In today's era of surging artificial intelligence, having a platform that integrates a variety of mainstream artificial intelligence services will undoubtedly greatly improve our work and creative efficiency. Today I will introduce to you a treasure project on GitHub. With it, you can easily create a "super mall" of AI applications and use a large number of large model products such as ChatGPT, Midjourney, Suno, etc., and as time goes by, the AI applications it supports will continue to expand.
To build this project, you first need a server, which can be a local computer or a cloud server. But from the perspective of speed and stability, cloud servers, especially overseas nodes, have more advantages. Here, I have to mention Amazon cloud technology. Newly registered users can enjoy free use of cloud servers for a certain period of time, and many cloud services even provide lifetime free rights.
The registration process is as follows: 1. Select an overseas account : For individuals, overseas account restrictions are relatively loose. 2. Fill in your email : Enter your email and get the verification code. 3. Complete information : Fill in the necessary information and provide a credit card with the UnionPay or Visa logo. Registration is now complete.
After registration is completed, start the server configuration journey. This time we tried using the LightCell server (the EC2 demo commonly used in past blogs and videos). 1. Log in to the console : Log in to the account you just registered and click on the console. 2. Create an instance : Search for "LightCell" in the console, and then create an instance. During this period, you can switch regional system types and pre-installed environments according to your needs. If you are unsure, you can keep the default settings. Then select an instance specification (different specifications have different fees), name the instance and complete creation. Just keep the defaults in subsequent steps, and finally download the certificate for local SSH login.
The server is ready. Next, configure the project environment. It is recommended that you download the OnePanel panel software. Its installation is very simple and can be done with just one command. After the installation is completed, OnePanel access information will be output. Be sure to find a place to record it to prevent subsequent logins from being blocked. After that, you need to release OnePanel's port and return to the console for simple configuration. Open a new browser window and enter the OnePanel address. If the corresponding interface appears, the installation is successful. The OnePanel panel tool plays a great role in subsequent server operation and maintenance monitoring, so I won’t go into details here.
After the environment is set up, go to the GitHub project address and copy the Docker command into the text editor. The content after each "E" is the operating parameter. You need to replace the content after the equal sign according to the AI service you need, such as ChatGPT, Midjourney, Suno, etc. In order to facilitate the centralized management and use of these AI services, it is recommended to use the transfer service website recommended by the project author. Log in to the website and recharge ten yuan for the first time to experience. Click "Get Key" and copy the key to replace all "xxx_key" values in the Docker command. At the same time, copy the service address of the transit website and replace the corresponding address in the Docker parameters. After the modification is completed, execute this command on the console and wait a moment for the project to start successfully. After starting, don't forget to release the 6015 port mapped by Docker, and then enter "IP + port" in the browser to access it. If you can see the corresponding interface, congratulations, you are done!
Enter the interface and first experience the AI conversation function. Its interaction with ChatGPT is basically the same, with extremely fast response to questions and model selection. Also supports GBDS function. However, it should be noted that some models have higher token fees, so you need to proceed according to your ability.
Then try to use AI to generate pictures. After entering the prompt word and waiting for a period of time, the AI will return the image. And the interaction here is the same as Midjourney’s native operation on Discord, making it very smooth to use.
As for functions such as Suno's song creation and Runway's video generation, due to limited time, this article will not demonstrate them. You can explore them on your own.
In this way, our own AI tool collection website is completed and deployed online. If you encounter any problems during the operation, please leave a message in the comment area and I will answer them one by one. The URLs involved in the video, including the GitHub project address, will be placed in the pinned comment. I hope you can use this toolset to start your own amazing AI journey!
Share on Twitter Share on Facebook
Indie123
Comments
There are currently no comments