🔥Telegram İfşa

Delighted to share that I participated in (RSNA) “Screening Mammography Breast Cancer Detection”, CV Kaggle Competition and have secured Bronze Medal 🥉 in the competition. It was a fulfilling experience dealing with huge volume of imbalanced dataset(314+ GBs) in dicom format.

Good news: Kaggle birdclef competition is the first to "support Python as a first-tier language, unlike other metrics that were traditionally implemented in C#" Now kaggle can share unusual metrics directly with kagglers, preventing a lot of metric related confusion. 🥳

Underrated things in Data Science: * Unit tests * Docker Overrated things in Data Science: * Data Cleaning * Domain Expertise

I’ll reach office faster walking as compared to cab. Maybe this is what people should vote for tomorrow. #bangalorecity #traffic #bengalurutraffic #bengaluru #KarnatakaAssemblyElection2023

Easily migrate your codebase from one language or framework to another github.com/0xpayne/gpt-mi… #ai #deeplearning #gpt #chatgpt #artificalintelligence

There should be equality of opportunities and not equality of outcomes. Think!! It’s important that society and world realise this difference. Maybe people would stop being proud and feeling entitled to handouts then. #equality #foodforthought

Pathetic service experience by airtel India. Scheduled conn relocation request on Sunday for today. 1. Service engg doesn’t show up without any info 2. Customer care has no resolve. on hold for 45mins Gonna disconnect my airtel black @Airt_Presence Reliance Jio JioCare

4 down, RCB doing RCB things 🤦#RCBvsGT

Amazing blog by NVIDIA AI on LLM pruning and distillation developer.nvidia.com/blog/how-to-pr…

Increasingly, docs are for LLMs to read, not just for humans to read. We're starting to explicitly design our docs to be great for *both* uses. As part of that, I'm aiming to include full markdown docs in our projects. FastHTML is the 1st! Here you go: github.com/AnswerDotAI/fa…

🛒 RexBERT: ModernBERT except for E-commerce was just released by RAHUL BAJAJ et al! 4 base encoders (17M, 68M, 150M, 400M) trained on 2.3T tokens (with 350B E-commerce related tokens), easily outperforming base models on E-commerce tasks! Details in 🧵

We're announcing a new update to MTEB: RTEB It's a new multilingual text embedding retrieval benchmark with private (!) datasets, to ensure that we measure true generalization and avoid (accidental) overfitting. Details in our blogpost below 🧵