part 1: books, algorithms, mathematics, architecture

LICENSES.json

@@ -0,0 +1,147 @@
+{
+ "aosa": {
+  "declared": "CC-BY-3.0",
+  "evidence": "https://aosabook.org/en/ 'License and Royalties' + repo LICENSE.md",
+  "topics": [
+   "design",
+   "architecture",
+   "performance"
+  ],
+  "commit": "12308deff812eb3a4fa4e210ae5496e154961311",
+  "url": "https://github.com/aosabook/aosabook",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "cp-algorithms": {
+  "declared": "CC-BY-SA-4.0",
+  "evidence": "GitHub licence API: CC-BY-SA-4.0",
+  "bucket": "algorithms",
+  "commit": "bbbbb6f04e5c9a665c8477ec68ee4f28b1367dd7",
+  "url": "https://github.com/cp-algorithms/cp-algorithms",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "discrete-book": {
+  "declared": "CC-BY-NC-SA-4.0",
+  "evidence": "repo LICENSE: 'fourth edition ... CC BY-NC-SA 4.0'; older editions differ, so this is treated as NC",
+  "bucket": "math",
+  "commit": "2d28f5dee1370beeed729d3712d53638597885dd",
+  "url": "https://github.com/oscarlevin/discrete-book",
+  "restricted": true,
+  "risk": "nc"
+ },
+ "miaow": {
+  "declared": "BSD-3-Clause",
+  "evidence": "GitHub licence API: BSD-3-Clause",
+  "bucket": "arch_code",
+  "commit": "dbc5d7cc6e5fd58828239b59491eeb4f66503074",
+  "url": "https://github.com/VerticalResearchGroup/miaow",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "ods": {
+  "declared": "CC-BY-2.5-CA",
+  "evidence": "https://opendatastructures.org/ 'released under a Creative Commons Attribution License' + repo COPYING",
+  "topics": [
+   "algorithms",
+   "data-structures"
+  ],
+  "commit": "9d22c44906dda2017b2ef0c762025bee644b58aa",
+  "url": "https://github.com/patmorin/ods",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "ods-code": {
+  "declared": "CC-BY-2.5-CA",
+  "evidence": "repo COPYING: 'released under a Creative Commons Attribution license'",
+  "bucket": "algorithms",
+  "commit": "9d22c44906dda2017b2ef0c762025bee644b58aa",
+  "url": "https://github.com/patmorin/ods",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "opencl-spec": {
+  "declared": "CC-BY-4.0",
+  "evidence": "repo LICENSE: 'asciidoctor sources ... CC BY 4.0'",
+  "topics": [
+   "gpu-architecture",
+   "api-design",
+   "specification"
+  ],
+  "commit": "0027b7f1a5e97b1bf5256888757479e012a5cb59",
+  "url": "https://github.com/KhronosGroup/OpenCL-Docs",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "opentitan": {
+  "declared": "Apache-2.0",
+  "evidence": "GitHub licence API: Apache-2.0",
+  "bucket": "arch_code",
+  "commit": "41049e9deb40e0602cd5272c40398be380ae2637",
+  "url": "https://github.com/lowRISC/opentitan",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "rocket-chip": {
+  "declared": "BSD-3-Clause OR Apache-2.0",
+  "evidence": "repo LICENSE.Berkeley (BSD-3-Clause) + LICENSE.SiFive (Apache-2.0), both read directly",
+  "bucket": "arch_code",
+  "commit": "ece7b9ad544b07df39cd13b6fd7236562a26355d",
+  "url": "https://github.com/chipsalliance/rocket-chip",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "sdxpy": {
+  "declared": "CC-BY-NC-4.0",
+  "evidence": "repo LICENSE.md: 'written material ... CC-BY-NC-4.0'",
+  "topics": [
+   "design",
+   "testing",
+   "profiling"
+  ],
+  "commit": "469c08a8807c631cf0f4deea675782775cc83ac4",
+  "url": "https://github.com/gvwilson/sdxpy",
+  "restricted": true,
+  "risk": "nc"
+ },
+ "spike": {
+  "declared": "BSD-3-Clause",
+  "evidence": "repo LICENSE: BSD 3-clause (Regents of the University of California)",
+  "bucket": "arch_code",
+  "commit": "4c6a9bc79dab80616932949707e7c0471d8e9e54",
+  "url": "https://github.com/riscv-software-src/riscv-isa-sim",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "thealgorithms-c": {
+  "declared": "GPL-3.0",
+  "evidence": "GitHub licence API: GPL-3.0",
+  "bucket": "algorithms",
+  "commit": "e5dad3fa8def3726ec850ca66a7f51521f8ad393",
+  "url": "https://github.com/TheAlgorithms/C",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "vexriscv": {
+  "declared": "MIT",
+  "evidence": "GitHub licence API: MIT",
+  "bucket": "arch_code",
+  "commit": "baf7dc82f855eddaf5b0a20a6802c526f127e38a",
+  "url": "https://github.com/SpinalHDL/VexRiscv",
+  "restricted": false,
+  "risk": "clean"
+ },
+ "vulkan-spec": {
+  "declared": "CC-BY-4.0",
+  "evidence": "repo LICENSE.adoc: 'CC-BY-4.0 ... for specification source documents'",
+  "topics": [
+   "gpu-architecture",
+   "api-design",
+   "specification"
+  ],
+  "commit": "b94f8bd2d004d2949469d8a5071ee10eb7d23566",
+  "url": "https://github.com/KhronosGroup/Vulkan-Docs",
+  "restricted": false,
+  "risk": "clean"
+ }
+}

README.md

@@ -0,0 +1,100 @@
+---
+license: other
+task_categories:
+- text-generation
+language:
+- en
+- code
+tags:
+- kernel
+- systems-programming
+- computer-architecture
+- algorithms
+pretty_name: kernel-llm pool (part 1, knowledge slice)
+---
+
+# kernel-llm pool — part 1: books, algorithms, mathematics, architecture
+
+Training text for a kernel/machine-level C specialist. This is **part 1 of a
+multi-part pool**, not the full 1T-token dataset: it is the knowledge slice
+(design, architecture, algorithms, maths), built from repositories whose licences
+were read rather than assumed.
+
+Volume is stated as measured: **15.6M tokens** across 6735 rows, where a token is
+estimated as bytes/3.6 because no tokenizer was run over this part. Byte counts are
+exact; the token figure is an estimate and is labelled as one in `stats.json`.
+
+## Contents
+
+| source | bucket | licence | rows | est. tokens | flag |
+|---|---|---|---:|---:|---|
+| `opentitan` | arch_code | Apache-2.0 | 4065 | 8034k | clean |
+| `vulkan-spec` | books | CC-BY-4.0 | 726 | 3118k | clean |
+| `aosa` | books | CC-BY-3.0 | 76 | 797k | clean |
+| `rocket-chip` | arch_code | BSD-3-Clause OR Apache-2.0 | 308 | 592k | clean |
+| `discrete-book` | math | CC-BY-NC-SA-4.0 | 132 | 577k | restricted |
+| `cp-algorithms` | algorithms | CC-BY-SA-4.0 | 165 | 457k | clean |
+| `miaow` | arch_code | BSD-3-Clause | 204 | 447k | clean |
+| `opencl-spec` | books | CC-BY-4.0 | 106 | 422k | clean |
+| `thealgorithms-c` | algorithms | GPL-3.0 | 398 | 275k | clean |
+| `sdxpy` | books | CC-BY-NC-4.0 | 115 | 255k | restricted |
+| `spike` | arch_code | BSD-3-Clause | 259 | 249k | clean |
+| `vexriscv` | arch_code | MIT | 97 | 228k | clean |
+| `ods` | books | CC-BY-2.5-CA | 19 | 135k | clean |
+| `ods-code` | algorithms | CC-BY-2.5-CA | 65 | 61k | clean |
+| **total** | | | **6735** | **15.6M** | |
+
+Every row carries `source`, `repo_license`, `spdx`, `commit`, `n_bytes`, `n_tokens`
+(`-1` when not measured), `restricted` and `path`, so a subset can be rebuilt exactly
+and every row can be traced to the commit it came from.
+
+## Licence flags — read before training
+
+This pool contains NonCommercial material. It is included and marked, but it is
+**not** free for commercial use:
+
+* `discrete-book` — CC-BY-NC-SA-4.0
+* `sdxpy` — CC-BY-NC-4.0
+
+Filter those rows out of any commercial training mix:
+
+```python
+ds = ds.filter(lambda r: not r["restricted"])
+```
+
+Publishing this data, or keeping it private, does not change the position: an NC
+licence restricts the *use*, not only the redistribution, so training a model you
+sell on NC text is a commercial use even if the text never leaves your disk.
+
+Publisher-copyrighted books (Operating Systems: Three Easy Pieces, Computer
+Architecture: A Quantitative Approach, Designing Data-Intensive Applications,
+Systems Performance, Linux Kernel Development, Refactoring, Design Patterns) are
+**not present at all** — no rows, no URLs. A book you can read online is not a
+book you may train on.
+
+## What is not here
+
+* kernel and driver trees (Linux, RTOS, GPU drivers) — part 2
+* other kernels and compute kernels — part 3
+* Stack v2 subsets — part 4 (gated)
+* traces, commits, synthetic and replay data — part 5
+
+## Dedup and splits
+
+`6735` distinct near-duplicate signatures; exact duplicates and MinHash near-duplicates
+were dropped, with sources deduplicated in priority order so a trusted copy wins over
+a mirror. `heldout` is 2% of groups, capped so no bucket is gutted.
+
+**The held-out split is public.** It is a development split for measuring
+contamination, not a secret holdout; encrypting public data does not create one.
+
+## How to load
+
+```python
+from datasets import load_dataset
+ds = load_dataset("ruthvikanne/kernel-pool-knowledge", split="train")
+```
+
+`mix.json` and `mix-manifest.parquet` hold a *plan* (tier weights, repeat counts)
+that joins on `path`; the text stays in the shards. `LICENSES.json` records, per
+source, the declared licence, the evidence it was read from, and the exact commit.

data/heldout-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:738db77c2fd5fbfd52399e75cd104ad53d9e7bb65cbff098a7ee52c3678f293e
+size 83495

data/train-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b1d4c4af6c48c2d67c34c90a7d266e20bbb3e65ecf1392f62f98d84d5df5308
+size 3442136

data/train-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:176ceaf20ec6823945c5f8eb0aa2ddcdb927b95a76178ef6813c9c33a91cce5c
+size 3767848

data/train-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de4146fc460d3a7732147fd3a6bec66bf3381e53b7234d8c878f3f4604018021
+size 5253426

data/train-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad6b1d377b88db38b0dcfe9071088b6de3d5a0493ee1f1c3d165ee35b2fe232d
+size 1790213

mix-manifest.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd12147a2b176323662e1dea5648fffb13c862a4489b6d98bd9a20ed413a8561
+size 108390

mix.json

@@ -0,0 +1,37 @@
+{
+ "target_tokens": 45000000,
+ "pool_train_tokens_available": 14775329,
+ "planned_tokens": 35388726,
+ "coverage_of_target": 0.786,
+ "restricted_excluded": true,
+ "restricted_rows_dropped": 218,
+ "max_repeat": 3,
+ "tiers": {
+  "core": {
+   "weight": 0.9375,
+   "target_tokens": 42187500,
+   "available_tokens": 10306698,
+   "repeat": 3,
+   "achieved_tokens": 30920095,
+   "shortfall_tokens": 11267404,
+   "files": 5376
+  },
+  "design": {
+   "weight": 0.0625,
+   "target_tokens": 2812500,
+   "available_tokens": 4468630,
+   "repeat": 1.0,
+   "achieved_tokens": 4468630,
+   "shortfall_tokens": 0,
+   "files": 926
+  }
+ },
+ "tiers_unavailable": [
+  "systems",
+  "traces",
+  "math",
+  "synthetic",
+  "replay"
+ ],
+ "note": "Shortfall is the honest gap: up to max_repeat copies of the domain core is not enough to reach the target, so breadth/replay must come from sources not in this pool."
+}

stats.json

@@ -0,0 +1,122 @@
+{
+ "config": {
+  "max_src_tok_M": 40.0,
+  "max_bucket_tok_M": 500.0,
+  "fresh_since": "2024-10-15",
+  "fresh_sources": "none",
+  "heldout_frac": 0.02,
+  "max_heldout_share": 0.2,
+  "near_dup_agree_of_32": 26,
+  "restrict_mode": "nc_copyleft",
+  "dedup_priority": [
+   [
+    "mcu",
+    0
+   ],
+   [
+    "linux",
+    1
+   ],
+   [
+    "kernels",
+    2
+   ],
+   [
+    "arch_code",
+    2
+   ],
+   [
+    "algorithms",
+    3
+   ],
+   [
+    "books",
+    3
+   ],
+   [
+    "docs",
+    3
+   ],
+   [
+    "math",
+    4
+   ]
+  ],
+  "tokenizer": "bytes/3.6 estimate"
+ },
+ "bytes_per_token_fallback": 3.6,
+ "heldout_groups": {
+  "arch_code": {
+   "groups": 14,
+   "held": 1,
+   "held_bytes": 138201,
+   "total_bytes": 38376749
+  },
+  "algorithms": {
+   "groups": 58,
+   "held": 1,
+   "held_bytes": 12409,
+   "total_bytes": 2858874
+  },
+  "books": {
+   "groups": 47,
+   "held": 1,
+   "held_bytes": 12346,
+   "total_bytes": 17077889
+  },
+  "math": {
+   "groups": 6,
+   "held": 1,
+   "held_bytes": 297782,
+   "total_bytes": 2301437
+  }
+ },
+ "near_dup": {
+  "n_sigs": 6735,
+  "near_dup_disabled_after": null
+ },
+ "by_bucket": {
+  "arch_code": {
+   "seen": 5690,
+   "kept_train": 4751,
+   "tok_train": 9515180,
+   "tok_estimated_train": 4751,
+   "near_dup": 199,
+   "exact_dup": 558,
+   "kept_heldout": 182,
+   "tok_heldout": 32867,
+   "tok_estimated_heldout": 182
+  },
+  "algorithms": {
+   "seen": 630,
+   "kept_train": 625,
+   "tok_train": 788947,
+   "tok_estimated_train": 625,
+   "kept_heldout": 3,
+   "tok_heldout": 3446,
+   "tok_estimated_heldout": 3,
+   "near_dup": 2
+  },
+  "books": {
+   "seen": 1055,
+   "kept_train": 1041,
+   "tok_train": 4723034,
+   "tok_estimated_train": 1041,
+   "near_dup": 13,
+   "kept_heldout": 1,
+   "tok_heldout": 3429,
+   "tok_estimated_heldout": 1
+  },
+  "math": {
+   "seen": 144,
+   "kept_train": 103,
+   "tok_train": 514275,
+   "tok_estimated_train": 103,
+   "kept_heldout": 29,
+   "tok_heldout": 63141,
+   "tok_estimated_heldout": 29,
+   "near_dup": 10,
+   "exact_dup": 2
+  }
+ }
+}

upload-manifest.json

@@ -0,0 +1,31 @@
+{
+ "local": "C:\\Users\\ruthvikanne\\AppData\\Local\\Temp\\opencode\\pool1\\out",
+ "repo": "ruthvikanne/kernel-pool-knowledge",
+ "private": false,
+ "files": 12,
+ "bytes": 15678278,
+ "largest": [
+  [
+   "train-00002.parquet",
+   5253426
+  ],
+  [
+   "train-00001.parquet",
+   3767848
+  ],
+  [
+   "train-00000.parquet",
+   3442136
+  ],
+  [
+   "train-00003.parquet",
+   1790213
+  ],
+  [
+   "lsh_index.parquet",
+   1220737
+  ]
+ ],
+ "include": null,
+ "exclude": "lsh_index.parquet"
+}