Are you using speech rec on @comcast devices? You might be using SpeechNet: ASR designed for label-scarce, compute-limited settings using weak supervision, Snorkel, & CUDA graphs. Join @ralph_tang at https://
future.snorkel.ai to learn how daily 12M+ smart TV queries are automated.
MACHINE LEARNING
-
SpeechNet: ASR Technology Powers 12M Daily Smart TV Queries
By
–
-
AI at Scale for Global South Social Impact
By
–
Do not miss the opportunity to hear @alpanrav
, our Chief Scientist, AI/ML talk about how we identify specific problems and apply AI at scale to solve those problems, with the ultimate goal of improving lives and livelihoods across the global south. #MachineLearning #Socialimpact -
Prompt Engineering’s Role in Fine-Tuning AI Models
By
–
prompt instructions in a fine tune mostly just sculpt completions outside the fine-tune distribution. i’d say only include a fixed prompt if you’re very constrained on examples (<~1k) and can swallow the inflated tuning and inference costs
-
Explanation-Based Learning: Leveraging Domain Theory for AI
By
–
Explanation-based learning is a powerful type of machine learning that utilizes a robust and accurate domain theory to generalize or generate concepts from training data. If you’re new to AI, learn more about explanation-based learning here: https://
bit.ly/3HFSjL0 -
RLHF: Reinforcement Learning from Human Feedback Guide
By
–
informative blog post by Chip Huyen on "RLHF: Reinforcement Learning from Human Feedback" This blog post covers:
⦿ Pre-training for completion
⦿ Supervised fine-tuning for dialogue
⦿ Reinforcement Learning from Human Feedback Blog link: https://
lnkd.in/gPuKbCiP Bonus RLHF -
Unlimiformer Enables Transformers to Process Unbounded Input Lengths
By
–
CMU’s Unlimiformer Augments Transformers to Enable Unbounded Input Lengths https://
syncedreview.com/2023/05/04/cmu
s-unlimiformer-augments-transformers-to-enable-unbounded-input-lengths/
… -
Privacy, ML, and Health Data Workshop in Ottawa
By
–
Thrilled to be in Ottawa, the nation's capital, for a workshop with @UWaterloo_CPI @uwaterlooAI @GovCanHealth @StatCan_eng @CIHI_ICIS
, on privacy, ML, and health data. Exciting to have convos between academics & policy makers on these important topics https://
uwaterloo.ca/cybersecurity-
privacy-institute/events/cybersecurity-privacy-and-artificial-intelligence-health
… -
Google Introduces MaMMUT Compact Multimodal AI Model
By
–
Introducing MaMMUT, a compact multimodal model that trains across contrastive, text generative, and localization-aware objectives, and outperforms the state of the art or achieves competitive performance on a range of tasks. Learn more → https://
goo.gle/42lZuk5 -
Download Data Analytics Essentials Guide with Case Studies and Insights
By
–
Download the "Data #Analytics Essentials Guide" (with Case Studies, Actionable Insights, Proven Techniques, and more): https://
rapidops.com/services/data-
analytics/
… by @rapid_ops =======
#DataStrategy #DataLeadership #BigData #DataFluency #DataInnovation #DataScience #AI #MachineLearning -
Machine Learning Innovation Driven by Building Useful Products
By
–
The real driver of #MachineLearning innovation is the need to build ever more useful products: https://
linkedin.com/posts/damienbe
nveniste_machinelearning-datascience-artificialintelligence-activity-7057378198384181248-FlJ7
…
————
#RecSys #Martech #BigData #Analytics #DataScience #AI #DataScientists #DeepLearning #NeuralNetworks