Merge branch 'main' into support-comyui-flux-loras

Author: sayakpaul

docs/source/en/api/pipelines/wan.md

@@ -14,6 +14,10 @@
 
 # Wan
 
+<div class="flex flex-wrap space-x-1">
+  <img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
+</div>
+
 [Wan 2.1](https://github.com/Wan-Video/Wan2.1) by the Alibaba Wan Team.
 
 <!-- TODO(aryan): update abstract once paper is out -->

docs/source/en/quantization/torchao.md

@@ -126,7 +126,7 @@ image = pipe(prompt, num_inference_steps=30, guidance_scale=7.0).images[0]
 image.save("output.png")
 ```
 
-Some quantization methods, such as `uint4wo`, cannot be loaded directly and may result in an `UnpicklingError` when trying to load the models, but work as expected when saving them. In order to work around this, one can load the state dict manually into the model. Note, however, that this requires using `weights_only=False` in `torch.load`, so it should be run only if the weights were obtained from a trustable source.
+If you are using `torch<=2.6.0`, some quantization methods, such as `uint4wo`, cannot be loaded directly and may result in an `UnpicklingError` when trying to load the models, but work as expected when saving them. In order to work around this, one can load the state dict manually into the model. Note, however, that this requires using `weights_only=False` in `torch.load`, so it should be run only if the weights were obtained from a trustable source.
 
 ```python
 import torch

examples/advanced_diffusion_training/README_flux.md

@@ -79,13 +79,13 @@ This command will prompt you for a token. Copy-paste yours from your [settings/t
 ### Target Modules
 When LoRA was first adapted from language models to diffusion models, it was applied to the cross-attention layers in the Unet that relate the image representations with the prompts that describe them. 
 More recently, SOTA text-to-image diffusion models replaced the Unet with a diffusion Transformer(DiT). With this change, we may also want to explore 
-applying LoRA training onto different types of layers and blocks. To allow more flexibility and control over the targeted modules we added `--lora_layers`- in which you can specify in a comma seperated string
+applying LoRA training onto different types of layers and blocks. To allow more flexibility and control over the targeted modules we added `--lora_layers`- in which you can specify in a comma separated string
 the exact modules for LoRA training. Here are some examples of target modules you can provide: 
 - for attention only layers: `--lora_layers="attn.to_k,attn.to_q,attn.to_v,attn.to_out.0"`
 - to train the same modules as in the fal trainer: `--lora_layers="attn.to_k,attn.to_q,attn.to_v,attn.to_out.0,attn.add_k_proj,attn.add_q_proj,attn.add_v_proj,attn.to_add_out,ff.net.0.proj,ff.net.2,ff_context.net.0.proj,ff_context.net.2"`
 - to train the same modules as in ostris ai-toolkit / replicate trainer: `--lora_blocks="attn.to_k,attn.to_q,attn.to_v,attn.to_out.0,attn.add_k_proj,attn.add_q_proj,attn.add_v_proj,attn.to_add_out,ff.net.0.proj,ff.net.2,ff_context.net.0.proj,ff_context.net.2,norm1_context.linear, norm1.linear,norm.linear,proj_mlp,proj_out"`
 > [!NOTE]
-> `--lora_layers` can also be used to specify which **blocks** to apply LoRA training to. To do so, simply add a block prefix to each layer in the comma seperated string:
+> `--lora_layers` can also be used to specify which **blocks** to apply LoRA training to. To do so, simply add a block prefix to each layer in the comma separated string:
 > **single DiT blocks**: to target the ith single transformer block, add the prefix `single_transformer_blocks.i`, e.g. - `single_transformer_blocks.i.attn.to_k`
 > **MMDiT blocks**: to target the ith MMDiT block, add the prefix `transformer_blocks.i`, e.g. - `transformer_blocks.i.attn.to_k` 
 > [!NOTE]

examples/advanced_diffusion_training/train_dreambooth_lora_flux_advanced.py

@@ -378,7 +378,7 @@ def parse_args(input_args=None):
         default=None,
         help="the concept to use to initialize the new inserted tokens when training with "
         "--train_text_encoder_ti = True. By default, new tokens (<si><si+1>) are initialized with random value. "
-        "Alternatively, you could specify a different word/words whos value will be used as the starting point for the new inserted tokens. "
+        "Alternatively, you could specify a different word/words whose value will be used as the starting point for the new inserted tokens. "
         "--num_new_tokens_per_abstraction is ignored when initializer_concept is provided",
     )
     parser.add_argument(
@@ -662,7 +662,7 @@ def parse_args(input_args=None):
         type=str,
         default=None,
         help=(
-            "The transformer modules to apply LoRA training on. Please specify the layers in a comma seperated. "
+            "The transformer modules to apply LoRA training on. Please specify the layers in a comma separated. "
             'E.g. - "to_k,to_q,to_v,to_out.0" will result in lora training of attention layers only. For more examples refer to https://github.com/huggingface/diffusers/blob/main/examples/advanced_diffusion_training/README_flux.md'
         ),
     )

examples/advanced_diffusion_training/train_dreambooth_lora_sd15_advanced.py

@@ -662,7 +662,7 @@ def parse_args(input_args=None):
         action="store_true",
         default=False,
         help=(
-            "Wether to train a DoRA as proposed in- DoRA: Weight-Decomposed Low-Rank Adaptation https://arxiv.org/abs/2402.09353. "
+            "Whether to train a DoRA as proposed in- DoRA: Weight-Decomposed Low-Rank Adaptation https://arxiv.org/abs/2402.09353. "
             "Note: to use DoRA you need to install peft from main, `pip install git+https://github.com/huggingface/peft.git`"
         ),
     )

examples/advanced_diffusion_training/train_dreambooth_lora_sdxl_advanced.py

@@ -773,7 +773,7 @@ def parse_args(input_args=None):
         action="store_true",
         default=False,
         help=(
-            "Wether to train a DoRA as proposed in- DoRA: Weight-Decomposed Low-Rank Adaptation https://arxiv.org/abs/2402.09353. "
+            "Whether to train a DoRA as proposed in- DoRA: Weight-Decomposed Low-Rank Adaptation https://arxiv.org/abs/2402.09353. "
             "Note: to use DoRA you need to install peft from main, `pip install git+https://github.com/huggingface/peft.git`"
         ),
     )
@@ -1875,7 +1875,7 @@ def compute_text_embeddings(prompt, text_encoders, tokenizers, clip_skip):
     # pack the statically computed variables appropriately here. This is so that we don't
     # have to pass them to the dataloader.
 
-    # if --train_text_encoder_ti we need add_special_tokens to be True fo textual inversion
+    # if --train_text_encoder_ti we need add_special_tokens to be True for textual inversion
     add_special_tokens = True if args.train_text_encoder_ti else False
 
     if not train_dataset.custom_instance_prompts:

examples/community/mixture_tiling_sdxl.py

@@ -1,4 +1,4 @@
-# Copyright 2025 The HuggingFace Team. All rights reserved.
+# Copyright 2025 The DEVAIEXP Team and The HuggingFace Team. All rights reserved.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -1070,32 +1070,32 @@ def __call__(
                     text_encoder_projection_dim = int(pooled_prompt_embeds.shape[-1])
                 else:
                     text_encoder_projection_dim = self.text_encoder_2.config.projection_dim
-                    add_time_ids = self._get_add_time_ids(
-                        original_size,
-                        crops_coords_top_left[row][col],
-                        target_size,
+                add_time_ids = self._get_add_time_ids(
+                    original_size,
+                    crops_coords_top_left[row][col],
+                    target_size,
+                    dtype=prompt_embeds.dtype,
+                    text_encoder_projection_dim=text_encoder_projection_dim,
+                )
+                if negative_original_size is not None and negative_target_size is not None:
+                    negative_add_time_ids = self._get_add_time_ids(
+                        negative_original_size,
+                        negative_crops_coords_top_left[row][col],
+                        negative_target_size,
                         dtype=prompt_embeds.dtype,
                         text_encoder_projection_dim=text_encoder_projection_dim,
                     )
-                    if negative_original_size is not None and negative_target_size is not None:
-                        negative_add_time_ids = self._get_add_time_ids(
-                            negative_original_size,
-                            negative_crops_coords_top_left[row][col],
-                            negative_target_size,
-                            dtype=prompt_embeds.dtype,
-                            text_encoder_projection_dim=text_encoder_projection_dim,
-                        )
-                    else:
-                        negative_add_time_ids = add_time_ids
+                else:
+                    negative_add_time_ids = add_time_ids
 
-                    if self.do_classifier_free_guidance:
-                        prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds], dim=0)
-                        add_text_embeds = torch.cat([negative_pooled_prompt_embeds, add_text_embeds], dim=0)
-                        add_time_ids = torch.cat([negative_add_time_ids, add_time_ids], dim=0)
+                if self.do_classifier_free_guidance:
+                    prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds], dim=0)
+                    add_text_embeds = torch.cat([negative_pooled_prompt_embeds, add_text_embeds], dim=0)
+                    add_time_ids = torch.cat([negative_add_time_ids, add_time_ids], dim=0)
 
-                    prompt_embeds = prompt_embeds.to(device)
-                    add_text_embeds = add_text_embeds.to(device)
-                    add_time_ids = add_time_ids.to(device).repeat(batch_size * num_images_per_prompt, 1)
+                prompt_embeds = prompt_embeds.to(device)
+                add_text_embeds = add_text_embeds.to(device)
+                add_time_ids = add_time_ids.to(device).repeat(batch_size * num_images_per_prompt, 1)
                 addition_embed_type_row.append((prompt_embeds, add_text_embeds, add_time_ids))
             embeddings_and_added_time.append(addition_embed_type_row)
 

examples/research_projects/anytext/README.md

@@ -0,0 +1,32 @@
+# AnyTextPipeline Pipeline
+
+Project page: https://aigcdesigngroup.github.io/homepage_anytext
+
+"AnyText comprises a diffusion pipeline with two primary elements: an auxiliary latent module and a text embedding module. The former uses inputs like text glyph, position, and masked image to generate latent features for text generation or editing. The latter employs an OCR model for encoding stroke data as embeddings, which blend with image caption embeddings from the tokenizer to generate texts that seamlessly integrate with the background. We employed text-control diffusion loss and text perceptual loss for training to further enhance writing accuracy."
+
+Each text line that needs to be generated should be enclosed in double quotes. For any usage questions, please refer to the [paper](https://arxiv.org/abs/2311.03054).
+
+
+```py
+import torch
+from diffusers import DiffusionPipeline
+from anytext_controlnet import AnyTextControlNetModel
+from diffusers.utils import load_image
+
+# I chose a font file shared by an HF staff:
+# !wget https://huggingface.co/spaces/ysharma/TranslateQuotesInImageForwards/resolve/main/arial-unicode-ms.ttf
+
+anytext_controlnet = AnyTextControlNetModel.from_pretrained("tolgacangoz/anytext-controlnet", torch_dtype=torch.float16,
+                                                            variant="fp16",)
+pipe = DiffusionPipeline.from_pretrained("tolgacangoz/anytext", font_path="arial-unicode-ms.ttf",
+                                          controlnet=anytext_controlnet, torch_dtype=torch.float16,
+                                          trust_remote_code=False,  # One needs to give permission to run this pipeline's code
+                                          ).to("cuda")
+
+# generate image
+prompt = 'photo of caramel macchiato coffee on the table, top-down perspective, with "Any" "Text" written on it using cream'
+draw_pos = load_image("https://raw.githubusercontent.com/tyxsspa/AnyText/refs/heads/main/example_images/gen9.png")
+image = pipe(prompt, num_inference_steps=20, mode="generate", draw_pos=draw_pos,
+             ).images[0]
+image
+```