|
Download README.md from zesen-kth/tiny-llm: direct link, hf CLI and curl.
- Browser
- Download file 2.45 kB
-
https://huggingface.co/datasets/zesen-kth/tiny-llm/resolve/main/README.md
- Command line
-
hf download hf://datasets/zesen-kth/tiny-llm/README.md
-
curl -L -o README.md https://huggingface.co/datasets/zesen-kth/tiny-llm/resolve/main/README.md
2.45 kB
| pretty_name: tiny-llm training logs | |
| license: mit | |
| tags: | |
| - language-model | |
| - pretraining | |
| - learning-rate-schedule | |
| - decentralized-training | |
| - training-logs | |
| size_categories: | |
| - 1K<n<10K | |
| # tiny-llm retained training evidence | |
| 1782 seed/horizon records across 594 configurations: cosine-to-zero and WSD | |
| schedules, synchronous and four- and eight-worker decentralized training, at 20 to 160 | |
| global tokens per parameter, for 20.4M-parameter models on C4. Snapshot 2026-09-17. | |
| This mirrors `doc/data/current-training/` in | |
| [WangZesen/tiny-llm](https://github.com/WangZesen/tiny-llm). The website built from it is | |
| at <https://wangzesen.github.io/tiny-llm/>. | |
| ## Layout | |
| Retained artifacts are grouped into gzipped JSONL containers, one line per artifact: | |
| ```json | |
| {"path":"logs/cosine/sync/h20/lr-0.01-b1-0.9-b2-0.99/s42/metrics.jsonl", | |
| "bytes":71410,"sha256":"...","text":"<the artifact verbatim>"} | |
| ``` | |
| `sources.json` is the index: it maps every artifact to its container and records the | |
| digest of the original file. Metrics containers concatenate one gzip member per line, so | |
| `offset` and `compressedBytes` address a single run's log directly; every other container | |
| is a single stream. Placement is a pure rule, so a misfiled container is detectable. | |
| ## What is retained, and what is not | |
| 9836 artifacts are byte-identical to the files written | |
| during training. The 1296 `metrics.jsonl` logs had two | |
| members deleted from every train event: `local_grad_norms`, `local_losses`. | |
| These are per-worker diagnostics for the packed models; no published figure, table or | |
| curve reads them, and they were 37% of the logged bytes. | |
| The transform is a deletion, not a re-serialization: no other byte of any line, and no | |
| line, was changed, added or reordered. `sources.json.sha256` remains the digest of the | |
| original file and can no longer be recomputed from this copy for those logs — but for the | |
| cosine runs it stays independently attested by the retained `receipt.json`, which records | |
| the trainer's own hash of `metrics.jsonl` and is checked on every verification run. The | |
| complete pre-repack bundle, with the arrays intact, is this repo at revision | |
| `per-file-original`. | |
| ## Verify and regenerate | |
| ```bash | |
| uv run python scripts/export_publication.py --check --bundle hf://zesen-kth/tiny-llm | |
| ``` | |
| No GPU, campaign directory or checkpoint is needed. Verification checks every retained | |
| byte against `sources.json` and every derived file against `checksums.json`. | |