12 papers
Training nGPT
Ilya Loshchilov, Boris Ginsburg
The normalized Transformer (nGPT) realizes hyperspherical representation learning by constraining model parameter vectors and activation vectors to the unit hypersphere. In this pa…
TurboBias 2.0: Streaming Context-Biasing for Production-Efficient ASR Systems
Vladimir Bataev, Lilit Grigoryan, Andrei Andrusenko +3
Contextualization is essential for production automatic speech recognition (ASR) systems, where user-provided phrases must be recognized accurately under strict latency constraints…
Normalized Architectures are Natively 4-Bit
Maxim Fishman, Brian Chmiel, Ron Banner +2
Training large language models at 4-bit precision is critical for efficiency. We show that nGPT, an architecture that constrains weights and hidden representations to the unit hype…
Extending Automatic Machine Translation Evaluation to Book-Length Documents
Kuang-Da Wang, Shuoyang Ding, Chao-Han Huck Yang +4
Despite Large Language Models (LLMs) demonstrating superior translation performance and long-context capabilities, evaluation methodologies remain constrained to sentence-level ass…
FlexCTC: GPU-powered CTC Beam Decoding With Advanced Contextual Abilities
Lilit Grigoryan, Vladimir Bataev, Nikolay Karpov +3
While beam search improves speech recognition quality over greedy decoding, standard implementations are slow, often sequential, and CPU-bound. To fully leverage modern hardware ca…
TurboBias: Universal ASR Context-Biasing powered by GPU-accelerated Phrase-Boosting Tree
Andrei Andrusenko, Vladimir Bataev, Lilit Grigoryan +2
Recognizing specific key phrases is an essential task for contextualized Automatic Speech Recognition (ASR). However, most existing context-biasing approaches have limitations asso…
Pushing the Limits of Beam Search Decoding for Transducer-based ASR models
Lilit Grigoryan, Vladimir Bataev, Andrei Andrusenko +3
Transducer models have emerged as a promising choice for end-to-end ASR systems, offering a balanced trade-off between recognition accuracy, streaming capabilities, and inference s…
NGPU-LM: GPU-Accelerated N-Gram Language Model for Context-Biasing in Greedy ASR Decoding
Vladimir Bataev, Andrei Andrusenko, Lilit Grigoryan +3
Statistical n-gram language models are widely used for context-biasing tasks in Automatic Speech Recognition (ASR). However, existing implementations lack computational efficiency…
WIND: Accelerated RNN-T Decoding with Windowed Inference for Non-blank Detection
Hainan Xu, Vladimir Bataev, Lilit Grigoryan +1
We propose Windowed Inference for Non-blank Detection (WIND), a novel strategy that significantly accelerates RNN-T inference without compromising model accuracy. During model infe…
L0-Reasoning Bench: Evaluating Procedural Correctness in Language Models via Simple Program Execution
Simeng Sun, Cheng-Ping Hsieh, Faisal Ladhak +3
Complex reasoning tasks often rely on the ability to consistently and accurately apply simple rules across incremental steps, a foundational capability which we term "level-0" reas…
HAINAN: Fast and Accurate Transducer for Hybrid-Autoregressive ASR
Hainan Xu, Travis M. Bartley, Vladimir Bataev +1
We present Hybrid-Autoregressive INference TrANsducers (HAINAN), a novel architecture for speech recognition that extends the Token-and-Duration Transducer (TDT) model. Trained wit…
nGPT: Normalized Transformer with Representation Learning on the Hypersphere
Ilya Loshchilov, Cheng-Ping Hsieh, Simeng Sun +1
We propose a novel neural network architecture, the normalized Transformer (nGPT) with representation learning on the hypersphere. In nGPT, all vectors forming the embeddings, MLP,…