[Hanwang Science and Technology: the first internal test version of the professional model being developed is expected to be launched in early 2024] Hanwang Science and Technology said in a survey that the company is developing a professional model with moderate scale parameters and moderate power. It is a professional model for specific knowledge data training. The model avoids the hundreds of billions of parameters and violent calculation needed by the general model, and the construction goal is "units can be deployed, data is privatized, capabilities are specialized, wiser and wiser, and investment is acceptable." there are some unknown risks in the research of the model. It may not be as expected, and the first internal test version is expected to be launched in early 2024. In terms of voice technology, the company combines the results of public papers and open source models to carry out research and training on speech synthesis and generation, which has made good progress and is expected to be applied to the company's related hardware products in half a year.