V2 has same triggers as V1.
Dataset size: 40 (V1) -> 404 (V2), V2 should be more flexible and less overstyled at same level consistency (hopefully)
Please comment if you feel V2 is generally worse than V1 in your use cases as I'm remaking some of my previous gesture LoRAs, feedback is always welcome.
Trigger: finger heart
Trigger (double hands): double finger heart
You may see missing digits in some cases.
It's a requested LoRA trained on PonyXL and tested on AutismMix and Momoiro, following Korean style finger heart pose.