Posts by Maxime Labonne

[ad_1] Create your own models easily, no GPU required!Image by authorModel merging is a technique that combines two or more LLMs into a single model. It’s a relatively new and…

[ad_1] Boost the performance of your supervised fine-tuned models10 min read·13 hours agoImage by authorPre-trained Large Language Models (LLMs) can only perform next-token prediction, making them unable to answer questions.…

[ad_1] Quantize and run EXL2 modelsImage by authorQuantizing Large Language Models (LLMs) is the most popular approach to reduce the size of these models and speed up inference. Among these…

[ad_1] The tokenizer, Byte-Pair Encoding in this instance, translates each token in the input text into a corresponding token ID. Then, GPT-2 uses these token IDs as input and tries…