| localGPT can use the GPT model to chat on the local device, the data is run locally, and it is 100% confidential. This project was inspired by the original privateGPT. This model replaces the GPT4ALL model with the Vicuna-7B model, using InstructorEmbeddings instead of the LlamaEmbeddings used in the original privateGPT. Both Embeddings and LLM will run on the GPU instead of the CPU. If you don’t have a GPU, there is also CPU support (see note below). Use the power of LLM to ask questions about documents without an internet connection. 100… |
#GPT #model #local #deployment #localGPT