GitLab

Fix issue with not loading the SSD1B clip correctly.

Use fp16 text encoder weights for CPU inference to lower memory usage.

Use a simple CLIP model implementation instead of the one from
transformers.

This will allow some interesting things that would too hackish to implement
using the transformers implementation.

More generic clip model class that can be used on more types of text
encoders.

Don't apply weighting algorithm when weight is 1.0

Don't compute an empty token output when it's not needed.

Do inference in fp32 to make sure quality stays the exact same.

load_model_gpu() is now used with the text encoder models instead of just
the unet.

This is faster on big text encoder models than running it on the CPU.

Update LoadClip to load clip models for SDXL refiner.

Large refactor of the model detection and loading code.

Make the CLIP model work on the GPU.