A minor enhancements to textual inversion ui

[patrickas]

• User can specify {name} in the preview prompt to use the name of the embedding

• User can specify multiple preview prompts which will create multiple preview images xxx_1.png xxx_2.png This is useful useful to make sure embedding works for multiple types of photos like a photo of {name}|a painting of {name} by renoir

[dfaker]

last_saved_image is pulling a few duties here, first it's a flag for not having saved, then it's a filename, then it's a list of concatenated prompts.

Nicer to a new variable for your list of prompts used and images generated and then format them clearly as what they are in the output html.

[patrickas]

@dfaker last_saved_image is already doing all that in the current code... Notice in my change i use last_saved_image just onece to create a new variable called current_file_name the rest of its duties are already being done in the current code so my changes do not touch that

[1blackbar]

Can You add ability to load yaml file with settings ? Current implementation is kinda broken and wont get same results as main repos... https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/2784

Having actual original code would fix this.Maybe a window to paste a path to original yaml file?

[AUTOMATIC1111]

I'd prefer to keep things simple.