@playtime_ai already had a Wan 2.1 Wednesday, but I wanted one for 2.2, so I trained one. Then they released theirs yesterday, which excited me because they are an awesome creator and I wanted to compare my outputs to theirs to see how my LoRAs perform in competition under pressure.

I will be uploading more of these with diverse prompts - these were preliminary and all used the same prompt (one of theirs).

Anyway, I'm quite pleased with the results. Both models seem highly capable and you might be hard pressed to sort outputs from the two.

It's notable, however, that my LoRA is only 150mb. And it is only one file that you can use in high and low. For Playtime's, you need two separate 300mb files to achieve the same results.

Playtime_AI's Wednesday is here: https://civitai.com/models/1768588/wednesday-addams

and mine is here: https://civitai.com/models/2050635/wednesday-addams

I will upload the comparison workflow if anyone wants it.

[EDIT: after more generations, it seems clear that without prompting for her appearance, Playtime's LoRA is not terribly strong, producing bland generic girls and even men and boys. So... lots of the content of those gens using those his prompts are in the base model. My LoRA just produces Wednesday with nothing other than "wednesday" for the prompt.]