Thank You
Acknowledgments
LemmaLoom was diesigned and built by Ryan S. Cooper, and special thanks go to Mary Lynne Gasaway Hill, PhD, FRSA for her guidance, feedback, and support of this project.
LemmaLoom also contains software, models, language data, and typefaces that other people made and shared under open licenses. A model here is a large set of numbers learned from many example recordings or texts. LemmaLoom uses its models to transcribe speech, tell speakers apart, and analyze grammar.
The credits below are the same as those under Licenses in LemmaLoom ▸ About LemmaLoom. Click an entry there to read the full text of its license or licenses (see Licenses). In that list, the Whisper models and their tokenizer files are credited under WhisperKit, the pyannote models under FluidAudio, and TypeScript, DOMPurify, and marked under Monaco Editor.
Software
| Name | License | Copyright holder |
|---|---|---|
| Factory (FactoryKit) | MIT License | Michael Long |
| GRDB.swift | MIT License | Gwendal Roué |
| WhisperKit | MIT License | argmax, inc. |
| swift-transformers, included with WhisperKit | Apache License 2.0 | Hugging Face |
| FluidAudio | Apache License 2.0 | Fluid Inference |
| fastcluster, included with FluidAudio | BSD 2-Clause License | Daniel Müllner and Google Inc. |
| VBx, included with FluidAudio | Apache License 2.0 | BUT Speech@FIT |
| Monaco Editor | MIT License | Microsoft Corporation |
| TypeScript, included with Monaco Editor | Apache License 2.0 | Microsoft Corporation |
| DOMPurify, included with Monaco Editor | Apache License 2.0 or Mozilla Public License 2.0 | Cure53 |
| marked, included with Monaco Editor | MIT License | MarkedJS and Christopher Jeffrey |
| Mozilla Readability | Apache License 2.0 | Arc90 Inc. |
| PCRE2 | BSD 3-Clause License | University of Cambridge |
Models and language data
| Name | License | Copyright holder or source |
|---|---|---|
| Whisper large-v3 and large-v3-turbo | MIT License | OpenAI |
| Whisper models, as adapted by Argmax | MIT License | argmax, inc. |
| Whisper tokenizer files | Apache License 2.0 | Hugging Face’s openai/whisper-large-v3 repository |
| pyannote speaker-diarization-community-1, as adapted by Fluid Inference | Creative Commons Attribution 4.0 | pyannote, WeSpeaker, and BUT Speech@FIT |
| spaCy en_core_web_md | MIT License | ExplosionAI GmbH |
| spaCy de_core_news_md | MIT License | ExplosionAI GmbH |
| spaCy zh_core_web_md | MIT License | ExplosionAI GmbH |
| spacy-pkuseg | MIT License | ExplosionAI GmbH and the pkuseg authors |
| WordNet 3.0 | WordNet 3.0 License | Princeton University |
| Berkeley Neural Parser: benepar_en3, benepar_de2, and benepar_zh2 | MIT License | Nikita Kitaev, Victor Huang, and Mitchell Stern |
| T5 (T5-small) | Apache License 2.0 | Google Research |
| XLM-RoBERTa | MIT License | Facebook, Inc. and its affiliates |
| Chinese ELECTRA (ELECTRA-180G-large) | Apache License 2.0 | Joint Laboratory of HIT and iFLYTEK Research |
| LemmInflect 0.2.3 | MIT License | Brad Jascob |
Typeface
LemmaLoom’s interface is set in the Ubuntu font family. Canonical Ltd. holds the copyright and publishes the fonts under version 1.0 of the Ubuntu font license.
The example project
“Climate Debate”, the project behind most examples and figures in this guide, is invented. Any resemblance to a real person, place, or publication is a coincidence.
Thank you for using LemmaLoom!