# Initial release of Roosterize, a tool for suggesting lemma names, for Coq 8.10

**URL:** <https://discourse.rocq-prover.org/t/initial-release-of-roosterize-a-tool-for-suggesting-lemma-names-for-coq-8-10/917>\
**Category:** Announcements\
**Tags:** serapi\
**Created:** [July 3, 2020, 8:25am UTC](https://discourse.rocq-prover.org/t/initial-release-of-roosterize-a-tool-for-suggesting-lemma-names-for-coq-8-10/917 "2020-07-03T08:25:15Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![palmskog](https://sea1.discourse-cdn.com/flex001/user_avatar/discourse.rocq-prover.org/palmskog/32/38_2.png) [@palmskog](https://discourse.rocq-prover.org/u/palmskog)\
**Post date:** [July 3, 2020, 8:25am UTC](https://discourse.rocq-prover.org/t/initial-release-of-roosterize-a-tool-for-suggesting-lemma-names-for-coq-8-10/917/1 "2020-07-03T08:25:15Z")

</div>

Dear Coq users and developers,

We are happy to announce the [initial release](https://github.com/EngineeringSoftware/roosterize/releases/tag/v8.10.0) of Roosterize, a tool for suggesting lemma names in Coq verification projects.

> **[EngineeringSoftware/roosterize](https://github.com/EngineeringSoftware/roosterize)**
>
> Tool for suggesting lemma names in Coq verification projects - EngineeringSoftware/roosterize

Roosterize learns and suggests lemma names using neural networks that take serialized Coq lemma statements and elaborated terms as input. The technique is described in a [research paper](https://arxiv.org/abs/2004.07761) published in the proceedings of IJCAR 2020, where it was evaluated on a [large dataset](https://github.com/EngineeringSoftware/math-comp-corpus) of Coq code from the Mathematical Components family of projects, with promising results.

The initial release supports Coq 8.10.2 and is based on using machine-readable representations of Coq code provided by [SerAPI](https://github.com/ejgallego/coq-serapi). See the [README](https://github.com/EngineeringSoftware/roosterize/blob/v8.10.0/README.md) for detailed instructions on installing Roosterize and its dependencies, such as the machine learning framework [PyTorch](https://pytorch.org/get-started/previous-versions/#v110).

Roosterize is released under the MIT license, so please don’t hesitate to report issues and contribute on GitHub.

Best regards,  
Karl and all contributors to the project