README.md: fix demo code.

README.md

@@ -124,7 +124,6 @@ pip install -U git+https://github.com/huggingface/diffusers
 Now, you can run the examples below (note that the upsampling stage is optional but reccomeneded):
 
 ### text-to-video:
-```
 ```py
 import torch
 from diffusers import LTXConditionPipeline, LTXLatentUpsamplePipeline
@@ -137,6 +136,11 @@ pipe.to("cuda")
 pipe_upsample.to("cuda")
 pipe.vae.enable_tiling()
 
+def round_to_nearest_resolution_acceptable_by_vae(height, width):
+    height = height - (height % pipe.vae_spatial_compression_ratio)
+    width = width - (width % pipe.vae_spatial_compression_ratio)
+    return height, width
+
 prompt = "The video depicts a winding mountain road covered in snow, with a single vehicle traveling along it. The road is flanked by steep, rocky cliffs and sparse vegetation. The landscape is characterized by rugged terrain and a river visible in the distance. The scene captures the solitude and beauty of a winter drive through a mountainous region."
 negative_prompt = "worst quality, inconsistent motion, blurry, jittery, distorted"
 expected_height, expected_width = 704, 512
@@ -145,6 +149,7 @@ num_frames = 121
 
 # Part 1. Generate video at smaller resolution
 downscaled_height, downscaled_width = int(expected_height * downscale_factor), int(expected_width * downscale_factor)
+downscaled_height, downscaled_width = round_to_nearest_resolution_acceptable_by_vae(downscaled_height, downscaled_width)
 latents = pipe(
     conditions=None,
     prompt=prompt,
@@ -154,7 +159,7 @@ latents = pipe(
     num_frames=num_frames,
     num_inference_steps=7,
     decode_timestep = 0.05,
-    guidnace_scale=1.0,
+    guidance_scale=1.0,
     decode_noise_scale = 0.025,
     generator=torch.Generator().manual_seed(0),
     output_type="latent",
@@ -178,7 +183,7 @@ video = pipe(
     num_inference_steps=10,
     latents=upscaled_latents,
     decode_timestep = 0.05,
-    guidnace_scale=1.0,
+    guidance_scale=1.0,
     decode_noise_scale = 0.025,
     image_cond_noise_scale=0.025,
     generator=torch.Generator().manual_seed(0),
@@ -205,6 +210,11 @@ pipe.to("cuda")
 pipe_upsample.to("cuda")
 pipe.vae.enable_tiling()
 
+def round_to_nearest_resolution_acceptable_by_vae(height, width):
+    height = height - (height % pipe.vae_spatial_compression_ratio)
+    width = width - (width % pipe.vae_spatial_compression_ratio)
+    return height, width
+
 image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/penguin.png")
 video = [image]
 condition1 = LTXVideoCondition(video=video, frame_index=0)
@@ -226,7 +236,7 @@ latents = pipe(
     height=downscaled_height,
     num_frames=num_frames,
     num_inference_steps=7,
-    guidnace_scale=1.0,
+    guidance_scale=1.0,
     decode_timestep = 0.05,
     decode_noise_scale = 0.025,
     generator=torch.Generator().manual_seed(0),
@@ -250,7 +260,7 @@ video = pipe(
     num_frames=num_frames,
     denoise_strength=0.3,  # Effectively, 4 inference steps out of 10
     num_inference_steps=10,
-    guidnace_scale=1.0,
+    guidance_scale=1.0,
     latents=upscaled_latents,
     decode_timestep = 0.05,
     decode_noise_scale = 0.025,
@@ -263,7 +273,6 @@ video = pipe(
 video = [frame.resize((expected_width, expected_height)) for frame in video]
 
 export_to_video(video, "output.mp4", fps=24)
-
 ```
 
 ### For video-to-video: 
@@ -281,9 +290,10 @@ pipe_upsample.to("cuda")
 pipe.vae.enable_tiling()
 
 def round_to_nearest_resolution_acceptable_by_vae(height, width):
-    height = height - (height % pipe.vae_temporal_compression_ratio)
-    width = width - (width % pipe.vae_temporal_compression_ratio)
+    height = height - (height % pipe.vae_spatial_compression_ratio)
+    width = width - (width % pipe.vae_spatial_compression_ratio)
     return height, width
+
 video = load_video(
     "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cosmos/cosmos-video2world-input-vid.mp4"
 )[:21]  # Use only the first 21 frames as conditioning
@@ -306,7 +316,7 @@ latents = pipe(
     height=downscaled_height,
     num_frames=num_frames,
     num_inference_steps=7,
-    guidnace_scale=1.0,
+    guidance_scale=1.0,
     decode_timestep = 0.05,
     decode_noise_scale = 0.025,
     generator=torch.Generator().manual_seed(0),
@@ -331,7 +341,7 @@ video = pipe(
     num_frames=num_frames,
     denoise_strength=0.3,  # Effectively, 4 inference steps out of 10
     num_inference_steps=10,
-    guidnace_scale=1.0,
+    guidance_scale=1.0,
     latents=upscaled_latents,
     decode_timestep = 0.05,
     decode_noise_scale = 0.025,
@@ -344,8 +354,8 @@ video = pipe(
 video = [frame.resize((expected_width, expected_height)) for frame in video]
 
 export_to_video(video, "output.mp4", fps=24)
-
 ```
+
 To learn more, check out the [official documentation](https://huggingface.co/docs/diffusers/main/en/api/pipelines/ltx_video). 
 
 Diffusers also supports directly loading from the original LTX checkpoints using the `from_single_file()` method. Check out [this section](https://huggingface.co/docs/diffusers/main/en/api/pipelines/ltx_video#loading-single-files) to learn more.