Hyperparameter (HP) tuning in deep learning is an expensive process, prohibitively so for neural networks (NNs) with billions of parameters.We show that, in the recently discovered Maximal Update Parametrization ($\mu$P), many optimal HPs remain stable even as model size changes. This leads to a new HP tuning paradigm we call *$\mu$Transfer*: parametrize the target model in $\mu$P, tune the HP indirectly on a smaller model, and *zero-shot transfer* them to the full-sized model, i.e., without directly tuning the latter at all.We verify $\mu$Transfer on Transformer and ResNet. For example, 1) by transferring pretraining HPs from a model of 13M parameters, we outperform published numbers of BERT-large (350M parameters), with a total tuning cost equivalent to pretraining BERT-large once; 2) by transferring from 40M parameters, we outperform published numbers of the 6.7B GPT-3 model, with tuning cost only 7% of total pretraining cost. A Pytorch implementation of our technique can be found at github.com/microsoft/mup.

Recent advances in large language models (LLMs) and the intensifying popularity of ChatGPT-like applications have blurred the boundary of high-quality text generation between humans and machines. However, in addition to the anticipated revolutionary changes to our technology and society, the difficulty of distinguishing LLM-generated texts (AI-text) from human-generated texts poses new challenges of misuse and fairness, such as fake content generation, plagiarism, and false accusations of innocent writers. While existing works show that current AI-text detectors are not robust to LLM-based paraphrasing, this paper aims to bridge this gap by proposing a new framework called RADAR, which jointly trains a $\underline{r}$obust $\underline{A}$I-text $\underline{d}$etector via $\underline{a}$dversarial lea$\underline{r}$ning. RADAR is based on adversarial training of a paraphraser and a detector. The paraphraser's goal is to generate realistic content to evade AI-text detection.RADAR uses the feedback from the detector to update the paraphraser, and vice versa.Evaluated with 8 different LLMs (Pythia, Dolly 2.0, Palmyra, Camel, GPT-J, Dolly 1.0, LLaMA, and Vicuna) across 4 datasets, experimental results show that RADAR significantly outperforms existing AI-text detection methods, especially when paraphrasing is in place. We also identify the strong transferability of RADAR from instruction-tuned LLMs to other LLMs, and evaluate the improved capability of RADAR via GPT-3.5-Turbo.

Data parallelism across multiple machines is widely adopted for accelerating distributed deep learning, but it is hard to achieve linear speedup due to the heavy communication. In this paper, we propose SAPipe, a performant system that pushes the training speed of data parallelism to its fullest extent.

Big AI companies courted controversy by scraping wide swaths of the public internet. With the rise of AI agents, the next data grab is far more private. For years, the cost of using "free" services from Google, Facebook, Microsoft, and other Big Tech firms has been handing over your data. Uploading your life into the cloud and using free tech brings conveniences, but it puts personal information in the hands of giant corporations that will often be looking to monetize it. Now, the next wave of generative AI systems are likely to want more access to your data than ever before. Over the past two years, generative AI tools--such as OpenAI's ChatGPT and Google's Gemini--have moved beyond the relatively straightforward, text-only chatbots that the companies initially released.

When you purchase through links in our articles, we may earn a small commission. Thinking of switching from MacBook? RTX 5070 laptops deliver faster creative performance, powerful AI features and next-level gaming - built for demanding workflows. Apple's MacBooks are icons of the creative arts, and are beloved by creatives for their performance and streamlined design. But as capable as they are, they don't offer the same kind of power and versatility as the latest RTX AI PCs equipped with NVIDIA's GeForce RTX 50 Series graphics cards. If you're considering a laptop upgrade this year, GeForce RTX 50 Series laptops, backed by the latest Blackwell architecture, are specifically designed to handle the most demanding creative projects, outperforming the competition in both speed and visual precision.

We introduce DROID-SLAM, a new deep learning based SLAM system. DROID-SLAM consists of recurrent iterative updates of camera pose and pixelwise depth through a Dense Bundle Adjustment layer. DROID-SLAM is accurate, achieving large improvements over prior work, and robust, suffering from substantially fewer catastrophic failures. Despite training on monocular video, it can leverage stereo or RGB-D video to achieve improved performance at test time.

The hippocampus plays a critical role in learning, memory, and spatial representation, processes that depend on the NMDA receptor (NMDAR). Inspired by recent findings that compare deep learning models to the hippocampus, we propose a new nonlinear activation function that mimics NMDAR dynamics. NMDAR-like nonlinearity shifts short-term working memory into long-term reference memory in transformers, thus enhancing a process that is similar to memory consolidation in the mammalian brain.

Deep learning models are computationally intense, and implementations often have to be highly optimized by experts or hardware vendors to be usable in practice. The DL compiler, together with Learning to Compile have proven to be a powerful technique for optimizing tensor programs. However, a limitation of this approach is that it still suffers from unbearably long overall optimization time. In this paper, we present a new method, called AdaTune, that significantly reduces the optimization time of tensor programs for high-performance deep learning inference. In particular, we propose an adaptive evaluation method that statistically early terminates a costly hardware measurement without losing much accuracy. We further devise a surrogate model with uncertainty quantification that allows the optimization to adapt to hardware and model heterogeneity better. Finally, we introduce a contextual optimizer that provides adaptive control of the exploration and exploitation to improve the transformation space searching effectiveness. We evaluate and compare the levels of optimization obtained by a state-of-the-art DL compiler and AdaTune. The experiment results show that AdaTune obtains up to 115% higher GFLOPS than the baseline under the same optimization time budget.

Bilevel optimization (BO) is useful for solving a variety of important machine learning problems including but not limited to hyperparameter optimization, meta-learning, continual learning, and reinforcement learning.Conventional BO methods need to differentiate through the low-level optimization process with implicit differentiation, which requires expensive calculations related to the Hessian matrix. There has been a recent quest for first-order methods for BO, but the methods proposed to date tend to be complicated and impractical for large-scale deep learning applications. In this work, we propose a simple first-order BO algorithm that depends only on first-order gradient information, requires no implicit differentiation, and is practical and efficient for large-scale non-convex functions in deep learning. We provide non-asymptotic convergence analysis of the proposed method to stationary points for non-convex objectives and present empirical results that show its superior practical performance.

Executing a family of Deep Neural Networks (DNNs) training jobs on the same or similar datasets in parallel is typical in current deep learning scenarios. It is time-consuming and resource-intensive because each job repetitively prepares (i.e., loads and preprocesses) the data independently, causing redundant consumption of I/O and computations. Although the page cache or a centralized cache component can alleviate the redundancies by reusing the data prep work, each job's data sampled uniformly at random presents a low sampling locality in the shared dataset that causes the heavy cache thrashing. Prior work tries to solve the problem by enforcing all training jobs iterating over the dataset in the same order and requesting each data in lockstep, leading to strong constraints: all jobs must have the same dataset and run simultaneously. In this paper, we propose a dependent sampling algorithm (DSA) and domain-specific cache policy to relax the constraints. Besides, a novel tree data structure is designed to efficiently implement DSA. Based on the proposed technologies, we implemented a prototype system, named Joader, which can share data prep work as long as the datasets share partially. We evaluate the proposed Joader in practical scenarios, showing a greater versatility and superiority over training speed improvement (up to 500% in ResNet18).