Readme update (mosaicml#2581)

* images * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md graphics * Update README.md * Update README.md * add key features image * Update README.md key features images * Update README.md * Update README.md ecosystem image * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md remove methods link * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md delete old README * Update README.md update careers page * Update README.md block quote citation * Update README.md * Update README.md * delete test_release_tests_reflect_readme * remove release test * update docs * lint * filter warning * import --------- Co-authored-by: Mihir Patel <[email protected]>
mvpatel2000 · Oct 3, 2023 · 9bffbcb · 9bffbcb
1 parent 00939a0
commit 9bffbcb
Show file tree

Hide file tree

Showing 12 changed files with 160 additions and 424 deletions.
diff --git a/.ci/build_conda.sh b/.ci/build_conda.sh
diff --git a/.ci/release_tests/example_1.py b/.ci/release_tests/example_1.py
diff --git a/.ci/release_tests/example_2.py b/.ci/release_tests/example_2.py
diff --git a/.ci/release_tests/print_world_size.py b/.ci/release_tests/print_world_size.py
diff --git a/.ci/run_release_tests.sh b/.ci/run_release_tests.sh
diff --git a/README.md b/README.md
diff --git a/docs/source/_static/images/ecosystem.png b/docs/source/_static/images/ecosystem.png
diff --git a/docs/source/_static/images/key_features.png b/docs/source/_static/images/key_features.png
diff --git a/docs/source/_static/images/logo.png b/docs/source/_static/images/logo.png
diff --git a/docs/source/_static/images/traning_loop.png b/docs/source/_static/images/traning_loop.png
diff --git a/docs/source/index.rst b/docs/source/index.rst
@@ -3,40 +3,28 @@
 Composer
 ========
 
-`Composer <https://github.com/mosaicml/composer>`_ is a library for training neural networks better, faster, and cheaper.
-It contains many state-of-the-art methods for accelerating neural network training and improving generalization, along with
-an optional :doc:`Trainer<trainer/using_the_trainer>` API that makes `composing` many different enhancements easy.
-
-We implemented dozens of papers and `trained thousands of models <https://www.mosaicml.com/blog/mosaic-resnet>`_ so you don't have to.
-
-With Composer, you can build recipes to:
-
-.. include:: ../../README.md
-   :parser: myst_parser.sphinx_
-   :start-after: <!-- start numbers -->
-   :end-before: <!-- end numbers -->
-
-.. include:: ../../README.md
-   :parser: myst_parser.sphinx_
-   :start-after: <!-- start main results -->
-   :end-before: <!-- end main results -->
-
-
-Composer features:
-
-- 20+ curated efficiency methods for training better language and vision models.
-- An easy-to-use Trainer interface written to be as performant as possible, with integrated multi-GPU training, mixed-precision training, and numerous `best practices <https://www.mosaicml.com/blog/5-best-practices-for-efficient-model-training>`_ implemented out of the box.
-- Standalone functional forms of training methods for easy integration into your own training loop
-- Strong, reproducible baselines to get you started as fast as possible.
-
-See :doc:`getting_started/installation` for
-how to install and start using Composer, :doc:`trainer/using_the_trainer` for an introduction
-to our trainer, and :doc:`trainer/algorithms` for details about our efficiency methods and how to use them in your code.
-
-At `MosaicML <https://mosaicml.com>`_, we are focused on making fast, high-quality ML training accessible to everyone. To do this,
-we continually productionize state-of-the-art academic research on efficient model
-training, and also study the `combinations` of these methods in order to ensure
-that model training is ✨ as efficient as possible ✨.
+`Composer <https://github.com/mosaicml/composer>`_  is an open-source deep learning training
+library by [MosaicML](https://www.mosaicml.com/). Built on top of PyTorch, the Composer library
+makes it easier to implement distributed training workflows on large-scale clusters.
+
+We built Composer to be **optimized for scalability and usability**, integrating best practices
+for efficient, multi-node training. By abstracting away low-level complexities like parallelism
+techniques, distributed data loading, and memory optimization, you can focus on training modern
+ML models and running experiments without slowing down.
+
+We recommend using Composer to speedup your experimentation workflow if you're training neural
+networks of any size, including:
+
+- Large Language Models (LLMs)
+- Diffusion models
+- Embedding models (e.g. BERT)
+- Transformer-based models
+- Convolutional Neural Networks (CNNs)
+
+Composer is heavily used by the MosaicML research team to train state-of-the-art models like MPT,
+and we open-sourced this library to enable the ML community to do the same. This framework is used
+by organizations in both the tech industry and the academic sphere and is continually updated with
+new features, bug fixes, and stability improvements for production workloads.
 
 If you have any questions, please feel free to reach out to us on `Twitter`_, `Email`_, or `Slack`_!
 

diff --git a/tests/test_docs.py b/tests/test_docs.py
@@ -65,38 +65,3 @@ def test_docker_build_matrix():
 
     with open(os.path.join(docker_folder, 'build_matrix.yaml'), 'r') as f:
         assert existing_build_matrix == f.read()
-
-
-@pytest.mark.doctest
-@pytest.mark.parametrize('example', [1, 2])
-def test_release_tests_reflect_readme(example: int):
-    """Test that example_1.py and example_2.py in release_tests reflect the README.md."""
-    with open(os.path.join(os.path.dirname(__file__), '..', 'README.md'), 'r') as f:
-        readme_lines = f.readlines()
-    example_code_lines = []
-    found_begin = False
-    started = False
-    for i, line in enumerate(readme_lines):
-        if f'begin_example_{example}' in line:
-            found_begin = True
-            continue
-        # Wait until we get the ```python for start of code snippet
-        if found_begin and not started:
-            if line == '```python\n':
-                started = True
-        # Reached end of code snippet
-        elif started and line == '```\n':
-            # Code snippet continues
-            if i + 2 < len(readme_lines) and '-->\n' == readme_lines[
-                    i + 1] and '<!--pytest-codeblocks:cont-->\n' == readme_lines[i + 2]:
-                started = False
-            # Code snippet ends
-            else:
-                break
-        # Add line
-        elif started:
-            example_code_lines.append(line)
-
-    example_file = os.path.join(os.path.dirname(__file__), '..', '.ci', 'release_tests', f'example_{example}.py')
-    with open(example_file, 'r') as f:
-        assert f.readlines() == example_code_lines