patrickvonplaten commited on
Commit
13c351d
1 Parent(s): f61a46b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -76,7 +76,7 @@ Here are some results:
76
  ## Long Video Generation
77
 
78
  You can optimize for memory usage by enabling attention and VAE slicing and using Torch 2.0.
79
- This should allow you to generate videos up to 10 seconds on less than 16GB of GPU VRAM.
80
 
81
  ```bash
82
  $ pip install git+https://github.com/huggingface/diffusers transformers accelerate
@@ -96,8 +96,8 @@ pipe.enable_model_cpu_offload()
96
  pipe.enable_vae_slicing()
97
 
98
  # generate
99
- prompt = "Spiderman is surfing"
100
- video_frames = pipe(prompt, num_inference_steps=25, num_frames=80).frames
101
 
102
  # convent to video
103
  video_path = export_to_video(video_frames)
 
76
  ## Long Video Generation
77
 
78
  You can optimize for memory usage by enabling attention and VAE slicing and using Torch 2.0.
79
+ This should allow you to generate videos up to 25 seconds on less than 16GB of GPU VRAM.
80
 
81
  ```bash
82
  $ pip install git+https://github.com/huggingface/diffusers transformers accelerate
 
96
  pipe.enable_vae_slicing()
97
 
98
  # generate
99
+ prompt = Spiderman is surfing. Darth Vader is also surfing and following Spiderman"
100
+ video_frames = pipe(prompt, num_inference_steps=25, num_frames=200).frames
101
 
102
  # convent to video
103
  video_path = export_to_video(video_frames)