Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations

Fangda Gu, Jiaqi Zhai, Rui Li, Yu Shi, Xing Liu, Lucy Liao, Yueming Wang, Xuan Cao, Leon Gao, Zhaojie Gong, Michael He, Yin-Hua Lu

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 42 functions out of this paper's own repositories and ran 15 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

Repositories linked to this paper

• facebookresearch/generative-recommenders — marked official in Papers with Code; named in the paper
• foreverYoungGitHub/generative-recommenders-pl
• glb400/Toy-RecLM
• 1hb6s7t/Gr-HSTU
• MindSpore-scientific-2/code-9
• MindSpore-scientific/code-12
• MindSpore-scientific/code-4
• alibaba/TorchEasyRec
• bailuding/rails
• pwc-1/Paper-9
• roman-dusek/GR-HSTU
• snapfinger/hstu-blair

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Large-scale recommendation systems are characterized by their reliance on high cardinality, heterogeneous features and the need to handle tens of billions of user actions on a daily basis. Despite being trained on huge volume of data with thousands of features, most Deep Learning Recommendation Models (DLRMs) in industry fail to scale with compute. Inspired by success achieved by Transformers in language and vision domains, we revisit fundamental design choices in recommendation systems. We reformulate recommendation problems as sequential transduction tasks within a generative modeling framework ("Generative Recommenders"), and propose a new architecture, HSTU, designed for high cardinality, non-stationary streaming recommendation data. HSTU outperforms baselines over synthetic and public datasets by up to 65.8% in NDCG, and is 5.3x to 15.2x faster than FlashAttention2-based Transformers on 8192 length sequences. HSTUbased Generative Recommenders, with 1.5 trillion parameters, improve metrics in online A/B tests by 12.4% and have been deployed on multiple surfaces of a large internet platform with billions of users. More importantly, the model quality of Generative Recommenders empirically scales as a power-law of training compute across three orders of magnitude, up to GPT-3/LLaMa-2 scale, which reduces carbon footprint needed for future model developments, and further paves the way for the first foundation models in recommendations.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2402.17152")
get_code_for_paper("2402.17152")
have("2402.17152")

Connect an agent — have() is free.