# Records

<!-- AUTO-GENERATED by scripts/update_leaderboard.py from records/records.json — do not edit by hand -->

Wall-clock record progression per track. Every verified record links its
submission folder and its public verification report.

## Track 1 — GSM8K · Qwen2.5-1.5B · target ≥ 57.0% · 1× L40S

**Current record: 1m 44s** by [@stared](https://github.com/stared) — Shortest-4k data pruning, 1 aggressive-LR epoch, <<...>> annotations stripped, custom GPU-resident packed loop, chunked completion-only CE (no full logits).

| # | Date | Author | Train time | GSM8K/EM | Δ | Technique |
|---|------|--------|-----------|----------|---|-----------|
| 0 | 2026-07-18 | [@Saivineeth147](https://github.com/Saivineeth147) | 11m 57s | 59.4% | — | Baseline: plain LoRA r=16 on all linear layers, 3 epochs, cosine LR. No tricks. ([report](records/verifications/000-baseline.md)) |
| 1 | 2026-07-18 | [@Saivineeth147](https://github.com/Saivineeth147) | 6m 05s | 61.1% | −49% | Sequence packing + completion-only loss masking, 2 epochs. Same LoRA config as #0; ~2x faster at higher accuracy. ([report](records/verifications/001-saivineeth147.md)) |
| 2 | 2026-07-20 | [@stared](https://github.com/stared) | 1m 53s | 59.6% | −69% | One 4e-4 epoch over 3,000 examples with 2x optimizer updates (integrated-LR), 512-token packs. First outside record. ([report](records/verifications/002-pmigdal.md)) |
| 3 | 2026-07-20 | [@stared](https://github.com/stared) | 1m 44s | 60.3% | −8% | Shortest-4k data pruning, 1 aggressive-LR epoch, <<...>> annotations stripped, custom GPU-resident packed loop, chunked completion-only CE (no full logits). ([report](records/verifications/003-pmigdal.md)) |

## Track 2 — SQuAD v1.1 · SmolLM2-1.7B · target ≥ 75.5% · 1× L40S

**Current record: 11m 08s** by [@Saivineeth147](https://github.com/Saivineeth147) — Track 2 baseline: plain LoRA r=16 on SQuAD, first 20k examples, 1 epoch, full-sequence loss. No tricks.

| # | Date | Author | Train time | GSM8K/EM | Δ | Technique |
|---|------|--------|-----------|----------|---|-----------|
| 0 | 2026-07-20 | [@Saivineeth147](https://github.com/Saivineeth147) | 11m 08s | 77.5% | — | Track 2 baseline: plain LoRA r=16 on SQuAD, first 20k examples, 1 epoch, full-sequence loss. No tricks. ([report](records/verifications/t2-000-baseline.md)) |

## Notable attempts

_None yet. A notable attempt is a merged submission that didn't take the record but taught us something — see CONTRIBUTING.md._

## How records are verified

Every entry above was re-run 3× with fresh seeds on spec hardware by a verifier
before merge — see [JUDGING.md](../JUDGING.md). Full reports live in
[verifications/](./verifications/).