---
license: odc-by
viewer: false
pretty_name: olmo3_lit50b_shuf
tags:
- pretraining
- token-stream
- dolma3
- olmo3
- mode-collapse
size_categories:
- 10B<n<100B
---

# olmo3_lit50b_shuf — story-filtered dolma3 literature stream, document-level reshuffle

A document-level global shuffle (seed 20260730) of `olmo3_lit50b` (story-filtered dolma3 literature partition; same 13.46M documents and token multiset), frozen **in the exact order it was consumed during training**. This is the stream behind the `olmo3_lit50b_shuf` run of the *explain-m0de-collapse* project — a 50B-token run on narrative-only data used to study name collapse under a story-dense distribution.

**Source:** `allenai/dolma3_mix-6T-1025-7B` @ revision `2ca900fbe14e86c5c83d064d9f0882f1c0b8c05b` · **Tokenizer:** `allenai/dolma2-tokenizer` (copy included under `tokenizer/`) ·
**Created by:** `scripts/shuffle_stream.py` @ git `c805e86e1b01` on 2026-07-30.
**Ordering policy:** `global_doc_shuffle` (`shuffle_during_training: false`; **not** the official OLMo-3 data order).
**Training runs that consumed this stream:** `olmo3_lit50b_shuf` (400k updates).

- Parent stream: [`Ibisbill/olmo3-lit50b`](https://huggingface.co/datasets/Ibisbill/olmo3-lit50b).

## Format — the order *is* the data

| | |
|---|---|
| Stored tokens | 52,428,800,001 (= 52,428,800,000 training tokens + 1 look-ahead token) |
| Shards | 2098 files `shards/tokens_000000.u32 … tokens_002097.u32` |
| Shard size | 25,000,000 tokens (100,000,000 bytes) each; last shard 3,800,001 tokens |
| Element type | `uint32`, little-endian, raw (no header); docs separated by EOS id 100257 |
| Vocab | dolma2, 100,278 raw ids, padded to 100,352 in the model |

Shard `i` holds global token offsets `[25,000,000·i, 25,000,000·(i+1))`. `shard_manifest.jsonl` lists every shard with
its `start_token`, `num_tokens` and `sha256`; `manifest.json` records provenance; `verification.json` (where present)
records the end-to-end check performed after freezing.

Training read the stream strictly sequentially with `global_tokens_per_update = 131,072` (128 sequences × 1,024
tokens per update, split across DDP ranks), 400,000 updates = 52,428,800,000 tokens. Hence **update `k` (0-based) trained on
tokens `[131,072·k, 131,072·(k+1))`**, i.e. shards `⌊131,072·k / 25,000,000⌋ … ⌊(131,072·(k+1) − 1) / 25,000,000⌋`.
The target for position `t` is token `t+1`, which is why one extra token is stored.

```python
import json, numpy as np

def tokens(root, start, n):
    """Return stream tokens [start, start+n) as a uint32 array."""
    shards = sorted((json.loads(l) for l in open(f"{root}/shard_manifest.jsonl")), key=lambda e: e["start_token"])
    out, need = [], n
    for e in shards:
        s0, s1 = e["start_token"], e["start_token"] + e["num_tokens"]
        if s1 <= start or need == 0: continue
        mm = np.memmap(f"{root}/{e['shard']}", dtype="<u4", mode="r")
        a = max(start, s0) - s0; b = min(a + need, e["num_tokens"])
        out.append(mm[a:b]); need -= b - a
        if need == 0: break
    return np.concatenate(out)

G = 131_072
k = 55_000                                  # what did the model see at update 55,000?
batch = tokens("path/to/olmo3_lit50b_shuf", G * k, G + 1)
x, y = batch[:-1].reshape(128, 1024), batch[1:].reshape(128, 1024)
```

Decode with `tokenizers.Tokenizer.from_file("tokenizer/tokenizer.json")` or
`transformers.AutoTokenizer.from_pretrained("tokenizer/")`.

## Verifying a download

```bash
python - <<'EOF'
import json, hashlib
for l in open("shard_manifest.jsonl"):
    e = json.loads(l)
    h = hashlib.sha256(open(e["shard"], "rb").read()).hexdigest()
    assert h == e["sha256"], e["shard"]
print("all shards match")
EOF
```

## License

The token stream is a derivative of `allenai/dolma3_mix-6T-1025-7B` and is redistributed under the same terms
(ODC-BY 1.0); per-subset terms of the underlying sources apply as documented by the source dataset.
The bundled tokenizer is `allenai/dolma2-tokenizer` (Apache-2.0).

Source machine path: `/mnt/nlpgpu-io1/data/maggie/explain-m0de-collapse/replicate/data/olmo3_lit50b_shuf`. Uploaded 2026-09-02.