docs: fix two docstrings that reference names the code does not have

src/diffusers/models/autoencoders/autoencoder_kl_ltx2.py

@@ -40,7 +40,7 @@ class PerChannelRMSNorm(nn.Module):
     def __init__(self, channel_dim: int = 1, eps: float = 1e-8) -> None:
         """
         Args:
-            dim: Dimension along which to compute the RMS (typically channels).
+            channel_dim: Dimension along which to compute the RMS (typically channels).
             eps: Small constant added for numerical stability.
         """
         super().__init__()

src/diffusers/models/transformers/transformer_ltx2.py

@@ -1446,7 +1446,7 @@ def forward(
                 (SDPA) operation. If `None` or empty, STG will not be applied to any block.
             perturbation_mask (`torch.Tensor`, *optional*):
                 Perturbation mask for STG of shape `(batch_size,)` or `(batch_size, 1, 1)`. Should be 0 at batch
-                elements where STG should be applied and 1 elsewhere. If STG is being used but `peturbation_mask` is
+                elements where STG should be applied and 1 elsewhere. If STG is being used but `perturbation_mask` is
                 not supplied, will default to applying STG (perturbing) all batch elements.
             use_cross_timestep (`bool` *optional*, defaults to `False`):
                 Whether to use the cross modality (audio is the cross modality of video, and vice versa) sigma when

src/diffusers/modular_pipelines/z_image/before_denoise.py

@@ -97,7 +97,7 @@ def calculate_dimension_from_latents(latents: torch.Tensor, vae_scale_factor_spa
     Args:
         latents (torch.Tensor): The latent tensor. Must have 4 dimensions.
             Expected shapes: [batch, channels, height, width]
-        vae_scale_factor (int): The scale factor used by the VAE to compress image spatial dimension.
+        vae_scale_factor_spatial (int): The scale factor used by the VAE to compress image spatial dimension.
             By default, it is 16
     Returns:
         tuple[int, int]: The calculated image dimensions as (height, width)

src/diffusers/optimization.py

@@ -173,7 +173,7 @@ def get_cosine_schedule_with_warmup(
             The number of steps for the warmup phase.
         num_training_steps (`int`):
             The total number of training steps.
-        num_periods (`float`, *optional*, defaults to 0.5):
+        num_cycles (`float`, *optional*, defaults to 0.5):
             The number of periods of the cosine function in a schedule (the default is to just decrease from the max
             value to 0 following a half-cosine).
         last_epoch (`int`, *optional*, defaults to -1):

src/diffusers/pipelines/deprecated/stable_diffusion_diffedit/pipeline_stable_diffusion_diffedit.py

@@ -882,10 +882,10 @@ def generate_mask(
                 not provided, `negative_prompt_embeds` are generated from the `negative_prompt` input argument.
             source_prompt (`str` or `list[str]`, *optional*):
                 The prompt or prompts to guide semantic mask generation using DiffEdit. If not defined, you need to
-                pass `source_prompt_embeds` or `source_image` instead.
+                pass `source_prompt_embeds` or `image` instead.
             source_negative_prompt (`str` or `list[str]`, *optional*):
                 The prompt or prompts to guide semantic mask generation away from using DiffEdit. If not defined, you
-                need to pass `source_negative_prompt_embeds` or `source_image` instead.
+                need to pass `source_negative_prompt_embeds` or `image` instead.
             source_prompt_embeds (`torch.Tensor`, *optional*):
                 Pre-generated text embeddings to guide the semantic mask generation. Can be used to easily tweak text
                 inputs (prompt weighting). If not provided, text embeddings are generated from `source_prompt` input