Murmur: Efficient Long-Context ASR Inference with Hierarchical Chunking
cs.LG, cs.AI, eess.AS
Submitted: 2026-05-31
Updated: 2026-11-04
Code: https://github.com/uw-syfi/Murmur
Terminology
Sources
- Whisper-CD: Accurate Long-Form Speech Recognition using Multi-Negative Contrastive Decoding
- pyannote.audio: neural building blocks for speaker diarization
- PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling
- GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio
- LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models
- Speech recognition for medical conversations
- Earnings-21: A Practical Benchmark for ASR in the Wild
- Distil-Whisper: Robust Knowledge Distillation via Large-Scale Pseudo Labelling
- TagSpeech: End-to-End Multi-Speaker ASR and Diarization with Fine-Grained Temporal Grounding
- VoxServe: Streaming-Centric Serving System for Speech Language Models
- Voxtral
- Lost in the Middle: How Language Models Use Long Contexts
- VIBEVOICE-ASR Technical Report
- Context-Driven Dynamic Pruning for Large Speech Foundation Models
- Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference
- Efficient Streaming Language Models with Attention Sinks
- Early Attentive Sparsification Accelerates Neural Speech Transcription
- SpeakerLM: End-to-End Versatile Speaker Diarization and Recognition with Multimodal Large Language Models
- H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models
- Tactic: Adaptive Sparse Attention with Clustering and Distribution Fitting for Long-Context LLMs
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis