{"id":107246,"date":"2023-03-13T18:48:30","date_gmt":"2023-03-13T16:48:30","guid":{"rendered":"https:\/\/prohoster.info\/?p=107246"},"modified":"2023-03-14T14:08:05","modified_gmt":"2023-03-14T12:08:05","slug":"opublikovan-openchatkit-instrumentarij-dlya-sozdaniya-chatbotov","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/opublikovan-openchatkit-instrumentarij-dlya-sozdaniya-chatbotov","title":{"rendered":"OpenChatKit has been released, a toolkit for creating chatbots.","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
The OpenChatKit toolkit has been introduced, aimed at simplifying the creation of chatbots for specialized and general use. The system is adapted to perform tasks such as answering questions, conducting multi-stage dialogues, summarizing, extracting information, and text classification. The code is written in Python and distributed under the Apache 2.0 license. The project includes a ready-made model, code for training your own model, utilities for testing the model's performance, tools for supplementing the model with context from external indexes, and adapting the base model to solve your own tasks. <\/p>\n
The foundation of the bot is a basic machine learning model (GPT-NeoXT-Chat-Base-20B), built using a language model that encompasses around 20 billion parameters and optimized for conversational interactions. The model was trained using data obtained from the collections of LAION, Together, and Ontocord.ai projects. <\/p>\n
To expand the existing knowledge base, a system is proposed that can extract additional information from external repositories, APIs, and other sources. For example, updating information using data from Wikipedia and news feeds is possible. Additionally, a moderation model is available, trained using 6 billion parameters, based on the GPT-JT model and designed to filter inappropriate questions or limit discussions within specific topics. <\/p>\n
Notably, the ChatLLaMA project offers a library for creating intelligent assistants similar to ChatGPT. The project is being developed with an eye toward the capability to run on one's own hardware and create personalized solutions aimed at covering niche areas of knowledge (such as medicine, law, gaming, scientific research, etc.). The ChatLLaMA code is released under the GPLv3 license. <\/p>\n
The project supports the use of models based on the LLaMA (Large Language Model Meta AI) architecture proposed by Meta. The full LLaMA model encompasses 65 billion parameters, but for ChatLLaMA, it is recommended to use versions with 7 and 13 billion parameters or the GPTJ (6 billion), GPTNeoX (1.3 billion), 20BOPT (13 billion), BLOOM (7.1 billion), and Galactica (6.7 billion) models. Initially, LLaMA models are available only to researchers upon special request, but since torrents were used for data delivery, enthusiasts prepared a script that allows anyone to download the model. \n Source: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"