GitHub - Lightning-AI/litgpt: Pretrain, finetune, deploy 20+ LLMs on your own data. Uses state-of-the-art techniques: flash attention, FSDP, 4-bit, LoRA, and more.

Fine-Tuning for LLM Research by AI Hero

This repo contains the code that will be run inside the container. Alternatively, this code can also be run natively. The container is built and pushed to the repo using Github actions (see below). You can launch the fine tuning job using the examples in the https://github.com/ai-hero/llm-research-examples pr... See more

2-5x faster 50% less memory local LLM finetuning

• Manual autograd engine - hand derived backprop steps.


• 2x to 5x faster than QLoRA. 50% less memory usage.


• All kernels written in OpenAI's Triton language.


• 0% loss in accuracy - no approximation methods - all exact.


• No change of hardware necessary. Supports NVIDIA GPUs since 2018+. Minimum CUDA Compute Cap

... See more