Other
Transformers
Safetensors
ldf_motion
image-feature-extraction
text-to-motion
motion-generation
diffusion-forcing
humanml3d
computer-animation
custom_code
Instructions to use ShandaAI/FloodDiffusion with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ShandaAI/FloodDiffusion with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ShandaAI/FloodDiffusion", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
Commit ·
b70b9d7
1
Parent(s): 82d5f99
stream dropout update
Browse files
ldf_models/diffusion_forcing_wan.py
CHANGED
|
@@ -226,12 +226,12 @@ class DiffForcingWanModel(nn.Module):
|
|
| 226 |
text_list, text_end_list
|
| 227 |
):
|
| 228 |
if np.random.rand() > self.drop_out:
|
| 229 |
-
single_text_list = [""]
|
| 230 |
-
single_text_end_list = [0, seq_len]
|
| 231 |
-
else:
|
| 232 |
single_text_end_list = [0] + [
|
| 233 |
min(t, seq_len) for t in single_text_end_list
|
| 234 |
]
|
|
|
|
|
|
|
|
|
|
| 235 |
single_text_length_list = [
|
| 236 |
t - b
|
| 237 |
for t, b in zip(
|
|
|
|
| 226 |
text_list, text_end_list
|
| 227 |
):
|
| 228 |
if np.random.rand() > self.drop_out:
|
|
|
|
|
|
|
|
|
|
| 229 |
single_text_end_list = [0] + [
|
| 230 |
min(t, seq_len) for t in single_text_end_list
|
| 231 |
]
|
| 232 |
+
else:
|
| 233 |
+
single_text_list = [""]
|
| 234 |
+
single_text_end_list = [0, seq_len]
|
| 235 |
single_text_length_list = [
|
| 236 |
t - b
|
| 237 |
for t, b in zip(
|