Using PeFT (Parameter efficient Fine Tuning) and the larger Google Gemma 7B model to generate a training set to customize the Gemma 2B model

[obriensystems]

Discovery

see https://github.com/ObrienlabsDev/blog/wiki/Lab:-DLIW62596-Efficient-Large-Language-Model-(LLM)-Customization-%E2%80%90-NVidia-Deep-Learning-Institute-at-GTC24-San-Jose
https://www.mercity.ai/blog-post/fine-tuning-llms-using-peft-and-lora

• LoRA has better accuracy over P-Tuning (but more GPU compute required)
• why PEFT over prompt engineering - (less token size)
• understand LoRA https://github.com/microsoft/LoRA
• PEFT - Parameter Efficient Fine Tuning - https://github.com/huggingface/peft
• NEMO
• RAG - DLI lab is
• 8h https://courses.nvidia.com/courses/course-v1:DLI+S-FX-15+V1/
• 1h https://courses.nvidia.com/courses/course-v1:NVIDIA+S-FX-16+v1/