Large language models (LLMs) have demonstrated remarkable performance across a wide range of industrial applications, from search and recommendations to generative tasks. Although scaling laws indicate that larger models generally yield better generalization and performance, their substantial computational requirements often render them impractical for many real-world scenarios at scale. In this paper, we present methods and insights for training small language models (SLMs) that deliver high performance and efficiency in deployment. We focus on two key techniques: (1) knowledge distillation and (2) model compression via quantization and pruning. These approaches enable SLMs to retain much of the quality of their larger counterparts while significantly reducing training, serving costs, and latency. We detail the impact of these techniques on a variety of use cases at a large professional social network platform and share deployment lessons - including hardware optimization strategies that enhance speed and throughput for both predictive and reasoning-based applications.

Ranking and recommendation systems are the foundation for numerous online experiences, ranging from search results to personalized content delivery. These systems have evolved into complex, multilayered architectures that leverage vast datasets and often incorporate thousands of predictive models. The maintenance and enhancement of these models is a labor intensive process that requires extensive feature engineering. This approach not only exacerbates technical debt but also hampers innovation in extending these systems to emerging problem domains. In this report, we present our research to address these challenges by utilizing a large foundation model with a textual interface for ranking and recommendation tasks. We illustrate several key advantages of our approach: (1) a single model can manage multiple predictive tasks involved in ranking and recommendation, (2) decoder models with textual interface due to their comprehension of reasoning capabilities, can generalize to new recommendation surfaces and out-of-domain problems, and (3) by employing natural language interfaces for task definitions and verbalizing member behaviors and their social connections, we eliminate the need for feature engineering and the maintenance of complex directed acyclic graphs of model dependencies. We introduce our research pre-production model, 360Brew V1.0, a 150B parameter, decoder-only model that has been trained and fine-tuned on LinkedIn's data and tasks. This model is capable of solving over 30 predictive tasks across various segments of the LinkedIn platform, achieving performance levels comparable to or exceeding those of current production systems based on offline metrics, without task-specific fine-tuning. Notably, each of these tasks is conventionally addressed by dedicated models that have been developed and maintained over multiple years by teams of a similar or larger size than our own.

Web-scale search systems typically tackle the scalability challenge As one of the largest professional networking platforms globally, with a two-step paradigm: retrieval and ranking. The retrieval step, LinkedIn is a hub for job seekers and recruiters, with 65M+ job also known as candidate selection, often involves extracting standardized seekers utilizing the search and recommendation services weekly entities, creating an inverted index, and performing term to discover millions of open job listings. To enable realtime personalization matching for retrieval. Such traditional methods require manual for job seekers, we adopted the classic two-stage paradigm and time-consuming development of query models. In this paper, of retrieval and ranking to tackle the scalability challenge. The retrieval we discuss applying learning-to-retrieve technology to enhance layer, also known as candidate selection, chooses a small set LinkedIn's job search and recommendation systems. In the realm of of relevant jobs from the set of all jobs, after which the ranking layer promoted jobs, the key objective is to improve the quality of applicants, performs a more computationally expensive second-pass scoring thereby delivering value to recruiter customers. To achieve and sorting of the resulting candidate set. This paper focuses on this, we leverage confirmed hire data to construct a graph that improving the methodology and systems for retrieval.