It is widely acknowledged that large models have the potential to deliver superior performance across a broad range of domains. Despite the remarkable progress made in the field of machine learning systems research, which has enabled the development and exploration of large models, such abilities remain confined to a small group of advanced users and industry leaders, resulting in an implicit technical barrier for the wider community to access and leverage these technologies. In this paper, we introduce PyTorch Fully Sharded Data Parallel (FSDP) as an industry-grade solution for large model training. FSDP has been closely co-designed with several key PyTorch core components including Tensor implementation, dispatcher system, and CUDA memory caching allocator, to provide non-intrusive user experiences and high training efficiency. Additionally, FSDP natively incorporates a range of techniques and settings to optimize resource utilization across a variety of hardware configurations. The experimental results demonstrate that FSDP is capable of achieving comparable performance to Distributed Data Parallel while providing support for significantly larger models with near-linear scalability in terms of TFLOPS.

Video piracy has become an increasing problem particularly with the proliferation of media sharing through the advancement of internet services and various storage technologies. Thus, research in copyright protection mechanisms, where one of which includes digital watermarking has been receiving an increasing interest from scientists especially in designing a seamless algorithm for effective implementation. Basically digital watermarking involves embedding secret symbols known as watermarks within video data which can be used later for copyright detection purposes. This paper presents the state of the art in video watermarking techniques. It provides a critical review on various available techniques. In addition, it addresses the main key performance indicators which include robustness, speed, capacity, fidelity, imperceptibility and computational complexity.