Product Design | Modern Product Design Style SD1.5-V1

- As a product design and industrial design practitioner, I waited all year for 2023 without waiting for a product design Lora model that satisfied my needs. (By the way, there's no relevant category even on Lib.)

So I decided to create a Lora model myself to meet the needs of my usual solution exploration.

- After more than half a year of learning, testing, and trial and error, I finally came up with this version. It classifies products with different temperament

styles, mainly to suit a sufficiently wide range of product categories.

- I marked more than 200 selected images using LEOSAM, the GPT4V Image Captioner image marking processing tool and the Little Cross Fish Boss's one-click package made by Mr. Rabbit.

I've tried N types of alchemy optimizers and different configurations, and currently there's a version that's quite usable. Although there are still some issues, it's already in a usable state. I'm sharing this version with you today, and I hope

you get more feedback.

- This is a more conservative version, intended to be slightly more versatile.

- Regarding the plotting settings, I haven't fully tested it yet, but what I want to say is that the purpose of industrial design of Lora is to generate enough solution proliferation without first considering the level of detail of the plot. Therefore, when Lora is combined with different large models and different samplers, it will generate different design styles and design methods, which brings great value to our solution expansion. At least, I'm not stuck with one parameter

.

Currently, the weight range I've tested is fine between 1.0 and 0.6, and it works well with some common large models.

- The product is a very large category. Currently, the amount of data in this model is still very small. If a certain type of item is specified to be generated, it is likely that the fit phenomenon has occurred. It is recommended to use it

with Controlnet or lower the weight.

- I'm still preparing a larger training set to prepare big models for industrial design.