
Acoustic Model Fusion for End-to-end Speech Recognition
Improving speech recognition through the fusion of an external acoustic model into an end-to-end system, resulting in a significant reduction in word error rate and enhanced named entity recognition.

User-level Differentially Private Stochastic Convex Optimization: Efficient Algorithms with Optimal Rates
Efficient algorithms for user-level differentially private stochastic convex optimization with optimal rates.

Investigating Salient Representations and Label Variance Modeling in Dimensional Speech Emotion Analysis
Investigating the impact of salient representations and label variance modeling on dimensional speech emotion analysis

Large-scale Training of Foundation Models for Wearable Biosignals
Large-scale training of foundation models for wearable biosignals using self-supervised learning on curated Apple Watch data

Co-ML: Collaborative Machine Learning Model Building for Developing Dataset Design Practices
Co-ML is a collaborative machine learning model building tool that fosters hands-on learning of dataset design practices through a distributed experience.

Modular: What is loop unrolling? How you can speed up Mojoš„ code with @unroll
Speed up Mojoš„ code by understanding and implementing loop unrolling.

Boost your data & AI skills with our latest offerings: Databricks Academy Labs and Blended Learning
Boost your data & AI skills with Databricks Academy Labs and Blended Learning, offering hands-on guided lab experiences in Databricks environments and a combination of self-paced and instructor-led sessions.
How to use -ed and -ing adjectives in English
Learn how to correctly use -ed and -ing adjectives in English to convey the intended meaning of a sentence.

āFall in love with your dataāāSnorkel AIās Enterprise LLM Summit
Empowering enterprise AI success through data development at Snorkel AIās Enterprise LLM Summit

āFall in love with your dataāāSnorkel AIās Enterprise LLM Summit
Building GenAI with Your Data - Snorkel AI's Enterprise LLM Summit addressed the importance of data development in enterprise AI, showcasing case studies, research, and practical examples.

Mixed-input matrix multiplication performance optimizations
Optimizing mixed-input matrix multiplication for AI-driven applications with memory-efficient data types and layout conformance on NVIDIA Ampere architecture

Architect defense-in-depth security for generative AI applications using the OWASP Top 10 for LLMs
Architect defense-in-depth security for generative AI applications using the OWASP Top 10 for LLMs