116 citations · 153 across the 15 of their papers we have counts for
15 papers
Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning
NVIDIA, :, Aaron Blakeman +572
We introduce Nemotron 3 Ultra, a 550 billion total and 55 billion active parameter Mixture-of-Experts Hybrid Mamba-Attention language model. We pre-trained Nemotron 3 Ultra on 20 t…
Not All NVFP4 QAT Recipes Are Equal: How Architecture and Scale Shape Model Quality for Anomaly Segmentation
Zijian Du, Oleg Rybakov
Real-time anomaly segmentation demands both high recall and efficient low-precision inference. We study the three-way interaction of model architecture, model scale, and FP4 quanti…
USM RNN-T model weights binarization
Oleg Rybakov, Dmitriy Serdyuk, Chengjian Zheng
Large-scale universal speech models (USM) are already used in production. However, as the model size grows, the serving cost grows too. Serving cost of large models is dominated by…
SimulTron: On-Device Simultaneous Speech to Speech Translation
Alex Agranovich, Eliya Nachmani, Oleg Rybakov +5
Simultaneous speech-to-speech translation (S2ST) holds the promise of breaking down communication barriers and enabling fluid conversations across languages. However, achieving acc…
USM-Lite: Quantization and Sparsity Aware Fine-tuning for Speech Recognition with Universal Speech Models
Shaojin Ding, David Qiu, David Rim +10
End-to-end automatic speech recognition (ASR) models have seen revolutionary quality gains with the recent development of large-scale universal speech models (USM). However, deploy…
2-bit Conformer quantization for automatic speech recognition
Oleg Rybakov, Phoenix Meadowlark, Shaojin Ding +4
Large speech models are rapidly gaining traction in research community. As a result, model compression has become an important topic, so that these models can fit in memory and be…
RAND: Robustness Aware Norm Decay For Quantized Seq2seq Models
David Qiu, David Rim, Shaojin Ding +2
With the rapid increase in the size of neural networks, model compression has become an important area of research. Quantization is an effective technique at decreasing the model s…
STEP: Learning N:M Structured Sparsity Masks from Scratch with Precondition
Yucheng Lu, Shivani Agrawal, Suvinay Subramanian +3
Recent innovations on hardware (e.g. Nvidia A100) have motivated learning N:M structured sparsity masks from scratch for fast model inference. However, state-of-the-art learning re…
Streaming Parrotron for on-device speech-to-speech conversion
Oleg Rybakov, Fadi Biadsy, Xia Zhang +3
We present a fully on-device streaming Speech2Speech conversion model that normalizes a given input speech directly to synthesized output speech. Deploying such a model on mobile d…
4-bit Conformer with Native Quantization Aware Training for Speech Recognition
Shaojin Ding, Phoenix Meadowlark, Yanzhang He +3
Reducing the latency and model size has always been a significant research problem for live Automatic Speech Recognition (ASR) application scenarios. Along this direction, model qu…
A Scalable Model Specialization Framework for Training and Inference using Submodels and its Application to Speech Model Personalization
Fadi Biadsy, Youzheng Chen, Xia Zhang +3
Model fine-tuning and adaptation have become a common approach for model specialization for downstream tasks or domains. Fine-tuning the entire model or a subset of the parameters…
Real time spectrogram inversion on mobile phone
Oleg Rybakov, Marco Tagliasacchi, Yunpeng Li +3
We present two methods of real time magnitude spectrogram inversion: streaming Griffin Lim(GL) and streaming MelGAN. We demonstrate the impact of looking ahead on perceptual qualit…
Pareto-Optimal Quantized ResNet Is Mostly 4-bit
AmirAli Abdolrashidi, Lisa Wang, Shivani Agrawal +4
Quantization has become a popular technique to compress neural networks and reduce compute cost, but most prior work focuses on studying quantization without changing the network s…
Real-time Speech Frequency Bandwidth Extension
Yunpeng Li, Marco Tagliasacchi, Oleg Rybakov +2
In this paper we propose a lightweight model for frequency bandwidth extension of speech signals, increasing the sampling frequency from 8kHz to 16kHz while restoring the high freq…
Streaming keyword spotting on mobile devices
Oleg Rybakov, Natasha Kononenko, Niranjan Subrahmanya +2
In this work we explore the latency and accuracy of keyword spotting (KWS) models in streaming and non-streaming modes on mobile phones. NN model conversion from non-streaming mode…