In time for FS24, 🆕 research➕capabilities built on 🌼 Flower for serving LLMs efficiently while still maintaining privacy: MeanCache! A contribution frm @virginia_tech, @Cisco, @UMNews. Waris Gil (@warisgil) lead author 🚀
📖 Paper: https://t.co/UH30VXYv36 (code avail soon)
Cricket legend Imran Khan is rotting to death in a tiny jail cell in Pakistan. It's time Britain used its influence to help free him
https://t.co/Yf5eCx2fAi
Imran Khan, former Captain of World Cup Cricket Champions,former President, rapidly going blind in a Pakistani jail - from neglect. A man who built & funded a first class cancer hospital that charged patients from all over the country ONLY what they could afford for amazing care.
The Man Who Saw (Cricket’s) Tomorrow, Imran Khan has lost his vision in the jail. They say his vision is now only 15 percent and that’s such a sad story
#Pakistan: Imran Khans solitary confinement and inhumane detention conditions must end - UN expert. Prolonged or indefinite solitary confinement is prohibited under int'l human rights law – and when it extends longer than 15 days, it's a form of #torture.
https://t.co/m2Re8Yhyc0
One of the game's all-round greats, he averaged 40.06 with the bat and 22.22 with the ball from his seventh to his 87th Test.
The finest cricketer from Pakistan, he led them to World Cup glory in 1992.
#OnThisDay The legendary Imran Khan turns 73
Everyone is starting to converge. AI is heading federated. Monoliths are what centralized single data centers generate. Agents are federated models collaborating. Same fundemental drivers with different terminology.
Below is a THREAD with some shorter clips of the most important messages from my boys’ interview regarding their father’s incarceration.
Please share if you can.
1) His conditions in jail
Advancing Semantic Caching for LLMs with Domain-Specific Embeddings and Synthetic Data
Redis demonstrates that fine-tuning smaller embedding models on domain-specific data outperforms larger models for semantic caching.
📝https://t.co/E1srpHWGCM
At @Redisinc we even show that fine-tuning smaller models purely on synthetic data can outperform OpenAI’s embedding models for semantic caching. More exciting updates coming soon : )
Redis is the world-class standard for caching. In this paper, we demonstrates that fine-tuning smaller embedding models on domain-specific data outperforms larger models for semantic caching.
https://t.co/BmFJEpSKyP
Promoting democracy, freedom, human rights, free & fair elections, & the rule of law in Pakistan, @RepJoeWilson introduces the Pakistan Democracy Act.