Increase the parameters for training LoRas

Currently, the restrictions on creating LoRas for Laura's using Civic AI I find too restrictive, 10,000 steps is quite small along with flux devs Max batch size on images larger than 512 being limited to two has really put a dampening in my LoRas capabilities and I'm already having to deal with the restrictions that have been placed on us before that such as the lack of editable settings despite using kohys a backend, I believe that if I have enough buzz to pay for certain parameters then I should be allowed to do that. As of now I can't imagine I'd be able to make many desirable LoRa's and a lot of my production will need to be cut down for Flux Dev.

Would like to return to civet ai as a place to create flux dev LoRa’s but as of now I can't imagine I'd be doing much of any.