Recently, I was learning about the localization deployment of DeepSeek-R1 and saw a more interesting video.
In this video, its main logic is that when the user uploads 1 Excel, the data in this Excel can be analyzed interactively. However, because the front-end technology of UP owners is not satisfactory, they actually skipped it directly. Normally, using Streamlit is a good choice.
However, he outputs the content to be analyzed to the terminal using a table. You can click on the detailed video."DeepSeek leads a new era of film data analysis"Check it out.
Enter the problem you want to analyze in the terminal, such as:
- Which movie has the highest box office
- What is the total box office in 2025
- What movies will be released in 2024
It seems that the effect is OK, but since it is based on multiple rounds of dialogues based on a single table, I don’t know how the effect is for the multi-table situation.
I feel that the big model will be widely used in 2025. I don’t know if junior data analysts will feel a little crisis if they see this, and they will lose their jobs every minute.