Chatgpt

You are a Douyin short video operations expert, especially skilled in the creation and optimization of copy for the openings of short videos. Next, I need you to help me write some opening sections for short videos. The following are several commonly used open... Read more

CriticGPT is a GPT-4-based model that critiques ChatGPT’s responses to help human trainers spot errors during RLHF. The current error problem OpenAI trained a GPT-4-based model called CriticGPT to find errors in ChatGPT’s code output. OpenAI found that when ... Read more

When preparing a local knowledge base, we usually deal with question-and-answer pairs and multi-turn dialogues, especially in customer service systems. We want RAG to handle rigorous one-question-one-answer interactions, while also hoping it can chat better an... Read more

When it comes to customer service chatbots, many people immediately think of fine-tuning large models, and then after getting started they find that it cannot achieve the expected chat experience. This is a misunderstanding caused by a lack of understanding of... Read more

ChatGPT and similar AI tools are still far from widespread adoption. For beginners, the impression of ChatGPT is vague; beginners generally think ChatGPT is just a chatbot, and this understanding is very misleading. If you really think it is just chatting, it ... Read more

When it comes to handling data input and output, we have 3 methods: prompt engineering (prompt), RAG knowledge base & plugins, and large model fine-tuning. They have different roles. When developing an application that fits actual business needs, it is often n... Read more

There are three ways to handle how large models answer specific content: prompt engineering, RAG, and fine-tuning. In many cases, when we face situations where prompt engineering cannot achieve the required answer logic, we think of fine-tuning the large model... Read more

I regret to tell you that your money-saving plan is going to fall through—you will almost certainly need to use it. We often use the web interface to explore, learn, and build workflows, and then use the API to implement automation. Specifically, we usually e... Read more

They both perform the same type of vector retrieval on the provided data. The Assistant can be called externally via the API and is billed by usage; GPT is limited to 50 messages/3 hours but requires paying for a Plus subscription. GPTS is simply an easy-to-us... Read more

ChatGPT's ability in summarization is outstanding. At work, summaries and meeting minutes are usually repetitive tasks that consume a lot of your working time. You can absolutely act as the commander and let ChatGPT be your assistant to complete these tasks. P... Read more

1. Writing Code We’ve all been there: sitting at the computer at 2 a.m., stuck on a piece of code. Or you just can’t seem to recall that particular syntax, and no matter how hard you try, Google searches don’t help. That’s where ChatGPT comes in. It can wr... Read more

When fine-tuning models using OpenAI's API, choosing the correct base model is crucial. Below is an explanation of the available models and some suggestions on how to choose a model that suits your specific needs. Models available for fine-tuning OpenAI provid... Read more

ChatGPT is an AI technology that can solve complex tasks. It is an NLP-based large natural language model, so chat-style commands are how it is used, but that does not mean it is a chatbot. ChatGPT's capability is content output and screening—output and scree... Read more

Recharging an API is indeed a rather troublesome issue, and few bloggers share this kind of content, mainly because recharge methods are unstable. After all, it involves money, and if the recharge fails, they will be blamed. It is truly a thankless task. The m... Read more

https://www.bilibili.com/video/BV1pM4m1r7hW/ Disclaimer: This content is for study and research only. No tools are provided, and no services are provided—please don't ask. The overall workflow is: crawl information from self-media such as videos and images ac... Read more

This livestream lasted 20 minutes and mainly introduced several new features of GPT-4o, including real-time voice conversation, real-time video conversation, and sentiment analysis, which gives ChatGPT eyes, ears, a mouth, and emotional intelligence. After the... Read more

GPTs are not equivalent to fine-tuned large models; the gap is huge. Although both are trained in the form of a knowledge base and output content, there are huge differences in support and functionality. Now we will introduce the differences between the two in... Read more

Necessity of Classification and Integration After the data collection process ends, the data must be organized. This includes two parts: classifying the data and integrating it. The reason for doing this is to facilitate management, as well as to address issue... Read more

The term API is not unfamiliar. Simply put, an API is a special string key made up of letters, numbers, and some symbols, for example yzwfwFefwf321joOjolj-jFjjo. After connecting to your program through an API string, you can use the functions behind the API. ... Read more

The role of the gpt-4-vision-preview model is image recognition. For video, a video is composed of multiple frames, and each frame is a picture, so as long as we extract the pictures from these frames at the right time and send them to the API, and give it a c... Read more