Hi @noble frigate. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
#Best current general-purpose model
stuck flintBOT
The GPT4All platform supports a variety of models that you can use for chatting. The best model for you would depend on your specific needs and resources. However, the platform is optimized to run large language models (LLMs) in the 3-13B parameter range on consumer-grade hardware, which includes your 24-core CPU and AMD GPU.
The GPT4All Chat UI supports models from all newer versions of llama.cpp with GGUF models including the Mistral, LLaMA2, LLaMA, OpenLLaMA, Falcon, MPT, Replit, Starcoder, and Bert architectures. You can find an exhaustive list of supported models on the GPT4All website (https://gpt4all.io) or in the models directory (https://raw.githubusercontent.com/nomic-ai/gpt4all/main/gpt4all-chat/metadata/models3.json).
If you want to use a model that is not listed, you can sideload it into GPT4All Chat if it is compatible with the gpt4all-backend. The process involves downloading your model in GGUF format, identifying your GPT4All model downloads folder, placing your downloaded model inside GPT4All's model downloads folder, and restarting your GPT4ALL app. Your model should then appear in the model selection list.
Please note that while GPT4All is optimized to run on consumer-grade hardware, some larger models may still require more RAM. Therefore, you should consider the resources of your machine when choosing a model.
Sources:
- GPT4All Documentation > Old > GPT4All Chat UI (https://docs.gpt4all.io/old/gpt4all_chat.html#gpt4all-chat-ui)
- GPT4All Documentation > Gpt4all_desktop > Models (https://docs.gpt4all.io/gpt4all_desktop/models.html#models)
- GPT4All Documentation > Old > GPT4All > Models (https://docs.gpt4all.io/old/index.html#models)