PPUNITHinbpe-tokenizer.hashnode.dev·22h ago · 5 min readHow I Built a BPE Tokenizer from Scratch (Part 1: Training)Okay so before we get into the code, let me tell you why I even did this. I could've just used a library. tiktoken exists. HuggingFace tokenizers exist. But I wanted to actually get it — like really u00
RRishikantinrishiii2.hashnode.dev·18h ago · 7 min readThe Architecture That Changed the World: Transformers, BERT, and the LLM RevolutionIn our previous post, we discovered that Recurrent Neural Networks (RNNs) and LSTMs suffer from a massive flaw: they must process data sequentially, one word at a time. This makes them agonizingly slo00
RRishikantinrishiii2.hashnode.dev·18h ago · 6 min readBreaking the Bottleneck: Seq2Seq Models and the Invention of AttentionIn our previous post, we used LSTMs to solve the Vanishing Gradient problem. By utilizing mathematical memory gates, LSTMs could successfully read a 200-word movie review and predict whether it was po00
RRishikantinrishiii2.hashnode.dev·18h ago · 6 min readMastering Time and Text: RNNs, LSTMs, and the Magic of Word EmbeddingsUp until this point, we have operated in a static world. Whether we were predicting housing prices or running convolutions over an image of a dog, the inputs were fixed in time and space. But what if 00
IMIshwari Mulyeinishwarimulye.hashnode.dev·2d ago · 8 min readFrom TF-IDF to Transformers: Rebuilding My Mental Health ClassifierIntroduction I finished a supervised machine learning course a few months ago and immediately wanted to build something with it. Not a tutorial project. Something that felt real. I am a third-year Com00
GBGracie Boltonincontain.hashnode.dev·4d ago · 6 min readNLP Models Are Becoming the Engine Behind Smarter PlatformsHow natural language processing is quietly reshaping the way people interact with digital products Modern users don't just want software that works. They want platforms that understand them — tools th00
KSKedareswari Siriinkedareswarisiri-tech.hashnode.dev·4d ago · 5 min readUnderstanding Natural Language Processing (NLP): How Computers Understand Human LanguageHave You Ever Thought About This? You wake up in the morning and say: "Hey Google, what's the weather today?" Within seconds, your phone understands your speech, converts it into text, figures out y00
KAKeshav Agarwalinbeyondthebenchmark.hashnode.dev·Jun 19 · 5 min readHow I Actually Choose an AI Model for Production (Not From a Leaderboard)Every time a client asks me "which AI model should we use," they expect a technical answer. A benchmark score. A leaderboard ranking. The model with the best number on Papers With Code. That's not how00
MAMuhanad Abulhusninmuhanadabulhusn.hashnode.dev·Jun 15 · 6 min readSocioRAG: Retrieval Built for Reading Social Dynamics, Not Just DocumentsSay you have two thousand pages of reporting on a contested election, a conflict, or a corruption case. You ask an ordinary retrieval system: "what does this corpus say about voter intimidation?" It f00
PSPenumudi Sreejainsreeja-penumudi.hashnode.dev·Jun 12 · 2 min readHow I Built a Real-Time ESG Risk Scorer using Python, NLP & StreamlitAs a final-year Computer Science student, I wanted to build something meaningful that combines AI and finance. That's how I built the ESG Risk Scorer — a real-time platform that analyzes Environmental00
