Hershel Don

Update and rescale CFG denoising step

2

Update as described by: https://twitter.com/jeremyphoward/status/1584667278733324288 https://twitter.com/jeremyphoward/status/1585009792787304448 This requires using X/Y plot and small CFG scale (i.e. 0.15 work well) until the GUI slider parameters are modified. The new CFG update...

Lora training step time >2x increase after checkpoint

4

' Python 3.9.13 (tags/v3.9.13:6de2ca5, May 17 2022, 16:36:42) [MSC v.1929 64 bit (AMD64)] Commit hash: 685f9631b56ff8bd43bce24ff5ce0f9a0e9af490 Installing requirements for Web UI [auto-sd-paint-ext] Attempting auto-update... [auto-sd-paint-ext] Fetch upstream. [auto-sd-paint-ext] Pull upstream....

Save between epochs is broken

2

Only getting a save and sample after the first epoch. For the rest of the epochs it only pauses, no save or sample. This applies for dreambooth and lora I...

This attempted fix results in the following error. Putting it here in case someone can see the problem with inject_trainable_lora. Any help appreciated, thanks! Traceback (most recent call last): File...