Merge branch 'site' into accelerators_table

Author: aradys

_community_blog/3d-rotations-and-spatial-transformations-made-easy-with-roma-356a495a20c4.md

@@ -0,0 +1,8 @@
+---
+title: '3D rotations and spatial transformations made easy with RoMa'
+author: Romain Brégier
+ext_url: https://medium.com/pytorch/3d-rotations-and-spatial-transformations-made-easy-with-roma-356a495a20c4
+date: Jan 25, 2024
+---
+
+Struggling with quaternions, rotation vectors, right-hand rules and all these stuffs? Try RoMa: an easy-to-to-use, stable and efficient library to deal with rotations and spatial transformations in PyTorch.

_community_blog/colossal-llama-2-low-cost-and-high-quality-domain-specific-llm-solution-using-llama-and-26d2e4b9fd92.md

@@ -0,0 +1,8 @@
+---
+title: 'Colossal-LLaMA-2: Low Cost and High-quality Domain-specific LLM Solution Using LLaMA and Colossal-AI'
+author: Yang You
+ext_url: https://medium.com/pytorch/colossal-llama-2-low-cost-and-high-quality-domain-specific-llm-solution-using-llama-and-26d2e4b9fd92
+date: Jan 29, 2024
+---
+
+The most prominent distinction between LLaMA-1 and LLaMA-2 lies in the incorporation of higher-quality corpora, a pivotal factor contributing to significant performance enhancements in LLaMA-2. This, coupled with its commercial availability, extends the potential for creative applications of large models within the open-source community.

_community_blog/distributed-training-with-pytorch-and-azure-ml-898429139098.md

@@ -0,0 +1,8 @@
+---
+title: 'Distributed training with PyTorch and Azure ML'
+author: Beatriz Stollnitz
+ext_url: https://medium.com/pytorch/distributed-training-with-pytorch-and-azure-ml-898429139098
+date: Jan 6, 2023
+---
+
+Suppose you have a very large PyTorch model, and you’ve already tried many common tricks to speed up training: you optimized your code, you moved training to the cloud and selected a fast GPU VM, you installed software packages that improve training performance (for example, by using the ACPT curated environment on Azure ML). And yet, you still wish your model could train faster. Maybe it’s time to give distributed training a try! Continue reading to learn the simplest way to do distributed training with PyTorch and Azure ML.

_community_blog/exploring-scientific-machine-learning-pipelines-through-the-simulai-toolkit-9fda42d6c6a0.md

@@ -0,0 +1,8 @@
+---
+title: 'Exploring scientific machine learning pipelines through the SimulAI toolkit'
+author: Joao Lucas de Sousa Almeida
+ext_url: https://medium.com/pytorch/exploring-scientific-machine-learning-pipelines-through-the-simulai-toolkit-9fda42d6c6a0
+date: Feb 15, 2024
+---
+
+SciML, short for Scientific Machine Learning, encompasses work that merges quantitative sciences with machine learning. It has gained significant traction over the past decade, driven by the widespread availability of specialized hardware (such as GPUs and TPUs) and datasets. Additionally, it has been propelled by the overarching influence of the machine learning wave, now ingrained in the zeitgeist of our times. In this context, we’d like to introduce SimulAI, an open-source toolkit under the Apache 2.0 license. SimulAI is designed to be user-friendly, providing a high-level Python interface for managing scientific machine learning pipelines. This article aims to showcase its current workflow and utility in constructing scientific experiments. We encourage feedback and potential contributions from the interested community, with plans to delve into more advanced topics in future articles.

_community_blog/how-activation-checkpointing-enables-scaling-up-training-deep-learning-models-7a93ae01ff2d.md

@@ -0,0 +1,8 @@
+---
+title: 'How Activation Checkpointing enables scaling up training deep learning models'
+author: PyTorch
+ext_url: https://medium.com/pytorch/how-activation-checkpointing-enables-scaling-up-training-deep-learning-models-7a93ae01ff2d
+date: Nov 9, 2023
+---
+
+Activation checkpointing is a technique used for reducing the memory footprint at the cost of more compute. It utilizes the simple observation that we can avoid saving intermediate tensors necessary for backward computation if we just recompute them on demand instead.

_community_blog/how-fashable-achieves-soa-realistic-ai-generated-images-using-pytorch-and-azure-machine-learning-2313c4cf5f44.md

@@ -0,0 +1,9 @@
+---
+title: 'How FASHABLE achieves SoA realistic AI generated images using PyTorch and Azure Machine Learning'
+author: Orlando Ribas Fernandes
+ext_url: https://medium.com/pytorch/how-fashable-achieves-soa-realistic-ai-generated-images-using-pytorch-and-azure-machine-learning-2313c4cf5f44
+date: Feb 10, 2023
+---
+
+Fashable is a company born at XNFY Lab (a joint initiative with Microsoft). The company’s main goal is to revolutionize the world of fashion with ethical Artificial Intelligence (AI) technologies built on PyTorch framework. Fashable is focused on developing AI models that generates synthetic contents for the global fashion industry. The Fashion industry has been criticized in recent years because it generates a lot of waste and is responsible for up to 10% of global carbon dioxide output. Fashable has stepped up to address this issue by introducing multiple AI solutions that generates realistic personalized consumer garments without actually producing them to help in reducing carbon footprint. This will help the fashion brands make informed decisions without investing in experimental products and also reducing the industry’s carbon footprint globally. Hence, in Fashable, our IP models utilize modern approaches, such as Generative Adversarial Networks (GANs), best seller analysis, custom dataset creation, and so on to resolve such problems.
+

_community_blog/introducing-torchopt-a-high-performance-differentiable-optimization-library-for-pytorch-37c4c0ef6ae1.md

@@ -0,0 +1,8 @@
+---
+title: 'Introducing TorchOpt: A High-Performance Differentiable Optimization Library for PyTorch'
+author: Benjamin Liu
+ext_url: https://medium.com/pytorch/introducing-torchopt-a-high-performance-differentiable-optimization-library-for-pytorch-37c4c0ef6ae1
+date: Jun 29, 2023
+---
+
+Explore TorchOpt, a PyTorch-based library that revolutionizes differentiable optimization with its unified programming abstraction, high-performance distributed execution runtime, and support for various differentiation modes.”

_community_blog/latest-colossal-ai-boasts-novel-automatic-parallelism-and-offers-savings-up-to-46x-for-stable-1453b48f3f02.md

@@ -0,0 +1,8 @@
+---
+title: 'Latest Colossal-AI boasts novel automatic parallelism and offers savings up to 46x for Stable Diffusion 2'
+author: Yang You
+ext_url: https://medium.com/pytorch/latest-colossal-ai-boasts-novel-automatic-parallelism-and-offers-savings-up-to-46x-for-stable-1453b48f3f02
+date: Jan 31, 2023
+---
+
+As a new PyTorch Ecosystem Partner, we at HPC-AI Tech look forward to working with the PyTorch community to advance AI technologies through our open source project, Colossal-AI. We are excited to join forces with the PyTorch community in this effort.

_community_blog/profiling-pytorch-language-models-with-octoml-profile-eda7ece6b7bd.md

@@ -0,0 +1,8 @@
+---
+title: 'Profiling PyTorch language models with octoml-profile'
+author: Team Octo
+ext_url: https://medium.com/pytorch/profiling-pytorch-language-models-with-octoml-profile-eda7ece6b7bd
+date: Apr 4, 2023
+---
+
+The recent launch of PyTorch 2.0 makes it clear that the community is heavily investing in a compiler-powered future for machine learning. The new OctoML Profiler can help any user realize the full potential of these shifts in the ML landscape.

_community_blog/pypose-a-library-for-robot-learning-with-physics-based-optimization-861bc0bb92f1.md

@@ -0,0 +1,8 @@
+---
+title: 'PyPose: A Library for Robot Learning with Physics-based Optimization'
+author: PyPose
+ext_url: https://medium.com/pytorch/pypose-a-library-for-robot-learning-with-physics-based-optimization-861bc0bb92f1
+date: Dec 6, 2023
+---
+
+We are excited to share our new open-source library PyPose. It is a PyTorch-based robotics-oriented library that provides a set of tools and algorithms for connecting deep learning with physics-based optimization.