[Oryx] Use poetry/pyproject.toml for building/dependency management

- Also updates README.md
- Also fixes matcher dispatch for Python3.9

PiperOrigin-RevId: 381165823

spinoffs/oryx/README.md

@@ -1,10 +1,51 @@
 # Oryx
 
 Oryx is a library for probabilistic programming and deep learning built on top
-of Jax. The approach is to expose a set of function transformations that
-compose and integrate with JAX's existing transformations (e.g. `jit`, `grad`,
-and `vmap`).
+of Jax. The approach is to expose a set of function transformations that compose
+and integrate with JAX's existing transformations (e.g. `jit`, `grad`, and
+`vmap`).
+
+## Installation
+
+You can install Oryx via `pip`:
+
+```bash
+$ pip install oryx
+```
 
 ## Documentation and Examples
 
-Coming soon!
+Documentation is available
+[on the Oryx website](https://www.tensorflow.org/probability/oryx/api_docs/python/oryx).
+
+### Guides
+
+*   [A tour of Oryx](https://www.tensorflow.org/probability/oryx/notebooks/a_tour_of_oryx)
+*   [Probabilistic programming](https://www.tensorflow.org/probability/oryx/notebooks/probabilistic_programming)
+
+## Development
+
+To develop and modify Oryx, you need to install
+[`poetry`](https://python-poetry.org/), a tool for Python packaging and
+dependency management.
+
+To install the development dependencies of Oryx, you can run
+
+```bash
+$ poetry install
+```
+
+and to enter a virtual environment for testing or debugging, you can run:
+
+```bash
+$ poetry shell
+```
+
+### Running tests
+
+Oryx uses [Bazel](https://bazel.build/) for building and testing. Once Bazel is
+installed, you can run tests by executing:
+
+```
+$ poetry run bazel test //oryx/...
+```

spinoffs/oryx/oryx/experimental/matching/matcher.py

@@ -230,6 +230,7 @@
 `Pattern` and implementing the  `match(self, expr, bindings, succeed)` method,
 the class is automatically registered with `matcher`.
 """
+import collections
 import functools
 
 from typing import Any, Callable, Dict, Iterator, Optional, Sequence, TypeVar
@@ -601,7 +602,7 @@ def __init__(self,
         Dot, name=name, accumulate=accumulate, plus=plus, greedy=greedy)
 
 
-@matcher.register(Sequence)
+@matcher.register(collections.abc.Sequence)
 def sequence_matcher(pattern: Sequence[Any]):
   """Returns a matcher for a given sequence pattern."""
 

spinoffs/oryx/pyproject.toml

@@ -0,0 +1,35 @@
+# Copyright 2021 The TensorFlow Probability Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+[tool.poetry]
+name = "oryx"
+version = "0.2.0"
+description = "Probabilistic programming and deep learning in JAX"
+authors = ["Google LLC"]
+license = "Apache 2.0"
+
+[tool.poetry.dependencies]
+python = "^3.6"
+jax = "0.2.16"
+jaxlib = "0.1.68"
+dataclasses = { version = "*", python = "<3.7" }
+tfp-nightly = {extras = ["jax"], version = "0.14.0-alpha.20210623"}
+
+
+[tool.poetry.dev-dependencies]
+inference-gym = "^0.0.4"
+
+[build-system]
+requires = ["poetry-core>=1.0.0"]
+build-backend = "poetry.core.masonry.api"