Direct Preference Optimization (DPO) and SimPO paper explanationIn this article, we will discuss the Direct Preference Optimization paper and the Simple Preference Optimization paper. Both of these are…3d ago3d ago
PPO to GRPO in Large Language Models AlignmentIn this article, I will discuss the theory behind Proximal Policy Optimization and Group Relative Policy Optimization (which was used in…Mar 5Mar 5
DoRA paper deep diveIn this article, we will be going through the paper DoRA, which came after LoRA and QLoRA. I have discussed AdaLORA and Representation…Feb 3Feb 3
ReFT: Representation Finetuning Paper deep diveThis article will dive deeper into the paper ReFT (Representation fine-tuning). It is a parameter-efficient finetuning (PEFT) method that…Jan 29Jan 29
JSON vs YAML function calling Finetuning comparisonWANDB TRAINING RUNS AND CHECKPOINTSNov 20, 2024Nov 20, 2024
Character.ai optimized inference blog post explainedRecently, character.ai, a role-playing based LLM startup, released a blog post on their inference pipeline. The blog posts mentioned three…Jun 30, 2024Jun 30, 2024
Adaptive LoRA (AdaLORA) paper explanationIn this article, we will dive deeper into the paper AdaLORA, which is based on Singular value decomposition (SVD) to dynamically choose low…May 6, 20241May 6, 20241
ColBERT: Contextualized Late Interaction BERT explained with a tutorialIn this article, we will go over the Colbert architecture, both v1 and v2. It is a neural Information Retrieval technique that can help us…Mar 9, 20241Mar 9, 20241
Neo4j: Analyzing the supplier's list of Apple and SamsungIn this article, we will review Neo4j basics by getting data about Apple and Samsung supplier lists. We are analyzing the supplier list of…Feb 3, 2024Feb 3, 2024
MAMBA and State Space Models ExplainedThis article will go through a new class of deep learning models called Structured State Spaces and Mamba.Feb 1, 20241Feb 1, 20241