File size: 2,668 Bytes
2f8be4b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f21226d
2f8be4b
f21226d
2f8be4b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
---
license: odc-by
tags:
- reproducibility
- verification
- synthetic-persona-pretraining
pretty_name: SPP Corpus Verification
---

# SPP Corpus Verification

Checksums and document-boundary indices for verifying a rebuilt copy of the
**Synthetic Persona Pretraining (SPP)** training corpus, byte for byte.

The Megatron token streams themselves are **2.17 TB** (`annotated.bin` 421 GB,
`compact.bin` 1.75 TB) and are fully derived from the published reflections, the
uid manifest, and the tokenizer recipe — so they are not published. These
`.idx` sidecars carry per-document boundaries and lengths, which is enough to
prove an independently rebuilt `.bin` matches ours at ~200× less data.

> 🔁 **Manifest (what to rebuild from):** [`dlab-spp/corpus-1T-manifest`](https://huggingface.co/datasets/dlab-spp/corpus-1T-manifest)
>
> 📄 **Reflections:** [`dlab-spp/reflection-50m`](https://huggingface.co/datasets/dlab-spp/reflection-50m)

## Files

| file | size | what |
| --- | --- | --- |
| `annotated.idx` | 2.06 GB | document boundaries for the annotated stream (102,772,028 docs) |
| `compact.idx` | 8.55 GB | document boundaries for the compact stream (925,065,551 docs) |
| `token_lengths.npy` | 0.41 GB | per-document token counts, annotated stream |
| `checksums.json` | — | sha256 + byte size for each of the above |

## How to use

Rebuild the streams per `REPRODUCTION.md` §6, then compare your `.idx` document
boundaries against these.

**The failure this is most likely to catch:** using
`transformers.AutoTokenizer` instead of the Rust `tokenizers` library. The two
disagree on `\n\n` — Rust emits a single token `1116`, `AutoTokenizer` emits
`[198, 198]` — so the wrong tokenizer shifts boundaries on essentially every
document containing a blank line.

Two other things worth asserting while you are here: the truncation cap is
**1919 content tokens** (`enable_truncation(max_length=1920)`, EOS counts toward
the limit), with 25.4% of annotated rows sitting exactly at it; and the measured
annotated-stream total is **107,007,683,660 tokens**, about 3% below the 110.30B
estimate in the subsample metadata, which was computed on untruncated text.

## License and attribution

Released under the **Open Data Commons Attribution License (ODC-BY 1.0)**,
inherited from the upstream source.

> Contains information from [`allenai/dolma3_mix-6T`](https://huggingface.co/datasets/allenai/dolma3_mix-6T),
> made available under the Open Data Commons Attribution License (ODC-BY 1.0).

Please cite Olmo 3 ([arXiv:2512.13961](https://arxiv.org/abs/2512.13961)) and observe AI2's
[Responsible Use Guidelines](https://allenai.org/responsible-use).