Spaces:
Running
on
Zero
Running
on
Zero
Commit
•
7187257
1
Parent(s):
c0b00f2
Try to use 2 models: one optimized for 25 f/s, another for 14 f/s (#17)
Browse files- Try to use 2 models: one optimized for 25 f/s, another for 14 f/s (4c56c8bedde0be0730d2ebb4814ee4b2ae8fdae1)
Co-authored-by: Fabrice TIERCELIN <[email protected]>
app.py
CHANGED
@@ -12,10 +12,15 @@ from PIL import Image
|
|
12 |
import random
|
13 |
import spaces
|
14 |
|
15 |
-
|
16 |
"vdo/stable-video-diffusion-img2vid-xt-1-1", torch_dtype=torch.float16, variant="fp16"
|
17 |
)
|
18 |
-
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
max_64_bit_int = 2**63 - 1
|
21 |
|
@@ -44,7 +49,10 @@ def sample(
|
|
44 |
base_count = len(glob(os.path.join(output_folder, "*.mp4")))
|
45 |
video_path = os.path.join(output_folder, f"{base_count:06d}.mp4")
|
46 |
|
47 |
-
|
|
|
|
|
|
|
48 |
export_to_video(frames, video_path, fps=fps_id)
|
49 |
|
50 |
return video_path, gr.update(label="Generated frames in *." + frame_format + " format", format = frame_format, value = frames), seed
|
|
|
12 |
import random
|
13 |
import spaces
|
14 |
|
15 |
+
fps25Pipe = StableVideoDiffusionPipeline.from_pretrained(
|
16 |
"vdo/stable-video-diffusion-img2vid-xt-1-1", torch_dtype=torch.float16, variant="fp16"
|
17 |
)
|
18 |
+
fps25Pipe.to("cuda")
|
19 |
+
|
20 |
+
fps14Pipe = StableVideoDiffusionPipeline.from_pretrained(
|
21 |
+
"stabilityai/stable-video-diffusion-img2vid", torch_dtype=torch.float16, variant="fp16"
|
22 |
+
)
|
23 |
+
fps14Pipe.to("cuda")
|
24 |
|
25 |
max_64_bit_int = 2**63 - 1
|
26 |
|
|
|
49 |
base_count = len(glob(os.path.join(output_folder, "*.mp4")))
|
50 |
video_path = os.path.join(output_folder, f"{base_count:06d}.mp4")
|
51 |
|
52 |
+
if 14 < fps_id:
|
53 |
+
frames = fps25Pipe(image, decode_chunk_size=decoding_t, generator=generator, motion_bucket_id=motion_bucket_id, noise_aug_strength=noise_aug_strength, num_frames=25).frames[0]
|
54 |
+
else:
|
55 |
+
frames = fps14Pipe(image, decode_chunk_size=decoding_t, generator=generator, motion_bucket_id=motion_bucket_id, noise_aug_strength=noise_aug_strength, num_frames=25).frames[0]
|
56 |
export_to_video(frames, video_path, fps=fps_id)
|
57 |
|
58 |
return video_path, gr.update(label="Generated frames in *." + frame_format + " format", format = frame_format, value = frames), seed
|