Sublime
An inspiration engine for ideas
Daan van Hulsen
@daan
vLLM: AI Server with up to 24x higher throughput than HuggingFace Transformers
🚀 3.5x higher throughput than HuggingFace TGI
🧠 Optimised Memory Usage
🔄 Parallel Processing
🔗 OpenAI Compatibility
🌐 Private Cloud or Local Setup
🖥️ Demo Integrating Qwen... See more
Mervin Praisonx.com
MoRA: High-Rank Updating for Parameter-Efficient Fine-Tuning
Outperforms LoRA on memoryintensive tasks and achieves comparable performance on other tasks
repo: https://t.co/EV3CSsYpKq
abs: https://t.co/4WpHBl4EPt https://t.co/Gl8yBzeobi
Eran Friedman
@cloudyb
Cory
@prtcl
'Universality for random matrix ensembles of Wigner type' — Terence Tao, UCLA.
Lecture series:
https://t.co/i1q1UM4syk
Alex Bilzerianx.com