Wang Xiaochuan's Baichuan intelligently released the Baichuan-53B model, proclaiming that "writing and text have reached the best level in the industry."

Thank you, Mr. Air, a netizen of CTOnews.com, for your clue delivery! CTOnews.com August 8 news, Wang Xiaochuan's Baichuan Intelligent today announced the release of large model product Baichuan-53B, officials said, "Baichuan-53B large model in writing, text creation ability, has reached the industry's best level."

However, CTOnews.com found that Baichuan-53B positioning "closed source large model", currently did not publish relevant information in GitHub, HuggingFace.

"Baichuan-53B mainly provides services to B-end users, and after the release, the big model will also start internal testing, and related components such as API are expected to be available next month," the official said.

Prior to this, Baichuan Intelligence has released two general language models, Baichuan-7B and Baichuan-13B, and interested friends can find relevant content in GitHub.

▲ Picture Source Baichuan Intelligent GitHub related pages according to official introduction, Baichuan-7B is a Chinese and English pre-training model with 7 billion parameters, which has been comprehensively evaluated on the three most influential Chinese evaluation benchmarks, C-Eval, AGIEval and Gaokao, and has achieved excellent results. It has become the best native pre-training model in Chinese under the same parameter scale.

Baichuan-13B is an open source and commercially available large-scale language model with 13 billion parameters developed by Baichuan Intelligence after Baichuan-7B. It achieves the best results in the same size model on both Chinese and English Benchmark.