Anchen·Apr 27, 2024Fine-tune Llama3 with function calling via MLX-LMWith the recent release of Llama 3, Meta has delivered a game-changing open-source model that combines impressive performance with a…A response icon3A response icon3
Anchen·Jan 25, 2024Qlora Mixtral 8x7B via mlxMixture of Experts (MoE) models, the latest in cutting-edge LLM model, Their unique architecture and gating mechanism necessitate more…
Anchen·Dec 22, 2023Running LLM with apple’s MLX frameworkMLX is a new ML framework for machine learning on Apple Silicon that was recently released. Considering that Apple Silicon devices…
Anchen·Oct 14, 2023Replace Grammarly with open source LLMI have been using the self-hosted llama 30B model for local grammar check and translation because most of the smaller llama models are not…A response icon5A response icon5
Anchen·Sep 14, 2023Host your own ChatGPT alternativeIn my previous posts, we fine-tuned some of the open-source LLM models and most of the time we just use text-generation UI or handcrafted…
Anchen·Sep 6, 2023Fine-Tuning CodeLlama 34B for ChatWhile Meta’s Llama2 captured significant attention in the AI landscape, however the 34b model was notably absent for quite a while. For…
Anchen·Aug 13, 2023Fine-tune Llama 2 with SFT and DPOIn my previous article, we discussed how to fine-tune the LLAMA model using Qlora script. However, with the latest release of the LLAMA 2…A response icon2A response icon2
Anchen·Jun 19, 2023Build your own copliot using open source LLMI first became aware of LLM through the GitHub Copilot release. I was impressed by how it improves developer productivity and always…A response icon1A response icon1
Anchen·Jun 10, 2023Fine-tune the llama model for your application use caseWhen developing an LLM related application, often, you may find that the model output doesn’t strictly follow the instructions, especially…
Anchen·Jun 1, 2023How to fine-tune an open-source LLaMa using QLoRaAs a developer who often turns to ChatGPT as my trusty pair-programming companion, we end up engaging in many detailed discussions on…A response icon1A response icon1