Roosterize v1.1.0 for Coq 8.10.2

Major changes since the initial release: improved command-line interface; added VSCode interface (using LSP); added functionality of utilizing .roosterizerc.

Supporting Coq version 8.10.2. The tool requires OCaml 4.07.1, SerAPI 0.7.1, Python 3.7 or later, PyTorch 1.1.0.

Attached to this release:
roosterize-dist-debian-cpu.tgz: binary distribution built under Debian system, using CPU only.
roosterize-model-t1.tgz: pre-trained model, trained on the tier-1 part of our MathComp corpus.
roosterize-model-ta.tgz: pre-trained model, trained on all tiers of our MathComp corpus.

Roosterize v1.1.0 for Coq 8.10.2 (beta)

Major changes since the last release: improved command-line interface, added VSCode interface (using LSP), added functionality of utilizing .roosterizerc.

Supporting Coq version 8.10.2. The tool requires OCaml 4.07.1, SerAPI 0.7.1, Python 3.7 or later, PyTorch 1.1.0.

Attached to this release:
roosterize-dist-debian-cpu.tgz: binary distribution built under Debian system, using CPU only.
roosterize-model-t1.tgz: pre-trained model, trained on the tier-1 part of our MathComp corpus.
roosterize-model-ta.tgz: pre-trained model, trained on all tiers of our MathComp corpus.

Roosterize for Coq 8.10

Initial release of the Roosterize tool for suggesting lemma names in Coq verification projects using neural networks, supporting Coq version 8.10.2. The tool requires OCaml 4.07.1, SerAPI 0.7.1, Python 3.7.7 or later, PyTorch 1.1.0, and pre-trained models (attached to this release). The tool release accompanies the research paper Deep Generation of Coq Lemma Names Using Elaborated Terms, published in the proceedings of IJCAR 2020.