Fixes for `git checkout`, MPS/macOS fixes and optimizations

Fix typo in launch_utils.py

existance -> existence

add new callback for scripts to be used before processing

For MPS, using a tensor created with `torch.empty()` can cause `torch.baddbmm()` to include NaNs in the tensor it returns, even though `beta=0`. However, with a tensor of shape [1,1,1], there should be a negligible performance difference between `torch.empty()` and `torch.zeros()` anyway, so it's better to just use `torch.zeros()` for this and avoid unnecessarily creating issues.

Even if this causes chunks to be much smaller, performance isn't significantly impacted. This will usually reduce memory usage but should also help with poor performance when free memory is low.

Support search and display of hashes for all extra network items

Clear sampler and garbage collect before decoding images to reduce VRAM

Encode batch items individually to significantly reduce VRAM

Update description of eta setting

Add `s_noise` param to more samplers

Restore `extra_params` that was lost in merge

Lora: output warnings in UI rather than fail for unfitting loras; switch to logging for error output in console

More significant VRAM reduction.

Significantly reduces VRAM.
This makes encoding more inline with how decoding currently functions.

Add Norm Module to lora ext and add "bias" support

Bring back CSV mode for XYZ grid

write out correct model name in infotext, rather than the refiner model