Stepfunction
Stepfunction
sd-scripts has added support for multi-resolution training with latent caching as documented here: https://github.com/kohya-ss/sd-scripts/tree/sd3?tab=readme-ov-file#flux1-multi-resolution-training Would it be possible to augment the current Max Resolution argument to provide a range of...
I think this would be a good place to discuss finetuning the new Flex.1 Alpha model created by Ostris: https://huggingface.co/ostris/Flex.1-alpha Initial tests I've tried on training LoRAs using ai-toolkit are...
As in the title, if you select a Lycoris LoRA type, enter a max grad norm value and then switch to a non-Lycoris LoRA type (I'm using Flux), the max...
### 🚀 The feature, motivation and pitch Currently, PagedAttention only supports specific head_size values. This prevents models like Magistral 2509 (with a head_size of 160) from running. vLLM resolves this...