During the decoding process of large language models, especially in Auto-regressive models, decoding must be performed step-by-step until the entire sequence is generated. Within this process, there are caching techniques that can help reduce computation and improve decoding speed; one such technique is known as the KV Cache.
Read More »KV Cache: A Caching Mechanism To Accelerate Transformer Generation
When implementing various services through LLMs, do you worry about uncontrolled language generation? Recently, at a critical juncture in wrapping up a project, I used tools like Outlines to constrain LLM decoding, which effectively controlled the model's output to follow the desired patterns. However, a colleague posed a deep question: What if I want it not to generate specific words?
Read More »Using Finite State Machine (FSM) and Rollback Mechanism to Restrict LLM from Generating Banned Words
bisect is a built-in Python module, primarily designed to maintain the order of a sorted list, allowing items to be inserted without the need to re-sort the entire list.
Read More »[Python] Array Bisection Algorithm bisect NoteHydra is an open-source Python framework designed to simplify the research and deployment process, especially for complex applications. Hydra dynamically creates hierarchical configuration files during deployment and allows command line-based overwriting of these configurations.
Read More »Note Of Hydra: Environment Configure Manager Package
Today, while developing a web application with React.js for the frontend and Python Flask for the backend, I pushed the project to my GitHub repository after reaching a satisfactory milestone. However, upon checking the repository, I was surprised to find that I couldn’t access the folder my-app created by npx create-react-app my-app.
I've always used rough formulas to estimate the relationship between the scale of my models and the GPU VRAM consumption; after all, there are too many variables involved—model architecture, number of layers, attention mechanism implementation, sequence length, batch size, data precision used in training or inference... all of these affect our final calculation results.
Read More »Note on Calculating VRAM Consumption for Training and Inference of AI Models
Today, while I was eating, I came across a video (the video is attached at the end of this article). Unlike many tech channels that jump straight into discussing AI, economics, and replacing humans, this video took a more careful approach. It explained in detail how hardware specifications have influenced algorithms (or AI model architectures) over time.
Read More »Here’s a thought: Will Transformers be replaced in the future?
We usually start various services on Linux servers, the most common being hosting a website or opening a port to allow us or users to test developing functionalities.
Read More »[Linux] How To Check The Port is Used?
I've been intermittently reading about a fine-tuning method called Kahneman-Tversky Optimization (KTO) from various sources like HuggingFace's official documents and other online materials. It's similar to DPO as a way to align models with human values, but KTO's data preparation format is much more convenient, so I'm quickly applying it to my current tasks before making time to study the detailed content in the related papers.
Read More »Note Of KTOTrainer (Kahneman-Tversky Optimization Trainer)
Podman is an open-source tool designed to manage containers and images. Its full name is Pod Manager tool (podman). While Podman is similar to Docker, there are a few key differences in its design.
Read More »[Linux] [Linux] Podman Basic Command Note
