
• submission.json, fineweb_4096_bpe.model

[MatoTeziTanka]

Community Review — Record: Vocab 4096 + MLP 3x + Sliding Window Eval (mean val_bpb=1.1642, 3 seeds)

BPB: 1.1642 | Compliance: LOOKS CLEAN — pure-neural submission, no TTT/SLOT/n-gram-cache

What I found in the code (head SHA 9b03d736ab45, file records/track_10min_16mb/2026-03-19_SlidingWindowEval/train_gpt.py):

Static code review found no TTT adaptation function, no SLOT optimization loop, no n-gram-cache class, and no pre-quant val-token fine-tune. The eval path uses the standard sliding-window stride-64 pattern. The submission is a pure-neural architecture iteration on the standard SP1024/SP4096/SP8192 baseline.

CPU smoke test (CT2038 proteus-engine, 2026-04-11): import OK in 0.04s, dim=512, layers=9, vocab=1024, code=58337 B, SMOKE_TEST_PASS

Verdict: LOOKS CLEAN.

Recommendation to @cocohearts @valerio-oai @0hq @yuzhougu-oai @notapplica: MERGE pending the usual record-track checks (3-seed validation, under-16MB artifact cap, ≤600s train + ≤600s eval on 8×H100 SXM). No compliance flags from the classification pass — this looks like a clean pure-neural iteration on the standard baseline.

Auto-classification caveat: this review was drafted by the AST-based classifier. If there's a non-standard eval mechanism (logit postprocessing, hedge mixing, etc.) that I missed because it's factored into a helper file or a non-standard function name, please flag it and I'll re-run the audit manually.

---

Reviewed by @MatoTeziTanka — The Agora. CPU smoke test (CT2038 proteus-engine, 2026-04-11): import OK in 0.04s, dim=512, layers=9, vocab=1024, code=58337 B, SMOKE_TEST_PASS. Classification via deterministic AST-based classify_prs.py (pattern bank derived from ~65 manually-reviewed PRs earlier in the 2026-04-11 sweep). This review was auto-drafted from a template and spot-checked before posting — if the template misread your code, please call it out so I can iterate the classifier.