py-rotary-embedding-torch: New package (#47059)

Co-authored-by: Benjamin Meyers <bsmits@rit.edu>

1 files changed, 27 insertions, 0 deletions

diff --git a/var/spack/repos/builtin/packages/py-rotary-embedding-torch/package.py b/var/spack/repos/builtin/packages/py-rotary-embedding-torch/package.py
new file mode 100644
index 0000000000..c307b827be
--- /dev/null
+++ b/var/spack/repos/builtin/packages/py-rotary-embedding-torch/package.py
@@ -0,0 +1,27 @@
+# Copyright 2013-2023 Lawrence Livermore National Security, LLC and other
+# Spack Project Developers. See the top-level COPYRIGHT file for details.
+#
+# SPDX-License-Identifier: (Apache-2.0 OR MIT)
+
+
+from spack.package import *
+
+
+class PyRotaryEmbeddingTorch(PythonPackage):
+    """A standalone library for adding rotary embeddings to transformers in Pytorch,
+    following its success as relative positional encoding. Specifically it will make
+    rotating information into any axis of a tensor easy and efficient, whether they
+    be fixed positional or learned. This library will give you state of the art
+    results for positional embedding, at little costs."""
+
+    homepage = "https://github.com/lucidrains/rotary-embedding-torch"
+    pypi = "rotary-embedding-torch/rotary-embedding-torch-0.5.3.tar.gz"
+
+    maintainers("meyersbs")
+
+    version("0.5.3", sha256="45db29b19bd7025f09d202752c26bf6921b05d8b5a977cfcdc625ce96ddf2b5c")
+
+    depends_on("py-setuptools", type="build")
+    depends_on("py-beartype", type=("build", "run"))
+    depends_on("py-einops@0.7:", type=("build", "run"))
+    depends_on("py-torch@2.0:", type=("build", "run"))