Junkun Chen

DreamBooth: Not applying accelerator.accumulate on text_encoder

4

### Describe the bug At [examples/dreambooth/train_dreambooth_lora_sdxl.py#L1618](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/train_dreambooth_lora_sdxl.py#L1618), only `unet` is applied with `accelerator.accumulate()`. ```python for step, batch in enumerate(train_dataloader): with accelerator.accumulate(unet): # HERE pixel_values = batch["pixel_values"].to(dtype=vae.dtype) prompts = batch["prompts"] ``` However,...