Author(s): Davide Gazzè – Ph.D. Originally published on Towards AI. A fast insight into this fascinating projectPhoto by Shubham Dhage on Unsplash Recently, there have been many articles about ChatGPT and GPT4 (some of mine are [3] and [4]). One of the drawbacks of these models is the necessity to perform a remote call to an API. I would use an LLM model, also with lower performance, but in your local machine. In the last few days, Google presented Gemini Nano that goes in this direction. Another initiative is GPT4All. GPT4All is an open-source software ecosystem created by Nomic AI that allows anyone to train and deploy large language models (LLMs) on everyday hardware. GPT4All is compatible with the following Transformer architecture model: Falcon;LLaMA (including OpenLLaMA);MPT (including Replit);GPT-J. For more details, refer to the technical reports for GPT4All [5] and GPT4All-J [6]. Similar to ChatGPT, these models can do: Answer questions about the worldPersonal Writing AssistantUnderstand documents (summarization, question answering)Writing code Moreover, the website offers much documentation for inference or training. In this post, I use GPT4ALL via Python. To install the package type: pip install gpt4all After the installation, we can use the following snippet to see all the models available: from gpt4all import GPT4AllGPT4All.list_models() The output is the: [{'md5sum': '81a09a0ddf89690372fc296ff7f625af', 'filename': 'ggml-gpt4all-j-v1.3-groovy.bin', 'filesize': '3785248281', 'isDefault': 'true', 'bestGPTJ': 'true', 'description': 'Current best commercially licensable model based on GPT-J… Read the full blog for free on Medium. Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor. Published via Towards AI
↧