HomeAI News
【Beginner Tutorial】Basic commands of Midjourney

【Beginner Tutorial】Basic commands of Midjourney

March 15th, 2023
View OriginalTranslated by Google

▣ Upload your own pictures, AI training ①Click the + sign in the input box, there will be an upload file, double-click it, you can select the picture to upload, and input. ②Right click on the picture and select copy link. ③ Paste it directly into the input box after imagining the prompt. ④ After pasting, leave a space (*important), and then enter text (commas can be used to segment) and parameters. ⑤Filled graphs are not limited to only one graph, but can also be mixed with more than 2 graphs (remember to add spaces between graphs) ⑥Keyword hints, what, who, when, where, how, 5W. NSFW (Not Safe For Work), the words mentioned in this part, will be rejected by AI. ▣Basic instructions /imagine The most basic drawing commands. Enter keywords after imagine, and it can be drawn by AI. /help The robot will push some guides and basic help information. /subscribe Enters a subscribed channel. Basic Membership ($10/month). 200 minutes of GPU usage per month (fast mode). Standard membership ($30/month). 15 hours of GPU usage per month (fast mode) with unlimited access to relaxed mode. /fast Express mode generates images. New users can generate 25 pictures; basic members can generate about 200 pictures/month; standard members can generate about 900 pictures/month. GPU time must be purchased separately thereafter. /relax Relax mode generates images, only available for Standard members and above. The usage time is not limited, but the GPU queue time must be shared with other members, and a maximum of 3 tasks can be executed. Upgrading to maximum functionality is not provided. /private Private creations will not be placed in public spaces. Only available to Private Creation members for an additional $20! Don't click after typing, clicks will pay. /public This is a predefined value. When you accept the terms of use, you are informed that all works are public. You can also refer to other people's creations. /show This is followed by the Job_ID, which is used to reproduce the painting. You can only enter your own creations, not someone else's Job_ID. ✉️:envelope:envelope symbol will send you the image's seed and Job_ID information. You can find messages sent to you by bots in your personal private messages. ⭐️:star: The star symbol will mark the picture as "favorite". Send the picture to the public #favorites channel and bookmark it on the official website. ❌ The :x:cross symbol will cancel or delete this picture message and also delete the picture from the official website's gallery. /settings sets the first line of the current preferences. MJ version, select the currently used version. Currently v4 is a beta and only produces 1:1 image second row. Test/Test Photo: Use generic testability algorithms, or photo-authenticity enhancement algorithms, adding more detail than the default. Quality level: Select the preset quality level of the image, which is 0.5/1 (default)/2 in turn. The third row. Style level: Select the preset style strength of the image, in order of 1250/2500 (default)/5000/20000, the larger the number, the more you can let yourself go. fourth row. Upscaling Level: Select the detail level of the preset image: Normal (default)/Highlight/Test, followed by Normal (default)/Highlight/Test. The fifth line. Public mode/Private mode: public mode and private mode. Remix mode: mix and match mode. When this mode is enabled, the keyword modification function will be activated when V is pressed. Fast mode/Relax mode. Quick mode and infinite mode. ▣Common parameters --beta. Use beta testing algorithms. --hd: Use a higher resolution algorithm. --aspect/-ar: Set the aspect ratio of the image. For example: --ar 16:9. --w --h. Set the width and height of the image. For example: --w 1024 --h 768. --seed. Set the seed. ---sameseed: Output images in a seed-like composition. --no: Exclude an object. For example: --- does not include people. --iw: Set the weight of the reference image, the larger the value, the closer to the reference image. (default 0.25, max 5). --stylize / --s: Set the style strength, the higher the value, the more common, can be set to any integer between 625 and 60000. --quality / --q: Set the quality of the image, the higher the value, the slower the image time, the lower the value, the faster the image comes out, the more money you can save. There are five settings to choose from (0.25, 0.5, 1, 2, 5). --chaos / --c: Set the randomness of the image, the larger the value, the greater the randomness (0-100). --fast. Output images quickly and save GPU usage time. --stop. At which percentage (10-100) the image will stop when calculating the image. For example: --stop 80. --video: Save a video of an algorithm. Add the reaction envelope symbol and the video will be sent to you privately. --v: Which version of the algorithm to use. For example: --v 4.


no dataCoffee time! Feel free to comment