Fix typos in docs and comments (#11416)

* Fix typos in docs and comments

* Apply style fixes

---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

examples/dreambooth/train_dreambooth_lora_flux.py

@@ -567,7 +567,7 @@ 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. E.g. - "to_k,to_q,to_v,to_out.0" will result in lora training of attention layers only'
+            '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'
         ),
     )
 

examples/dreambooth/train_dreambooth_lora_hidream.py

@@ -596,7 +596,7 @@ 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. E.g. - "to_k,to_q,to_v" will result in lora training of attention layers only'
+            'The transformer modules to apply LoRA training on. Please specify the layers in a comma separated. E.g. - "to_k,to_q,to_v" will result in lora training of attention layers only'
         ),
     )
 

examples/dreambooth/train_dreambooth_lora_lumina2.py

@@ -514,7 +514,7 @@ 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. E.g. - "to_k,to_q,to_v" will result in lora training of attention layers only'
+            'The transformer modules to apply LoRA training on. Please specify the layers in a comma separated. E.g. - "to_k,to_q,to_v" will result in lora training of attention layers only'
         ),
     )
 

examples/dreambooth/train_dreambooth_lora_sana.py

@@ -513,7 +513,7 @@ 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. E.g. - "to_k,to_q,to_v" will result in lora training of attention layers only'
+            'The transformer modules to apply LoRA training on. Please specify the layers in a comma separated. E.g. - "to_k,to_q,to_v" will result in lora training of attention layers only'
         ),
     )
 

examples/dreambooth/train_dreambooth_lora_sd3.py

@@ -576,7 +576,7 @@ def parse_args(input_args=None):
         type=str,
         default=None,
         help=(
-            "The transformer block layers to apply LoRA training on. Please specify the layers in a comma seperated string."
+            "The transformer block layers to apply LoRA training on. Please specify the layers in a comma separated string."
             "For examples refer to https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/README_SD3.md"
         ),
     )
@@ -585,7 +585,7 @@ def parse_args(input_args=None):
         type=str,
         default=None,
         help=(
-            "The transformer blocks to apply LoRA training on. Please specify the block numbers in a comma seperated manner."
+            "The transformer blocks to apply LoRA training on. Please specify the block numbers in a comma separated manner."
             'E.g. - "--lora_blocks 12,30" will result in lora training of transformer blocks 12 and 30. For more examples refer to https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/README_SD3.md'
         ),
     )

examples/dreambooth/train_dreambooth_lora_sdxl.py

@@ -664,7 +664,7 @@ 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/flux-control/train_control_lora_flux.py

@@ -329,7 +329,7 @@ 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. E.g. - "to_k,to_q,to_v,to_out.0" will result in lora training of attention layers only'
+            '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'
         ),
     )
     parser.add_argument(

examples/kandinsky2_2/text_to_image/train_text_to_image_lora_decoder.py

@@ -400,7 +400,7 @@ def main():
 
     image_encoder.requires_grad_(False)
 
-    # For mixed precision training we cast all non-trainable weigths (vae, non-lora text_encoder and non-lora unet) to half-precision
+    # For mixed precision training we cast all non-trainable weights (vae, non-lora text_encoder and non-lora unet) to half-precision
     # as these weights are only used for inference, keeping weights in full precision is not required.
     weight_dtype = torch.float32
     if accelerator.mixed_precision == "fp16":

examples/research_projects/consistency_training/train_cm_ct_unconditional.py

@@ -1147,7 +1147,7 @@ def main(args):
         tracker_config = dict(vars(args))
         accelerator.init_trackers(args.tracker_project_name, config=tracker_config)
 
-    # Function for unwraping if torch.compile() was used in accelerate.
+    # Function for unwrapping if torch.compile() was used in accelerate.
     def unwrap_model(model):
         model = accelerator.unwrap_model(model)
         model = model._orig_mod if is_compiled_module(model) else model

examples/research_projects/flux_lora_quantization/README.md

@@ -69,7 +69,7 @@ accelerate launch --config_file=accelerate.yaml \
   --seed="0"
 ```
 
-We can direcly pass a quantized checkpoint path, too:
+We can directly pass a quantized checkpoint path, too:
 
 ```diff
 + --quantized_model_path="hf-internal-testing/flux.1-dev-nf4-pkg"

examples/research_projects/intel_opts/inference_bf16.py

@@ -13,7 +13,7 @@ args = parser.parse_args()
 
 
 device = "cpu"
-prompt = "a lovely <dicoo> in red dress and hat, in the snowly and brightly night, with many brighly buildings"
+prompt = "a lovely <dicoo> in red dress and hat, in the snowly and brightly night, with many brightly buildings"
 
 model_id = "path-to-your-trained-model"
 pipe = StableDiffusionPipeline.from_pretrained(model_id)

examples/research_projects/intel_opts/textual_inversion_dfq/README.md

@@ -80,7 +80,7 @@ export INT8_MODEL_NAME="./int8_model"
 
 python text2images.py \
   --pretrained_model_name_or_path=$INT8_MODEL_NAME \
-  --caption "a lovely <dicoo> in red dress and hat, in the snowly and brightly night, with many brighly buildings." \
+  --caption "a lovely <dicoo> in red dress and hat, in the snowly and brightly night, with many brightly buildings." \
   --images_num 4
 ```
 

examples/research_projects/pixart/pipeline_pixart_alpha_controlnet.py

@@ -664,7 +664,7 @@ class PixArtAlphaControlnetPipeline(DiffusionPipeline):
         # &amp
         caption = re.sub(r"&amp", "", caption)
 
-        # ip adresses:
+        # ip addresses:
         caption = re.sub(r"\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}", " ", caption)
 
         # article ids:

examples/research_projects/pixart/train_pixart_controlnet_hf.py

@@ -612,7 +612,7 @@ def main():
     # See Section 3.1. of the paper.
     max_length = 120
 
-    # For mixed precision training we cast all non-trainable weigths (vae, text_encoder) to half-precision
+    # For mixed precision training we cast all non-trainable weights (vae, text_encoder) to half-precision
     # as these weights are only used for inference, keeping weights in full precision is not required.
     weight_dtype = torch.float32
     if accelerator.mixed_precision == "fp16":

examples/research_projects/pytorch_xla/inference/flux/flux_inference.py

@@ -120,11 +120,11 @@ if __name__ == "__main__":
     parser.add_argument("--schnell", action="store_true", help="run flux schnell instead of dev")
     parser.add_argument("--width", type=int, default=1024, help="width of the image to generate")
     parser.add_argument("--height", type=int, default=1024, help="height of the image to generate")
-    parser.add_argument("--guidance", type=float, default=3.5, help="gauidance strentgh for dev")
+    parser.add_argument("--guidance", type=float, default=3.5, help="guidance strength for dev")
     parser.add_argument("--seed", type=int, default=None, help="seed for inference")
     parser.add_argument("--profile", action="store_true", help="enable profiling")
     parser.add_argument("--profile-duration", type=int, default=10000, help="duration for profiling in msec.")
-    parser.add_argument("--itters", type=int, default=15, help="tiems to run inference and get avg time in sec.")
+    parser.add_argument("--itters", type=int, default=15, help="items to run inference and get avg time in sec.")
     args = parser.parse_args()
     if args.schnell:
         ckpt_id = "black-forest-labs/FLUX.1-schnell"

examples/research_projects/realfill/train_realfill.py

@@ -759,7 +759,7 @@ def main(args):
         unet, text_encoder, optimizer, train_dataloader
     )
 
-    # For mixed precision training we cast all non-trainable weigths (vae, non-lora text_encoder and non-lora unet) to half-precision
+    # For mixed precision training we cast all non-trainable weights (vae, non-lora text_encoder and non-lora unet) to half-precision
     # as these weights are only used for inference, keeping weights in full precision is not required.
     weight_dtype = torch.float32
     if accelerator.mixed_precision == "fp16":

examples/research_projects/scheduled_huber_loss_training/dreambooth/train_dreambooth_lora_sdxl.py

@@ -661,7 +661,7 @@ 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/textual_inversion/textual_inversion.py

@@ -789,7 +789,7 @@ def main():
         text_encoder, optimizer, train_dataloader, lr_scheduler
     )
 
-    # For mixed precision training we cast all non-trainable weigths (vae, non-lora text_encoder and non-lora unet) to half-precision
+    # For mixed precision training we cast all non-trainable weights (vae, non-lora text_encoder and non-lora unet) to half-precision
     # as these weights are only used for inference, keeping weights in full precision is not required.
     weight_dtype = torch.float32
     if accelerator.mixed_precision == "fp16":

examples/textual_inversion/textual_inversion_sdxl.py

@@ -814,7 +814,7 @@ def main():
         text_encoder_1, text_encoder_2, optimizer, train_dataloader, lr_scheduler
     )
 
-    # For mixed precision training we cast all non-trainable weigths (vae, non-lora text_encoder and non-lora unet) to half-precision
+    # For mixed precision training we cast all non-trainable weights (vae, non-lora text_encoder and non-lora unet) to half-precision
     # as these weights are only used for inference, keeping weights in full precision is not required.
     weight_dtype = torch.float32
     if accelerator.mixed_precision == "fp16":

scripts/convert_flux_to_diffusers.py

@@ -220,7 +220,7 @@ def convert_flux_transformer_checkpoint_to_diffusers(
             f"double_blocks.{i}.txt_attn.proj.bias"
         )
 
-    # single transfomer blocks
+    # single transformer blocks
     for i in range(num_single_layers):
         block_prefix = f"single_transformer_blocks.{i}."
         # norm.linear  <- single_blocks.0.modulation.lin