Due to my busy work schedule recently, I haven’t shared anything about artificial intelligence with you for a long time. But recently, the AI field has been really exciting. Everyone should be aware of the popular ChatGPT, which is a natural language generation model that has attracted widespread attention. In addition, GPT-4 also undergoes rapid iterative updates. At the same time, many interactive chat language models have emerged, some open source and some not. In addition, SAM (Segment Anything Model) is a just-released image segmentation model that can help you segment any object you want from a picture without the need for samples. This has had a major impact on computer vision and has a huge impact on all walks of life.
Many companies are actively utilizing artificial intelligence technology to improve productivity and business innovation. For example, the Bing search engine has been integrated with GPT-4 to make it more intelligent. At the same time, Github's Copilot X is also connected to GPT-4, which can assist you in writing code and improve coding efficiency. Even our company organized an AIGC competition, hoping to use artificial intelligence technology to achieve business breakthroughs. For wage earners, such changes can be unsettling. However, if we can maintain a proactive attitude and continuously improve our skills and competitiveness, we will be able to meet the challenges of the future.
In addition, I have previously shared a content called Whisper model and made a small tool using it. Recently I made some modifications to its interface and added a file language recognition function, as shown in Figure 1-1 below. At the same time, I also want to try some new open source interesting models, such as VITS, which is an acoustic model that can train a more natural vocal model based on speech data sets. Combining VITS with Whisper and ChatGPT may produce some interesting results, and I will share the results of these experiments with you in future articles.
1-1
Finally, since many students have asked some questions about the whisper gadget before, I will say more here. If you want to use these models in GPU mode, you need to install a PyTorch version that matches the CUDA driver version of your computer. At the same time, since there may be missing modules in requirements.txt, you can install the missing modules yourself according to the prompts. I will also share the updated version of the application below, which you can also download and use directly. I hope you can try it and enjoy the process of using it. If you have any questions, please contact me. I look forward to seeing you next time.
Link:
https://pan.baidu.com/s/1nj174s3tyHUAsazi4tHW0g
Extraction code: 3ugi
Code gihub address:
https://github.com/PythonerLeague/PythonerLeague_translate