Datasets:
Copy NOTICE from Nickyang/RazorCal
Browse files
NOTICE
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
RAZOR: Pruning Replaceable Experts in LLMs
|
| 2 |
+
Copyright 2026 Mingyang Song, Mao Zheng
|
| 3 |
+
|
| 4 |
+
This product includes software developed by the RAZOR authors, licensed under
|
| 5 |
+
the Apache License, Version 2.0 (see LICENSE).
|
| 6 |
+
|
| 7 |
+
================================================================================
|
| 8 |
+
Third-party software
|
| 9 |
+
================================================================================
|
| 10 |
+
|
| 11 |
+
REAP (vendored under third_party/reap/)
|
| 12 |
+
Licensed under the Apache License, Version 2.0.
|
| 13 |
+
Copyright Cerebras Systems.
|
| 14 |
+
Upstream: https://github.com/CerebrasResearch/reap
|
| 15 |
+
Snapshot: 1970473c51ca3caeb98c10392f15b3a08a672974
|
| 16 |
+
Full license text: third_party/reap/LICENSE
|
| 17 |
+
Upstream README: third_party/reap/README-upstream.md
|
| 18 |
+
|
| 19 |
+
Included unchanged as a reference implementation; only a source subset is
|
| 20 |
+
vendored and the Python package does not import or distribute it.
|
| 21 |
+
|
| 22 |
+
REAM
|
| 23 |
+
NOT vendored in this repository. REAM is distributed by Samsung under the
|
| 24 |
+
MIT License. If you wish to compare against it, install it separately from
|
| 25 |
+
its upstream source.
|
| 26 |
+
|
| 27 |
+
================================================================================
|
| 28 |
+
Dataset attribution (data/)
|
| 29 |
+
================================================================================
|
| 30 |
+
|
| 31 |
+
RazorCal records remain subject to their applicable upstream terms. The
|
| 32 |
+
following attribution is retained from the recorded source information;
|
| 33 |
+
record-level provenance and license applicability are not fully established
|
| 34 |
+
for this release. See data/LICENSE-DATA and data/FIDELITY_AUDIT.md.
|
| 35 |
+
|
| 36 |
+
Recorded upstream attribution:
|
| 37 |
+
|
| 38 |
+
Nemotron-Post-Training-Dataset-v1 / v2 / v3 (NVIDIA)
|
| 39 |
+
CC-BY-4.0 https://creativecommons.org/licenses/by/4.0/
|
| 40 |
+
Verify the applicable source release and slice.
|
| 41 |
+
|
| 42 |
+
Nemotron-SFT-OpenCode-V1 (NVIDIA)
|
| 43 |
+
CC-BY-4.0
|
| 44 |
+
Domain: general coding
|
| 45 |
+
|
| 46 |
+
smoltalk2 (HuggingFaceTB)
|
| 47 |
+
ODC-BY-1.0 https://opendatacommons.org/licenses/by/1-0/
|
| 48 |
+
plus the licenses of its own upstream slices
|
| 49 |
+
Includes the recorded OpenHermes-2.5 attribution.
|
| 50 |
+
|
| 51 |
+
OpenHermes-2.5 (Teknium)
|
| 52 |
+
See the collection's individual source licenses.
|
| 53 |
+
Domain: world knowledge
|
| 54 |
+
|
| 55 |
+
tulu-3-sft-personas-instruction-following (Allen AI)
|
| 56 |
+
ODC-BY-1.0
|
| 57 |
+
Mapping to included records is not established here.
|
| 58 |
+
|
| 59 |
+
Dolci-Precise-IF / Dolci-Python-Algorithms
|
| 60 |
+
Recorded source labels for instruction-following and coding records.
|
| 61 |
+
Exact source releases and terms remain to be verified.
|
| 62 |
+
|
| 63 |
+
Nemotron-Cascade-2-SFT-Data / math (NVIDIA)
|
| 64 |
+
CC-BY-4.0 was recorded. Source of the 160 records labeled
|
| 65 |
+
nemotron-cascade2-math. The math slice was taken whole rather than
|
| 66 |
+
sampled per record, so mapping to individual upstream records is not
|
| 67 |
+
established. See data/README.md for the source-label inventory.
|
| 68 |
+
|
| 69 |
+
The sub_source field retains recorded attribution, not proof of complete
|
| 70 |
+
record-level provenance. See data/FIDELITY_AUDIT.md for the current status.
|
| 71 |
+
|
| 72 |
+
================================================================================
|
| 73 |
+
Models
|
| 74 |
+
================================================================================
|
| 75 |
+
|
| 76 |
+
Base-model weights are not redistributed here. Obtain them from their
|
| 77 |
+
respective providers under their own licenses.
|