images -> animation

It’s about time to move into the planning phase.

Ideally, ’d like to be able to add voiceovers as well.

This is going to be a much more difficult task than had anticipated.

With a configuration featuring an RTX 5090 (32GB VRAM) and 128GB of RAM, it is worth attempting LoRA training for Wan 2.2.

However, training a 14B model imposes strict VRAM constraints, making settings like quantization and offloading crucial.

Personally, would consider something like Musubi Tuner as the top candidate.