![xAI also said that the model was trained in October 2023 [File]](https://th-i.thgim.com/public/incoming/1gz6qm/article67963387.ece/alternates/LANDSCAPE_1200/2024-03-11T170932Z_1053685659_RC2SJ6AI759W_RTRMADP_3_MUSK-AI.JPG "xAI also said that the model was trained in October 2023 [File]")
Elon Musk’s AI company xAI has released the weights and architecture of its large language model Grok-1, per an official blog post on the xAI website on Sunday.
“Grok-1 is a 314 billion parameter Mixture-of-Experts model trained from scratch by xAI. This is the raw base model checkpoint from the Grok-1 pre-training phase, which concluded in October 2023. This means that the model is not fine-tuned for any specific application, such as dialogue,” noted xAI.
The company added that the base model had been trained on a “large amount of text data, not fine-tuned for any particular task” and shared a GitHub link so that others could use the model.
xAI also said that the model was trained in October 2023.
(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)
Musk had promised last week to make Grok open source. The move comes as he filed a lawsuit against OpenAI for allegedly deviating from its founding values. The ChatGPT-maker labelled the allegations “incoherent.”
The Tesla billionaire has repeatedly criticised OpenAI, claiming that it does not transparently share its tech for the betterment of humanity. However, OpenAI pointed out that its basic chatbot is free and that millions are using its services and products worldwide.
