This version is much better than previous flux version. Chances of extra finger or missng one are reduced. However, you may still experience some weirdness occasionally.
You can try model with or without trigger word(s) and see yourself which works better for you.
In addition, try starting with low lora weight. Start with 0.1 and work you way up till 1.
You'll notice some generated images are much better with low lora weight while others are great with higher weight. It's tricky, but you'll need to adjust the weight till you reach best spot.
This version is much better than previous flux version. Chances of extra finger or missng one are reduced. However, you may still experience some weirdness occasionally.
You can try model with or without trigger word(s) and see yourself which works better for you.
In addition, try starting with low lora weight. Start with 0.1 and work you way up till 1.
You'll notice some generated images are much better with low lora weight while others are great with higher weight. It's tricky, but you'll need to adjust the weight till you reach best spot.