SD 1.6 is not a stable diffusion model. It's local software version of automatic 1111 that runs sd models.

SDXL is a base model, an upgraded version of SD1.5 model that use 1024x1024 standard images instead of 512x512. (anything less will result bad images)

realistic stock photo model (Or any other sdxl base model on Civitai) are an upgraded or more like an enhanced version of SDXL base model.

A lora is a small size addon to the base model. (It complements the base model)

First, you need to select an sdxl base model, like "realistic stock photo" model for example. (There are many base models available on Civitai, each has it's own mojo) Make sure base model is correct 6.46 GB file.

Stable diffusion 2.0\webui\models\Stable-diffusion

After selecting base model. Make sure you add lora file in lora folder in stable diffusion automatic 1111 directory.

Stable diffusion 2.0\webui\models\Lora

Then adjust settings, like image dimensions should be at minimum of 1024x1024 (for starters)

Sampling method, play with each till you figure out which one is the best when generating images.

Sampling steps differ from a standard base model to a turbo base model. (watch YouTube tutorials to know more)

CFG Scale is more of a finetune of the generated images, depending on base model, you may adjust it till you get the best results you desire.

Anything else would be image description in both positive and negative prompt.

Follow sample images with lora file as a startup guide.

Finally, a part from automatic 1111 there is another software that runs sd models named ComfyUI.

You can check YouTube tutorials on both to know and understand more about them both.

@hikaru056321 I don't have a YouTube channel, not at moment. I don't use Hires.fix. I use photoshop and or gigapixel when needed.