HKUST Xunfei AI model is too fierce!

In recent months, AIGC represented by ChatGPT has risen rapidly, and domestic top technology companies have also set off a hundred-model war. Among the many large-scale models, I am more optimistic about one: iFLYTEK.

I got an experience account through an old friend of HKUST iFLYTEK very early, and briefly experienced it. Let me talk about my specific feelings.
(I asked a friend of HKUST Xunfei Spark Model to apply for an exclusive application link : Click Apply to apply for experience through this link, and the review will be faster)

Due to work reasons, I often show Party A the PPT of the project, and sometimes it takes time to make PPT. Xunfei just has a PPT generation function, so I tried it with the Xunfei Xinghuo large model.

plug-in function

What I am most interested in is the new plug-in function this time. This is really a good thing. There is a 选择插件function after the new dialog of the Spark model

picture

You can choose plugins such as document question and answer, resume generation, PPT generation, or resume generation. Let me try these functions.

PPT generation

Sometimes the company leaders are in a hurry to ask for the project PPT. We have time to polish it slowly. We can use tools to help generate it in seconds. It must be generated first, and then we will slowly optimize and modify it later.

For example, if you receive a small project for a small program e-commerce, you need a PPT outline first, and directly input it to the Spark big model: "Help me generate a PPT for an e-commerce small program. It must have the basic functions of e-commerce, and also have distribution"

Then in less than 3 seconds, a PPT outline was generated for me, the core process was all there, and I added some special functions to it.

Let’s start Xunfei AI’s performance, blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah, download it yourself

Download it to your computer and enjoy it locally

Look at this outline is quite professional (the content is relaxed, with rich pictures and texts, and the layout is comfortable), when encountering more urgent needs, it is really nice to use it to automatically generate PPT first

resume generation

I find that many fresh graduates or those who have just worked for one year have ugly resume templates. I often ask me if I have a better resume template. Here I recommend another new function of iFlytek, resume generation. It is recommended that you input complete and detailed information, the better the effect of resume generation.

My input information below is still relatively simple: "My name is Zhang Pangpang, male, I have 1 year of front-end development internship experience, and I have developed an e-commerce applet project with technology vue react angular. Please help me generate a resume template."

Wait for a few seconds and it will be generated for me. The information items that should be included in the resume are very complete. Friends whose resumes are messy can try this function. The resume is the facade of the interview and must be neat and tidy.

Document Q&A

Finally, I tried 文档问答the function. After selecting it, an upload component popped up directly, allowing me to upload a document. I uploaded a document. The content in the document was written by me a long time ago. It was: "How to reduce overtime for programmers?", very It's a headache for programmers. I'm quite satisfied with the answer after looking at the answer.

multimodal functionality

Xunfei Xunhuo Cognitive Big Model V2.0 conference released the code and multi-modal capabilities. Here is a brief introduction to everyone. Multi-modal refers to information in multiple modes, including: text, image, video, audio etc., you can even broadly think that some items that can convey information are multimodal.

The new multi-modal function of HKUST Xunfei means that it already has functions such as image generation and image understanding. I quickly tried it, and I asked the Spark model to generate a beautiful girlfriend for me.

It turned out well, much better than I thought and it turned out to be a beautiful girlfriend.

Then I uploaded a photo of a panda eating bamboo to ask it, and it did pretty well too.

The Xinghuo model successfully recognized that it was a giant panda sitting on the grass and eating bamboo. Also added some remaining descriptive sentences "It has black and white fur and a round face, with a forest in the background and a tree in the distance."

Java code interpreter

When programmers look at new projects or learn the source code created by great masters, the code is very headache. Many masters don’t like to write comments, which causes a headache for Xiaobai. Let’s share another interesting function [Java code interpreter] , directly on the effect demonstration

The following is the generated comment, Xiaobai looks much more comfortable

The large model of Xunfei Xinghuo has many functions, I will briefly demonstrate a few, and there are many other functions that everyone can experience.

By the way, I asked a friend of HKUST Xunfei Spark Model to apply for an exclusive application link : Click Apply to apply for experience through this link, the review will be faster (5 minutes at the fastest, 4 hours at the slowest), if you have not experienced it yet If you have passed the Xinghuo model of iFLYTEK, then I strongly recommend applying to experience it. Applying through my exclusive link can also have a higher Xinghuo large model API test quota, which is 30% more than ordinary channel applications, and can apply for up to 500w Tokens .

Guess you like

Origin blog.csdn.net/Design407/article/details/132428425